Professional Documents
Culture Documents
Hypothesis
Hypothesis
Characteristics of hypotheses:
1. It is conceptual in nature,
Importance of hypothesis:
5. It provides a clear and particular objectives to the research worker and facilities
him by offering a base for sample selection and research method for achieving the
set objectives.
6. It acts as the key function of clubbing together all the associated facts and
information into one.
7. It helps in avoiding a blend search and random collection of data which might
later on tense to be unnecessary for the problem which is being studied.
8. It acts as one of the powerful tool which helps in showing the right path for the
research work.
Types of hypotheses:
Types of Hypothesis
Question Hypothesis
Declaration
Hypothesis
Directional
Hyothesis
Null/Non-directional
Hypothesis
1.
Question Hypothesis:
It is regarded as the simplest form of empirical observation. Under this type, the
hypothesis is usually represented in a simple question form. It does not serve as
the definition of hypothesis that is it does not convey the meaning of the
hypothesis. This type of hypothesis is mostly used in those research projects which
needs to find out an answer to a simple question.
Ex: is there any effect of reinforced advertisement on the product sales volume?
2. Declaration statement:
Ex: there is a direct effect of reinforced advertisement upon the product sales
volume.
3. Directional hypothesis:
Directional hypothesis basically emphasis upon drawing and expected direction for
the existing relationship between different variables. The key attributes of this
hypothesis are:
a. The researcher who uses this type of hypothesis is usually confident about the
assumed evidence.
(ii) researcher instill the variables very clearly and that the existing evidence
supports the directional statement.
Ex: product sales volume can be increased by using excessive rain forced
advertisement whereas the failure to initiate such an advertisement would greatly
affect the product sales volume.
The non directional hypothesis is explained in the negative form that is it takes
that there exist no relationship between two different variables this type of
hypothesis is statistical in nature and can be measured within the framework of
the probability theory. It usually gives a negative outcome to the researcher.
Therefore it is considered or called as non directional in nature.
Ex: there is no impact of reinforced advertisement upon the product sales volume
and null hypothesis refers to zero hypothesis and researcher needs nothing to
develop anal hypothesis that is there is no need of a research to develop anal
hypothesis. It is the best suitable form for a competitive object of the research for
finding out the research information.
Statistical hypothesis:
a. Null hypothesis:
Hypothesis which is tested for possible under the assumption that is true. A
definite statement about the population parameters for applying the test of
significance is called the null hypothesis which is usually a hypothesis of an
difference.
Ex: H 0 μ=μ H 0
Ex: decide whether a given computer performs well we form the hypothesis that the
computer is doing well that is P = 100% where P is the probability of success.
b. Alternate hypothesis:
For example if we want, to test the null hypothesis that the average height of
soldiers is 165 cms
H a μ ≠ μ H a , μ > H 0 , μ< H 0
Example:
Therefore, in order to test the null hypothesis that the population has a specified
mean, μ0 that is H 0 =μ=μ0
1. Simple hypothesis:
Examples:
( i ) H 0 : μ=μ 0 that is
( ii ) H 0 :μ 1−μ2=δ
2. Composite hypothesis:
(i) H 0 : μ> μ0
(ii) H 0 : μ< μ0
(iii) H 0 : μ≠ μ0
(iv)
3. Test statistic:
To test whether the null hypothesis setup should be accepted or rejected be used at
a statistics which is based on appropriate probability distribution. For this
purpose z distribution and the normal curve is used for large samples that is in
greater than or equal to 30 and T distribution is used for small samples that is n
less than or equal to 30.
4. Level of significance:
Over of attes is the probability of rejecting h not given that the specific alternative is
true the power of a test can be completed as 1 - beta often different types of test are
compared by contrasting power properties.
6. Critical region
The area under the probability curve is divided into two regions
However the entire critical region lies in one tale of the distribution of the test
statistic.
The critical region for all alternative hypothesis containing the symbol greater than
lies entirely in the right tail of the distribution while the critical region for an
alternative hypothesis containing the symbol less than lies entirely in the left tail.
The symbol indicates the direction where the critical region lies.
