Professional Documents
Culture Documents
PYC3704 Previous Exam and Assignment Answers: Thorn
PYC3704 Previous Exam and Assignment Answers: Thorn
Assignment Answers
written by
Thorn
Buy and sell all your summaries, notes, theses, essays, papers, cases, manuals, researches, and
many more...
www.stuvia.co.za
Assignments Covered:
• Feedback Assignment 01 of 2016 Semester 2
• Feedback Assignment 02 of 2016 Semester 2
• Feedback Assignment 01 of 2019 Semester 2
• Feedback Assignment 01 of 2019 Semester 2
written by
ThorO
www.stuvia.co.uk
Oct/Nov 2015
QUESTION A PAGE QUESTION A PAGE NR
1. A psychological theory is best defined as - - - - - N
2 NR S
Study 2.ANSA master’s student in psychology plans to study human N2 S
S guide: S
consciousness, following the recent publication of
1. a set of observations of human behavior P99-100
Watson's theory about the nature of consciousness. The
2. postulated relationships among constructs
main aim of the research will probably be to empirically -
3. statistical inferences based on measurements
----
1. test predictions based on Watson's theory
2. test Watson's theory so that it can be accepted
or rejected as a whole
3. study consciousness with a view to
understanding, predicting and controlling it
3. Which of the following best describes "latent"? 3 Study 4. The (a) - - -variable has an effect on the (b) variable Study
guide: P7 guide: P8-9,
1. observable 1. (a) dependent (b) independent 24
2. manifest 2. (a) operational (b) measured
3. hidden
3. (a) independent (b) dependent
5. A psychologist conducts a study in which she measures the 3 Study 6. In a study, the relationship between anxiety (high and 2 Study
guide: low) and test performance (measured in terms of pass guide: P15,
reaction times of students doing a psychometric test. She
P7,11,18 and fail) is considered A suitable hypothesis for the 71,99
proceeds from the assumption that good reaction time is an
study can be viewed as - - - - -
indication of intelligence in this study 'intelligence' is the
variable 1. a rule suggesting how the values of 'anxiety' may be
related to the values of 'test performance'
1. 0.07
2. 0.15
3. 0.32
17. Consider the following Table 2 Study 18. The mean of the standard normal distribution equals - - 2 Study
guide: - while its standard deviation equals - - guide: P52-
Subject P55-60 55
Mean for Mean for Standard deviation for
Student X class class 1.
A 50% 40% 5%
B 55% 50% 5% 2.
C 60% 50% 10%
3.
In which subject did a student (indicated as Student X in the
second column) do best, relative to the class?
1. A
2. B
3. C
19. Study the histogram below of the exam marks (out of 60) of 1 Study 20. When a sample is randomly selected from a population, 2 Study
a group of students in the same class. (Note that the values guide: the sampling error depends on the - - - - - guide: P60-
on the horizontal axis are the class limits) P29, 61
40,47
1. the size of the sample
130-132
142-144 2. the size of the population
3. the population mean
1. 0.3
2. 0.5
3. 0.7
Questions 22 and 23 are based on the following scenario
The mean mark obtained for a psychology assignment by all the students who completed it is 35, and the standard deviation 15. John received a mark of 45 for this psychology
assignment
1. 0.7486 1. 0.16
2. 0.6667 2. 0.84
3. 0.2514 3. 0.34
25. Study the following figure of a normal curve for the 1 Study 26. Sibu is doing research on the IQ scores of the students 2 Study
distribution of a specific set of data guide: Assuming that this IQ will on average be greater than guide: P81
P163 that of the overall population, he states the following
hypothesis
1. Around 100
For a random measurement x from this population, what is 2. Around 107
p(x<155)? 3. More than 100
1. 0 84
2. 0.16
3. 0.50
29. Which of the following is likely to be specified in advance by 3 Study 30. When two population means are compared, the p-value 2 Study
the researcher, before any actual data are collected? guide: expresses the probability that a (n) guide: P87,
P82-86 P139
1. The effect size 1. observed difference between the means is due to
2. The level of significance sampling error
3. The expected p-value under Ho 2. difference found between the means is due to the
alternative hypothesis
3. difference will be found between the means
31. When doing a one-sample statistical test, the failure to 3 Study 32. The level of significance is the - - - - - 2 Study
reject H0 implies that a difference between the calculated guide: guide: P85
sample mean and its expected value as predicted under Ho P86 1. maximum allowable probability of Type II error
is due to - - - - - 2. maximum allowable probability of Type I error
3. probability that not rejecting the null hypothesis would
1. the dependent variable be an error
2. the independent variable
3. chance
33. The null hypothesis is rejected if - - - - - 3 Study 34. The power of a statistical test refers to the - - - - - 2 Study
guide: guide: P85-
1. the p-value under H0 is smaller than the level of P86-87 1. ability of the test to give small p-values 86
significance 2. ability of the test to detect significant results
2. the p-value under H1 is larger than the level of significance 3. the sample size
3. the p-value is smaller than 0.5 or 0. 1
Peter is a human resource consultant and his boss asks him to test the employees in their company to determine if the level of negative attitude
towards people with AIDS of these workers in some way differs from that of the general worker in South Africa. Suppose that on the basis of previous
studies it is accepted that the mean negative attitude score on a particular scale of the population of workers in South Africa is 55 (the higher the score
the more positive) and that the population standard deviation
39. Which of the following statements translates the research 3 Study 40. Peter finds that the workers in his company have a 2 Study
hypothesis into the appropriate statistical hypotheses? guide: mean attitude score of 50 Which of the statistical test guide: P6
P11-18, procedures below would be most appropriate to use?
21-26
1. H0 µ = 55 H1 µ < 55
1 A t-test for a single sample mean
2. H0 µ = 55 H1 µ > 55 2 A z-test for a single sample mean
3 A t-test for two sample means
3. H0 µ = 55 H1 µ at 55
An educational psychologist compares two groups of schoolchildren from urban and semi-urban areas on a language comprehension test. She manages to get a sample of
600 urban and 400 semi-urban children (matched for gender and age) to complete a test. She finds mean language comprehension scores of for
the urban and semi-urban groups respectively.
42. The educational psychologist decides to do a t-test to 2 Study 43. The psychologist calculates the appropriate t-test to 2 Study
compare the two sample means. Which of the following is guide: compare the two means, and the result is a t-statistic guide: P6
the appropriate t-test statistic that she should calculate? value of 2.67 She determines that this is significant
on the 1% level. The researcher is however, concerned
1. that the actual difference between the two-group
means is fairly small, and that the significant result may
2. be the consequence of the large sample sizes. What
could she do to check whether this is significant results
3. is meaningful in a practical sense?
An educational psychologist investigates what effect extra classes where learners are taught reasoning skills would have on the performance of pupils in mathematics. The
grade twelve learners in a specific school are divided into two groups One group of 20 pupils receives the extra classes in reasoning skills while a control group of 20 pupils
receives singing lessons in the same period. The teacher calculates the increase or decrease in his or her mathematics performance for each of the 40 pupils based on a
mathematics test before the extra classes and a second mathematics test after the extra classes for both groups.
55. Which research design is the teacher most likely to use? 2 Study 56. The most appropriate statistical test to use to compare 3 Study
guide: the two groups of learners with regard to the change in guide: P100-
1. A matched-pairs design P100-106 their mathematics test results would be the - - - - - 106
2. A two-sample design with independent groups
3. A two-sample design with dependent groups 1. t-test for dependent groups
2. chi-square test
3. t-test for independent groups
1. Ho can be rejected.
2. H0 cannot be rejected.
3. No conclusion is possible before the relevant test
statistic and p-value are calculated
1 problem-solving style
2 exam marks
3 accomplishment
61. A politician asks his audience of 100 whether they will vote 1 Study 62. A researcher predicts that a motivational talk will 1 Study
for him, and 60 say yes, He then delivers his speech and guide: improve the work performance of men. However, when guide: P75-
repeats the question Now 70 persons say yes. When P112 testing this hypothesis, he finds that the mean work 76
analyzing these results, the two sets of answers should be performance of his sample of 20 men is actually poorer
regarded as - - - - - after the motivational talk than before Based on these
results, what statistical test should the researcher do to
1. dependent make a conclusion regarding the hypothesis possible?
2. independent
3. normally distributed 1. A one-tailed statistical test is required
2. A two-tailed statistical test is required
3. No statistical test is required
63. Which of the following does not represent a valid value for a 2 Study 64. As the sample size (n) increases - - - - - 1 Study
Pearson's r ? guide: guide: P60-
P133 1. a smaller value of the Pearson's correlation coefficient 61
1. 0.00 r will reach significance
2. 1.09 2. a larger value of the Pearson's correlation coefficient r
3. -0.99 is required before the result will be significant
3. there are no implications for the significance of the
value of the Pearson's correlation coefficient r
May/June 2017
QUESTION A PAGE COMMENTS
1. The term 'inference’ in psychological research refers to N2 Study
NR. An inference is a conclusion that follows from existing information, by
S guide: P2 - generalising from the specific information to the general type of
1. the process of setting up a hypothesis as a relationship among variables phenomenon, where the conclusion is not absolutely certain. So, in
summary inferential statistics are techniques for making generalisations
2. making a prediction or generalization based on existing information
based on imperfect numeric data, where the conclusions have a high
3. the procedures for making a construct visible so that a measurement can probability of being true, but you can never be completely certain.
be made
2. In psychological research, a construct may be a (n) 3 Study Constructs and their interrelations (how they affect each other, their
guide: P4 patterns of interaction) are used in this way to develop theoretical
1 measurement based on the careful observation of aspects of humans or explanations of why people behave in certain ways in certain contexts,
human behavior or why mental phenomena appear to be as they are.
2. observation of an aspect of humans or human behavior which was
operationalized in some way
3. hypothetical aspect of humans or human behavior which we wish to
investigate
3. Which of the options below provides the best description of the main purpose 3 Study Psychology is a discipline that endeavors to collect information and
of quantitative research in psychology? its purpose is to guide: P1- develop theories about human behavior and mental processes. The
2 aim is to establish facts that are related to psychological phenomena,
1. develop theories that explain the relationships among observed aspects of that are valid and can be justified on scientific grounds.
human behavior and mental processes
2. develop predictions about human behavior of which we can be applied with
absolute certainty
3. develop hypotheses about relationships that may exist among various
constructs
4. Operationalizing a construct means to 3 Study 'Operationalization' is where you make the construct (which is usually
guide: an abstract concept, so it is difficult to observe it clearly) visible by
1 find an explanation for the construct to explain why it appears as it is P25 finding some suitable way to measure it. You need it to be able to test a
2. make an educated guess on how it relates to other constructs 1.2.1& P4 hypothesis, but it is not in itself 'the process of forming a hypothesis'.
3. devise a systematic procedure to make the construct observable, in such a 1.2.2 The primary aim of operationalization is to describe a construct clearly
way that we can measure it and unambiguously so that it can be measured and tested in a research
study.
“Generalised anxiety disorder (GAD) refers to a pattern of almost constant worry or tension, even when there is little or no apparent cause. Both genetic predisposition and
stressors in the life of a particular patient is believed to contribute to this condition. The research will investigate whether the level of anxiety of persons diagnosed with GAD is
actually reduced by psychotherapy. it is expected that patients receiving therapy will score lower on the Manifest Anxiety Scale than patients not receiving therapy "
6. Both genetic predisposition and stressors in the life of a particular patient is 2 Study A theory is a well-established principle that has been developed to
believed to contribute to this condition' is guide: explain some aspect of the natural world. A theory arises from repeated
P4, 15, observation and testing and incorporates facts, laws, predictions, and
18-19,
1. the research hypothesis 21-26 tested hypotheses that are widely accepted. In science, a theory is a
2. a theory about the causes of GAD framework for facts. It is some kind of description that tells you how the
3. a postulated relation between two constructs facts are connected, and why the facts are as they are (where the word
'facts' refers to things or events that were observed and described in a
careful way). A theory is a network of relations among facts that were
proposed to be true and explanations for observed phenomena in terms
of constructs.
7. "Whether the level of anxiety of persons diagnosed with GAD is actually 3 Study
reduced by psychotherapy" describes guide: P4
P15
1. a theoretical prediction about the effect of psychotherapy P18-19
2. the operationalization of the construct 'anxiety’ P21-26
3. the hypothesis to be investigated
8. The dependent variable is ______ and the independent variable is ______ 3 Study The dependent variable is the one that is predicted or explained, and
guide: P8- the independent variable is manipulated to see how it affects the
1. whether psychotherapy is received, the level of anxiety experienced by 9, 24,113- dependent variable.
116
patients
2. the effectiveness of psychotherapy, the level of anxiety The independent variable is that variable which affects the dependent
3. the level of anxiety experienced by patients, whether or not psychotherapy variable; or, conversely, the dependent variable depends on the
is received independent variable.
10. A researcher would use a ______ to make a(n) ______ about the nature of 2 Study A construct that has been measured in some way produces a variable.
the ______ guide: P7 A variable refers to a number that can take on any one of a range of
possible values. They can be discrete (when only whole numbers like 1,
1. sample, inference, population 2, 3 are allowed) or continuous (what mathematicians refer to as 'real
3. variable, prediction, construct numbers'). In some cases, variables also take on values smaller than
4. population, inference, sample zero to produce negative numbers. So, the (visible) variable reflects the
intensity of the underlying (invisible) construct, in terms of how it was
measured. We say that the variable is manifest (it is visible in the sense
that we can observe it) and the construct is latent (it is invisible in the
sense that we need some way to make it appear). So, the latent
construct is made manifest by the use of an appropriate measurement
procedure. The dependent variable is the one that is predicted or
explained, and the independent variable is manipulated to see how it
affects the dependent variable.
11. A measurement that summarizes an aspect of a population is called a_____ 2 Study Parameters are values that summarize aspects of population data.
while a measurement that describes the same aspect of a sample is called guide: While the word 'parameters' does refer to descriptive statistics, it does
P14, 23
not refer to all descriptive statistics. it is used only for those descriptive
1. construct, variable statistics that relate to the population, not to those that describe aspects
2. parameter, statistic of the sample.
3. statistic, parameter
A statistic is a sample measurement characteristic.
A test statistic is the quantity you calculate (often by making use of
sample statistics) to test a statistical hypothesis.
When we refer to these test quantities, we always refer to the name in
full -'test statistic', and when we use the term ‘statistic' on its own it
refers to a descriptive statistic that describes an aspect of the sample
data.
18. During the interpretation of psychological measurements, the normal 2 Study Many of the scores that we use are also clustered around the average
distribution is often ______ guide: P50 and tail off to the ends of the distribution. Because it can be used to
-51, 52-53 describe the distribution of many naturally or 'normally' occurring
1. adapted to fit the observed frequency distribution of scores continuous variables, this type of symmetrical probability distribution is
2. used as a theoretical model for interpreting the observed distribution of called a normal distribution. it is also commonly referred to as the
scores normal curve, because the distribution can be plotted by a bell-shaped
3. used to calculate the relative frequency of observed scores curve, (see section 2.3.3 on p52-53)
19. The scale along the x-axis of the standard normal distribution indicates____ 3 Study
guide:
1. probabilities P52-53
2. the number of standard deviations below and above the mean
3. the p-values
21. Suppose the height of military recruits is distributed normally with a mean of 2 Study The standard error is an extremely valuable measure because we can
1750 mm and a standard deviation of 50 mm. Drawing repeated samples of guide: use it to estimate how well a sample mean approximates its population
25 recruits each, we expect the standard deviation of the sample means to be P61-62, mean in general, that is, how much error you can expect on average
109
about ______ mm between the sample mean that you calculated from your sample and
the population mean (μ) that you are trying to estimate. In other words,
1. 2 it is an indication of the size of the error that you make by using a
2. 10 sample of a particular size (n) to determine the population mean.
3. 50 This amount of error will decrease as the size of the sample increases.
22. Which of the following formulas give an appropriate way in which the 3 Study
probability of a specific event E can be calculated? guide:
P28-31, 46-
50
24. Consider the following Table which represents the mark of a student (X) on 1 Study Use explanation in study guide from section 2.3.4
each of three subjects and the class averages for those subjects. guide:
P55-60 The marks should first be converted to z-values, to make it possible to
Subject Student X Mean of Standard deviation
compare them across the different means and standard deviations:
class of class
History 50% 40% 5% _ _
Geography 62% 62% 5% Z = X – X or Z= (X - X) / S
Mathematics 60% 50% 10% S
History = (50 - 40) / 5 = 10/5 = 2
In which subject did Student X do best, relative to his class? Geography = (62 - 62) / 5 = 5/5 = 1
Mathematics = (60 - 50) / 5 = 10/10 = 1
1. History
2. Geography
3. Mathematics So, it is clear that in the case of subject A, the student’s marks are 2
standard deviations above the mean. In the other subjects the student’s
marks are 1 standard deviation or less above the mean.
OR
To investigate her hypothesis, she obtains a random sample of n=100 persons older than 65. Each member of this sample is given a daily dose of supplements over a period of six
months. At the end of this time, each person is tested on the NPS test and a mean of ẋ = 86 is found. The researcher plans to test the hypothesis at α = 0.05.
26. The appropriate research hypothesis suggested by the scenario above is as 3 Study A psychological hypothesis formulates a testable empirical claim
follows guide: (something that can in principle be observed), and this usually involves
P15, 71 postulating a relationship between two or more variables
1. Cognitive functioning declines with age
2. The cognitive functioning of elderly persons is related to their perceptual
speed
3. Rate of object recognition will be better for elderly persons who take the
dietary supplement than for those who do not
28. The mean of the sampling distribution of the mean is ______ 1 Study The sampling distribution of means refers to the distribution of the
guide: means of all possible samples of a particular size randomly selected
1. 80 P60-61 from the same population
2. 86
3. 20
29. The standard error is ______ 2 Study We can estimate the size of the error we would make if we used the
guide: P60-
sample mean as an estimate of the population mean. This is referred to
61, 103-104
1. 20 as the standard error, and it is specified in the central limit theorem.
2. 2 The standard error is denoted by σẋ. The σ indicates that we are
3. 0.05 describing a population, and the subscript ẋ informs us that we are
dealing with a population of sample means. The standard error is given
by dividing the population standard deviation by the square root of the
sample size
σẋ = σ / √n
if σ = 20 and n = 100, then
σẋ = 20 / √100
= 20 / 10
=2
30. With the information as given in the scenario, what would be the appropriate 1 Study The t-distribution is a statistical distribution with a probability distribution
statistical test to test hypothesis? guide: that can be determined, which means that we can use it to predict the
P100-106 chances of obtaining specific outcomes when testing for comparisons
1. A one sample z-test of means when the population standard deviation σ is unknown. So, we
2. A two sample t-test have to use the t-test (tx) when the population standard deviation (σ) is
3. A test of correlation r for relationship between considered to be unknown - because the given standard deviation
variables comes from the sample.
