Research Methodology

Aspects of Sample Design

• Type of Universe: Finite and Infinite
• Sampling Unit: Individual/ Family/ Locality/ District
• Source List: Sampling Frame
• Sample Size: Thumb Rule
• Parameters of Interest: Aspects of Interest, Important Groups,
Inclusion/ Exclusion
• Budget
Sources of Sampling Bias
• Inappropriate Sampling Frame
• Defective measuring device
• Non-respondents
• Indeterminacy Principle
• Natural Bias (Acquiescence Bias (or Agreement Bias), Social
Desirability Bias, Experimenter Bias)
• Questionnaire Bias
Types of Questionnaire Bias (Wording)
• Ambiguous Question: Is your work made more difficult because you are expecting a baby?
• Complex Question: Has it happened to you that over a long period of time, when you neither practiced
abstinence, nor used birth control, you did not conceive?
• Double-barrelled Question: Do you agree that acquired immunodeficiency syndrome (AIDS) can be
transmitted by shaking hands with a person with AIDS or through other means of physical contact?
• Short Question: Have you had bad sore throats?
• Technical Jargon: What was your age at menarche?
• Uncommon Word: A) In the past week how many times have you felt worried? (or vexed)
• Vague Word: How often do you exercise? A) Regularly, B) Occasionally
• Belief/ Behaviour: A) Do you think that it is a good idea to have everyone’s chest regularly checked by X-ray?
B) Have you ever had yours checked?
• Start Time: A) In the last 12 months, have you had an accident causing head injury? B) From January 1 to
December 31 of last year, did you have an accident causing head injury?
• Insensitive Measure: A) How important is health to you, on a scale of 1 to 3? B) How important is health to
you, on a scale of 1 to 10?
Types of Questionnaire Bias (Leading
Questions)
• Framing: Which operation would you prefer? A) An operation that
has a 5% mortality. B) An operation in which 90% of the patients
will survive
• Leading: A) Do you do physical exercise, such as cycling? B) Don't
you agree that . . . ? C) We know that ….
• Mind-set: How many times do you order food a week? How many
times do you order food a month?
Types of Questionnaire Bias (Intrusiveness)

• Reporting: In the past month, how many times have you lied to
your loved ones?
• Sensitive Question: A) How old are you? B) What is your DOB?
Types of Questionnaire Bias (Inconsistency)

