Research Methods: Hypothesis Testing, Surveys, and Sampling

This document provides a comprehensive overview of various research methods, starting with the basics of hypothesis testing, differentiating between null and alternative hypotheses. It then delves into descriptive research designs, exploring the use of surveys and observations for data collection, and contrasts descriptive and causal research. The paper examines different data collection methods like questionnaires, interviews, and surveys, including their advantages and disadvantages. Various measurement scales (nominal, ordinal, interval, and ratio) are also discussed, emphasizing their importance in preventing errors and selecting appropriate statistical techniques. The document highlights the factors evaluating the goodness of a questionnaire, reliability, and validity of research. It further explains various sampling methods (probability and non-probability) and different types of surveys (cross-sectional and longitudinal), and discusses the importance of designing questionnaires appropriately. The document concludes with a discussion on open vs hidden and direct vs indirect observation methods, emphasizing their application and significance in research.

1
1160655

2
14. In hypothesis testing, a researcher always takes into account whether to reject the null
hypothesis. A null hypothesis is a statement that the researcher takes into account with the
proposition that no significant statistical relationship exists in a set of given observations. On
the other hand, an alternative hypothesis is a statement which states that there exist statistical
relationship in a set of given observations. In a hypothesis testing a researcher rejects a null
hypothesis and accepts an alternative hypothesis.
15. Descriptive research design is one of the most common methods of research design. In
this type of research design, the main objective is to find out what’s happening in a given
situation rather than trying to explain relationships or cause and effect. In this type of
research design, the information is collected by asking a set of predefined questions in a
structured questionnaire to a sample of individuals acting as a representative of the
population. The research methods in a descriptive research design are survey and observation.
While conducting a survey, a researcher has to deal with sampling, design of questionnaire,
administration of questionnaire and analysis of data. The questionnaires are administered
through interviews. The typical objective of a survey is to learn from an on-going activity
through the process of studying changes in the behavioural patterns of the subject. The
surveys are descriptive in nature (Sreejesh, Mohapatra & Anusree, 2013).
16. Descriptive research basically focuses on the study of what exists within the field of
study. In other words, descriptive research tells us what’s going on or what actually exists.
On the other hand, causal research helps in determining whether one or more variables affect
the other variables. Causal research is also known as exploratory research or experimental
research. In a causal research, within the data generating process, the researcher himself
becomes an active participant. He manipulates diverse treatment to examine their effects on a
variable of interest (McNabb, 2004).
17. The methods of descriptive research include questionnaires, personal interviews, phone
surveys and normative surveys.
Questionnaire method is one of the oldest and most reliable descriptive research methods.
This is designed to gather information about the attitudes, characteristic and behaviours of a
population by imbuing a standard set of questions. The questions are distributed to the
respondents for data collection (Clifford, French & Valentine, 2010).

3
Personal interviews are conducted when there is involvement of a specific target population.
The method of interview is face to face and here the researcher possesses the flexibility to
build rapport with the respondents and take more in depth information. However, personal
interview method is costly method.
A phone survey is a data collection procedure over telephone. It is a very cost effective
means to collect data. In this process, the telephone numbers are utilized by interviewers for
gathering data from the respondents.
The normative survey basically uses statistics and values considered normal for the group
that is being surveyed. This is done in order to comprehend and gather data on a particular
subject.
18. Nominal scale is also known as categorical variable scale. It is used to label variables into
distinct classifications. In a nominal scale there does not include any quantitative value.
Examples of nominal scales include gender, place of residence and political preference.
Ordinal scale is simply used to depict the order of the variables. Examples of ordinal scales
include status at workplace, order of satisfaction, and order of quality of product and so on.
Interval scale is a numerical scale and the order of the variable is known as the difference
between the variables. Examples include temperature scales, family income and so on.
Ratio scale is a variable measurement that makes the difference between variables apart from
producing the order of the variable. Examples of ratio scale are the boy’s current height is
between 5 feet to 6 feet, the weight of boy is between 71 to 90 kilograms.
19. The importance of differentiating between different measurement scales lies in the fact
that it helps in preventing mistakes. The researcher needs to see the distinctive differences by
differentiating different scales of measurement and also help in selecting the appropriate
statistical technique.
20. There are 7 factors that evaluate the goodness of a questionnaire. The factors are
simplicity, precision, distinctiveness, comprehensiveness, optionality, impartiality and
balance.