When the population standard deviation (σ) is known we use the z-test
(zx)
33. A statistical hypothesis is a formal statement about ______ 1 Study Take note that a research hypothesis always translates into two
guide: P74 mutually exclusive hypotheses (1.e. both cannot be true at the same
1. population parameters time): a null and an alternative hypothesis. Also remember that, in
2. sample statistics Topic 1, we referred to quantities such as parameters (population
3. test statistics parameters). These particular statistical hypotheses are, thus,
statements about the value of a particular population parameter.
34. The sampling distribution of a statistic (e g of the sample mean) can be 2 Study
calculated if we assume that the ____ hypothesis is true, but not if we assume guide:
that the ____ hypothesis is true P58, 77-79
1. alternative, null
2. null, alternative
3. research, statistical
You are in fact comparing the observed relationships in the data with
what you would expect if the null hypothesis is true by calculating a
relevant test statistic. This test statistic can then be used to find the p-
value if we know the probability distribution of the test statistic. if this
probability is small, it implies the null hypothesis is probably not true.
46. After findings that a significant difference exists between male and female 1 Study One way that statisticians have suggested to deal with this problem is
participants on a test which tests level of creativity, a researcher decides to guide: by the notion of effect size. Different procedures exist to determine the
also calculate an effect size, using Cohen's d. This is used to determine P86-87 effect size of a result. In the case of a comparison between means, one
way of calculating this is by the use of Cohen's d.
1. whether a significant effect is sufficiently great to be meaningful from a
practical point of view Cohen’s d test.
2. the size of the error that would be made if the null hypothesis is rejected
3. the ability of a statistical test to detect a significant relationship between
variables
She uses a random sample of n=64 drawn from the population of diagnosed
patients and finds that ẋ = 127 and s = 24
Which of the values below is the closest to the correct value of sẋ?
1. 0.37
2. 3.0
3. 24
48. Suppose H0 μ = 100 is tested against H1μ > 100 with α=0.05. if the t-statistic 3 Study H0: m = 100 H1: m > 100
is found to be tẋ = 1.20 and the two tailed p-value is 0.06, what decision guide: P75 Here, He specifies that only sample means greater than 100 will be
regarding the statistical hypothesis can be taken? considered as possible evidence for rejecting H0. Only sample results
different from 100 in a particular direction will thus be considered. For
this reason, H1here indicates that a directional or one-tailed test of H0 is
1. Reject Ho and accept H1 required.
2. Reject H1and accept Ho
3. Do not reject Ho, and accept H1
49. Suppose the alternative hypothesis states that μ > 60. The researcher should 1 Study H1: μ > 60
test H0 against H1 if the ______ guide: μ = sample mean
P106 μ > 60 is directional indicating larger than.
1. sample mean is larger than 60 if the sample mean is greater than 60, a test should be performed.
2. sample mean is smaller than 60
3. sample mean differs from 60
Two of the statements above are true. Choose the correct set of true
statements from the list below
1. (a) and (b)
2. (a) and (c)
3. (b) and (d)
51. An educational psychologist wants to establish if raised expectations has an 2 Study Also, section 4.2.1 in study guide for clarity on single sample. This
effect on how students do in a math’s test. She selects 100 students at guide: questions is two groups, which refers to correlation between the two
random, Half of the students (the control group, referred to as Group 1) were P110,112,1 groups.
14. 103-
told that they have to do 20 math’s problems, but that this just an exercise to Two-group designs: When two different groups from a sample (which is
106, 120
practice their skill. The remaining 50 students (the treatment group, Group 2) 129-130 one of many samples from a population) are chosen. Usually these two
are told to complete the same problems, but that the marks would count groups are characterised by the variety of the independent variable
towards their final exam mark. She decides to use a t-test to compare the (such as ‘gender”) and their influence on the dependent variable, such
results. Which of the test statistics given below is the appropriate test for her as “aggression.
to use?
The sample statistics below are the average reaction limes (in milliseconds) for each condition.
53. Given the scenario above, what type of statistical test is best suited to confirm 2 Study
the relevant statistical hypothesis? guide:
P113-116
1.A one-tailed t-test for a single group
2.A two-tailed t-test for independent groups
3.A one-tailed t-test for dependent groups
54. After doing the appropriate test, the researcher gets a statistically significant 1 Study
result. He notices however that the difference between the means of the sets guide: P86
of reaction times is quite small He is unsure whether this difference is large
enough to be of practical importance Which of the following strategies would
be the most appropriate to get a better idea of the practical usefulness of the
result?
To test the efficacy of a workshop aimed at improving people's interpersonal skills, a researcher applies a scale which rates the interpersonal skills of 30 participants before and
after they participate in the workshop.
59. The interpersonal skill measurements from before the workshop and those 3 Study Participants before and after they participate in the workshop"
after the workshop should be regarded as samples which comes from guide: The same group is used which make it dependent.
groups. P112
1. random
2. independent
3. dependent
60. Which is the appropriate test statistic to calculate? 2 Study Independent groups refers to the structure where one sample in no
guide:
way affects, the other sample. The two samples come from two groups
P102-106,
1. The t-statistic for the mean of a single sample. 112, 120- that have no obvious relationship.
2. The t-statistic for the difference between the means of two dependent 121
groups. Dependent groups refers to situations where the samples are related,
3.The t-statistic for the difference between the means of two independent implying that each subject in one group can be paired off with a subject
groups from the other group, affects, the other sample, often referred to as a
matched-pairs design.
61. When studying correlations in research, one investigates the relation between 3 Study The notion of the relationship between two continuous variables and how
______ guide: the size of the relationship can be expressed in terms of a correlation
P129-130 between them (the index of association is the Pearson product-moment
1. the mean of a single sample of subjects and a population mean correlation coefficient). This coefficient can also be used as a test
2. two dependent groups of subjects, with respect to a single variable statistic.
3. two variables measured on the same group of subjects
65. What is the most likely value of the correlation coefficient between the following 3 Study A perfect positive linear relationship exists (the dots in the scatter plot
values of variables X and Y? guide: run from lower left to upper right in a perfectly straight line) We use 'r'
P132-133 as the symbol that represents a correlation coefficient (as in the case of
X 2 7 4 5 1
the Pearson product-moment correlation coefficient), and the following
Y 2 7 4 5 1
applies:
• r = +1 implies a perfect positive linear relationship (the dots in a
1. -1
scatter plot will run from lower left to upper right in a perfectly straight
2. 0
3. +1 line)
• r = 0 implies no linear relationship at all (the dots may be scattered all
over the place)
• r = -1implies a perfect negative linear relationship (the dots will run
from upper left to lower right in a straight line)
1. r = 0
2. μ = 0
3. p = 0
67. A number of psychiatric patients are classified into one of four categories as: 1 Study
schizophrenic, severely depressed, bipolar disorder and others. Which of the guide:
following is suitable for representing this information versus the gender of P142
these patients?
1. A contingency table
2. A scatter plot
3. A histogram
Base your answers to Questions 68 to 70 on the following scenario
As part of an employment equity programme, a researcher Is asked to investigate the distribution of male and female staff members across different types of work in a particular IT
company. After collecting some data, she creates the table below, which represents the observed frequencies of managers, administrative staff and staff that provides technical
services m the company, cross- classified with genders of the staff members.
68. Based on the contingency table above, what would the expected frequency of 2 Study It is the differences between these expected and observed frequencies that
observations in the cell for female staff who provide technical services be, if guide: 143- interest us, that is, we want to know how far the actual (observed) results are
no interactions exist between the variables gender and type of work? 144 removed from the expected situation, if there is no interaction effect.
Managerial staff Administrative staff Technical staff Total
1. 2
Female 4 6 2 12
2. 4
Male 6 4 8 18
3. 12
Total 10 10 10 30
Row total = 12, Column total = 10
Sample total (size) = 30
E11 = (Row total x Column total) / Sample total
E11 = (12 x 10) / 30 = 120 / 30 = 4
May/June 2017-END
Oct/Nov 2017
QUESTION A PAGE COMMENTS
1. The process of finding a way to measure a construct so that it can be 2N NR.
Study
represented by a variable is referred to as - - - - - S guide:
P24-26
1. inference
2. operationalization
3. setting a hypothesis
2. Mean, range, variance and standard deviation are examples of - - - - - 2 Study
guide:
1. variables P10-11
2. descriptive statistics
3. inferential statistics
3. A theory is a (n) - - - - - 3 Study
guide:
1. process of careful observation P4
2. method to make a construct observable
3. explanation of why facts are as they are observed to be
4. The process of selecting a subset of a population for a survey is known as - 2 Study
--- guide:
1. triangulation P11-12
2. sampling
3. operationalization
5. An inference is - - - - - 3 Study
guide:
1. an explanation of why certain things are as they are observed to be P2
2. an educated guess about how certain phenomena may be interrelated
3. a generalization from a sample to a population which have a high
probability of being true
6. "Empirically" means "based on " 2 Study
guide:
1. theory P2
2. observations
3. facts
11. The variable manipulated by the researcher in an experiment is called the 2 Study
variable guide:
P2
1. empirical
2. independent
3. dependent
13. 1 Study
guide:
1. sample mean, population mean P13,
P161
2. sample mean, sample standard deviation
3. population mean, population variance
14. In probability theory, the number of distinct events that could possibly occur 1 Study
during the performance of an experiment is referred to as the - - - - - guide:
P28
1. probability distribution
2. frequency distribution
3. sample space
15. The table below gives frequency distribution of the exam marks of a number 1 Study
of students in a psychology exam guide:
P55,
143-144
What is the probability that a student will get a mark of between 51% and
70%? (Rounded off)
1. 0.45
2. 0.82
3. 0.18
1. 0.0668
2. 0.0228
3. 0.0062
17. Suppose the weights of the population of military recruits are distributed 1 Study
normally with a mean of 64 kg and a standard deviation of 8 kg Different guide:P
samples of these recruits, each with a sample size of 16, are drawn 60-61
repeatedly We would expect the standard deviation of the sample means
(the standard error) to be about-----kg
1. 2
2. 8
3. 4
18. A marble is drawn at random from a box containing 6 red marbles, 4 green 2 Study
marbles and 5 blue marbles If one marble is drawn at random from the box, guide:
what is the probability that it will be red? (Rounded off) P29
1. 0.07
2. 0.4
3. 0.17
19. There are 19 female and 8 male students in a group of psychology students 2 Study
of the 19 females, only 4 has no brothers or sisters, while 3 of the males are guide:
only children. If a student is selected at random from the group, what is the P29
probability that it will be a female with no siblings? (Results are rounded off).
1. 4/19 = 0.21
2. 19/27 = 0.70
3. 4/27 = 0.15
22. By convention, the total area under the standard normal curve is set equal 3 Study
to - - guide:
P53
1. zero
2. its standard deviation
3. one
23. The mean and standard deviation of a set of test scores are 20 and 8 2 Study X-μ 14 - 20 -6
respectively if the z-score which corresponds to a test score of 14 is guide: Z=σ= 8= 8= -0.75
calculated, in which of the intervals listed below would it fall? P55
• x represents the variable (test score),
1. 0.5 • μ is the population mean,
2. -0.75 • σ the standard deviation of the population from which x was
3. 0.75 obtained.
1. No relationship
2. A positive relationship
3. A significant relationship
27. Statistical hypotheses are statements about - - - - - 1 Study
guide:
1 population parameters P75
2 sample statistics
3 the z-distribution
28. Suppose we have stated Ho µ = 10, and H1 µ < 10, and find that the 3 Study Refer to Appendix D in study guide.
sample mean corresponds to a z-score of -3 This means that the guide:
corresponding p-value - - - - - P75,162
38. When two population means are compared, the p-value is calculated to 3 Study P-Value
represent the probability of observing a specific difference between the guide:
sample means given that - - - - - P87, Zx – X
P139 s
1. Ho is true
2. H1 is true
3. Ho is false
1. 20
2. 25
3. Not equal to 20
41. When two means are compared, the p-value expresses the probability that a 3 Study
difference which is observed between the means in a sample of guide:
measurements P76-77
1. will be significant
2. is due to the alternative hypothesis
3. is due to chance or sampling error
42. Which symbol is conventionally used to indicate the value of the maximum 1 Study
probability that an error would be made if the null hypothesis is rejected guide:
which a particular researcher is willing to allow? P83,85,
29,79
1. Less than 0
2. Between 0 and 0.1
3. Greater than 0.1
46. Under which condition would a researcher use a t-statistic to test a 1 Study
hypothesis about an unknown population mean µ? guide:
P100-
106
The value of the (a) - - - - - rs (b) - - - - -
Suppose that the memory span of adults is normally distributed with a mean of 7 items and a standard deviation of 2 items. A researcher is investing the impairment of memory
among persons who have been diagnosed as suffering from Korsakoff's syndrome (a neurological disorder linked to chrome alcohol abuse). He intends to test his prediction on a
sample of 50 persons who were diagnosed as suffering from this syndrome.
1. Ho µ < 7
2. Ho µ = 7
3. Ho µ ≠ 7
48. Which of the following is an appropriate alternative hypothesis for testing the 1 Study
above prediction, regarding the mean memory span of the population of guide:
persons suffering from Korsakoff's syndrome? P73-75
1. Ho µ < 7
2. Ho µ = 7
3. Ho µ ≠ 7
49. Testing the above prediction on a sample will require a statistical test 3 Study
guide:
1. non-directional P15,75
2. two-tailed
3. directional
50. A pharmaceutical company claims that a new sleeping pill which they are 2 Study
marketing will put people to sleep in less than 15 minutes. A researcher guide:
wants to test see if the average time before people fall asleep after using P81
this pill matches this claim. She uses the following hypothesis
Ho µ = 15 H1 µ < 15
Suppose she tests this on a random sample of n = 40 research participants
who suffer from insomnia She finds that the mean time before members of
the sample fall asleep after using the pill is 14.3 minutes with a standard
deviation of 3.2. A subsequent t-test produces a two-tailed p- value of
0.0345 and the level of significance was set at 0.05 What is the value of
the one-tailed or directional p-value?
1. 0.03450
2. 0.01725
3. 0.06900
1. 0.4
2. 10
3. 2
52. What does it mean to say, "the difference between the means of groups A 1 Study Refer to Study Unit 3.3 in study guide 3.3.1
and B is statistically significant"? guide:
P82-83
1. The alternative hypothesis is valid
2. The alternative hypothesis should be rejected
3. The null hypothesis cannot be rejected
Suppose the researcher finds that the mean and standard deviations for each group in the sample is as follows:
53. Which is the appropriate way to indicate the researcher's hypothesis which 3 Study
is to be tested? guide:
P11-18,
P21-26
A researcher compares a sample of children from a special school for gifted children with a group of children randomly drawn from other schools on a test which measures the
creativity of the children on a 9-point scale She wants to know whether the children from the school for gifted children is likely to have greater levels of creativity, planning to test
at of a = 0 01
59. What is an appropriate alternative hypothesis for this research scenario? 3 Study
guide:
P73-75
The table below can be used as a guide to Judge the practical importance of
the result
Based on the calculated effect size, the researcher can conclude that the
practical implication of her finding is - - - - -
1. Small
2. Medium
3. Large
1. low, low
2. low, high
3. high, high
64. A researcher wants to establish whether the type of employment category 3 Study
that is filled by employees of a particular company is significantly related to guide:
their gender. The employees can be categorized as manager, human P140
resources, administrative, maintenance or information technology worker
and the genders are male or female. Which would be the most appropriate
test to use?
1. 0.37
2. 6.32
3. 7.00
68. If there is no relationship at all between two variables X and Y, what would 3 Study
be the most likely value of Pearson's correlation coefficient r, out of the guide:
following? P132
1. -1 0
2. 05
3. 00
69. A contingency table represents - - - - - 3 Study Refer to Appendix B in study guide
guide:
1. the distribution of the frequencies for a variable P142-
144
2. frequency counts for each of a number of possible outcomes of an
experiment
3. the frequency counts if each outcome measured on two nominal-scale
variables when they are cross-classified
70. Which of the values given below is the closest to the probable value of the 1 Study
Pearson's product moment correlation coefficient for the variables X and Y? guide:
P132-
133
1. -1.0
2. 0
3. 1.0
1. theoretical
2. empirical
3. inferential
3. In the context of research in social science, measurement means Study
to - - - - - guide:
P18
1. determine the extent to which a specific phenomenon can be
represented on an appropriate numeric scale
2. specify the relationship that is believed to exist between two (or
more) constructs or variables
3. calculate a summary value which describes an aspect of a
specific construct or phenomenon
4. A variable is described as 'manifest' because it is a[n] (a) __ 3 Study So, the (visible) variable reflects the intensity of the underlying (invisible) construct, in terms
measurement of a construct which is (b) guide: of how it was measured. We say that the variable is manifest (it is visible in the sense that
P7,9,23 we can observe it) and the construct is latent (it is invisible in the sense that we need some
1. (a) latent (b) observable way to make it appear). So, the latent construct is made manifest by the use of an
2. (a) independent (b) dependent appropriate measurement procedure. To say that a construct is 'latent' is another way of
3. (a) observable (b) latent saying it is hidden from direct observation
1. descriptive statistic
2. variable
3. test statistic
6. Operational definitions of a construct are definitions that define 1 Study Operational definitions link constructs to observable phenomena so that we can observe
the construct in terms of - - - - - guide: them. Operational'' refers to practical procedures by which constructs are made
P6-7, 5- visible.
16,24-
1. phenomena which can be observed In essence: What a researcher must do to measure a construct
26,
2. measurable variables
3. summarizing parameters
7. Which of the following is appropriate as a research or operational 1 Study
hypothesis? guide:
P9
(a) Gender plays a role in determining employees' salaries in
Winston & Johnson inc
(b) Male employees earn more than female employees in Winston
& Johnson inc
(c) The company Winston & Johnson inc has recently been
declared bankrupt
1. (a) and (b) but not (c)
2. (b) and (c) but not (a)
3. (c) only, but not (a) and (b)
8. Quantities that summarize aspects of a population are called (a) 2 Study
, while quantities that summarize aspects of a sample are guide:
referred to as (b) - - - - - P11
10. A researcher believes that people who make eye contact with 1 Study
others when they speak to them are generally perceived to be guide:
more trustworthy than those who do not. She sets up an P8-9, 10-
11
experiment where each member of a group of 100 research
participants is interviewed by a research assistant in half of the
cases, the interviewer makes a lot of eye contact with the
participants during the interview, and in half of the cases no or
very little eye contact is made. Afterwards participants are asked
to rate the research assistant for level of trustworthiness in this
scenario, whether eye contact was made or not is the (a)
____variable, while perceived level of trustworthiness is the (b)
____ variable
12. Which of the descriptions given below is the most accurate 3 Study
description of the meaning of the word 'theory' in scientific guide:
research? It refers to a - - - - - P4, 15,
18 – 19,
21-26
1. reasonable guess about a relationship that may exist among
two or more variables
2. best explanation of why a specific relationship that is observed
among variables is as it is observed to be
3. careful description of the facts that have been observed in a
specific situation
1. 0 60,030
2. 0 05, 0 60
3. 0 30,060
1. 0 005
2. 0 067
3. 0 138
18. The value of x for the management test scores is equal to - - - - - 2 Study
guide:
1. 12 00 P59 - 60
2. 10 00
3. 9 00
19. The standard deviation of the distribution of test scores calculated 1 Study Formula:
and it is found that S = 2.11. If the data in the table above is guide: (Score X) – (Sample S) = 1 ÷ ( Deviation σ 2.11) = 0.47
transformed to z-scores, what would the z-score for individual E P53
be equal to? 9 – 10 = 1 ÷ 2.11 = 0.47
1. 0 47
2. 1 42
3. -0 47
20. In a population with a normal distribution, with a mean of µ and a 3 Study
standard deviation of μ, approximately- - - - - of the scores will fall guide:
below the exact value of µ P50-54
1. 34%
2. 95%
3. 50%
1. 0 208
2. 1 457
3. 10 200
26. Given the scenario, which of the following gives the best Study
description of the alternative hypothesis and type of statistical guide:
test that would be required? P113-
116
1. A non-directional hypothesis tested with a two-tailed test
2. A non-directional hypothesis tested with a one-tailed test
3. A directional hypothesis tested with a one-tailed test
27. Given the data above, what would be the most appropriate 1 Study
statistical test procedure to establish whether there is a guide: P
statistically significant difference between the average weight of 102-106
the women in the sample and the weight of the women recorded
in the census?
1. Accept H1
2. Reject Ho
3. Do not reject Ho
29. The normal distribution is useful for interpreting psychological 1 Study
measurements because - - - - - guide: P
51
1. many psychological variables are approximately normally
distributed
2. it has a mean of zero and a standard deviation of 1
3. it is symmetrical in shape
30. The examination scores of a group of students are approximately 3 Study Utilize the z score calculation. Refer to Appendix D: Z-scores and areas under the normal
normally distributed with a mean of 60% and a standard deviation guide: curve
of 8% Suppose Lebo has a mark of 70% This implies that she did P53-56, μ = population mean
162
better than about of the students σ = standard deviation
z = standard normal distribution (= test statistic)
1. 40% x = a particular individual score
2. 89%
3. 11% The standard normal distribution is sometimes called the z-distribution
• The measures on the x-axis of the curve are called standard scores/ z-scores
• The z-scores reflect the number of standard deviations that a particular score lies above or
below the mean
• Because all distributions of z-scores have a mean of 0 and a deviation of 1, they can be
used to compare individual scores across different distributions
statistic
2. enables one to compare variables with different means and
standard deviations from scores with different original units
3. can be used to determine whether a score is normally
distributed
33. A probability of an event occurring which depends on something 1 Study
else occurring, such as passing a test when you do not guide: P
understand your course, can be described as - - - - - 36-37
1. a conditional probability
2. an independent event
3. dependent probabilities
34. The sampling error of the mean will be smaller in cases where - - 1 Study
--- guide: P
61-62
1. the sample is larger and the standard deviation of the
population smaller
2. the population is larger and the variability of the scores in the
sample is smaller
3. the sample mean is smaller, and the variance is larger
Suppose that the memory span of adults is normally distributed with a mean of 7 ,terns and a standard deviation of 2 ,terns A researcher predicts that "dyslexic adults have a shorter
memory span than adults in general"
1. 1 2
2. 0 24
3. No estimate is possible because μ is unknown
49. Two samples can be considered independent when - - - - - 1 Study
guide:
1. the composition of one sample is not systematically related to P110
1. dependent
2. independent
3. both normally distributed
55. Which variables are related to one another by the research 2 Study
hypothesis? guide:
P117-
118
1. Before or after workshop for assertiveness training, self-esteem
2. Self-esteem in the treatment group, self-esteem m the general
population
3. Level of assertiveness, level of self-esteem
56. Which is an appropriate null hypothesis for the analysis of the 1 Study
results? guide:
P73-75
1. µ = 30
2. µ2 > µ1
3. µ1 - µ2 = 0
Group 1 (urban) X1 = 52 5, S1 = 4 7
Group 2 (rural) X2 = 55 9, S2 = 5 2
X -2 -1 0 1 2
Y 2 1 0 -1 -2
1. r = -1
2. r = 0
3. r = 1
Which is the most appropriate way to state the null hypothesis for
this that is being investigated?
1. Ho p = 0
2. Ho r = 0
3. Ho µ1 = µ2
64. A researcher wants to determine whether a relationship exists 2 Study
between students' general level of anxiety and their exam marks. guide:
He presents each student from a random sample with a general P130
anxiety scale just before they are to write an important exam.
Which of the following is the most appropriate test statistic to use
to determine whether a relationship exists between the two
variables (anxiety level and exam results)?
1. r = 0
2. r < 0
3. r > 0
A sample of 300 clients are drawn from three community mental health centers (indicated in the table as A, B and C) Counts are made of those clients who are diagnosed as having
social adjustment problems, those with problems related to anxiety, and the remaining clients are classified under 'other problems ' Counts of the number of clients from the different
centers which fall in each of the categories are supplied below.
1. 20
2. 26 7
3. 33 3
70. Which of the following would be the most appropriate statistical 1 Study
test to use to test the hypothesis that the distribution of diagnoses guide:
across the three centers may be significantly different? P110
5. Which best describes 'research hypothesis·? 2 Study A research hypothesis is formed as a clear statement in terms of a
guide: relationship among the constructs (and the variables by which they are
1. A proven relation between two constructs P1, 18- measured). It is a statement about a possible relationship among constructs
19
2. A proposed relation between two or more variables that may explain some set of observations that one intends to investigate
3. A network of all the possible relations between constructs
7. The (a) - - - - - variable can be said to 'be caused by' the (b) variable 3 Study The dependent variable is the one that is predicted or explained, and the
guide: independent variable is manipulated to see how it affects the dependent
1. (a) dependent (b) independent P8-9, 24, variable.
2. (a) operational (b) measured The independent variable is that variable which affects the dependent
3. (a) independent (b) dependent variable; or, conversely, the dependent variable depends on the independent
variable.
When a researcher focuses on the interaction of only two variables at a time,
the dependent variable is usually the one that the researcher is interested in,
the variable that is the focus of the research. The independent variable is
something that the researcher manipulates, to see how this affects the
dependent variable (in other words, the dependent variable is dependent on
the independent variable).
The dependent variable is the one that is predicted or explained, and the
independent variable is manipulated to see how it affects the dependent
variable.
8. A construct may be a - - - - - 2 Study Constructs and their interrelations (how they affect each other, their patterns
guide: of interaction) are used in this way to develop theoretical explanations of
1. hypothetical relation between two or more variables P4 why people behave in certain ways in certain contexts, or why mental
2. hypothetical aspect of humans that we wish to investigate phenomena appear to be as they are.
3. testable prediction derived from a theory of human behavior
14. Which of the following does NOT represent a probability? 3 Study The probability value tells us at a glance how frequent or infrequent the event
guide: is, and what the likelihood is of obtaining a favorable outcome associated with
1. 99% P33 it.
2. 0 Probabilities can be expressed as percentages (e.g. a 10% probability), as
3. -0 05 fractions (e.g. a 1/10 probability), or as a decimals (e.g. a 0.10 probability).
A probability value represents a proportion (i.e. the proportion of outcomes
supporting the event). A proportion is a decimal number between 0 and 1 and
indicates the fraction of the total.
We often refer to the probability of an event (or statistic) as its p-value.
When decimal notation is used to describe probabilities, they fall in a range
between 0 and 1, with values closer to 1 indicating a greater likelihood (or
chance of success) than values close to zero.
Because probabilities fall in a range from 0.0 to 1.0 when expressed
decimally, a probability can never be higher than 1 or lower than 0. The
general rule is written symbolically as follows: 0 ≤ p ≤ 1.
15. Use the following frequency distribution of 100 scores on a subscale of an 1 Study Frequency of X larger than or equal to 18:
intelligence test to answer the question guide: 18 numbers = 6
P30-33, 18 numbers = 0
53,143- Number of participants (N) = 100
Score (X) 11 12 13 14 15 16 17 18
144
4 11 13 22 18 17 9 6 Formula is :
Relative
So:
frequency(%)
The value of p(X larger than or equal to 18) is - - - - - μ = Σxi / N
= (6+0) / 100
1. 0 06 = 6 / 100
2. 0 = 0.06
3. 94%
17. The expression '0 05 ≤ p ≤ 0 1 0' denotes a probability value - - - - - 3 Study Larger than or equal to 0.05 and smaller than or equal to 0.10. Because
guide: probabilities fall in a range from 0.0 to 1.0 when expressed decimally, a
1. somewhere ,in the range from 0 05 to 0 10 P33-34 probability can never be higher than 1 or lower than 0. The general rule is
2. larger than or equal to 0 10, or smaller than or equal to 0 05 written symbolically as follows: 0 ≤ p ≤ 1. Note that a probability can be 0, but
3. larger than 0 05 and smaller than 0 10 to say that a probability is 0 is actually the same as saying that the event is
impossible and can never happen. Likewise, to say that the probability of an
event is 1 is to assert that it is an absolute certainty. In actual practice,
probabilities fall within these two extremes. You will typically encounter
reference to probabilities in expressions such as ''p > 0.05''. This statement is
interpreted as ''the probability value is higher than 0.05''.
20. Which of the following represents the best estimate of the population mean? 2 Study
guide:
P60-61,
1. The sample mean 103-104
2. The sample standard deviation
3. The mean of several sample means
Approximately 0.477 (i.e. 0.3413 + 0.1359) of the area lies between the
mean and 2 standard deviations in each direction.
Approximately 0.998 (i.e. 0.954 + (0.0215 x 2)) of the area lies within three
standard deviations from the mean.
22. The mean and standard deviation of a set of test scores are 20 and 8 Study X-μ 14 - 20 -6
respectively What is the z-score corresponding to a test score of 14? guide: Z=σ= 8= 8= -0.75
P55
1. 6 • x represents the variable (test score),
2. 0 75 • μ is the population mean,
3. -0 75 • σ the standard deviation of the population from which x was
obtained.
23. Which of the following statements are true? 3 Study These summary quantities are sometimes referred to as parameters (when
guide: they refer to the whole collection or population of data You should take
1. Parameters describe samples and statistics describe populations P11, 161 careful note of the following important distinctions between samples and
2. Statistics describe samples and populations populations. Summary values for populations are called 'parameters' and are
3. Parameters describe populations and statistics describe samples usually denoted by Greek letters, while summary values for samples are
called 'statistics' and are denoted by Roman letters
26. Which of the following is an appropriate null hypothesis for testing the above 2 Study
prediction? guide:
P73-75
1. The mean memory span of the population of dyslexic adults is smaller than
7
2. The mean memory span of the population of dyslexic adults equals 7
3. The mean memory span of the population of adults equals 7
27. Testing the above prediction will require a ______ statistical test 3 Study
guide:
1. non-directional P15,75
2. two-tailed
3. directional
Two important elements you need to note on the central limit theorem:
(a) “This theorem gives the sample distribution of the sample means for any
population, irrespective of the shape, mean or standard deviation of the
original population.”
(b) “The distribution of sample means (equal to population mean) will
become more normal as sample size (n) increases.
30. The sampling distribution of a statistic (e g of the sample mean) can be 1 Study
guide:
calculated if we assume that the ---- hypothesis is true, but not if we assume
P61-62
that the ---- hypothesis is true
1. null, alternative
2. alternative, null
3. statistical, research
34. The standard error of the mean is · · - - - 1 Study the standard error of the mean tells us by what average amount the sample
guide: means deviate from the mean of the sampling distribution. It is an estimate of
1. the standard deviation of the population mean P60-61, the size of the error we shall make if we use the mean of the distribution of
103-104
2. the standard deviation of the sampling distribution of the mean for samples sample means as an estimate of the true population mean.
of a specific size
3. the mean of the standard deviations of repeated samples of a specific size Sections 2.4.1 and 2.4.2 of the PYC3704 Guide
35. When a statistical test yields a very small p-value, we know that the sample 3 Study The p-value gives the probability of obtaining the sample result under H0.
result is very - - - - - guide: If the p-value is very small, the probability is very small that the sample result
P81 would occur under H0, and one should consider rejecting H0 in favor of H1.
1. likely under the null hypothesis The smaller the p-value, the more likely that the null hypothesis is false and
2. unlikely under the alternative hypothesis should be rejected in favor of the alternative hypothesis. So, if the p-value is
3. unlikely under the null hypothesis very large, the probability is very big that the sample result would occur
under H0, and one should consider accepting H0 in favor of H1. The null
hypothesis is then probably true
A researcher hypothesises that babies born prematurely will be somewhat less intelligent as young adults than their peers. She uses the records of various maternity hospitals to
identify a random sample of 25 persons who are now young adults, but who were born more than four weeks prematurely. She measures the IQ of each, using the SAWAIS (IQ
scores on this test are distributed normally in the general population, with a mean of 100 and a standard deviation of 15 ). Suppose she finds that the mean IQ score of her
sample is 97 9 and the standard deviation of the scores is 17
37. How would you describe the population investigated in this research? 1 Study 25 persons who are now young adults, but who were born more than four
guide: weeks prematurely
1. Adults that were born more than four weeks prematurely P71-76
2. All adults
3. Adults that completed the SAWAIS
38. Which research design did the researcher use? 2 Study
guide:
1. Correlational design P100-
106
2. Single-sample groups design
3. Two-sample groups design
39. Which of the following are appropriate null and alternative hypotheses? 1 Study H0 always has = sign
guide:P
1. Ho µ = 100, H1 µ < 100 73-
80,82-
2. Ho µ = 100, H1 µ ≠ 100
83,137
3. Ho µ ≠ 100, H1 µ < 100
40. Which is the correct value of the standard deviation of the sampling distribution Study See Section 2.4.2
of the mean of 25 IQ scores? guide:
P60-61
1. 3 0
2. 34
3. 15
48. Which of the following formulas would you use to calculate the T-test for the 3 Study
comparison of means for independent samples? guide:
P104-
106
49. Susan sets up the following hypotheses 3 Study The decision rule for H0 is simply as follows:
guide: If the p-value of the sample result is smaller (less) than α (level of
Ho µ = 100 P71-82 significance), the null hypothesis is rejected. If the p-value is not smaller than
H1 µ < 100 α, the null hypothesis (H0) is not rejected.
She draws a random sample of 50 persons. Given that the null hypothesis is
false, what can Susan expect the mean of the sample to be?
1. dependent
2. independent
3. having been drawn from the same population
57. A researcher wants to test the hypothesis that boys are generally more 2 Study
aggressive than girls He draws a sample of 100 boys and a sample of 100 guide:P
girls and gives each child a test that measures their general level of 118
aggression. Which of the following would be the most appropriate statistical
test to use?
58. The difference score (d = x2 – x1) is used in the calculation of the t-test 1 Study Section 6.1.2 in study guide
statistic in the case of - - - - - guide:
P112,13
4-136
1. dependent samples
2. independent samples
3. both of the above
1. p = 0
2. µ=0
3. r = 0
63. Which of the following can never be exactly zero? Study
guide:
1. a probability P86-129-
130
2. a level of significance
3. a correlation coefficient
A researcher studying possible sex-linked inheritance of three psychiatric disorders (denoted by A, B, and
C) tabulated the gender (male/female) of 100 psychiatric patients against their diagnosis
A B C
Male 20 20 10
Female 30 10 10
Totalresearch
64. Which 50 design
30 did20
the researcher use? 1 Study
guide:
1. A correlational design P100-
106
2. A two-sample groups design
3. A three-sample groups design
65. What are the requirements with regard to the statistical test to be performed? 1 Study
guide:
1. A directional statistical test is required P75-76,
106
2. A non-directional statistical test is required
3. No statistical test is required
69. Which of the following is an appropriate null hypothesis to test relationships 3 Study The decision rule for H0 is simply as follows:
given the data above? guide: If the p-value of the sample result is smaller (less) than α (level of
P60- significance), the null hypothesis is rejected. If the p-value is not
61,75-
1. The particular community mental health centers that the clients visit have smaller than α, the null hypothesis (H0) is not rejected.
75, 82-
no relationship to the type of problem that clients present with 86,143- Similarly,
144 the decision rule for H1 is simply as follows:
2. There is a correlation between the type of problems that clients have and If the p-value of the sample result is larger than α, the alternative
the particular community mental health centers that they visit hypothesis is accepted. If the p-value is not larger than α, the alternative
hypothesis is not accepted.
3. There are no significant difference among clients having social adjustment
problems, those with problems related to anxiety, and those with other
problems
70. Given the data above, which test statistic will need to be calculated to test the 2 Study The chi-square test is usually used when you have a cross tabulation of
null hypothesis above against an appropriate alternative hypothesis? guide: frequency counts of events which are nominal scale measurements. This
P110 table is referred to as a contingency table. It is used to compare an observed
1. The Pearson product-moment correlation coefficient (r) frequency distribution (frequency counts based on a sample of observation)
2. The chi-square (X2) test statistic with the frequency distribution which we would expect to find if the null
3. The t-test statistic for independent groups (tc) hypothesis of no relationship between two cross-tabulated variables were
true.
Oct/Nov 2018
QUESTION A PAGE COMMENTS
1. Inferential statistics refers to - - - - - NS NR S
2 Study Inferential statistics refers to the use of statistical techniques to make
guide: generalisations about the relationships among (two or more) variables. Here
1. the techniques whereby data from samples are used to summarize aspects P10-11 the patterns that may exist in the data are carefully investigated
of a population
3. Male employees at Winston & Johnson inc earn higher annual salaries than
female employees at the same company, at corresponding post levels
8. Select the most appropriate definition of a psychological theory Psychological 2 Study Refer to Appendix B in study guide
theories are - - - - - guide:
P99-100
1. careful descriptions based on observations of human behavior
2. statements about possible relationships which may exist among psychological
constructs
3. statements which aim to explain psychological phenomena which were
observed
10. For ratio or interval measurements, one would consider the - - - - - of a 2 Study Refer to Appendix B in study guide.
construct, while for nominal measurements, the ------ will be considered guide: On a nominal scale, numbers show category membership, but are otherwise
P7, arbitrary. They do not represent a size or intensity of something, but are only
51,156-
1. category membership, quantity or intensity used as labels to distinguish among qualities or characteristics. They can
157
2. quantity or intensity, category membership also be referred to as categorical variables, or qualitative variables. This is
3. operationalization, frequency distribution because differences in the numbers represent differences in quality,
character or type, but not in amount
11. In a study, the relationship between level of physiological arousal (high and low) 1 Study hypothesis is formed as a clear statement in terms of a relationship among
guide:
and mood (measured on three levels) is investigated Suitable hypotheses for the constructs (and the variables by which they are measured).
