LESSON 17: T-TEST FOR TWO-INDEPENDENT SAMPLES
Research designs that are used to obtain the two sets of data
can be classified in two general categories:
 two sets of data could come from two completely
separate groups of participants
o ex. study that involves a sample of men
compared with a sample of women
 two sets of data could come from the same group of
participants
o ex. one set of scores measuring depression
for a sample of patients before they begin
therapy and then obtain a second set of data
by measuring the same individuals after 6
weeks of therapy
INDEPENDENT-MEASURES RESEARCH DESIGN
(BETWEEN-SUBJECTS DESIGN)
 research design that uses a separate group of
participants for each treatment condition (or for each
population)
Repeated-Measures Research Design (Within-Subjects
Design)
 two sets of data are obtained from the same group of
participants
I. Hypotheses for an Independent-Measures Test
 goal of an independent-measures research study: to
evaluate the mean difference between two
populations (or between two treatment conditions)
 null hypothesis: there is no change, no effect, or, in
this case, no difference.
o H0: μ1 − μ2 = 0 (No difference between the
population means)
 could also be stated as μ1 = μ2
 alternative hypothesis: there is a mean difference
between the two populations.
o H1: μ1 − μ2 ≠ 0 (There is a mean difference.)
 two population means are not
equal: μ1 ≠ μ2
II. Formulas for an Independent-Measures Hypothesis Test
Two points to remember:
 The basic structure of the t statistic is the same for
both the independent-measures and the single-
sample hypothesis tests
 The independent-measures t is basically a two-
sample t that doubles all the elements of the single-
sample t formulas.
 value for μ1 − μ2 comes from the null hypothesis: 0
Estimated Standard Error
 represented by the symbol S(M -M )
1 2
Interpreting the Estimated Standard Error
Can be interpreted in 2 ways:
 a measure of the standard or average distance
between a sample statistic (M1 − M2) and the
corresponding population parameter (μ1 − μ2).
 measures the standard, or average size of (M 1 − M2)
if the null hypothesis is true. That is, it measures how
much difference is reasonable to expect between the
two-sample means.
III. Calculating the Estimated Standard Error
Three points:
 Each of the two-sample means represents its own
population mean, but in each case there is some error
o M1 approximates μ1 with some error
o M2 approximates μ2 with some error
 Thus, there are two sources of
error.
 estimated standard error for each sample mean is
computed as follows:
 resulting formula for standard error is:
o limited to situations in which the two samples
are the same size (that is, n1 = n2)
IV. Pooled Variance
 variance obtained from a large sample is a more
accurate estimate of σ2 than the variance obtained
from a small sample.
 Pooled variance – one method for correcting the bias
in the standard error is to combine the two sample
variances into a single value.
o Obtained by averaging or “pooling” the two
sample variances using a procedure that
allows the bigger sample to carry more
weight in determining the final value.
V. Estimated Standard Error
 Resulting formula for ESE
 measures how accurately the difference between two
sample means represents the difference between the
two population means.
 how much difference is expected on average between
the two sample means.
VI.Final Formula
 complete formula for the independent-measures t
statistic:
 df
HYPOTHESIS TESTS WITH HE INDEPENDENT-MEASURES
T STATISTIC
I. Assumptions Underlying the Independent-Measures t
Formula
 The observations within each sample must be
independent
 The two populations from which the samples are
selected must be normal.
 The two populations from which the samples are
selected must have equal variances.
II. Hartley’s F-Max Test
 Used to check homogeneity of variance
EFFECT SIZE AND CONFIDENCE INTERVALS FOR THE
INDEPENDENT-MEASURES T
 Cohen’s d – produces a standardized measure of
mean difference
 formula for estimating Cohen’s d becomes:
 r2 measures how much of the variability in the scores
can be explained by the treatment effects.
LESSON 18: T TEST FOR TWO RELATED SAMPLES PT1
INTRODUCTION TO REPEATED-MEASURES DESIGN
 Repeated-measures designs (within-subject design) –
one in which the dependent variable is measured two
or more times for each individual in a single sample.
 The same group of subjects is used in all the
treatment conditions.
I. Matched-Subjects Design
 each individual in one sample is matched with an
individual in the other sample.
 matching is done so that the two individuals are
equivalent (or nearly equivalent) with respect to a
specific variable that the researcher would like to
control.
THE T STATISTIC FOR A REPEATED-MEASURES
RESEARCH DESIGN
I. Difference Scores: Data for a Repeated-Measures Study
 difference scores are obtained by subtracting the first
score (before treatment) from the second score (after
treatment) for each person:

II. Hypotheses for a Related-Samples Test I. Assumptions of the Related-Samples t Test

 repeated-measures study: null hypothesis: the mean  The observations within each treatment condition
difference for the general population is zero must be independent. Notice that the assumption of
independence refers to the scores within each
treatment. Inside each treatment, the scores are
obtained from different individuals and should be
 alternative hypothesis: there is a treatment effect that independent of one another.
causes the scores in one treatment condition to be  The population distribution of difference scores (D
systematically higher (or lower) than the scores in the values) must be normal.
other condition.
EFFECT SIZE AND CONFIDENCE INTERVALS FOR THE
REEATED-MEASURES T

III. t Statistic for Related Samples  Estimated d

 T stats

 Percentage of variance
 First, Variance or SD

 Then, ESE  Confidence intervals for estimating 𝛍D

COMPARING REPEATED- AND INDEPENDENT-MEASURES

 Df in t stats: n-1 DESIGNS
LESSON 19: T TEST FOR TWO-RELATED SAMPLES PT 2  Number of subjects:
HYPOTHESIS TEST FOR THE REPEATED-MEASURES o repeated-measures design typically requires
DESIGN fewer subjects
o repeated-measures design uses the subjects
 Ex. more efficiently because each individual is
measured in both of the treatment conditions
 study changes over time
o RMD is well suited for studying learning, or
other changes that take place over time
  individual differences
o primary advantage of a repeated-measures
design is that it reduces or eliminates
problems caused by individual differences
o individual differences: age, IQ, gender,
personality
I. Time-Related Factors and Order Effects
 primary disadvantage of a repeated-measures
design: structure of the design allows for factors other
than the treatment effect to cause a participant’s
score to change from one treatment to the next.
o outside factors that change over time may
be responsible for changes in the
participants’ scores
 participation in the first treatment influences the
individual’s score in the second treatment.
o Order effects – changes in scored that are
caused by participation in n earlier
treatment; can distort the mean differences
found in repeated-measures research
studies.
 Counterbalancing – one way to deal with time-related
factors and order effects
o Participants are randomly divided into two
groups with one group receiving treatment 1
followed by treatment 2; other group
receiving treatment 2 followed by treatment
1.
o Goal: to distribute any outside effects evenly
over the two treatments
LESSON 20: ANALYSIS OF VARIANCE (ANOVA) PT 1
 hypothesis-testing procedure that is used to evaluate
mean differences between two or more treatments (or
populations).
 Major advantage: can be used to compare two or
more treatments.
o provides researchers with much greater
flexibility in designing experiments and
interpreting results.
 Goal: to determine whether the mean differences
observed among the samples provide enough
evidence to conclude that there are mean differences
among the three populations.
 Two interpretations:
o There really are no differences between the
populations (or treatments). The observed
differences between the sample means are
caused by random, unsystematic factors
(sampling error) that differentiate one
sample from another.
o The populations (or treatments) really do
have different means, and these population
mean differences are responsible for
causing systematic differences between the
sample means.
I. Terminology in ANOVA
 Quasi-independent variable – non-manipulated
variable
 Factor – the variable that designated the groups
being compared
 Levels – individual groups or treatment conditions that
are used to make up a factor.
 ANOVA can be used with either an independent-
measures or a repeated-measures design.
 ANOVA can be used to evaluate the results from a
research study that involves more than one factor
 two-factor design or a factorial design - study that
combines two factors
o provides researchers with the flexibility to
develop studies that address scientific
questions that could not be answered by a
single design using a single factor
II. Statistical Hypotheses for ANOVA
 Null hypothesis: no effect; the population means are
all the same; there is no treatment effect
 Alternative hypothesis: population means are not all
the same; there is a real treatment effect
o One alternative: would be that the first two
populations are identical, but that the third is
different.
o Another: the last two means are the same,
but the first is different.
III. Type I Errors and Multiple-Hypothesis Tests
 Often a single experiment requires several hypothesis
tests to evaluate all the mean differences. However,
each test has a risk of a Type I error, and the more
tests you do, the greater the risk.
 Testwise alpha level – risk of a Type I error, or alpha
level, for an individual hypothesis test.
 Experimentwise alpha level – involves several
different hypothesis tests; total probability of a Type I
error that is accumulated from all of the individual
tests in the experiment; greater than the value of
alpha used for any one of the individual tests.
IV. Test Statistic for ANOVA
 we first computed the standard error, which measures
the difference between two sample means that is
reasonable to expect if there is no treatment effect
(that is, if H0 is true). Then we computed the t statistic
with the following structure:
  we can use variance to measure sample mean
differences when there are two or more samples. The
test statistic for ANOVA uses this fact to compute an
F-ratio with the following structure:
LOGIC OF ANOVA
 goal: measure the amount of variability and to explain
why the scores are different
 first step: determine total variability for the entire set
of data.
o Combine all scores
o Break it apart into separate components
 Analysis process divides the total variability into two
basic components:
o Between-Treatments Variance – We will
calculate the variance between treatments to
provide a measure of the overall differences
between treatment conditions; the variance
between treatments is really measuring the
differences between sample means.
o Within-Treatment Variance – variability within
each sample; provides a measure of the
variability inside each treatment condition.
I. Between-Treatments Variance
 Measures how much difference exists between the
treatment conditions.
 Two possible explanations:
o The differences between treatments are not
caused by any treatment effect but are
simply the naturally occurring, random and
unsystematic differences that exist between
one sample and another. That is, the
differences are the result of sampling error.
o The differences between treatments have
been caused by the treatment effects.
II. Within-Treatments Variance
 provides a measure of how big the differences are
when H0 is true.
III. F-Ratio: Test Statistic for ANOVA
 Comparison (between and within treatments) is made
by computing an F-ratio.
 For independent measures ANOVA:
 value obtained for the F-ratio helps determine
whether any treatment effects exist. Consider the
following two possibilities:
o When there are no systematic treatment
effects, the differences between treatments
(numerator) are entirely caused by random,
unsystematic factors; numerator and
denominator are both measuring ransom
differences and should be roughly the same
size.
 Roughly equal numerator and
denominator: it should have a value
around 1.00
 F-ratio near 1.00: the differences
between treatments (numerator)
are random and systematic, same
sa denominator; there is no
evidence to suggest that the
treatment has any effect.
o When the treatment does have an effect,
causing systematic differences between
samples, then the combination of systematic
and random differences in the numerator
should be larger than the random differences
alone in the denominator.
 Error term – the denominator of the F-ratio measures
only random and unsystematic variability
ANOVA NOTATION AND FORMULAS
 k – used to identify the number of treatment
conditions – number of levels of the factor.
o For IMS, it is the number of separate
samples.
 n – number of scores in each treatment
 N – total number of scores in the entire study
 T – sum of the scores ( ΣX) for each treatment
condition
 G – sum of all the scores in the research study (grand
total) G = ΣT
I. ANOVA Formulas
 Final calculation for ANOVA is the F-ratio (composed
of 2 variances)
 Each of the two variances in the F-ratio is calculated
using the basic formula for sample variance.
  the entire process of ANOVA requires nine
calculations: three values for SS, three values for df,
two variances (between and within), and a final F-
ratio
II. Analysis of Sum of Squares (SS)
 Total Sum of Squares (SStotal) – the sum of squares
for the entire set of N scores
o Computational formula:
 Within-Treatments Sum of Squares (SSwithin) – the
variability inside each of the treatment conditions
 Between-Treatments Sum of Squares (SSbetween)
III. Analysis of DF
 two important considerations to keep in mind:
o Each df value is associated with a specific
SS value.
o Normally, the value of df is obtained by
counting the number of items that were used
to calculate SS and then subtracting 1.
 Total Degrees of Freedom (dftotal)
 Within-Treatments Degrees of Freedom (dfwithin)
 Between-Treatments Degrees of Freedom (dfbetween)
 Total – entire set of scores
 Within – differences that exist inside the individual
treatment conditions.
 Between – differences from one treatment to another.
IV. Calculation of Variances (MS) and the F-Ratio
 MSbetween
 MSwithin
 F-ratio
LESSON 21: ANOVA PT 2
I. Distribution of F-Ratios
 F-ratio is constructed so that the numerator and
denominator of the ratio are measuring exactly the
same variance when the null hypothesis is true.
o If the null hypothesis is false, the F-ratio
should be much greater than 1.00
 two obvious characteristics:
o Because F-ratios are computed from two
variances (the numerator and denominator
of the ratio), F values always are positive
numbers. Remember that variance is always
positive.
o When H0 is true, the numerator and
denominator of the F-ratio are measuring the
same variance. In this case, the two sample
variances should be about the same size, so
the ratio should be near 1. In other words,
the distribution of F-ratios should pile up
around 1.00.
II. F Distribution Table
 For ANOVA, we expect F near 1.00 if H0 is true.
 F-ratio that is much larger than 1.00 is an indication
that H0 is not true.
POST HOC TEST
 additional hypothesis tests that are done after an
ANOVA to determine exactly which mean differences
are significant and which are not.
 These tests are done after ANOVA when:
o You reject H0 and
o there are three or more treatments (k ≥ 3)
 Rejecting H0 indicates that at least one difference
exists among the treatments.
I. Posttests and Type I Errors
 Pairwise comparisons – compare the individual
treatments two at a time.
o Ex. with k = 3, we would compare μ 1 vs. μ2,
then μ2 vs. μ3, and then μ1 vs. μ3.
 Experimentwise alpha level – as you do more and
more separate tests, the risk of a Type I error
accumulates
II. Tukey’s Honestly Significant Difference (HSD) Test
  allows you to compute a single value that determines
the minimum difference between treatment means
that is necessary for significance.
 HSD – used to compare any two treatment
conditions.
 If the mean difference exceeds Tukey’s HSD, you
conclude that there is a significant difference between
the treatments.
 Otherwise, you cannot conclude that the treatments
are significantly different.
 Formula:
 Yung q, galing sa F Dist table
III. Scheffè Test
 uses an F-ratio to evaluate the significance of the
difference between any two treatment conditions.
 “safety factor” for the Scheffé test comes from the
following two considerations:
o Although you are comparing only two
treatments, the Scheffé test uses the value
of k from the original experiment to compute
df between treatments. Thus, df for the
numerator of the F-ratio is k – 1.
o The critical value for the Scheffé F-ratio is
the same as was used to evaluate the F-
ratio from the overall ANOVA. Thus, Scheffé
requires that every posttest satisfy the same
criterion that was used for the complete
ANOVA.
IV. Relationship Between ANOVA and t Tests
 relationship between pooled variance and MS within is
only part of the general relationship between the
independent-measures t test and the corresponding
ANOVA.
 basic relationship between t statistics and F-ratios
can be stated in an equation:
o t statistic compares distances: the distance
between two sample means (numerator)
and the distance computed for the standard
error (denominator).
o F-ratio compares variances. You should
recall that variance is a measure of squared
distance. Hence, the relationship: F = t2.
 Several other points to consider in comparing the t
statistic to the F-ratio:
o you will be testing the same hypotheses
whether you choose a t test or an ANOVA.
Hypotheses for either test are:
o The degrees of freedom for the t statistic and
the df for the denominator of the F-ratio
(dfwithin) are identical.
o The distribution of t and the distribution of F-
ratios match perfectly if you take into
consideration the relationship F = t2.
LESSON 22: OVERVIEW OF CHI-SQUARE STATISTIC
I. Parametric and Nonparametric Statistical Tests
 Parametric tests – tests all concern parameters and
require assumptions about parameters
o they require a numerical score for each
individual in the sample.
o require data from an interval or a ratio scale
o require numerical scores
 nonparametric tests – alternatives
o use sample data to evaluate hypotheses
about the proportions or relationships that
exist within populations.
o do not state hypotheses in terms of a
specific parameter and they make few (if
any) assumptions about the population
distribution.
o Distribution-free tests
o participants are usually just classified into
categories
o involve measurement on nominal or ordinal
scales
o simply frequencies
 situations for which transforming scores into
categories might be a better choice.
o it is simpler to obtain category
measurements
o original scores may violate some of the
basic assumptions that underlie certain
statistical procedures
o original scores may have unusually high
variance
o an experiment produces an undetermined,
or infinite, score
II. Chi-Square Test for Goodness of Fit
 uses sample data to test hypotheses about the shape
or proportions of a population distribution
 determines how well the obtained sample proportions
fit the population proportions specified by the null
hypothesis.
III. Null Hypothesis for the Goodness-of-Fit Test
 specifies the proportion (or percentage) of the
population in each category.
 H0 falls into one of the following categories:
o No Preference, Equal Proportions – the
population is divided equally among the
categories.
 o No difference from a Known Population –
the proportions for one population are not
different from the proportions than are
known to exist for another population.

 alternative hypothesis (H1) simply states that the

population proportions are not equal to the values
specified by the null hypothesis.
IV. Data for the Goodness-of-Fit-Test
 select a sample of n individuals and count how many
are in each category
 observed frequencies (fo)– resulting values; number
of individuals from the sample who are classified in a
particular category.
V. Expected Frequencies
 frequency value that is predicted from the proportions
in the null hypothesis and the sample size (n).
 define an ideal, hypothetical sample distribution that
would be obtained if the sample proportions were in
perfect agreement with the proportions specified 9in
the null hypothesis.

VI.Chi-Square Statistic
 measures how well the data ( fo) fit the hypothesis ( fe).

 steps:
o Find the difference between fo (the data) and fe
(the hypothesis) for each category.
o Square the difference. This ensures that all
values are positive.
o Next, divide the squared difference by fe.
o Finally, sum the values from all the categories.
 numerator measures how much difference there is
between the data and the hypothesis. (first 2 steps)
 final step – add the values to obtain the total discrepancy
between the data and the hypothesis
o large value = data do not fit the hypothesis: reject
H0
 third step: obtained discrepancy between fo and fe is
viewed as relatively large or relatively small depending on
the size of the expected frequency.