4
21. Reliability basically means the extent to which the results of an experiment can be
produced when the research study is repeated given under the same conditions. It cannot be
always assured about the validity of a reliable measurement. The results can reproducible and
they are not necessarily correct. Validity on the other hand represents the extent to which the
results actually measures. It is a fact that validity guarantees reliability.
22. There are four methods of evaluating reliability of a scale. They are inter-rater reliability,
test-retest reliability, parallel forms reliability and internal consistency reliability. Inter-rater
reliability explains the extent to which the observers respond the same way to a particular
given phenomenon. Test-retest reliability is basically obtained through managing the same
test over a period of time to a set of respondents twice. Parallel forms reliability uses a set of
questions that is divided into two equal parts for measuring same construct, knowledge and
skill. The two sets are given to a sample to estimate reliability from the two sets. Internal
consistency reliability is a way to find out how satisfied the respondents or customers are by
giving them three different set of questions.
23. It is only necessary for questionnaires to be valid or reliable. It is also necessary to make
it fool-proof when we are using it in a cross-cultural setting. If a questionnaire is used in a
cross cultural setting, then it is necessary to do a form of cognitive interviewing of some
participants with the help of instruments to comprehend whether the questions makes sense to
study population. Once calibration in the language is done, then it has to be treated with a
pilot test in a small sample of population. Once we are sure that it is working, then the actual
data collection procedure must be started.
24. Unlike survey, in observation method, the researcher is required to engross themselves in
a culture. In typical type of observation methods like anthropological research, the researcher
has to do months of fieldwork. On the other hand, surveys can be done quickly via different
media.
25. Open observation is observation collected directly from the respondents via interview.
While on other hand, hidden observations are observation collected via hidden methods. The
hidden method raises ethical questions like violation of privacy.
Direct observation includes beholding at the actual behaviour and the occurrence of a study
rather than focusing on the results of the study. Indirect observation actually circumscribes

5
around analysis of textual materials. The materials can be generated from transcripts or
directly from narratives.
In natural observation, the study of subjects occurs within real life environments and
circumstances. The subjects may or may not know that they are being watched or observed.
In contrived observation, the study of subjects takes place in controlled settings like in a
research lab. The subjects usually know that they are being observed.
26. The main advantage of observation is that it helps in formulating the hypothesis which is
the basis of research. Observation has greater universality of practice.
The disadvantage of observation is that because social phenomena cannot be controlled, the
generalizations made by observations may not be reliable. Thus, the issue of reliability crops
up.
27. Survey research is basically the amalgamation of data obtained by asking individuals
questions. The questions are asked in person, via phone or via online. Opinions, beliefs and
feelings of the respondents are collected through survey design.
28. Survey method is the best method as because it utilizes a wide array of data collection
methods to obtain information. Nowadays with the full-fledged spread of internet, online
surveys are proving to be the best method of data collection.
29. The considerations which the researcher takes into account while designing a
questionnaire are the mode of data collection, ordering of questions, using language that is
easy to understand and that also determine the accuracy of the answers, clear question
structure, simple design and short and simple analysis plan.
30. There are basically three types of questions. They are factual, interpretive and evaluative.
Factual questions are questions in which the answers are that which everyone will agree on.
An interpretive question is one that possesses answer which can be supported with evidence
from the text. An evaluative questions is a question that gives the reader the option to agree
or not with the author’s viewpoint.
31. A personal interview is a survey method often called face-to-face interview. It is
conducted when there is involvement of a target population. The main objective of personal
interview is gather responses in a deeper manner as compared to other means like online or

6
telephonic interview. A researcher may also develop rapport with the respondent and gather
deeper insights regarding the study in question.
A self-administered questionnaire basically refers to questionnaires that have been designed
in such a way that the respondent alone will be able to complete it without active intervention
of the researcher.
Mail survey is also a part of self-administered questionnaire. In this method, the respondents
are given the questionnaires via envelope through the postal system. The respondents are
asked to answer the questions and revert back.
Telephonic interview are interview method in which the respondents are asked questions by
the interviewer via telephone. It is a cost effective method of survey.
32. Cross sectional survey is a survey method where the information is collected from a
sample. The sample is basically drawn from a prearranged population. The information
collected is just made at one point in time. In a longitudinal survey, collection of information
takes place at different points of time. Trend studies, cohort studies and panel studies are
significant methods in longitudinal studies.
33. Probability sampling technique is a technique wherein the subjects within the population
have the equal probability of getting selected. On the other hand, non-probability sampling is
a sampling method the selection of the subjects from the population does not have equal
probability of selection. Probability sampling is known as random sampling and non-
probability sampling is known as non-random sampling.
34. A simple random sample is a subset of a population wherein each member of the sample
has equal probability of selection. A simple random sample is an unbiased representation of a
group.
Systematic sampling is a probability sampling. In this type of sampling, elements are chosen
from a population by a starting point and selecting other members after making fixed
sampling interval. The sampling interval is obtained by the dividing the population by
sample.
Stratified sampling is a sampling method wherein the population is segmented to various
groups or strata. The strata are developed based on the common features of the population
data and then the researcher selects proportionally the data.

7
Convenience sampling is a type of non-probability sampling wherein the sample is drawn
from a part of the population. It is done as per the convenience of the researcher.
Judgment sampling is an authoritative sampling. It is also a non-probability sampling
technique in which the researcher selects the sample based on his own judgment and
knowledge. When using judgment sampling, the researcher can indulge in various activities
like interview conduction as a result of lower subject volume.
35. Qouta sampling is a sampling method wherein a sample is obtained by identifying
subgroups that is to be included and then by establishing qoutas for individuals through
convenience from each group. Stratified sampling is a sampling process wherein a sample is
obtained by dividing the population into various subgroups and then randomly selecting equal
numbers from each of the subgroups. Example of a quota sample is for instance grouping
population by the states they live and by income. Example of a stratified sample is like
dividing the age groups into different brackets like 18-30 years, 30 – 45 years and so on.
36. First the data has to be determined like whether the dependent and independent variables
are numerical or categorical. The numeric is again divided into parametric and otherwise.
Same is for categorical data. For parametric tests for numerical data, 2 groups unpaired t-test
will be done of it is greater than 2 groups, then 2 groups Analysis of Variance or F test is
done. In otherwise case, 2 groups mann-Whitney U test, Wilcoxon’s rank and greater than 2
groups Krushall-Wallis H-test is done. For categorical data, for two groups, Chi square test or
Fisher’s exact test will be done and for greater than 2 groups Chi-square test is to be done.
37. If the independent variable in a dataset contains more than two levels and we wish to
analyse whether the dependent variable has impact on them, then we do an ANOVA analysis.
If there is one independent variable with three or more levels and we have only one
dependent variable then a one-way ANOVA is done.
A chi-square test is done when we count the incidents of some occurrences and then we
compare the actual data and compare them with the expected data.

8
38. P-value is a very important parameter while accepting or rejecting a hypothesis. P-value
is basically the probability of gaining the experimental results of a test with the assumption
that null hypothesis is true. There are various criteria. Suppose at 5% level of significance,
the p-value is less than 0.05 then the researcher will reject the null hypothesis and accept the
alternative hypothesis. At 1% level of significance, if the p-value is less than 0.01 then the
researcher will reject the null hypothesis and accept the alternative hypothesis.
44. a. The characteristics that must be there in a well written report are conciseness,
simplicity, consistency, preciseness and accuracy, comparability and promptness.
b. The contents of a research report are Title page, Abstract, Introduction, Methodology,
Discussion, Conclusion, References, Appendices.
45. The importance of research report lies in the fact that it helps in developing the ability to
finish a whole research piece with inclusion of research design. The objective is to
summarize what exactly has to be obtained by the study. The objectives also include various
answers to the research questions and testing the hypotheses.
Case Study
First we will formulate the null and alternative hypothesis
Case I
H0: The employees do not perceive that their compensations are lower than their counterparts
in similar companies
H1 : The employees do perceive that their compensations are lower than their counterparts in
similar companies
Case II
H0: The working hours are not so long
H1 : The working hours are long
Case III

9
H0: The nature of work is not exhausting
H1: The nature of work is exhausting
Then we will collect data via questionnaires and we analyse the hypotheses and try to reject
the null hypotheses.
Reference
Clifford, N., French, S. & Valentine, G. (2010). Key Methods in Geography. SAGE.
McNabb, D. E. (2004). Research Methods for Political Science. M.E. Sharpe.
Sreejesh, S., Mohapatra, S. & Anusree, M. R. (2013). Business Research Methods: An
Applied Orientation. Springer Science & Business Media.