P15,
this are to be regarded as - - - - - 71,99
1. The level of alcohol consumption among the students has an effect on their
motor performance
2. Comparing two groups of students on alcohol consumption and motor
performance
3. Some students can consume more alcohol than others before their motor
coordination is affected
13. A Jar contains 5 red, 8 blue, 3 green and 4 yellow marbles. What is the 2 Study Calculations:
probability that a person would choose either a red marble or a yellow one? guide: 5 + 8 + 3 + 4 = 20
P29 5+4=9
1. 0 90 9 ÷ 20 = 0.45
2. 0 45
3. 0 25
14. Suppose the following sequences made up out of five random letters each are 2 Study Question indicates letters A, E. O, U, I
written on separate pieces of paper and placed in a box. You are asked to draw guide: Number of blocks = 20
one piece of paper from the box at random What is the probability that the P27-57 A = 2 ÷ 20 = 0.1
sequence of letters on the paper will contain at least one vowel? (I e a letter E = 2 ÷ 20 = 0.1
from the sequence A, E, O, U, I) O = 2 ÷ 20 = 0.1
U = 1 ÷ 20 = 0.5
I = 3 ÷ 20 = 0.15
What is the probability that both Mary and John will be selected? 1 ÷ 10 = 0.1
1 ÷ 12 = 0.083
1. /10+1/12=0183
2. 2/22 = 0 091 0.1 + 0.083 = 0.183
3. 1/10x1/12=0008
16. The central limit theorem states that - - - - - 3 Study Study unit: 2.4.2
guide:
1. the sampling error of repeated measurements of a random variable will P60
increase as the population size increases
2. the sampling distribution of any random variable will come closer to a normal
distribution as the sample size increases
1. 0 60, 0 03
2. 0 05, 0 60
3. 0 03, 0 60
18. In a population there are 450 people of whom 150 smokes. What is the 1 Study (450-150 = 300) - 300 ÷ 450 = 0.67
probability of randomly selecting a non-smoker from this population? guide: 150 ÷ 450 = 0.33
P29,35- There are only 150 people out of 450 who do not smoke.
36, 55,
1. 0 67
67
2. 0 33 Sections 2.2.2 and 2.2.3 in the PYC3404 Guide
3. 05
1. 68%
2. 05
3. 1
20. There are 12 girls and 8 boys in a specific class at school. Two children who 1 Study 1 ÷ 12 = 0.083
happen to be ,in this same class are Tom and Elizabeth. The children in the guide: 1 ÷ 8 = 0.125
class have to choose two class representatives, one from the boys and one from P29
the girls. The children make the choice by writing the names of boys and girls on Number of possible outcomes = Total kids = 21
slips of paper, putting the names of boys in one box and girls' names m another, Number of favourable events = Either Mary or Elizabeth = 2
and then asking their teacher to draw one name blindly from each box. What are
the chances of Tom or Elizabeth being chosen, purely at random? p(E) = Number of favourable events
Number of possible outcomes
1. 1/12x1/8=001 =2
2. 1/12+ 1/8=021 21
3. 2/20 = 0 10 = 2/21
Questions 21 to 24 are based on the scenario below
A psychologist is conducting research on xenophobia (hatred of foreigners). She makes use of a Xenophobia Scale which measures attitudes towards foreign language speakers,
and which consists of 60 ,terns, each one scored 0 – 4. This scale is applied to a random sample of n=100 citizens. The results for each research participant are added to
produce an overall xenophobia score which falls ,in a range from a minimum score of 0 to a maximum score of 240
The researcher calculates descriptive statistics for this sample and finds a mean of x = 120 and a standard deviation of 10.Since the sample data is roughly normally
distributed, she draws the graph below.
21. Based on the data in the scenario, what would the variance of the distribution of 1 Study
the sample of scores be? guide:
P53-56
1. 100
2. 10
3. 1
1. 120
2. 10
3. 1
23. If a specific person has a xenophobia score of 130 on the scale in the scenario 2 Study
above, what would the person's z-score be? guide:
P53, 55-
56
1. 0 0
2. 1 0
3. 05
24. What is the probability that a person, chosen at random, will get a score of 2 Study Refer to Standard normal distribution
greater than 130 on the xenophobia scale in the scenario above? guide:
P34,52,
55-56
1. 0 023
2. 0 159
3. 0 977
25. Study the histogram below of the exam marks of a group of students in the 3 Study Also see, Appendix B in the PYC3704
same class. Note that the values on the horizontal axis are the class (category) guide:
limits P29,
40,47
130-132
142-144
1. 0 40
2. 0 20
3. 0 60
1. decrease
2. increase
3. not be affected
37. When applying a z-test to compare a sample mean to a known population 3 Study When the σ is known, we use the z-test.
mean, the p-value represents the probability of - - - - - guide:P
82, 100,
112,114
1. rejecting the null hypothesis if it is false
2. failing to reject the null hypothesis when It is in fact true
3. rejecting the null hypothesis if it is actually true Also refer to Section 4.1.2 in study guide
Use the t-test (tẋ) when the population standard deviation (σ) is unknown as
the given standard deviation is derived from the sample.
1. 0 04
2. 0 05
3. 0 01
39. The hypothesis "H1 µ < 30" is a - - - hypothesis and requires a statistical test 3 Study
guide:
1. non-directional, one-tailed P75 -76,
81
2. directional, two-tailed
3. directional, one-tailed
Questions 40 to 42 are based on the following scenario
An educational psychologist investigates whether a series of special interactive exercises presented on a computer will improve the performance of pupils in mathematics. A
group of 100 pupils are selected at random from a number of schools to receive the computer-aided mathematics exercises. The researcher uses a general test for mathematics
which has been standardized to a mean score of 50 and a standard deviation of 16 to test each of the pupils in the sample after they have completed the computer-aided
mathematics exercises.
40. Which of the following statements translates the research hypothesis into the 1 Study
correct statistical hypotheses? guide:
P11-18,
21-26
1. Ho µ = 50 H1 µ < 50
2. Ho µ = 50 H1 µ > 50
3. Ho µ = 50 H1 µ ≠ 50
41. Based in the information in this scenario (above), what is the value of the 3 Study Deviation ÷ √𝑆𝑎𝑚𝑝𝑙𝑒
standard error of the mean? guide:
P58-62
Formula:
1. 1 60
2. 0 16
3. 16
43. When the results are statistically significant, this means that - - - - - 3 Study
guide:
(a) the obtained probability is equal to or less than alpha P82 - 83
(b) the independent variable has had a large effect
(c) we can reject Ho
44. What can a researcher do while performing statistical analysis of data to limit the 3 Study In a hypothesis test, a type I error occurs when the null hypothesis is rejected
risk of a Type I error? guide: when it is in fact true = H0 is wrongly rejected.
P84 - 85
1. Increase the sample size
2. Decrease sampling error, measurement error, etc.
3. Set a smaller level of significance
45. The power of a statistical test refers to the - - - - - 2 Study
guide:
1. test's ability to give small p-values P85 -86
2. test's ability to detect significant results
3. probability that an error of Type I will not be made when the test is used
49. The standard error of the mean for samples of a specific size is the - - - - - 2 Study
guide:
1. standard deviation of the sample mean P103 -
104
2. standard deviation of the sampling distribution of the mean
3. mean of the standard deviations of repeated samples of this specific size
Suppose the two-tailed p-value is 0 0345 and the level of significance is set at 0
05 The sample mean was found to be 55 What is the value of the one-tailed or
directional p-value?
1. 0 0250
2. 0 0173
3. 0 0690
51. When applying a t-test for the difference between the means of two independent 1 Study
samples, the probability of obtaining the calculated t-statistic under the null guide:
hypothesis is compared to the to reach a decision P86 –
87, 105,
116
1. level of significance
2. two-tailed probability
3. effect size
52. A researcher plans to use the t-test to compare two independent samples of 3 Study
data of only 15 individuals each. What minimum assumption needs to be met guide:P
before she may proceed? 115
1. Only (b)
2. Either (a) or (b)
3. Both (a) and (b)
1. only (a)
2. only (b)
3. both (a) and (b)
56. A researcher plans to compare a group of male students with a group of female 1 Study
students on a test that measures problem-solving style. Which is the guide:
independent variable? P7-9, 24
1. gender
2. problem-solving style
3. students
Base your answers to Questions 57 and 58 on the follow scenario
A psychologist develops a series of workshops providing assertiveness training to a group of persons who suffer from low self-esteem. To test the efficacy of the workshops, she
applies a psychometric test which measures level of self-esteem to 50 persons at the start and again after the end of the series of workshops, predicting that the latter scores will
be higher (reflecting higher self-esteem). The self-- esteem scale was standardised on the general population with a mean score of 30 and a standard deviation of 10
1. p = 0
2. µ=0
3. r = 0
64. A contingency table is used to summarise the relationship between two 1 Study
variables measured on at least ---------- scale guide:P
144
1. a nominal
2. an ordinal
3. an interval or ratio
65. What is the closest to the most likely value of the correlation coefficient between 3 Study
the following values of variables X and Y? guide:P
132-133
1. -1
2. 0
3. +1
A psychologist reads an article in which the author claims that playing computer games leads to higher levels of aggression in children. She decides to test this by asking a
sample of children to report the number of computer games they play per month and measuring the aggression level of each child with an appropriate psychometric test. She
expects to find that a positive correlation will exist in her sample between level of aggression and number of computer games played.
66. If she should draw a graph of the relationship between aggression and number 1 Study A scatter plot is a graph showing the relationship between two numerical
of computer games, which of the scatterplots below would give the most guide: P variables. In such a graph the data of the one variable are plotted on the
probable representation of the data, If the relationship exists as expected by the 130-132 horizontal axis (usually referred to as the X axis), and the data of the other
researcher, according to the scenario above? variable on the vertical (or Y) axis. It is not a comparison of sample and
population, nor has it to do with spread of data or the independence of
variables.
The closer the dots in the plot are to a straight line, the closer the correlation
coefficient is to 1 (it can be either a positive number (+1) or a negative
number (-1)). The more arbitrary or spread out the dots, the closer the
correlation coefficient is to 0. If the plot seems to form a line from lower left to
upper right, the correlation is positive. On the other hand, if the line runs from
upper left to lower right, the correlation is negative. She expects to find that a
positive correlation will exist in her sample between level of aggression and
1. Graph A number of computer games played. Therefore, the plot must form a line from
2. Graph B lower left to upper right
3. Graph C
67. Which of the following expresses the appropriate way to formulate the 1 Study
alternative hypothesis implied by the scenario above? guide:
P76-92
1. r > 0
2. µ≠ 0
3. p> 0
68. If the researcher wants to establish the relationship that exists between playing 3 Study
computer games and levels of aggression in children (in the scenario) is in fact guide:P
significant, which of the following would be the most appropriate test to use? 140
The contingency table below reflects a sample of people from drawn from rural, semi-urban and urban regions of the country, cross-classified by sex or gender.
69. What would the expected value be for the cell indicating 'rural males' in the 2 Study Male = 10
contingency table above, if in fact no relation exists between gender and guide:P Row = 6 ÷
whether a person comes from a rural, semi-urban or urban area in this particular 142 Column = 25
sample of data? (Choose the value closest to the correct estimate from the
options below) 10 x 6 ÷ 25 = 2.4
1. 2
2. 24
3. 04
70. Which of the following would be the most appropriate test to use to establish 3 Study
whether a relationship exists between type of environment (rural, semi-urban or guide:P
urban) and gender, in the scenario above? 110, 140
Oct/Nov 2018-END
1. develop appropriate statistical tests which can be used to determine the relationships among psychological variables that occur at a level greater than chance
2. develop theories that helps us to explain human experience and behaviour
3. formulate clear hypotheses based on insights about human experience and behaviour
4. convert theoretical constructs into measurable variables through operationalisation
The goal of research is to develop theories which can explain aspects of human behaviour and experience. Options 3 and 4 refer to stages in the process of doing quantitative
research but these are not the goals of the research. The goal of the research is also not to develop statistical tests, as implied in option 1. These tests are developed by
statisticians and are used by researchers in social and other scientific research, but developing the tests is not part of the goal of the research.
Question 2
Consider the following statement: "That phase of sleep during which brain rhythms resemble those of an alert person is called paradoxical sleep or rapid eye movement (REM)
sleep". This statement is a - - - - -
The statement gives a definition of REM sleep in terms of the measurement of brain waves (which can be done with an EEG machine). Since it is simply a statement about
what the expression ‘paradoxical sleep or rapid eye movement (REM) sleep’ means, it is not a conclusion, hypothesis or inference, as suggested by the other options.
Question 3
In scientific research, the word theory refers to a(n) - - - - -
As explained on page 4 in the Guide for PYC3704, a theory is a framework for facts: it is the explanation of why the facts (i.e. observations, measurements) are as they are, or
are related in the way in which they are related, based on empirical investigations. Option 1 is a description of a hypothesis, but this is often how the word ‘theory’ is used in
informal conversation. Option 2 refers to the process of measurement and option 3 is a description of a process such as an experiment by which a theory can be evaluated but
not to a theory as such
.
A psychologist wants to study how aspects of motivation can influence people’s productivity in their work. She reads an article which claims that an important aspect of
motivation is locus of control, which distinguishes people who are driven by their own personal ambition from those who act by conforming to a social group. To test this idea,
she draws a sample of 100 workers from a number of companies in the information technology sector. She divides the workers into two groups, those with an internal locus of
control and those with an external locus of control, based on an appropriate psychometric test. She then uses job evaluation forms to assess the actual work performance of
the workers in the two groups.
Question 4
Which of the following gives the best expression of the hypothesis that the researcher wishes to test?
Option 3 is the clearest expression of an appropriate research hypothesis. Option 1 is too vague: ‘motivation’ is probably too complex to be captured in a single construct.
Option 2 is too specific: it is not clear from the information given in the scenario whether internal or external locus of control can be expected to lead to higher productivity.
‘Locus of control’ is regarded as an aspect of motivation (according to the scenario), not as a separate construct (or variable) that can be compared to it, so Option 4 is also
wrong.
Question 5
Given that research is an investigation of a relationship between (two or more) constructs, which of the following constructs have to be compared to do this research?
Locus of control’ is the specific aspect of ‘motivation’ that is being studied, which excludes Option 2. In this particular context, ‘people who work in the information technology
sector’ is part of the definition of the population that the researcher chooses to study, not a construct that is being compared to another, so Option 3 is incorrect. Option 4 is
incorrect because ‘job evaluation forms’ are the instruments used to measure the construct ‘job performance’, and not a construct as such.\
Question 6
The dependent variable in the study is - - - - - and the independent variable is - - - - -
The dependent variable is the one that is predicted or explained, and the independent variable is manipulated to see how it affects the dependent variable. In this study the
researcher tries to predict job performance with the aid of locus of control: so, the researcher is trying to see if job performance depends, to a significant degree, on locus of
control. Option 2 is wrong because locus of control is the aspect of motivation that is being studied. Option 3 is incorrect because ‘productivity’ is a term indicating a
measurement of job performance, and it therefore refers to the dependent variable and not the independent variable
Question 7
Which is of the following statements gives the best description of the goal of psychological research?
As explained on pages 3 – 5 in the Study Guide for PYC3704, psychological research is mainly concerned with the testing of theories. The other three options are all part of
the process, but none of them are the goal.
Question 8
Empirical knowledge is knowledge that is based on - - - - -.
1. careful reasoning
2. appropriate theories
3. the observation of events
4. published research
Empirical knowledge refers specifically to knowledge that is based on the observation of physical objects and events in the world. The term is defined on p. 2 of the Guide to
PYC3704. The other options all refer to other ways in which knowledge can be gained.
Question 9
A ball is drawn at random from a box containing 6 red balls, 4 white balls and 5 blue balls. What is the probability that it is red?
1. 0.333
2. 0.400
3. 0.500
4. 0.667
Question 10
The probability that Jaime will fail his research methodology examination is 0.37. Find the probability that he will pass this examination.
1. 0.50
2. 0.37
3. 0.05
4. 0.63
There are only two possibilities: Jaime can either pass or fail. Therefore, the probability that he can pass is the probability that he can fail subtracted from 1 (where '1'
represents 100% certainty). So, we can use the formula p(not A) = 1- p(A) (see the PYC3704 Guide, p. 34). Since the probability that Jaime will fail is given, we have: p(Jaime
will pass) = 1 - p(Jaime will fail) = 1 - 0.37 = 0.63
Question 11
Two pupils, one girl and one boy, will be chosen to go on a field trip from a class of 12 girls and 15 boys. What is the probability that Mary and her brother John (who happen to
be in the same class) will both be selected if boys and girls are selected separately at random?
1. 0.0056
2. 0.15
3. 0.0741
4. 0.037
You have to use the basic formula for probability: P(A) = (No. of favourable events)/(No. of possible outcomes) separately for boys and girls. In the case of the girls, we know
that there are 13 possible outcomes (any one of the 12 girls can be selected) and we also know that there is only one ‘favourable’ event (Mary being selected). Therefore, we
have: p(Mary) = 1/12 = 0.0833 (rounded off) Similarly, for the boys: p(John) = 1/15 = 0.0667. To combine the probabilities for Mary and John, we should use the multiplicative
rule (see p. 35 of the PYC3704 Guide): p(Mary AND John) = p(Mary) X p(John) = 0.0833 X 0.0667 = 0.0056.
Note that since the scenario states 'boys and girls are selected separately at random', this is not a conditional probability (the choice of Mary from the group of girls have no
effect on the choice of John from the boys) so the formula for conditional probabilities (p. 36 in the Guide) does not apply.
Question 12
Danny flips four well-balanced coins. What are the odds that two of the coins will land 'heads' up?
1. 0.125
2. 0.25
3. 0.375
4. 0.5
The probability can be calculated as Number of Favourable Events / Number Possible Outcomes. (See p. 29 of the Study Guide). To get the 'number of possible outcomes',
we need to establish the sample space, which is the list of all possible elementary outcomes in this experiment. We can expect that since the outcome for each coin is
independent of the other, and each coin can have one of two outcomes ('Heads' or 'Tails'), the size of the sample space is 2 X 2 X 2 X 2 = 16 (this can also be written as 24).
But what does this sample space consist of? This list of all possible elementary outcomes of the experiment is presented in the table below. It indicates all possible
configurations when four coins are flipped (with outcome 'Heads' indicated as H and 'Tails' as T):
This table actually reflects the probability distribution of the possible outcomes in this experiment, with the probability of each outcome given in the row at the bottom. If you
count them, you can see that there are 16 possible outcomes, as we predicted. The middle column contains the favourable outcome, as required in this question. This refers to
those possible outcomes where two of the coins fall Heads and the other two fall Tails. You can see there are 6 of these possible outcomes. So, the probability (the 'odds') of
getting this result is 6/16 = 0.375
The probability of getting 'Heads' is 50%, so we should set p = 0.5 in the formula. We are interested in two successes in four trails, so we can substitute x = 2 and n = 4 in the
formula. Keeping in mind that 4! is the factorial of four, defined as 4! = 4 X 3 X 2 X 1 (see p. 42 in the Guide), the calculation would proceed as follows:
A third way in which this probability can be determined is by making use of a computer program, such as the 'BINOMDIST' function in Microsoft Excel. Follow the instructions
on p. 43 of the Guide, making the following substitutions:
The result should look like this, with the probability calculated as 0.375 (inside the diagram):
Question 13
A standard normal distribution has a mean of - - - - - and a standard deviation of - - - - - and the probability of getting a standardized z-score of greater than 1 standard
deviation is - - - - -
1. 0; 1; 0.3413
2. 1; 0; 0.5
3. 0; 1; 0.1587
4. 1; 0; 0.3413
The standard normal distribution is also referred to as the z-distribution, and it is a normal distribution which was transformed so that the mean is zero and the standard
deviation is 1. Options 2 and 4 are therefore ruled out. To determine the probability of getting a standardized z-score of greater than 1 standard deviation you can look at
Figure 2.7 on p. 53 of the Guide to PYC3704. It would be equivalent to the area starting at 1 up to the far right of the figure, which adds up to 0.1359 + 0.0215 + 0.0013 =
0.1587. An alternative way to determine this would be to look it up in the z-tables, in Appendix D of the Guide. Look for the row where z = 1 and then at the column under
'Smaller Portion'. It is this smaller portion that is relevant because (as you can see from Figure 2.7) it is the small region at the far right of the distribution (beyond the point
where z = 1) that is relevant, and this gives you a probability of p(z ≥ 1) = 0.1587. Note that the 'larger portion' would give the probability of getting a standardized z-score of
less than 1 standard deviation, that is, the large area to the right of z = 1 in the graph. The value of 0.3413 given in Option 1 is the region between 0 and 1 on Figure 2.7. It is,
in other words, the probability of a z-score of between 0 and 1, which is not what was asked for in this question
.
Question 14
Which best describes the frequency distribution of the ages of students attending a particular class?
1. A graph of the ages of the students showing the number of students in each of a number of categories, arranged from young to old.
2. The total of all ages, divided by the number of students.
3. The standard deviation of ages, indicating the width of the age distribution.
4. A graph of the ages of the students showing the way that the data is distributed around the mean age of the students relative to the standard deviation.
The expression ‘frequency’ implies the count of observations in each of a number of categories. A frequency distribution of ages will represent the number of students falling in
each of a number of age categories, which can be represented graphically in a histogram. Options 2 and 3 are not suitable for representing a frequency distribution. The
frequency distribution will actually give an approximation of the general distribution of the raw scores around the mean as implied in Option 4, but this is relevant only if the raw
scores can be assumed to be normally distributed, which is not necessarily true.
Question 15
The normal curve is asymptotic. This means that - - - - -.
1. it is symmetrical
2. it is bell shaped
3. the two tails never touch the horizontal axis
4. it follows the z-distribution
The term ‘asymptotic’ is defined on page 52 in the Guide for PYC3704. It refers to the fact that the two tails of a perfect normal distribution move closer and closer to the
horizontal axis without ever touching it.
Question 16
The examination marks of a statistics course yielded a normal distribution with a mean of 62 and a standard deviation of 10. Determine the probability that the score would be
50 or below, if you were to select the score of one student at random. Into which of the four intervals listed below would the probability fall?
1. 0 to 0.25
2. 0.26 to 0.50
3. 0.51 to 0.75
4. 0.76 to 1.00
To determine the probability, the given value of x must first be transformed into its equivalent in the z-distribution. The z-score is calculated by using the formula for the z-
transformation (from p. 55 in the Guide):
So, the z-value associated with 50 is equal to -1.2. The probability of a score of x ≤ 50 is therefore equivalent to the probability of a z-value of z ≤ -1.2. This can be illustrated
by looking at the graph below. The probability that we are interested in is the grey area on the left-hand side of this graph.
The probability can be determined from the table of the areas under the standard normal distribution (given in Appendix D of the PYC3704 Guide). Note however that since the
table represents only positive values of z, you have to look at the probability that z ≥ 1.2, which will be the same probability as p(z ≤ -1.2). The standard normal distribution is
symmetrical around the mean of zero, so the smaller region at the far left represents the same probability as the smaller region on the far right of the graph (see p. 163 of the
PYC3704 Guide). you have to look in the table for the probability associated with z = 1.2 under the column marked ‘Smaller portion,’ because we are interested in the area
under the normal curve from the point where z = 1.2 and above (the mirror image of the darker area in the graph); that is, the area in the tail of the normal distribution graph on
the far right hand side, which is the same as the area we are interested in, on the far left. (Note that if the question related to the probability of x > 50, we would have looked at
the ‘Larger portion’ given in the table). You should find that p(z ≥ 1.2) = 0.1151, which is equivalent to p(z ≤ -1.2), and lies between 0 and 0.25, as specified in Option 1.
Question 17
1. 0.05 or 0.01
2. the level of significance
3. the calculated value of the test statistic
4. the p-value of the test statistic under the null hypothesis
The p-value is the probability that researchers would obtain the result which they observe in the sample of data if, in fact, the null hypothesis is correct (which is what is
referred to with 'under the null hypothesis'). This is also a direct indication of the probability that the researchers would make a mistake if they rejected the null hypothesis
based on this sample of data, which is referred to as an error of Type I. (See p. 85 in the Guide: "… the p-value gives the probability of an error of Type I exactly …").
Question 18
During the interpretation of psychological measurements, the normal distribution is often - - - - -.
Data from the real world are unlikely to follow an exact normal distribution. However, we often find that the normal distribution gives an acceptable approximation for the mean
values of any particular distribution of scores as long as the sample is not too small (over 30 or so). This is a consequence of the central limit theorem (see pp. 60 - 62 in the
PYC3704 Guide). So, if the measurement level is acceptable and the sample size adequate, the normal distribution can be used as a theoretical model for interpreting the
observed distribution of scores. Option 1 does not make much sense, although one could put it the other way around, to say that an observed frequency distribution of scores
is often adapted to fit the normal distribution. Option 3 is wrong because the relative frequency of observed scores cannot be calculated from the normal distribution. The mean
and standard deviation of a sample is calculated from the measurements, not from the normal distribution, so Option 4 is also invalid.
Question 19
The size of the level of significance depends on - - - - -.
The researcher chooses the greatest risk of making an error of Type I (rejecting the null hypothesis in error) that she is willing to make, and compares the p-value with this to
see whether this p-value is lower than this chosen level of significance (α). While there are conventional values for α such as 0.01 or 0.05, the researcher can in fact use any
value he/she deems appropriate, which is why Option 2 is not strictly correct. Unlike the p-value, this level of significance is not dependent on the size of the calculated test
statistic (as implied in Option 3).
A researcher suspects that the addition of certain food supplements to the diet of elderly people will reduce the decline in cognitive functioning that comes about because of
aging. She decides to test this using a neuropsychological test that measures the speed with which objects are identified (the Neuropsychological Perceptual Speed or NPS
test). It is known that the distribution of scores on this test is approximately normal and that a mean of μ=80 and σ=20 was found in the population of persons older than 65.
To investigate her hypothesis, she obtains a random sample of n=100 persons older than 65. Each member of this sample is given a daily dose of supplements over a period
of six months. At the end of this time, each person is tested on the NPS test and a mean of x¯ = 76 is found. The researcher plans to test the hypothesis at α = 0.05.
Question 21
The appropriate alternative hypothesis to be tested is- - - - -.
1. H1: μ < 80
2. H1: μ < 84
3. H1: x> 80
4. H1: μ ≠ 80
Perceptual speed is the time it takes to react to a visual stimulus. If someone has greater perceptual speed, this reaction time will become smaller (this, according to the
scenario, is measured by a test referred to as the NPS). If the supplement is effective, one would expect this measurement to decrease, not to increase. It follows that in order
to reject the null hypothesis, a researcher would expect the sample of elderly people who takes the supplement to have a significantly smaller mean measurement than the
population mean of μ = 80. Option 3 uses the sample mean symbol and in any case points in a direction that would imply elderly people who use the supplement do worse
(react more slowly) than the population on the test. Option 4 implies two-tailed testing, but the researcher wants to know whether cognitive functioning gets better. Whether it
gets worse is not relevant, given the research question and the way in which the hypothesis was set up (see Question 20 above). Option 2 is irrelevant.
Question 22
The mean of the sampling distribution of the mean is - - - - -.
1. 80
2. 84
3. 20
4. unknown
According to the central limit theorem, the population mean can be used as the mean of the sampling distribution of the mean. (See Section 2.4.2 on p. 60 of the PYC3704
Guide).
Question 23
The standard error is - - - - -.
1. 20
2. 2
3. 0.05
4. unknown
We know from the scenario that the population standard deviation σ = 20 and that the sample size is n = 100. These can be substituted into the formula for the standard error
(on p. 61 of the PYC3704 Guide) as follows:
Question 24
With the information as given in the scenario, what would be the appropriate statistical test to test the hypothesis?
The scenario and the hypothesis to be tested (identified in Question 21) indicate that the mean of a single sample is to be compared with a known population mean. Because
the population standard deviation is known (it is specified in the scenario above as σ = 20) the most appropriate test here would be the single sample z-test (zx¯). (See pp. 100
– 102 of the PYC3704 Guide). Using a single-sample t-test (Option 1) would not be totally wrong, but it is not the best test to use of the options given. When the population
standard deviation is known, the z-test is in fact more powerful than a t-test; i.e. it is more sensitive to statistical effects (see the reference to the 'power' of a test on p. 86 of the
PYC3704 Guide).
Question 25
The test statistic is calculated and, based on this, a computer program is used to determine that the two-tailed p-value = 0.055. What conclusion can be drawn?
1. The null hypothesis can be rejected, so the supplement improves cognitive functioning.
2. The null hypothesis cannot be rejected, so the supplement does not improve cognitive functioning.
3. The alternative hypothesis can be rejected, so the supplement does not improve cognitive functioning.
4. The null hypothesis can be rejected, so the supplement does not improve cognitive functioning.
The computer program supplied a two-tailed p-value, while the scenario requires a one-tailed test (consider the hypothesis, as indicated in Question 21). Therefore, the one-
sided result should be divided by 2: One-tailed p-value = two-tailed p-value/2 = 0.055/2 = 0.0275 (see p. 81 in the Guide).
Since this value is lower than the chosen level of significance (given in the scenario as α = 0.05), the null hypothesis can be rejected. Consequently, we can conclude that the
food supplement improves cognitive functioning. Note that while the rejection of the null hypothesis in Option 4 is correct, the wrong conclusion is drawn.
Question 1
Why would a researcher calculate a test statistic? It is a calculation - - - -.
1. which determines how far the observed measurements deviate from what we may expect by chance
2. which shows directly whether or not we can accept that the null hypothesis is true
3. to get a measurement by which we can calculate the level of significance
4. to determine whether or not we can reject the alternative hypothesis
Calculating the test statistic is the first step in a process of comparing the observed data with what may be expected by chance (i.e., if the null hypothesis were
true). Option 2 is not really appropriate because the emphasis is wrong here. The test statistic is calculated to see how far the effect which was observed in
the data (our measurements) deviate from what we may expect by chance if the null hypothesis was true. Not rejecting the null hypothesis would imply the
observation is too close to the situation predicted by the null hypothesis for a researcher to feel confident in rejecting the null hypothesis. So even if we were
hoping to find no effect (for example, no difference between groups or no relationship among variables), the formal aim of calculating the test statistic is still to
determine whether it is safe to reject the null hypothesis or not. For the same reason, Option 4 is not really correct (accepting the alternative hypothesis would
follow only as a consequence of the null hypothesis being rejected). The level of significance (α) is not calculated but chosen by the researcher as the
maximum risk of rejecting the null hypothesis in error that he or she is willing to take, so Option 3 is false
Question 2
Two samples may be regarded as dependent when - - - - - -.
1. each measurement in one sample is correlated with a measurement in the other sample
2. they were drawn from the same population
3. there is a systematic relationship between the measurements in one sample and the other
Answer: Option 3 is correct and is a definition of dependent samples, where the measurements form matched pairs.
Option 1 is not really valid as a definition of dependence. Correlations are usually calculated from two measurements (that is, two variables) in the same
sample, not from different samples. It is however sometimes possible to correlate measurements from two samples, and finding a correlation would probably
imply dependence of some kind. In such a case one first would have to pair off the specific measurements in the one sample with those in the other before you
would be able to calculate the correlation coefficient. Examples would be husband and wife pairs, or mother and infant pairs etc., compared on some
characteristic (construct) such as ‘empathy.’ This implies that you must already suspect that a dependence exists (based on the principle in Option 3) to be
able to match pairs of measurements, and the correlation is calculated to measure the extent of this correlation for some variable under investigation. A
significant correlation would imply that as one specific measurement changes, a corresponding change takes place in the other measurement in the pair,
above a level one may expect purely by chance. Being drawn from the same population does not imply dependence, so Option 2 is not correct. Randomness
of the samples is not relevant, and is in any case likely to imply a lack of dependence, so Option 4 is also not correct.
Base your answers to Questions 3 to 5 on the following scenario: Sally, an educational psychologist, predicts that the mean IQ score of a group of 50
children in a special school for gifted children will be higher than the expected population average of 100.
Question 3
Indicate which null hypothesis Sally should specify from the options below:
1. H0: x¯ = 100
2. H0: μ = 0
3. H0: x¯ > 100
4. H0: μ = 100
Sally must compare the mean of her sample of 50 children with a population mean, and she knows this is equal to the expected population average of 100. A
null hypothesis is a hypothesis which states that there is no effect, and in this case, it would mean no difference between the observed mean (calculated after
testing the sample of children) and the population mean of 100. Option 1 is wrong because hypotheses are always stated in terms of population parameters
(not sample statistics), which in the case of a mean is symbolised by μ. This is also an error in Option 3, but that error is made worse by the fact that ‘greater
than’ is not how we would formulate a null hypothesis, because ‘greater than’ implies an effect, not the lack of an effect which is implied in a null hypothesis.
Option 2 is wrong because the mean IQ in the population used in the comparison is 100, not 0.
Question 4
From reading the psychometric test manual, Sally knows that the IQ test was standardized on a mean of 100 and a standard deviation of 15. Which of the
options below would be the most appropriate statistical test which she could use to test the hypotheses?
1. tc
2. zx¯
3. tx¯
4. td
Answer: Option 2 gives the most appropriate test statistic to use.
Sally needs to compare a sample mean with a constant population mean which is already known to be 100. This implies a single sample test for comparing a
mean with a given value (100 in this case). It would therefore be either a single sample z-test or a single sample t-test. Because the population standard
deviation is known (it is specified in the question above as σ = 15) the most appropriate test here would be the single sample z-test, which is written zx¯. (In
such a case the z-test is in fact more powerful than a t-test; i.e. it is more sensitive to statistical effects).
Option 1 and Option 4 both refer to two-sample t-tests which are to be used to compare two sample means and not one sample mean with a constant
population mean.
Question 5
Sally decides to perform her statistical test at a level of significance of α = 0.05. Based on the data from the gifted children the computer calculates a test
statistic and reports a two-tailed p-value of p = 0.082. What can Sally conclude with regard to the hypotheses?
1. She should reject H0, and can conclude that the IQs of the gifted children is significantly higher than the population average
2. She cannot reject H0 and so she cannot conclude that the IQs of the gifted children is significantly higher than the population average
3. She should reject H1 and can therefore conclude that the IQs of the gifted children is not higher than the population average
4. She cannot make a conclusion because the exact value of the calculated test statistic is not provided
Question 6
A researcher wants to test the efficacy of psychotherapy aimed at relieving anxiety. She applies a test which measures level of anxiety to 30 patients before
their therapy begins, and again three months later, predicting that the latter scores will be lower (reflecting less anxiety). Which description of the research
design below is the most appropriate?
This is a repeated measures design, with the same people being measured twice, before and after an event. From the point of view of formal research design,
this is to be regarded as two samples with highly dependent measurements. (See the top of p. 118 in the Guide for PYC3704
Question 7
Suppose you are comparing two means with a t-test for independent samples, and you find that the value of the statistic calculated for your research results is
zero. Which conclusion is appropriate?
Look at the formula for the t-test for two independent samples:
The only time that the value of tc can be zero, is when there is no difference between the means, which implies that the null hypothesis must be true. A similar
argument works for any of the other forms of the t-test.
This also implies that Option 3 must be false. If the means are the same the t-statistic will be zero, so Option 4 is also not true. Option 2 is false because the
probability of a Type I error will actually be very large: if there is no difference between the means you are sure to make an error if you reject the null
hypothesis in spite of this. The whole point of the test is to determine whether the means differ significantly. If they do not differ at all there is actually no need
to do a statistical test
Question 8
The probability under the null hypothesis of obtaining a t-value of 2.5 or higher in the case of a one-tailed test is - - - - - that for a two-tailed test.
1. the same as
2. twice
3. half
4. unrelated to
Note that the question actually relates not to the t-statistic value as such, but to the probability of getting that particular value for the t-statistic; in other words, to
the p-value. A one-tailed p-value (used in the case of a directional hypothesis) is half the size of a two-tailed probability. Conversely, a two-tailed p-value (used
in the case of a non-directional hypothesis) is twice the size of a one-tailed p-value. (See page 81 of the PYC3704 Study Guide).
Question 9
Which is the most appropriate research hypothesis for the teacher to test?
1. The mean mathematics score for the treatment group will be greater than that of the control group
2. The mean mathematics difference score (difference scores of treatment group minus control group) will differ significantly from zero
3. The mean mathematics score for the control group will be greater than that of the treatment group
4. There will be a significant difference in the mean scores for mathematics between the treatment and control groups.
Question 10
Which is the appropriate test statistic to be calculated when testing this hypothesis (as specified in Question 9)?
1. The t-statistic for the difference between the means of two independent samples
2. The t-statistic for the difference between the means of two dependent samples
3. The t-statistic for the mean difference score of a single sample
4. The r-statistic for the correlation between two variables
There are two samples (treatment and control groups) which can be regarded as independent (there is no particular systematic relationship between them).
Since two independent samples of different sizes are being compared, rather than two variables from a single sample, the correlation coefficient is not really
relevant, so Option 4 is also incorrect.
Question 11
When calculating the test statistic referred to above, the psychologist uses a computer program to calculate the p-value. The computer presents her with a two-
sided p-value of p = 0.054. What can she conclude?
1. The outcome is significant, so she can conclude that the alternative hypothesis can be rejected and that learning to play chess does have a positive effect
on the mathematics ability of the pupils
2. The outcome is significant, so she can conclude that the null hypothesis can be rejected and that learning to play chess does have a positive effect on the
mathematics ability of the pupils
3. The outcome is not significant, so she can conclude that the null hypothesis cannot be rejected, and that chess has no effect on the mathematics ability of
the pupils
4. The outcome is not significant, so she can conclude that the null hypothesis cannot be rejected, and that chess has a negative effect on the mathematics
ability of the pupils
Note that we need to do a directional test (as shown in the answer to Question 9 above). The two-sided p-value given by the computer must therefore be
divided by 2 before it can be compared to the level of significance. This gives us a one-tailed p-value of 0.054/2 = 0.027. This is smaller than the level of
significance which was given in the scenario above as α = 0.05, and the outcome is therefore significant. For this reason, Options 3 and 4 can be eliminated. A
significant result means that there is only a small probability that the null hypothesis is true, so it is relatively safe to reject this null hypothesis. This in turn
implies that we can accept the alternative hypothesis, that is, we can conclude that learning to play chess does have a positive effect on the mathematics
ability of the pupils.
Note that Option 1 is false because significance implies that the null hypothesis can be rejected, not the alternative hypothesis. In any case, hypothesis testing
is always concerned with whether the null hypothesis can be rejected or not. If the p-value was larger than the level of significance, we would have concluded
that the probability that the null hypothesis is true is too large for us to reject it (in other words, the improvement in score for the pupils who play math is not
greater than what we would expect purely as a result of chance). By convention, this not referred to as 'rejecting the alternative hypothesis' even though this is
what is implied (see Section 3.3.1 in the PYC3704 Guide for an explanation of this rule).
Question 12
A scatter plot is a graphical representation of - - - - -.
1. the relationship between two variables measured on a nominal scale within a single group
2. the frequency distribution of a sample of measurements
3. relationship between two groups of subjects with regard to a single variable measured on an interval or ratio scale
4. the relationship between two variables measured on a ratio or interval scale within a single group
This type of graph is described on pp. 130 – 133 of the PYC3704 Study Guide. It would make sense to use such a graph only if the data represented actual
measurements and the distances between measurements are equal, that is to say, for ratio or interval scale measurements (see Appendix B in the Guide).
Such a graph can only be used when there are matched pairs of data, which implies two variables relating to a
relevant to frequency distributions.
Question 13
Which of the following does not represent a valid value for Pearson's product-moment correlation coefficient?
1. -0.72
2. 0.00
3. -1.01
4. 1.00
Answer: Option 3 is the correct answer, since a number smaller than -1 is not a possible value for Pearson’s product-moment correlation coefficient r.
Pearson's correlation coefficient can range from -1 to 1, so Options 1, 2 and 4 all represent possible values. (See p. 133 in the PYC3704 Guide).
Question 14
The - - - - - of the correlation coefficient r for two variables indicates the direction of the relationship, while the - - - - - indicates the strength of the relationship.
1. value; sign
2. linearity; magnitude
3. sign; value
4. size; slope
A correlation coefficient is a measurement of the linear relationship between two variables and it can vary between -1 and +1. The absolute value of the
number (ignoring the sign) indicates the strength of the relationship, while the sign (+ or -) shows the direction of the relationship. If the sign is positive the
variables vary in the same direction, but if it is negative it implies that as one variable gets bigger, the other tends to become smaller (See pp. 130 - 133 of the
PYC3704 Guide).
Question 15
A contingency table represents - - - - -
A contingency table is a two-dimensional table used to represent the cross classification, or cross tabulation, of the responses relating to two nominal or
categorical variables. It is basically a way to display and record the relationship between the two variables. The frequency counts of one variable are presented
in the rows of the table and the frequency counts of the other variable in the columns, as shown in table 6.4 on p. 142 and Table 6.5 on p. 144 of the PYC3704
Guide. (See also the table in the scenario for Questions 24 and 25 below).
Question 16
The type of graph reproduced below is called a - - - - -?
1. scatterplot
2. histogram
3. normal curve
4. probability distribution
A histogram is a graphical display which shows what proportion of cases fall into each of several categories. The categories are usually specified as non-
overlapping intervals of some variable and they have to be adjacent to each other. (An example is given on p. 40 of the PYC3704 Guide). A scatter plot (Option
1) is a display of data points for sets of variables (see the Guide, pp. 126-8). A normal curve (Option 3) is a graphical representation of the normal distribution
(see p. 42 of the Guide).
Question 17
A researcher hypothesizes that a relationship should exist between spatial ability and general aptitude for mathematics. She collects the results of a sample of
n = 100 school children for a mathematics test and measures the spatial ability of each with a test that represents a person’s ability to rotate objects mentally
on a 10-point scale.
Which of the following is the most appropriate way to express the null hypothesis for this research?
1. r = 0
2. μ = 0
3. ẋ = 0
4. ρ = 0
A null hypothesis implies that there is no effect, which in this case implies no relationship between the two variables spatial ability and general aptitude for
mathematics. If the value of a Pearson's correlation coefficient is calculated for two variables with no relationship between them, one would expect it to be
close to or equal to zero. Formal hypotheses are set in terms of population parameters, and the symbol for a population correlation is the Greek letter ' ρ' (or
rho), not the Roman letter 'r' (see p. 137 in the PYC3704 Study Guide).
Key:
Empathy and Assertiveness were tested on 5-point scales (ranging from 1 equalling ’a small amount’ to 5 equalling ’a large amount’). For Assertiveness,
Assert1 indicates the test result before the workshop, and Assert2 is the test result thereafter
Question 18
The researcher wants to determine whether the workshop is effective by comparing the measurements on the assertiveness test before the workshop with the
measurements on this test after the workshop. Which of the following is an appropriate test statistic to calculate?
The problem implies that you would like to compare the two mean values for assertiveness, for matched pairs of measurements, before and after the
workshop. Therefore, independent group testing (Option 2) is eliminated. Do not let the fact that the samples being compared refer to the same people being
tested twice confuse you (see the top of p. 118 in the Study Guide for PYC3704, where the use of the td test for test-retest situations is explicitly mentioned).
Calculating Pearson's correlation statistic r (as suggested in Option 1) is technically possible but is not a suitable choice for this particular problem. The
correlation coefficient would tell you whether people in the same group who had a high assertiveness score before the workshop still have a high score
thereafter. A high correlation could however occur even if the workshop was not effective at all (for example, if each individual somehow obtained nearly the
same score as before; or even if their assertiveness became systematically less rather than more, as the researcher is likely to hope for). This is also why a
directional hypothesis would be advisable here. The variables being compared are quantitative measurements, not categories or nominal level measurements,
so the chi-square test (Option 4) is not applicable here.
Question 19
The researcher calculates the value of the appropriate test statistic (in Question 23 above). In which of the four intervals below will the absolute value of the
test statistic fall (i.e., ignoring a plus or minus sign)?
1. Between 0 and1.0
2. Between 1.0 and 2.0
3. Between 2.0 and 3.0
4. Above 3.0
The mean and standard deviations of d can be calculated (using the formulas in Appendix C of the Guide, substituting x with d):
The mean of d will be:
Using these values in the formula on p. 119 of the Guide, the calculation of the td statistic proceeds as follows:
Note that the symbol D¯ in the formula refers to the expected value of the mean differences if the null hypothesis were true, which would be 0. If the null
hypothesis is true, there would be no difference in measurements, so the mean difference score would be zero. (See p. 119 in the PYC3704 Guide). Also keep
in mind that whether the sign of the calculated td value is positive or negative will depend on which data set you subtract from which, and whether or not it
matters will depend on whether you are testing a one- or two-tailed alternative hypothesis. We would expect leadership ability to increase (which implies
directional testing), so the mean leadership ability after the workshop should be greater than mean leadership ability before the workshop if the alternative
hypothesis is valid.
Question 20
The researcher finds that the test statistic (calculated in Question 19) yields a p-value of p = 0.000572 (as calculated by a computer). What conclusion can be
drawn from this? (Use a level of significance of 1%).
1. There is a significant improvement in the level of assertiveness as measured before and after the training
2. There is no significant improvement in the level of assertiveness as measured before and after the training
3. The probability value is too low, so the null hypothesis cannot be rejected with confidence
4. There is a significant difference in the level of assertiveness as measured before and after the training, in favour of the former
The p-value is smaller than the level of significance (that is, p-value 0.000572 < 0.01). The small p-value implies that there is a very small probability that the
difference between the two means (before and after the workshop) is the result of chance or random measurement error, so the null hypothesis (which would
imply no change) can be rejected.
Option 2 is incorrect, and Option 3 implies that a low p-value is associated with not rejecting the null hypothesis, which is the wrong way around. The smaller
the p-value, the less the chance is that the null hypothesis is valid. The answer given in Option 4 is in the wrong direction: it implies assertiveness is greater
before than after the assertiveness training.
Question 21
The researcher is also interested to know whether the difference between the two assertiveness scores before and after the workshop is of practical
significance; that is to say, irrespective of its statistical significance, whether the effect can be regarded as big or small. Calculate the relevant measurement of
the size of the effect to answer the question below.
The size of the effect was found to be - - - - -
1. small
2. medium
3. large
4. impossible to determine from this data
To get an indication of the practical usefulness of the result the effect size should be calculated. This would give an indication of the absolute size of the
statistical effect, independent of the sizes of the samples that were used (see Section 3.3.3 in the PYC3704 Study guide). In the case where two means are
being compared, this effect size can be found by calculating Cohen’s d. The general form of this measurement is to subtract one mean from the other and then
to divide this by the overall standard deviation for all the data (see p. 87 in the PYC3704 Study Guide). The actual way in which Cohen's d is to be calculated
can however differ depending on which mean is to be compared with which (i.e., it differs slightly depending on which statistical test was used for this data).
The specific form of Cohen's d which a researcher should use when those means being compared comes from two dependent samples is based on the mean
and standard deviation of the difference score d. (The fact that 'd' can refer to a difference score and that Cohen chose 'd' for his effect size index is just a
coincidence; do not confuse them). The difference score values which were calculated before (in Question 19) can be inserted directly into the formula for
Cohen’s d.
The version of the formula which is specifically designed for dependent means is given on p. 120 of the PYC3704 Guide:
The explanation on why we can assume D¯ = 0 is the same as that given in the feedback to Question 19, above).
The rule of thumb given on p. 88 of the PYC3704 Guide can be used to judge the relative practical importance of the result. An effect of greater than 0.8 can
be regarded as ‘large’.
Question 22
The researcher is also interested to determine whether a relationship exists between empathy and assertiveness before the workshop commences. Which of
the following is the most appropriate test statistic to use?
Question 23
Which of the following is the result if this test statistic (in Question 22) is calculated? The value of the test statistic will fall - - - - -
1. below -0.4
2. between -0.4 and 0
3. between 0 and 0.4
4. above 0.4
To answer the question, you will have to calculate r. The best way to do this would be to first create a table with a list of all the variables and their derivatives
which you need (similar to the table on p. 135 of the Guide).
Note the difference between Σxy and ΣxΣy. In the first case you multiply each x and y-value, then add the results up afterwards, while in the second case you
add all x-values, add all y values, then multiply the two results.
The high negative correlation implies that the more assertive a person (in this sample of data) is, the less empathy they are likely to have. Since this value of r
= –0.838 is smaller than –0.4, the result fits into the category given in Option 1
them as Male or Female, and according to the type of post which they occupy. From this, she creates
the table below.
Question 24
What is the kind of table that is given in the scenario above referred to as?
1. A frequency distribution
2. A correlation matrix
3. A contingency table
4. A cross tabulation table
Question 25
If gender has no effect on the posts that employees occupy in the organization, how many males would one expect to find in the human resources
department?
1. 5.62
2. 20
3. 4.44
4. 3
To get the expected frequency (the way data would be distributed purely by chance) you need to multiply the total number of people in human resources (total
of row 4) with the total number of males (total of column 1), and divide this by the overall total, using the formula:
Therefore: e41 = (10 x 20)/45 = 200/45 = 4.4444 (which can be rounded off to 4.44)
Note that Option 4 is incorrect because it gives the observed frequency for tall followers, while the question asks for the expected frequency if the frequencies
were distributed equally with no interaction between gender and job category. Option 1 is the total divided by the number of possible categories (i.e. 45/8) but
this does not take account of the fact that the expected distribution of the classifications should be proportional to the overall distribution of the data. Option 2 is
the number of males. (See pp. 143-5 of the PYC3704 Guide for details on calculating expected frequencies from a contingency table).
Question 1
The goal of quantitative research in psychology is best described as aiming to - - - - -
1. develop appropriate statistical tests which can be used to determine the relationships among psychological variables that occur at a level greater than
chance
2. develop theories that helps us to explain human experience and behaviour
3. formulate clear hypotheses based on insights about human experience and behaviour
4. convert theoretical constructs into measurable variables through operationalisation
The goal of research is to develop theories which can explain aspects of human behaviour and experience. Options 3 and 4 refer to stages in the process of
doing quantitative research but these are not the goals of the research. The goal of the research is also not to develop statistical tests, as implied in option 1.
These tests are developed by statisticians and are used by researchers in social and other scientific research, but developing the tests is not part of the goal of
the research.
Question 2
A psychologist believes that personality factors such as ability to get along with other people in a team may influence their success in a team sport. To
investigate this, she draws a sample of participants in team sport, comprising of soccer players, hockey players and cricketers from various sports clubs. After
evaluating their performance on an appropriate psychometric test, she divides players from the three categories of sports into two groups: those who have
good social skills and those who measure low on the test for social skills. She then uses evaluation forms, based on interviews with different sports experts, to
assess the actual sports performance of the members of the sample.
The researcher wants to determine whether sports performance (a measurement of how well a specific sports participant performs) depends on social skills.
This implies that the construct ‘social skills’ is the independent variable which is varied to see how it affects the dependent variable, the measurement of ‘sports
performance.’ The category of sport (soccer, hockey or cricket) gives us information about the population from which the sample was drawn (the types of
sports participants that were involved), but in the context of this scenario this variable is not being investigated here. It would be possible to compare the three
groups, but this would not answer the research question.
Question 3
Empirical knowledge is defined as - - - -
Empirical knowledge is the knowledge that you have obtained because you observed what has happened 6 during an event. Option 1 is wrong because
empirical knowledge derives from experience rather than theory. Option 2 is and 3 also wrong because neither of these fit to the definition of ‘empirical’
knowledge.
Question 4
Operationalizing a construct means to - - - - -
A construct is some event or entity which the researcher regards as being of importance in an explanation of the phenomenon under investigation. In
quantitative research, where data are treated in numeric form, the problem is how to measure it, and the procedure used to achieve this is referred to as
operationalization. For example, one may suppose that persons with concentration problems are affected by anxiety. To research this, some way will have to
be found to measure each of the constructs 'ability to concentrate' and 'anxiety', perhaps by developing suitable tests or observation methods. Option 1 refers
to the process of developing theories and Option 2 to the process of developing hypotheses. Option 3 is fairly ambivalent (it is not clear what ‘level’ refers to)
but is not part of a definition of the process called ‘operationalizing’
Question 5
Which one of the definitions below is FALSE?
1. The term construct is used to refer to an aspect of human behaviour or experience which is abstracted from observations for study in psychological research
2. Measurement is a process whereby numbers are allocated to a construct according to a rule
3. When a psychological variable is measured, the result is referred to as a statistic
4. When a construct is measured, the resulting quantity is referred to as a variable
Option 3 is the only description that is false. A statistic is some kind of summary value calculated from a set of measurements across a sample of data, not one
specific measurement. This would include, for example, means, variances and standard deviations that are calculated from a sample of cases which were
measured on some variable. These are collectively referred to as sample statistics (see the PYC3704 Guide, section 1.4.3, especially p. 13). The term ‘test
statistic’ is also used for the test value (a value with a known probability distribution) calculated in various statistical tests. The other three options are all true.
Question 6
A statistician studying measurement theory would assume that a measurement x is made up of the ‘true’ measurement x0 and an error term e. Which of the
following assumptions are usually made regarding the error component (e)?
It is assumed that the error term is normally distributed around a mean of zero. This is because positive and negative deviations from the 'true' mean (the
errors) are assumed to be equally probable, and are likely to cancel each other out (see pp. 14-15 of the PYC3704 Study Guide).
Question 7
A measurement that summarizes an aspect of a population is called a - - - - - while a measurement that describes the same aspect of a sample is called - - - -
1. construct; variable
2. parameter; statistic
3. statistic; parameter
4. variable; construct
This option makes use of the definitions of ‘parameter’ (for a population) and ‘statistic’ for a sample. See Section 1.4.3 of the PYC3704 Guide (especially the
part that begins on the bottom of p. 12). The other options are incorrect.
Question 8
A variable that we may not be aware of during an experiment or research process that may influence the outcome is called a - - - - -
1. manifest variable
2. hidden variable
3. latent variable
4. random variable
A hidden variable affects the dependent variable in ways that we are not aware of, or that we choose to ignore. A manifest variable is an observable variable
while a latent variable cannot be directly observed. The latent variable is made manifest by using an appropriate procedure to measure it. It is also referred to
as a latent construct. A random variable is a variable with a value that cannot be predicted before the experiment has been performed (see p. 9 of PYC3704
Study Guide).
Question 9
A ball is drawn at random from a box containing 6 red balls, 4 white balls and 5 blue balls. What is the probability that it is red?
1. 0.333
2. 0.400
3. 0.500
4. 0.667
Question 10
Suppose the height of military recruits is distributed normally with a mean of 1750 mm and a standard deviation of 50 mm. Drawing repeated samples of 25
recruits each we expect the standard deviation of the sample means to be about - - - - - mm.
1. 2
2. 10
3. 50
4. 25
If repeated samples of recruits are drawn, the mean for each of these samples can be calculated. The standard deviation of the sample means is the way in
which these calculated means vary across the repeated samples. It is an indication of how accurately these sample means reflect the true population mean.
This measurement of the variability of the sample means is referred to as the standard error of the mean (see pp. 58 – 62 in the PYC3704 Guide). It can be
estimated from the population standard deviation and the size of the samples using the formula on p. 61 of the Guide, as follows:
Question 11
If a coin is flipped three times, the sample space of possible outcomes is:
This question requires that you determine the sample space for three flips of a coin, which in turn means that you must determine all the possible outcomes if a
coin is flipped for three times. Option 2 gives the correct answer, because this is the only option that lists all the outcomes correctly. Option 1 does not list THH,
Option 3 does not list HHT and THH, and Option 4 does not list TTH while HTT is listed twice.
Question 12
The probability that Jaime will pass his research methodology examination is 0.53. Find the probability that he will fail this examination.
1. 0.57
2. 0.35
3. 0.054
4. 0.47
We use the formula p(not A) = 1- p(A) (see the Guide, p. 34). Since it is given that the probability that Jaime will pass is 0.53, we have:
p(Jaime will pass) = 0.53. Therefore, p(Jaime will not pass) = 1-0.53 = 0.47
Question 13
Danny flips four well-balanced coins. What are the odds that two of the coins will land ‘heads’ up?
1. 0.125
2. 0.25
3. 0.375
4. 0.5
The probability can be calculated as Number of Favourable Events / Number Possible Outcomes. (See p. 29 of the Study Guide). To get the 'number of
possible outcomes', we need to establish the sample space, which is the list of all possible elementary outcomes in this experiment. We can expect that since
the outcome for each coin is independent of the other, and each coin can have one of two outcomes (‘Heads’ or ‘Tails’), the size of the sample space is 2 X 2 X
2 X 2 = 16 (this can also be written as 24). But what does this sample space consist of? This list of all possible elementary outcomes of the experiment is
presented in the table below. It indicates all possible configurations when four coins are flipped (with outcome 'Heads' indicated as H and 'Tails' as T):
This table actually reflects the probability distribution of the possible outcomes in this experiment, with the probability of each outcome given in the row at the
bottom. If you count them, you can see that there are 16 possible ways in which four coins can fall (that is, 16 possible outcomes), as we predicted. The middle
column contains the favourable outcome, as required in this question. This shows all the possible outcomes where two of the coins fall Heads and the other
two fall Tails. You can see, there are 6 of these possible outcomes. So, the probability (the 'odds') of getting this result is 6/16 = 0.375
A third way in which this probability can be determined is by making use of a computer program, such as the ‘BINOMDIST’ function in Microsoft Excel. Follow
the instructions on p. 43 of the Guide, making the following substitutions:
• Number of successes to determine outcomes for (two heads): Number_s = 2
• Number of trails to consider (four coins): Trials = 4
• Probability for a single outcome ('Heads'): Probability_s = 0.5
• Cumulative probabilities are not required, so Cumulative is set to ‘False’
The result should look like this, with the probability calculated as 0.375 (inside the diagram):
Question 14
Consider the probability experiments given below:
In which of these experiments are the different possible outcomes NOT equally likely?
1. Only (b)
2. (b) and (c)
3. Only (c)
4. (a) and (c)
Question 15
In which subject did Patrick do best, relative to his class?
1. Mathematics
2. Science
3. Geography
4. History
So, it is clear that in the case of Mathematics, Patrick’s mark is higher than any of the others, and it is in fact 2.3 standard deviations above the mean. In each
of the other subjects his marks transform into a smaller value than this, relative to the z-distribution.
Question 16
Find the subject where Patrick did worst relative to the rest of his class. What is the probability of any other one of Patrick's classmates getting a mark of equal
to this or better for this particular subject? Choose the option closest to the correct answer.
1. 0.1
2. 0.3
3. 0.6
4. 0.9
From the calculations for the previous question above, you can see that Patrick's mark of 68% for History translates to zH= -1.4 on the z distribution, which is
lowest in comparison to any of the other scores. The fact that the z value is negative shows that it falls below the class mean. Now you have to find what the
probability is of a random classmate getting a better mark than this for History. This is indicated by the grey area in the graph below.
You can see that the probability of getting a mark of 68 or greater is equivalent to the area from the point where z = -1.4 to the far right of this on the graph of
the standard normal distribution. If you look this up in the tables for the z-distribution (Appendix D in the PYC3704 Guide), you will see that the probability of
getting a mark greater than 68% (which is the same as the probability of getting a mark greater than z = -1.4) is 0.9192. This can be written as p(z ≥ -1.4) =
0.9192 and rounded off to 0.9.
Note how the graph helps you to see that it would be the 'larger portion' indicated in the tables that is relevant. It is always useful to draw a graph like this if you
get a question where you have to look up the probability of a certain range of z-values.
Question 17
Suppose we have stated H0: μ = 10, and H1: μ < 10, and find that the sample mean corresponds to a z-score of -3. This means that the corresponding p-value
Question 18
The standard error is a measurement of - - - - -
Question 19
Look at the graph based on the frequency distribution of a measurement of a variable x from normally distributed data which is reproduced below. Using the
information provided in the graph, calculate the probability that a measurement made at random on this particular scale will fall in the area under the curve
coloured in grey.
Select the answer closest to the calculated probability from the options given below:
1. 0.933
2. 0.775
3. 0.159
4. 0.5
Answer: Option 2 is correct. [Note that to make this discussion easier to follow, we added a light grey section at the right-hand side of the graph below]
We want to determine the probability of a value of x which falls within the dark grey area in the graph, that is, the probability of getting a value of x which falls
above 25 but below 50. To find this, we first have to find the value on the z-distribution which is equivalent to an x-value of 25. In other words, the value of 25
must first be transformed into a z-score. The formula which should be used is given on p. 55 of the Guide (we can presume that we are working with sample
data, so the population mean symbol μ can be substituted by the sample mean symbol x¯ and the population standard deviation by the sample standard
deviation symbol s).
This tells us that getting a x-value of 25 or greater is equivalent to getting a z-value of -1.5 or greater. We can now look up the probability of getting this from
the tables of the standardized normal distribution (the z distribution) in Appendix D in the Guide. You should find that the probability of an x-value of 25 and
greater is p(x>25) = p(z > -1.5) = 0.9332. Note that you can see from the graph that you would be interested in the larger portion as indicated in the table, and
that the negative sign can be ignored (since the normal distribution is symmetrical). Since we are interested in the probability of z falling below 50 only, we
have to subtract the area of a probability of a value greater than 50 from this result (the area is indicated in light grey in the graph). To find out what the
equivalent z-value is when x = 50, we again make use of the z-transformation formula
Use the scenario below to answer Questions 20 and 21. An educational psychologist measures the general knowledge of 100 learners on a questionnaire
with 50 questions. She finds the mean and standard deviation of the scores are 20 and 8 respectively for the sample of data.
Question 20
What is the z-score corresponding to a test score of 14 on the general knowledge test?
1. 1.33
2. 0.75
3. – 0.75
4. – 1.33
Question 21
What is the probability that a specific learner will obtain a score of 14 or less on the questionnaire in the scenario above?
1. 0.2266
2. 0.7734
3. 0.0918
4. 0.9082
Question 22
When two means are compared, the p-value expresses the probability that an observed difference - - - - -
1. is statistically significant
2. is due to the alternative hypothesis
3. is due to chance or sampling error
4. will be found between the means
The p-value gives the probability that this result (for example, a difference between means which is observed in the sample data) would be obtained if the null
hypothesis was true. It reflects the probability that the result is due to chance. If the chance of getting this particular result is small, the result is probably NOT
due to chance. Then you can reject the null hypothesis.
Question 23
If you decided to set the level of significance to α = 0.001 rather than a more conventional level of α = 0.05, what would the consequences be? Choose the
correct option(s).
A smaller level of significance also means that there is an overall smaller chance of rejecting H0, so (b) is generally valid. The smaller the level of significance
the less likely one is to reject H0, but that could mean that H0 may not be rejected even when it should be. This would be an error of Type II, so (a) is true. A
smaller level of significance makes the risk of a Type I error smaller, so (c) is false.
Use the scenario below to answer Questions 24 to 27. Mpho is a psychology student, who wants to understand whether participation in sports is in any way
related to the cognitive abilities of children at an age of 10 years. In order to determine the level of cognitive abilities of the children, she decides to use a test
for cognitive abilities which was standardized on the general population of ten-year-old children to a mean of μ=100 and a standard deviation of σ=20.
To test her hypothesis, she obtains a random sample of n=100 children aged 10 years, all of whom actively partook in soccer and netball. Each of the children
was given the cognitive test to complete and a mean of 𝑥̅=95 and a standard deviation of s=15 was found for the sample. The significance level to test the
hypothesis is set at α = 0.05.
Question 24
Choose the most appropriate way to formulate the operational hypothesis based on the scenario above from these options:
1. Children aged 10 who play sports have cognitive abilities which differ from the average cognitive abilities of children in this age group.
2. Children aged 10 who play sports have higher than average cognitive abilities.
3. Children aged 10 who play sports have lower than average cognitive abilities.
4. The cognitive abilities of children aged 10 who play sports will differ from the cognitive abilities of ten-year olds who do not play sports.
The researcher wants to determine if playing sports have an impact on the cognitive abilities of the children, but she did not indicate the direction (positive or
negative) of the relationship between the two variables. Option 2 and 3 state relationships that are directional (because it is clear which mean is likely to be
greater or lesser to the other one), so they do not fit with the scenario. Option 4 implies that ten-year-olds who do sports are to be compared with a second
group who do not play sports. There is however no sample of non-sports-playing children specified in the scenario, so this cannot be tested for directly. The
children who play sports are in fact compared with ten-year-old children in general
Question 25
What would be the alternative hypothesis that is to be tested?
Answer: Option 4 gives the most appropriate formulation based on the scenario.
The answer to Question 1 implies that the hypothesis is non-directional, and a two-tailed test should be performed. Therefore, the sample mean must be either
greater than or less than 100 if the alternative hypothesis is valid. Option 1 and 2 are one-tailed hypothesis, so they do not fit in with the question as put in the
scenario. Option 3 refers to the sample mean, which is not even known when the hypothesis is set up. The null and alternative hypotheses are set up before
sampling takes place, and is calculated based on the data in the sample. This sample mean (of 𝑥̅ = 95) must then be compared with the expected population
mean (of μ = 100) to see if it differs significantly, so to evaluate the hypothesis
Question 26
Mpho decides to use a z-test to test her hypothesis. The calculated value of the relevant test statistic is - - - -
1. 2.5
2. - 2.5
3. 2
4. 0.012
Note that since it is a one-sample mean that has to be compared with a population mean and the population standard deviation is known, the z-test was
chosen. The formula for this test is given on p. 80 of the PYC3704 Study Guide. 18
In order to calculate the test statistic for the z-test, the standard error first needs to be calculated. This proceeds as follows:
Option 1 is not quite correct as it gives the positive or absolute value. Note that the sample mean is smaller than the population mean, so the change in
cognitive ability is in a negative direction. Option 3 is not correct as it is the value of the standard error. Option 4 is the equivalent p-value as derived from the z-
score table.
Question 27
Based on the scenario and the test statistic calculated above, what conclusion can Mpho make based on the results?
1. The alternative hypothesis is supported, so playing sport does have a relationship to cognitive abilities.
2. The alternative hypothesis must be rejected, so playing sport does not have any relationship to cognitive abilities.
3. The null hypothesis can be rejected, because ten-year-old children playing sport have higher than average cognitive abilities.
4. The null hypothesis can be accepted, because children playing sports have higher than average cognitive abilities.
We determined in the previous question that the test statistic zx¯ = -2.5, and the relevant p-value can be found in the z-tables (Appendix D in the PYC3704
Guide). It is the ‘smaller portion’ The table in Appendix D gives p-values for one-sided testing, so it is necessary to adjust this for two-sided testing, so p-value
= 0.0062 x 2 = 0.012 (see p. 81 in the Guide).This p-value is less than the 0.05 level of significance (indicated in the scenario). Therefore, we reject the null
hypothesis and accept the alternative hypothesis, meaning playing sport has an impact on or is related to cognitive abilities of children.
Use the scenario below to answer Questions 28 and 29 It is known that the distribution of scores on a particular test is approximately normal and that the
population has a mean of μ=50 and a standard deviation of σ=15. A random sample of 100 persons is tested on this test, producing a sample mean of x¯ =
58.4 and a sample standard deviation of s = 16.8.
Question 28
After comparing the sample with the population using a z-test, based on the scenario above, a significant result was found. The researcher now wants to
determine whether the difference between the performance of the sample and the population in general is also reasonably important in terms of its practical
implication.
What does she need to determine the practical importance of the result, irrespective of the size of the sample?
4. The p-value
To find whether a statistical result has practical significance (irrespective of the sample size, where larger samples are more likely to produce significant
results), the effect size should be considered. (See the PYC3704 Guide, Section 3.3.3, pp. 86 – 88).
Question 29
Determine the practical importance of the test, based on the information in the scenario and the answer to the previous question. Given the available
information, it can be concluded that the effective difference between the performance of the sample and the population is - - - - -
To find whether a statistical result has practical significance (irrespective of the sample size, where larger samples are more likely to produce significant
results), the effect size should be considered. So, in this particular scenario, the researcher should calculate the effect size of the difference between the
sample mean and the population mean. Cohen’s d can be used, as follows:
An effect size of around 0.5 can be considered as more or less a medium effect (see the PYC3704 Guide, pp. 86 – 88. A rule of thumb for the interpretation of
effect sizes is given on p. 88).
Question 30
The power of a statistical test refers to the - - - - -
‘Power’ refers to the sensitivity of a statistical test, which is its ability to detect significant results for a particular sample size (see the PYC3704 Guide, p. 86).
1. determine whether or not we can accept that the null hypothesis is true
2. determine how far the observed measurements deviate from what we may expect by chance
3. get a measurement by which we can calculate the level of significance
4. determine whether or not we can reject the alternative hypothesis
Calculating the test statistic is the first step in a process of comparing the observed data with what may be expected by chance (i.e., if the null hypothesis were
true). Option 1 is not really appropriate because the emphasis is wrong. The test statistic is calculated to see how far the effect that was observed in the data
(our measurements) deviate from what we may expect by chance if the null hypothesis was true. Not rejecting the null hypothesis would imply the observation
is too close to the situation predicted by the null hypothesis for a researcher to feel confident in rejecting the null hypothesis. So even if we were hoping to find
no effect (for example, no difference between groups or no relationship among variables), the formal aim of calculating the test statistic is still to determine
whether it is safe to reject the null hypothesis or not. For the same reason, Option 4 is not really correct (accepting the alternative hypothesis would follow only
as a consequence of the null hypothesis being rejected). The level of significance (α) is not calculated but chosen by the researcher as the maximum risk of
rejecting the null hypothesis in error that he or she is willing to take, which is why Option 3 is false.
Question 2
The p-value represents the probability of getting the observed results if the - - - - -
The p-value shows you the probability of seeing some relationship among the variables based on your calculations (such as a difference between means or a
high correlation between variables), if in fact this observed relationship is merely the consequence of chance (in other words, if the null hypothesis was true).
You are in fact comparing the observed relationships in the data with what you would expect if the null hypothesis is true by calculating a relevant test statistic.
This test statistic can then be used to find the p-value if we know the probability distribution of the test statistic. If this probability is small, it implies the null
hypothesis is probably not true.
Note that Options 1 and 2 imply the opposite: i.e., that the p-value refers to the alternative hypothesis being true, which is equivalent to the null hypothesis
being false. Even though Option 3 implies that the null hypothesis is true, it is not correct, because the p-value is calculated in relation to the null hypothesis.
The alternative hypothesis covers an infinite range of possibilities and its test statistic cannot be calculated directly (see p. 78 of the Guide for PC3704).
Question 3
Samples can be considered independent when - - - - -
1. the sample comes from the assignment of subjects to a treatment or experimental group and this is varied to see how it affects certain measurements
2. care was taken that the samples are drawn under different experimental conditions
3. the samples are drawn from more than a single population of subjects
4. the composition of one sample is not systematically related to the composition of the other one
In the other options the variables could be independent, but they need not be. Option 1 is a reference to random allocation of subjects to treatment or
experimental groups, and while it is reasonable to expect such a process to lead to independence of the samples, this is not part of the definition of
'independence' between samples. Some systematic connection between the groups may still exist. Similarly, the fact that the samples are drawn under
different experimental conditions (Option 2) or that they come from more than a single population (Option 3) do not guarantee independence.
Question 4
From reading the psychometric test manual, Sally knows that the IQ test was standardized on a mean of 100 and a standard deviation of 15. Which of the
options below would be the most appropriate statistical test which she could use to test the hypotheses?
1. tc
2. zx¯
3. tx¯
4. td
Answer: Option 2 gives the most appropriate test statistic to use
The likely hypothesis being tested is that Sally has to compare a sample mean with a constant population mean which is already known to be 100. This implies
a single sample test for comparing a mean with a given value (100 in this case). It would therefore be either a single sample z-test or a single sample t-test. 4.
Because the population standard deviation is known (it is specified in the question above as σ = 15) the most appropriate test here would be the single sample
z-test, which is written zx¯. In such a case using the t-test is not 'wrong'; it is just that the z-test will be more powerful than a t-test – i.e. it is more sensitive to
statistical effects. Option 1 and Option 4 both refer to two-sample t-tests which are to be used to compare two sample means and not one sample mean with a
constant population mean.
Question 5
By setting a low level of significance, we reduce the probability of - - - - -
The level of significance is the maximum risk we are willing to take of rejecting the null hypothesis in error, which would be a type I error (see pages 84-85 in
the Guide). By making it small, we reduce the risk of making this error. This by implication makes the probability of accepting the alternative hypothesis
smaller, not the null hypothesis, as suggested in Option 1. Option 4 implies the same as Option 1, so it is also wrong. By setting a low level of significance we
unfortunately increase the probability of not rejecting a null hypothesis when it should be rejected because it is false. This would be a type II error, so Option 3
is also wrong.
Question 6
A human resources manager in a large information technology company notices that there seems to be a greater number of resignations among clerical staff
than among technical staff in the company. She suspects that this may indicate that there is greater dissatisfaction with their jobs in one group than in the
other. She decides to use a t-test to compare the job satisfaction of a sample of clerical workers with that of a sample of technical workers drawn from the staff
of the company. Which would be the independent variable?
1. type of job
2. number of resignations
3. workers
4. job satisfaction
Answer: Option 1 is correct. The hypothesis is that job satisfaction is influenced by the type of job that is performed (i.e., technical or clerical); so, type of job or
job category would be the independent variable affecting job satisfaction (the dependent variable).
Question 7
To test the efficacy of psychotherapy aimed at relieving depression, a researcher applies a depression scale to 50 depressed patients at the start and again at
the end of their treatment, predicting that the latter scores will be lower (reflecting less depression). Scores on his depression scale among the general
population have a mean of 30 and a standard deviation of 10. Which research design is appropriate to test the research hypothesis?
The same persons are tested repeatedly, and the first test (before the therapy) is matched with the second test (after the therapy) for each person, which
produces matched pairs of measurements. This is a typical example of dependent measurements (see pp. 117 – 118 in the Study Guide for PYC3704). Since
the groups are not independent, Option 1 is wrong. Note that even though it is one group being tested (a sample of 50 depressed patients), there are two
samples of measurements being compared, so it is treated as a two-sample design, and Options 3 and 4 are therefore wrong. (From a technical point of view,
the purpose of the statistical test here is to determine whether two samples of measurements come from the same underlying population of measurements,
which is why two samples of measurements can come from repeated measurements on the same group of persons).
Question 8
What does it mean to say, "the difference between the means of groups A and B is statistically significant"?
The null hypothesis states that there is no difference in the means calculated from samples of data from each of groups A and B. When we calculate the two
means from sample data (which we regard as making an observation) we may find a difference in the two calculated means, but at least part of this difference
could be due to measurement errors. We calculate the p-value (based on a test statistic with a known probability distribution) to find out what the probability is
that that these observed differences in the sample data are just a consequence of measurement error if the null hypothesis is assumed to be true. If this
probability is low (lower than a pre-determined cut-off level, α), we conclude that the difference in the two means is statistically significant because the
probability that the null hypothesis is true is very small. In other words, we conclude that the size of the difference between means found in the sample data
would not be likely if the null hypothesis were true.
Option 1 implies the result is not significant, and so does Option 2 (if we take ‘rejecting’ the alternative hypothesis as implying accepting the null hypothesis).
We test the null hypothesis, not the alternative hypothesis so Option 4 is also wrong.
Question 9
The probability under the null hypothesis of obtaining a t-value of 2.5 or higher in the case of a one-tailed test is - - - - - that for a two-tailed test.
1. the same as
2. twice
3. half
4. unrelated to
Note that the question actually relates not to the t-statistic value as such, but to the probability of getting that particular value for the t-statistic; in other words, it
refers to the p-value. A one-tailed p-value (used in the case of a directional hypothesis) is half the size of a two-tailed probability. Conversely, a two-tailed p-
value (used in the case of a non-directional hypothesis) is twice the size of a one-tailed p-value. (See page 81 of the PYC3704 Study Guide).
Question 10
A positive correlation between variables X and Y implies that persons scoring - - - - - on X will generally score - - - - - on Y.
A positive correlation implies that two variable co-vary in the same direction: as one variable changes, the other changes in the same direction. A higher value
on X will imply a higher value on Y, while a lower value on X will be matched by a lower value on Y most of the time.