It is of any statistical hypothesis where the alternative is written with the symbol
not equal to is called to tell test however region is split into two equal parts one in
each tale of the distribution of the test statistic.
The critical region lies on both sides of the right and left tails of the curve such that
the critical region of area alphabet to lies on the right tail and critical region of area
alphabet to lies on the left tail as shown in the two tail test.
Two tail test physics the null hypothesis if say the sample mean is significantly
higher or lower than the hypothesised value of the mean of the population.
Such a test is appropriate when the null hypothesis is some specified value and
alternate hypothesis is a value not equal to the specified value of null hypothesis.
Even tail test would be used when we are to test say whether the population mean
is either lower than or higher than some hypothesised value when we are interested
in knowing whether population mean is lower than the hypothesised mean then it
is a left tail test. But when we want to know whether population mean is higher
than the hypothesised mean then it is a right tail test.
By comparing the critical value and the calculated value we reject or accept the
null hypothesis at 5% or 1% level of significance as follows:
Select a random
sample and workout
an appropriate value
from sample data
Calculate the
probability that
sample result would
diverge as widely as it
has from expectations,
if H0 were true.
is this probability
equal to or less than
and for one and two
tail respectively
Step 4: selected random sample and workout and appropriate value from sample
data
Step5: calculate the probability that sample result would diverge as widely as it is it
has from expectations if H0 were true.
Step 6. Is this probability equal to or less than alpha and alphabet to for 1 and 2
tail respectively.
These tests pertain to judgements made about the population mean. The choice of
technique to be used is influenced by factors like normality of population, size of
sample, whether the variance of population is known or unknown and whether it is
one-tailed or two tailed.
If the population variance is known, then sample size becomes irrelevant because
the normal distribution and related Z-tables can be used. However, when
population variance is unknown then sample size dictates the choice of the
probability standard distribution.
(ii)Population normal and finite, sample size may be large or small, Ha may
be one sided or two sided
X −μ
¿
Z
( )
σ
√n
x
√
(N−n)
(N −1)
(iii) Population may not be normal, population infinite but sample size large,
Ha may be one sided or two sided, then
X−μ
Z=¿ σp
√n ( )
b. When population variance is unknown:
(i) Population normal and infinite, sample size small, Ha
may be one sided or two sided
X−μ
t=
( )
σ p with d.f =(n-1)
√n
(ii) Population normal and finite, sample size small, Ha
may be one sided or two sided,
X −μ
Problems:
x−μ
t=
σs = -1.58
√n
Since the calculated value of t is less than 2.262, Ho
is accepted. We concluded that the average
thickness of washer is 0.025 cms only.
3. The mean price of shares of Andhra bank during
2004 was $s. 64 . in the year 2005 the mean price
of Andhra Bank for 20 randomly selected days is
found to be Rs. 84 with a standard deviation of $s.
4.50. Test whether there is significant difference in
the price of the shares for the two years at 5% level
of significance.
H0 : μ = 64
Ha : μ ≠ 64
x−μ
t=
σs = 19.877
√n
df =n−1=20−1=19
For 19 df t0.05 = 2.093
Since t calculated is greater than t table , null hypothesis
is rejected.
4. A random sample of male employees is taken at the
end of a year and the mean number of hours of
absenteeism for the year is found to be 63 hours. A
similar sample of 50 female employees has a mean
of 66 hours. Could these samples be drawn from a
population with same mean and standard deviation
of 10 hours? State clearly the assumption you
make?
Ho: Samples are drawn from same population
Ha: Samples are drawn from population with
different means
x 1−x 2
z= =−1.5
σ
√
2
n
Z values lies in the acceptance region , Ho accepted.
√
2 2
S1 S2
+
n1 n2
Mean SD N
Girls 75 15 150
Boys 70 20 250
Is there a significant difference in the mean scores
obtained by boys and girls.
Ho : μ1=μ2
Ha: μ1 ≠ μ2
√
2 2
S1 S2
+
n1 n2
√
2 2
S1 S2
+
n1 n2
Z = -3.87420
Reject Ho.
Chi square test:
The square of a standard normal variable is called a chi
square which variate with one degree of freedom. It is
expressed as χ 2 .
A chi square test can be used when the data satisfies four conditions: