TESTING OF HYPOTHESIS

RESEARCH METHODOLOGY PROJECT

Sitar Solomon
LL.M 10223: NUALS
1|Page

INTRODUCTION

The word Hypo’ means less than and ‘thesis’ means a generally held view. Ordinarily, a
hypothesis means a mere assumption or some supposition to be proved or disproved. For a
researcher, a hypothesis is a formal question he/she intends to resolve. ‘Hypothesis’ is a
temporary proposition, based on observation of certain occurrences or happenings.
It stands to be modified or corrected during the process of the investigation if the facts
discovered in the course of investigation demand it so. The hypothesis is a link in the process of
investigation, which presupposes the cause and its relationship1
A hypothesis is referred to as an essential tool in research. For a researcher, the
hypothesis is vital in terms of bringing clarity to the research problem. Hypothesis guides socio-
legal research by offering directions to the structure and operation. It offers a temporary answer
to the research question. The hypothesis makes the inquiry more specific and to the point. In the
absence of hypothesis, the researcher is like a sailor on the wide sea without compass or rudder.2
It helps in deciding the direction in which to proceed. It is the starting point of research. It helps
in selecting all available and pertinent facts. P.V> Young remarks, “The use of hypothesis
prevents a blind search and indiscriminate gathering of masses of data which may later prove
irrelevant to the problem under study.3

TESTING OF HYPOTHESIS

We cannot admit anything as valid knowledge until a satisfactory test of the validity has been
completed. It is a mistake to consider hypotheses as facts, no matter how talentedly and
systematically they have been arrived at. Indeed, very exact proof and measurements are
demanded. The function of the hypothesis is to state a specific relationship between phenomena
in such a way that this relationship can be empirically tested. That is, the hypothesis must be
empirically demonstrated as either probably or not probable. 4
Hypothesis testing is a kind of statistical inference that involves asking a question, collecting
data, and then examining what the data tells the researcher about how to proceed. In a formal
hypothesis test, hypotheses are always statements about the population.
For example, the researcher may want to know if the average time that first-year college students
spend studying each week is 20 hours per week. The researcher may want to compare this
average time spent studying for first-year students that earned a GPA of 3.0 or higher and those
that did not.

Developing Null and Alternative Hypotheses

In statistical hypothesis testing, there are always two hypotheses. The hypothesis to be tested is
called the null hypothesis and given the symbol H0. The null hypothesis states that there is no
difference between a hypothesised population mean and a sample mean. It is the status quo

1
Khan, J.A., Research Methodology, A.P.H. Publishing Corporation, New Delhi, 2008, p.57
2
Myneni, S.R., Legal Research Methodology, Allahabad Law Agency, Faridabad, 2004. P.95
3
Ibid.
4
Ahuja, Ram, Research Methods, Rawat Publications, 2003, pp. 84-85
2|Page

hypothesis. For example, if the researcher were to test the hypothesis that first-year college
students study 20 hours per week, he would express our null hypothesis as:

H0 : µ = 20

The researcher tests the null hypothesis against an alternative hypothesis, which is given by the
symbol H1. The alternative hypothesis is often the hypothesis that the researcher believes. It
includes the outcomes not covered by the null hypothesis. In this example, the researcher’s
alternative hypothesis would express that; the first-year students do not study 20 hours per
week:

H1: µ ≠ 20

REJECTION OF NULL HYPOTHESIS

The alternative hypothesis can be supported only by rejecting the null hypothesis. To reject the
null hypothesis means to find a significant enough difference between the sample mean and the
hypothesised (null) mean that it raises real doubt that the actual population mean is 20. If the
difference between the hypothesised mean and the sample mean is very large, the researcher may
reject the null hypothesis. If the difference is minimal, the researcher may not. In each hypothesis
test, the researcher must decide in advance what the magnitude of that difference must be to
allow him to reject the null hypothesis. Notice that if the researcher fails to find a large enough
difference to reject, he fails to reject the null hypothesis. Those are the researcher’s only two
alternatives. When a hypothesis is tested, the researcher must decide on how much of a
difference between means is necessary in order to reject the null hypothesis. He first chooses a
level of significance or alpha (α) level for the hypothesis test. Similar to the significance level the
researcher used in constructing confidence intervals, this alpha level tells how improbable a
sample mean must be for it to be deemed "significantly different" from the hypothesised mean.
The most frequently used levels of significance are 0.05 and 0.01. An alpha level of 0.05 means
the sample mean would be significantly different from the hypothesised mean if the chances of
observing that sample mean are less than 5%. Similarly, an alpha level of 0.01 means that the
researcher considers the sample mean to be significantly different from the hypothesised mean if
the chances of observing that sample mean are less than 1%.

Two-tailed Hypothesis Tests

A hypothesis test can be one-tailed or two-tailed. The examples above are all two-tailed
hypothesis tests. We indicate that the average study time is either 20 hours per week or it is not.
Computer use averages 3.2 hours per week, or it does not. We do not specify whether we believe
the true mean to be higher or lower than the hypothesised mean. We believe it must be different.
3|Page

In a two-tailed test, the researcher will reject the null hypothesis if their sample mean falls in
either tail of the distribution. For this reason, the alpha level (assume .05) is split across the two
tails. The curve below shows the critical regions for a two-tailed test. These are the regions under
the normal curve that, together, sum to a probability of 0.05. Each tail has a probability of 0.025.
The z-scores that designate the start of the critical region are called critical values.
If the sample mean taken from the population falls within these critical regions, or "rejection
regions," the researcher would conclude that there was too much of a difference and would reject
the null hypothesis. However, if the mean from the sample falls in the middle of the distribution
(in between the critical regions), the researcher would fail to reject the null hypothesis.

One-Tailed Hypothesis Test

A single-tail hypothesis test would only be used when the direction of the results is anticipated or
the researcher is only interested in one direction of the results. For example, a single-tail
hypothesis test may be used when evaluating whether or not to adopt a new textbook.
When performing a single-tail hypothesis test, the alternative hypothesis looks a bit different.
Symbols of greater than or less than are used here. For example, the researcher claims that the
average SAT score of graduating seniors was GREATER than 1,110. The researcher’s personal
hypothesis is the Alternative Hypothesis. Then the null and alternative hypothesis could look
something like:
H0: µ ≤ 1100
H1: µ > 1100
In this scenario, the null hypothesis states that the mean SAT scores would be less than or equal
to 1,100 while the alternate hypothesis states that the SAT scores would be greater than 1,100. A
single-tail hypothesis test also means that there is only one critical region because entire critical
region is put into just one side of the distribution. When the alternative hypothesis is that the
4|Page

sample mean is greater, the critical region is on the right side of the distribution. When the
alternative hypothesis is that the sample is smaller, the critical region is on the left side of the
distribution.

To calculate the critical regions, the researcher must first find the critical values or the cut-offs
where the critical regions start.
The Level of Significance

The level of significance of a hypothesis test refers to the probability of rejecting the null
hypothesis when it is true. In statistical jargon this is known as a Type I error. A Type II error
occurs when the researcher fails to reject the null hypothesis when it’s actually false.
The smaller the level of significance, the less likely there is to be a Type I error, but the more
likely there is to be a Type II error. The level of significance is selected based on the relative
importance of avoiding a Type I error compared with a Type II error. In many applications in
finance and economics, the level of significance is chosen to be 0.05 (5 percent).

Type I and Type II Errors

There will be some sample means that are extremes – that is going to happen about 5% of the
time since 95% of all sample means fall within about two standard deviations of the mean. What
happens if a hypothesis test is ran and an extreme sample mean is got? It will not look like the
hypothesized mean, even if it comes from that distribution. The researcher would be likely to
reject the null hypothesis, but they could be would be wrong.
5|Page

When the researcher decide to reject or not reject the null hypothesis, they have four possible
scenarios:
a. A true hypothesis is rejected.
b. A true hypothesis is not rejected.
c. A false hypothesis is not rejected.
d. A false hypothesis is rejected.

If a hypothesis is true and the researcher does not reject it or if a false hypothesis is rejected, they
have made the correct decision. But if they reject a true hypothesis or a false hypothesis is not
rejected then an error is made. Overall, one type of error is not necessarily more severe than the
other. Which type is more serious depends on the specific research situation, but ideally, both
types of errors should be minimised during the analysis.
The general approach to hypothesis testing focuses on the Type I error: rejecting the null
hypothesis when it may be true. The level of significance, also known as the alpha level, is the
probability of making a Type I error. At the 0.05 level, the decision to reject the hypothesis may
be incorrect 5% of the time. Calculating the probability of making a Type II error is not as
straightforward as calculating a Type I error.

The researcher should be able to recognise what each type of error looks like in a particular
hypothesis test. For example, suppose the researcher is testing whether listening to rock music
helps improve their memory of 30 random objects. He must assume further that it does not. A
Type I error would be concluding that listening to rock music did help memory (but you are
wrong). A Type I error will only occur when the null hypothesis is false. Assume that listening to
rock music does improve memory. In this scenario, if the researcher concluded that it did not,
they would be wrong again. However, this time they would be making a Type II error — failing
to find a significant difference when one in fact exists.

It is also crucial that the researcher realises that the chance of making a Type I error is under
their direct control. Often we establish the alpha level based on the severity of the consequences
of making a Type I error. If the consequences are not that serious, we could set an alpha level at
0.10 or 0.20. In other words, we are comfortable making a decision where we could falsely reject
the null hypothesis 10 to 20% of the time. However, in a field like medical research, we would
set the alpha level very low (at 0.001 for example) if there was potential bodily harm to patients.
6|Page

THE TEST STATISTIC

A test statistic is a numerical measure that is computed from sample data to determine whether or
not the null hypothesis should be rejected. The form of the test statistic depends on the type of
null hypothesis that is being tested. For example, if a hypothesis about the mean of a population
is being tested, and a small sample of data (fewer than 30 observations) is drawn from the
population, then the appropriate test statistic is as follows:

In this formula,
t indicates that this test statistic follows the Student’s t-distribution.
is the sample mean.
μ0 is the hypothesized value of the population mean.
s is the sample standard deviation.
n is the sample size.
If the sample size is large (30 or more observations), and the population standard deviation is
unknown, the test statistic becomes as follows:

Z indicates that this test statistic follows the standard normal distribution.
THE CRITICAL VALUE
In order to determine whether a hypothesis should be rejected, the test statistic is compared with
one or two critical values. The critical values depend on the type of hypothesis being tested as
well as the alternative hypothesis being used.
For example, when testing a hypothesis about the mean about the population with a small
sample, the critical values are determined as follows:

Two-tailed test: Critical value =

Right-tailed test: Critical value =

Left-tailed test: Critical value =

7|Page

In these formulas, α is the level of significance; n represents the sample size. These critical
values are drawn from the Student’s t-distribution with n-1 degrees of freedom (df). The number
of degrees of freedom used with the t-distribution depends on the particular application. For
testing hypotheses about the population mean, the appropriate number of degrees of freedom is
one less than the sample size (that is, n-1).

The critical value or values represent tail areas under the student’s t-distribution. For a two-tailed
test, the value of the level of significance (α) is split in half; the area in the right tail equals α/2
and the area in left tail equals α/2, for a total of α.
When testing a hypothesis about the mean about the population with a large sample, the critical
values are determined as follows:
Two-tailed test: Critical value = ±Za/2

Right-tailed test: Critical value = Za

Left-tailed test: Critical value =-Za

MAKING THE DECISION

The researcher can decide whether or not to reject the null hypothesis in one of two ways:
 Compare the test statistic with the critical value(s).
 Compare the probability value (p-value) to the level of significance.

Comparing the test statistic with the critical value(s)

The researcher can determine whether the null hypothesis should be rejected as follows:

Right-tailed test: If the test statistic is greater than the critical value, reject the null hypothesis
H0:µ= µ0 in favour of the alternative hypothesis H1:µ> µ0; otherwise, do not reject the null
hypothesis.

Left-tailed test: If the test statistic is less than the critical value, reject the null hypothesis H0:µ=
µ0 in favour of the alternative hypothesis H1:µ< µ0; otherwise, you do not reject the null
hypothesis.
8|Page

Two-tailed test: If the test statistic is less than the negative critical value, reject the null
hypothesis H0:µ= µ0 in favour of the alternative hypothesis H1:µ< µ0. If the test statistic is
greater than the positive critical value, reject the null hypothesis H0:µ= µ0 in favour of the
alternative hypothesis H1:µ> µ0. Otherwise, the researcher may not reject the null hypothesis.

Comparing the p-value to the level of significance

As an alternative to comparing the test statistic with critical values, the researcher can compare
the p-value with the level of significance. The p-value represents the likelihood that the test
statistic has a specified value when the null hypothesis is true. A very small p-value indicates
that the null hypothesis is unlikely to be true.
The decision rule when using the p-value is as follows:
If the p-value is below the level of significance, reject the null hypothesis; otherwise, do not
reject the null hypothesis.

If the researcher rejects the null hypothesis when using a two-tailed test, the sign of the test
statistic determines which alternative hypothesis to accept. If the test statistic is negative, the left
tailed alternative hypothesis is accepted; if the test statistic is positive, the right-tailed alternative
hypothesis is accepted.
9|Page

BIBLIOGRAPHY

1. Ahuja, Ram: “Research Methods”, Rawat Publications (2003).

2. Bast, Carol, ‘Legal Research and Writing’, London Delmar Publishers
3. Chapin F. Steward, ‘Experimental Design in Sociological RResearch’, Harper and Bros.,
1955.
4. Myneni, SR: Legal Research Methodology, Allahabad Law Agency
5. Singh, Rattan: Legal Research Methodology, LexisNexis
6. Alan Anderson and David Semmelroth: Statistics for Big Data For Dummies