Question 11
Suppose you calculate the correlation between two variables, X and Y, and you find a correlation of r = 0.87. Can you infer from this that one variable is the
cause of the other one, and why?
1. Yes, because a positive correlation indicates that the one variable has a direct impact on the other one
2. No, because you first have to establish which variable is dependent on which
3. Yes, because such the correlation is high enough to imply that a causal relationship is probable
4. No, because correlation need not imply a causal relationship and may depend on hidden variables affecting both X and Y
A positive correlation just tells you that as one variable becomes greater, so does the other, but you cannot see the reasons for this. There may be one or more
hidden variables that affect both of the variables that you are observing in indirect ways. Note that while the issue of dependence (which variable depends on
which) is relevant when we suspect causality, causality is not necessarily implied by it, so the answer in Option 2 is not sufficient. (See Section 6.1.4.2 on p.
140 of the Guide).
Base your answers to Questions 12 to 14 on the following scenario: An educational psychologist believes that the performance of pupils in mathematics
can be improved by teaching them to play chess. She tests this on two samples of children. She gives chess lessons to a treatment group of 20 pupils while a
control group of 24 pupils are given their normal classes. The psychologist intends to test her hypothesis at a significance level of α = 0.05.
Question 12
Which is the most appropriate research hypothesis for the teacher to test?
1. The mean mathematics score for the treatment group will be greater than that of the control group
2. The mean mathematics difference score (difference scores of treatment group minus control group) will differ significantly from zero
3. The mean mathematics score for the control group will be greater than that of the treatment group
4. There will be a significant difference in the mean scores for mathematics between the treatment and control groups
Given the fact that the psychologist expects to find an improvement in mathematics score, one would expect a greater score for the treatment group (which
were exposed to the chess lessons) than for the control group (which was not). Option 3 is the wrong way around and Option 4 indicates two-sided
hypotheses, where the expected direction of the change (an improvement in mathematics scores) is not taken into account. Difference scores can only be
calculated for matched pairs of data, so Option 2 is not appropriate
Question 13
Which is the appropriate test statistic to be calculated when testing this hypothesis (as specified in Question 12)?
1. The t-statistic for the difference between the means of two independent samples
2. The t-statistic for the difference between the means of two dependent samples
3. The t-statistic for the mean difference score of a single sample
4. The r-statistic for the correlation between two variables
There are two samples (treatment and control group) which can be regarded as independent (there is no particular systematic relationship between them).
Since two independent samples of different sizes are being compared, rather than two variables from a single sample, the correlation coefficient is not really
relevant, so Option 4 is also incorrect.
Question 14
When calculating the test statistic referred to above, the psychologist uses a computer program to calculate the p-value. The computer presents her with a two-
sided p-value of p = 0.054. What can she conclude?
1. The outcome is significant, so she can conclude that the alternative hypothesis can be rejected and that learning to play chess does have a positive effect
on the mathematics ability of the pupils.
2. The outcome is significant, so she can conclude that the null hypothesis can be rejected and that learning to play chess does have a positive effect on the
mathematics ability of the pupils.
3. The outcome is not significant, so she can conclude that the null hypothesis cannot be rejected, and that chess has no effect on the mathematics ability of
the pupils.
4. The outcome is not significant, so she can conclude that the null hypothesis cannot be rejected, and that chess has a negative effect on the mathematics
ability of the pupils.
Note that we need to do a directional test (as shown in the answer to question 12 above). The two-sided p-value given by the computer must therefore be
divided by 2 before it can be compared to the level of significance. This gives us a one-tailed p-value of 0.054/2 = 0.027. This is smaller than the level of
significance which was given in the scenario above as α = 0.05, and the result is therefore significant. A significant result implies that the null hypothesis should
be rejected, and this in turn implies that we can accept the alternative hypothesis; that is, we can conclude that learning to play chess does have a positive
effect on the mathematics ability of the pupils.
Question 15
A researcher hypothesizes that a relationship should exist between spatial ability and aptitude for art among children. She collects the results of a sample of n
= 100 school children for a test which measures artistic ability and also measures the spatial ability of each with a test that represents a person’s ability to
rotate objects mentally, scored on a 10-point scale.
Which of the following is the most appropriate way to express the null hypothesis for this research?
1. r = 0
2. μ = 0
3. x¯ = 0
4. ρ = 0
Two continuous quantitative measurements are to be compared for a single sample of data, so it is possible to calculate a Pearson correlation coefficient. The
appropriate statistical test will need to establish whether this correlation is significant in a positive direction, while the null hypothesis would state that no such
relationship exists. The symbol ρ (the Greek letter ‘rho’) would be used in the formal hypothesis statement to indicate that it is the population parameter which
is being tested for, based on a sample correlation, which would be indicated by r. (See the PYC3704 Guide, p. 137).
Question 16
A human resources manager in a large information technology company notices that there seems to be a greater number of resignations among clerical staff
than among technical staff in the company. She suspects that this may indicate that there is greater dissatisfaction with their jobs in one group than in the
other. She decides to use a t-test to compare the job satisfaction of a sample of clerical workers with that of a sample of technical workers drawn from the staff
of the company. Which would be the independent variable?
1. type of job
2. number of resignations
3. workers
4. job satisfaction
Answer: Here we accidentally repeated Question 6 above (did you notice?). The answer is still Option 1, as explained above. We will omit this question when
the assignment is marked.
Question 17
Whether the p-value for directional or non-directional testing on a hypothesis should be used depends on - - - -
The alternative hypothesis states that certain parameters differ, and when appropriate, gives an indication of the direction is which they differ. The null
hypothesis states that there is no difference, so there can obviously not be a direction (greater or smaller) in this statement, which implies that neither Options
1 nor 3 is correct. Option 2 is incorrect because the level of significance is chosen by the researcher and is not dependent on whether it is a directional or non-
directional test.
Question 18
A researcher wants to investigate the aptitudes and personality factors that may be involved when students choose their direction of study. She draws a
sample of 300 final year students and classifies them according to the college or faculty where they are enrolled. She creates a variable ‘direction of study’ and
codes this as science=1; economics = 2; social sciences = 3; and arts = 4. Which level of measurement would this variable represent?
1. nominal
2. ordinal
3. interval
4. ratio
The numbers used to classify people into groups according to 'direction of study' are just used to indicate the different categories into which people are
classified, but their numerical value ('sizes') are irrelevant. Since these values do not represent specific measurements, the measurement level is no stronger
than nominal (see Appendix B in the PYC3704 Guide).
Question 19
A scatter plot is a graphical representation of - - - - -
1. the relationship between two variables measured on a nominal scale within a single group.
2. the frequency distribution of a sample of measurements.
3. relationship between two groups of subjects with regard to a single variable measured on an interval or ratio scale.
4. the relationship between two variables measured on a ratio or interval scale within a single group.
Scatter plots are graphs which show the distribution of data from a single sample of data on two axes, each of which represents a measurement related to a
particular individual case (See PYC3704 Guide pp. 130 - 132). This is also why Option 3 (referring to two groups of subjects) is not valid.
Scatter plots are not used to indicate frequencies or counts, so Option 2 is not correct either. It is not possible to draw such a graph if any of the measurements
represents a measure on a nominal scale, since the actual categories (the numbers which are allocated) in such a measurement would be arbitrary, therefore
Option 1 cannot be correct (see Appendix B in the Guide). For a graph showing a scatter plot, see Question 20 below.
Question 20
The graph below shows the measurements of a sample of 20 research participants on two variables, job satisfaction (X) and depression (Y).
If a Pearson product-moment correlation (r) between the two variables is calculated, which of the following values is most reasonable to expect?
1. 0.3
2. 0
3. -0.6
4. 1.0
It is possible to imagine a line running roughly through the data points from upper left to the lower right. This implies that there is some kind of correlation (so r
= 0, as suggested in Option 2 is not likely), but also that the correlation is negative (as one variable (x) gets greater, the other one (y) gets smaller). If you use
the graph to work out the approximate values of the x to y data points (based on the coordinates), and did a calculation, you should get a result of close to r = –
0.5, but this calculation is not really necessary to answer the question. You should be able to tell from the graph that Option 3 is the only one of the options
given that is even plausible.
Question 21
A researcher employed by the traffic department takes blood samples from 100 subjects who have been drinking variable amounts of alcohol. She then asks
each of the participants to try and stand on one foot for as long as they possibly can. The blood alcohol level of the participants and length of time they can
balance on one foot is then correlated. Which of the following results would you hypothesise as being the most likely outcome?
1. A positive correlation
2. A negative correlation
3. No correlation
4. A Pearson product moment correlation
Answer: Option 2 is the most likely outcome.
One would expect to find that alcohol affects the motor functions, so that the higher the level of alcohol in the blood, the less time a person would be able to
balance on one foot. A negative correlation is implied because as one variable becomes larger, the other diminishes (see the answer to Question 20 above).
Option 4 refers to the name of a commonly used coefficient used to calculate the magnitude of a correlation, not to a kind of result which one is likely to expect.
It is not a result, but a way of calculation the result.
Question 22
The chi-square (X2) test statistic is used to compare - - - - -
1. the distribution of observed frequency data with the way that the data should be distributed if the null hypothesis is true
2. the variance or spread of observed data with the variance of the data as expected if the null hypothesis is true
3. the way in which each of two variables vary (their variances) with the covariance (how they vary together) of both
4. the extent to which the mean of a variable differs from the mean of another when both are nominal scale measurements
The chi-square test is usually used when you have a cross tabulation of frequency counts of events which are nominal scale measurements. This table is
referred to as a contingency table. It is used to compare an observed frequency distribution (frequency counts based on a sample of observation) with the
frequency distribution which we would expect to find if the null hypothesis of no relationship between two cross-tabulated variables was true. (See p. 140 of the
PYC3704 Guide for details).
Base your answers to Questions 23 to 28 on the following scenario: Alfred is a psychologist who is involved in a series of workshops which provide
exercises designed to improve the self-esteem of the participants. Alfred wants to establish the effectiveness of the training programme, so he tests a group of
15 participants on a questionnaire which measures their self-esteem before the training commences and again after the training programme is completed. He
also tests them on an anxiety scale before the workshop. The table below shows the results of these measurements.
Key:
Case is a unique number for each participant
Gender is coded with 1 = Male and 2 = Female.
Anxiety was measured on a 5-point scale (ranging from 1=’little or no anxiety’ to 5=’high anxiety’).
Self-esteem was measured on a 7-point scale (with 1=’very low self-esteem’ up to 7=’very high self-esteem’). SelfEst1 indicates the measurement before the
training programme, while SelfEst2 represents the measurement thereafter.
Question 23
Alfred wants to determine whether there is a gender difference in self-esteem after the workshop. Which hypotheses?
The two groups indicated in the variable Gender are to be compared (Male group coded as 1 and Female group coded as 2) on the dependent variable
SelfEst2 (which refers to self-esteem after the workshop), and no prediction is implied as to which group is likely to have higher self-esteem. Therefore, the
one-sided testing specified in Option 4 is false. Options 1 and 3 are invalid 12 because formal hypotheses should refer to the population parameters, not the
sample statistics (you use the sample data to make inferences about the state of affairs in the population).
Question 24
Alfred wants to determine whether the workshop is effective by comparing the measurements on the self-esteem test before the workshop with the
measurements on this same test after the workshop. Which of the following is an appropriate test statistic to calculate?
This is a situation where repeated measurements are used: the same sample of individual participants is measured twice. Each measurement can be matched
with the subsequent measurement for the same person. Even though it is the same sample of people being measured twice, you can think of it as two groups
of measurements in a matched-pairs design with dependent groups (see Study Unit 5.2 from p. 117 in the PYC3704 Guide). Therefore, a test for dependent
groups is indicated and comparing the two group means for the matched groups with a td test is an appropriate way to do it. Testing for independent groups
(Option 2) is not a good choice for matched pairs of measurements. The correlation coefficient (Option 1) is not very useful here, even though it would be
possible to calculate it. We want to determine whether the groups differ, not how well the sets of measurements agree. (It is actually technically possible to
convert a correlation coefficient into a measurement for differences between groups, but that would be a roundabout way of doing it, which would require
further calculation). The chi-square (Option 4) test is used mainly for nominal scale data and it is used to compare frequency distributions, so it is not the best
option in this particular scenario.
Question 25
Alfred calculates the value of the appropriate test statistic (as selected in Question 24 above). In which of the four intervals below will the absolute value (i.e.,
ignoring a plus or minus sign) of the test statistic fall?
1. Between 0 to 1.5
2. Between 1.5 and 3.0
3. Between 3.0 and 4.0
4. Above 4.0
In Question 24 above we decided that the appropriate test statistic is the td-test statistic and we will have to calculate this value in order to answer the
question. To calculate the td-test statistic, the difference scores (d) must first be found by subtracting, for each matched pair, the self-esteem score before the
programme (SelfEst1) from the self-esteem score after it (SelfEst2). These scores are indicated in the table below.
The mean and standard deviations of the difference scores can now be computed by using the values of the difference scores (d) in the appropriate formulas:
Keep in mind, however, that Alfred wants to know whether self-esteem increases, so he should be performing one-sided testing only. If the observed mean of
SelfEst2 in the sample was smaller than the observed mean of SelfEst1, the probability of SelfEst2 being greater than SelfEst1 would be zero and we would
not need to continue with statistical testing. This would imply that self-esteem got worse after the workshop, which is not what Alfred hopes to determine. (See
p. 77 and also p. 119 in the Guide). The mean and standard deviation of the difference scores must now be substituted in the formula for the td-test statistic:
This falls in the category specified in Option 2, as it is between 1.5 and 3.0.
Note how we subtracted the self-esteem score before the workshop (SelfEst1) from the self-esteem score after the workshop (SelfEst2) to calculate the
differences scores d (so d = SelfEst2 – SelfEst1 for each case). This makes sense, because we would expect a higher score after the workshop than before if
the workshop was effective (i.e. directional testing), given the hypothesis as implied in Question 24. This would imply that the mean of the d-scores (d¯) and
also the calculated t-value should be positive. It would however not matter if we did the calculation of d the other way round. In such a case we would expect
the mean of the d-scores (d¯) to be negative (to test whether the mean of SelfEst1 is smaller than mean of SelfEst2), and in this case we would expect a
negative value for the calculated t-value. In such a case you would have to use the absolute value of td (ignoring the minus sign) to answer the question. (see
'absolute value' on p. 171 of the PYC3704 Guide).
Question 26
Alfred is also interested to know whether the difference between the two self-esteem scores before and after the workshop is of practical significance; that is to
say, irrespective of its statistical significance, whether the effect can be regarded as big or small. Calculate the relevant measurement of the size of the effect
to answer the question below. The size of the effect was found to be - - - - -
1. small
2. medium
3. large
4. impossible to determine from this data
To test the practical significance of the result when two means are compared, irrespective of sample size, we have to calculate an effect size measurement
such as Cohen's d. The appropriate formula in the case where difference scores are being considered is given on p. 120 in the PYC3704 Guide, and can be
applied as follows (using the calculations of the mean and standard deviations of the differences scores 'd' from the previous question):
Note that D¯ refers to the expected population mean if the null hypothesis is true. In such a case, one would expect no difference between the two group
means (self-esteem before and after the workshop) so the mean difference should be zero. This result can be interpreted as a fairly large effect. (See p. 88 in
the PYC3704 Guide for a guideline to interpreting the effect size).
Question 27
Alfred is also interested to determine whether a relationship exists between anxiety and self-esteem after the workshop is completed. Which of the following is
the most appropriate test statistic to use?
Two measurements were performed on the same group but on different variables, and Alfred wants to know whether one variable (self-esteem) is affected by
the other (anxiety) to a significant degree. This implies calculating a correlation coefficient and testing it for its statistical significance.
The t-test (Options 3 and 4) implies two groups are to be compared, not two variables, so they would not be effective here. Note that the two variables are not
even measured on the same kind of scale, so comparing means would not make sense, unless you transformed them to a common scale first (for example, a
z-scale). The chi-square test is used to compare distributions, usually for nominal scale data, so Option 2 is not appropriate.
Question 28
Which of the following is the result if this test statistic (in Question 27) is calculated? The value of the test statistic will fall - - - - -
1. below – 0.5
2. between – 0.5 and 0
3. between 0 and 0.5
4. above 0.5
In the answer to the previous question (Question 27), we indicated that the statistical test for Pearson’s correlation coefficient (r) will have to be used. This test
is based on the probability distribution of Pearson's r, so we will have to calculate this correlation coefficient before we are able to answer this particular
question.
The best way to do this would be to first create a table such as the following (similar to the table on p. 135 of the Guide):
Note that in the fourth column Anxiety2 refers to Anxiety squared (multiplied with itself) while in the fifth column SelfEst22 refers to the square of Self-esteem
after the workshop. The last column is Anxiety multiplied with SelfEst2. The formula on p. 134 of the Guide must now be used, with values from this table
substituted to calculate r, as follows:
This is between –0.5 and 0. Note that the negative sign implies a negative correlation; that is, as Anxiety gets bigger, SelfEst2 (after workshop self-esteem)
tend to become smaller. Since –0.4755 is smaller than zero but greater than – 0.5, it follows that the Pearson correlation coefficient r falls in the interval
indicated in Option 2.
category which a worker falls into. The frequency distributions for each gender category are shown in the graph below (with the job categories on the horizontal
axis).
Question 29
Which of the options listed below would be the most appropriate to test whether a relationship exists between gender and category of staff?
While managers could be thought of as ‘more senior’ and manual workers as ‘least senior’ in an organization, there is no obvious differences in status among
administrative, technical and HR staff members which can be taken for granted. The order of the arrangement of the categories is therefore arbitrary, which
implies that the variables on the horizontal axis are nominal scale variables (see Appendix B in the PYC3704 Study Guide). In the case of the chi-square test
this does not matter. A variation of the t-test would be appropriate if the means on some measurement scale (stronger than nominal scale) were to be
compared, which is not the case here, so Options 3 and 4 are not valid. The nominal data on the job category variable makes it impossible to calculate a
Pearson product-moment correlation coefficient, so Option 2 is also not valid.
Question 30
Calculate the statistic chosen in the previous question and indicate the closest result from the list of options below.
1. – 0.02
2. 3.56
3. 9.55
4. 14.0
The chi-square statistic (from the previous Question) is to be calculated. The formula for this test is given on p. 141 of the Guide:
To use this formula, it is necessary to work out the expected frequencies for each cell in the contingency table. This is done by multiplying the row total with the
column total and dividing by the overall total for each cell (see p. 143 - 4 in the PYC3704 Study Guide). This would lead to the following contingency table
(using the observed frequencies from the scenario above with expected frequencies added in brackets):
It is useful to create a table like the one on p. 145 of the PYC3704 Guide if you have to do this calculation by hand. In the present case, the table will look like
this:
The value of the chi-square statistic is the total of the numbers in the right-hand column:
Note that Option 1 can be disqualified even without doing the calculation, because a chi-square statistic will never have a negative result (a sum of squared
values can never add up to less than zero). If the chi-square statistic is zero or very close to zero, it follows that the observed frequencies and the expected
frequencies did not differ, so the null hypothesis is probably true.