• Change of Scale/ Change of Wording: A) Compared to other persons

your age, would you say your health is excellent, good, fair, or poor?
B) Would you say your health in general is excellent, very good,
good, fair, or poor?
• Diagnostic Vogue: Do you feel irritable? Do you feel annoyed?
Types of Questionnaire Bias (Formatting)
Horizontal vs Vertical
Types of Questionnaire Bias (Formatting)
Juxtaposed Scale
Types of Questionnaire Bias (Other Sources)
• Response Fatigue
• Skipping Question
Types of Research Questions
• Research questions state a research problem in a form that can be investigated and define its nature and scope. One
or more research questions provide the foundation on which a research project is built. Their selection and wording
determines what, and to some extent, how, the problem will be studied.
• “What” questions require a descriptive answer; they are directed toward discovering and describing the
characteristics of and regularities in some social phenomenon (e.g., categories of individuals, social groups of all
sizes, and social processes).
• “Why” questions ask for either the causes of, or the reasons for, the existence of characteristics or regularities. They
are concerned with understanding or explaining the relationships between events, or within social activities and
social processes.
• “How” questions are concerned with bringing about change, and with practical outcomes and intervention (Blaikie,
2000, pp. 60–62)
• What -> Why -> How
• Should all research have hypotheses?
Sampling
• Purposive Sampling (Non-Probability Sampling): Judgement/ Convenience
• Simple Random Sampling
• Systematic Random Sampling
• Stratified Sampling (~homogenous exclusive layers)
• Stratified Random Sampling
• Stratified Systematic Sampling
• Quota Sampling
• Proportional Allocation ((Stratum Size/ Population Size) X Total Sample Size)
• Forced Equal Representation (Comparison)
• Variance-based Representation ((Total Sample Size X Stratum Size1 X SD1)/ Stratum Size1 X SD1 + Stratum Sizen X SDn)
• Cluster Sampling
• Random Clusters
• Area Sampling
• Multi-Stage Sampling
• Multi-Stage Random Sampling
Data Collection - Observation
• Observation becomes a scientific tool and the method of data collection for the researcher, when it serves a formulated
research purpose, is systematically planned and recorded and is subjected to checks and controls on validity and reliability.
• Types:
• Structured Observation: Clear identification of what is being observed, where, for how long, who is being observed,
etc.
• Unstructured Observation: No clear identification of what aspects need to be explicitly observed.
• Participatory Observation: Taking part in the activities of the observed
• Non-Participatory Observation: Detached from the group
• Disguised Observation: Observing without knowledge of the subject
• Controlled Observation: Observation in a controlled pre-arranged setting
• Uncontrolled Observation: Naturalistic observation
• Merits/ Demerits?
Data Collection - Interview
• Personal Interviews
• Structured: Such interviews involve the use of a set of predetermined questions and of highly
standardised techniques of recording. The interviewer in a structured interview follows a rigid procedure
laid down, asking questions in a form and order prescribed.
• Unstructured: Unstructured interviews do not follow a system of pre-determined questions and
standardised techniques of recording information. In an unstructured interview, the interviewer is
allowed to ask supplementary questions or at times he may omit certain questions if the situation so
requires. He may even change the sequence of questions. He has relatively greater freedom while
recording the responses to include some aspects and exclude others.
• Semi-structured: Middle ground between the above two that allows interviewers greater freedom than a
structured interview while providing some amount of structure and direction. Interviewers are able to
approach novel topics in the conversation that relate to their research questions, they are also able to
ask differential questions from a larger basket of questions.
Data Collection - Interview
• Merits of Interview Method:
• There is greater flexibility under this method as the opportunity to restructure questions is always there, specially in
case of unstructured interviews.
• Observation method can as well be applied to recording verbal answers to various questions
• Samples can be controlled more effectively as there arises no difficulty of the missing returns; non-response generally
remains very low.
• The interviewer may catch the informant off-guard and thus may secure the most spontaneous reactions than would
be the case if mailed questionnaire is used
• The interviewer can collect supplementary information about the respondent’s personal characteristics and
environment which is often of great value in interpreting results
• The language of the interview can be adopted to the ability or educational level of the person interviewed and as such
misinterpretations concerning questions can be avoided.
Data Collection - Interview
• Demerits of Interview Method:
• It is a very expensive method, specially when large and widely spread geographical sample is taken.
• There remains the possibility of the bias of interviewer as well as that of the respondent
• Certain types of respondents such as important officials or executives or people in high income groups
may not be easily approachable
• Under the interview method the organisation required for selecting, training and supervising the field-
staff is more complex
Data Collection – Semi Structured Interviews (SSIs)
• Designing and Conducting SSIs:
• Large Groups: Random/ Stratified Random Sampling
• Niche Groups: Purposive Sampling
• Small Groups: Complete sampling
• If staff members are being interviewed, HR/ managers can assist in setting up slots (Only true if the participants are taking part in
interview in formal capacity)
• Ensure contact is made with participant in step-by-step manner in order of increasing contact: Email, Telephone, Interview. Initial
contact MUST include researcher’s credentials, study purpose, outcomes, time required, consent and participant’s role in the
study.
• Drafting Questions:
• Interview guide or schedule
• Ensure adequate proof-reading and pilot tests before going to participants
• Identify Tier 1 and Tier 2 questions, to avoid interrogating the participant. Follow-up interviews can be used where feasible.
• Avoid questions, answers for which can be gathered from other static resources.
• Closed-ended vs Open-ended vs Closed + Open ended
• Be cautious when translator is being used. Preferable to have a length discussion about the translator to understand their
position on the topic, so that possible biases can be accounted for, if required.
• Sensitive questions require adequate rapport formation to generate truthful answers.
Data Collection – Semi Structured Interviews (SSIs)
• Question Order:
• Pleasantries
• Throwaway Questions/ Icebreakers
• General overarching questions
• Move from positives to negatives – Will allow those who are suspicious of your motives to feel safe as they have
already praised, and those critical are forced to think of positives
• Difficult/ problematic questions towards the end to ensure a lot of comfort has been established
• Ensure demographics are only essentials
• End by asking overall general question, request the participant to bring up anything they deem necessary and
finally as for feedback
Data Collection – Interview Guide
• Questions to ask yourself:
• What is the specific purpose of the interviews?
• What information am I interested in exploring?
• Who needs this information, and what are they going to do with it?
• There needs to be a balance between the number of questions you want to ask, the number of questions you
can ask in a certain period of time, and the number of questions that are needed to capture a topic entirely
• This creates a need for highly refined and efficient questions, allowing the researcher to capture larger
amounts of data with lesser number of questions
• Aspects to consider:
• Open-ended questions should be prioritised wherever possible due to their ability to gain larger amounts of data
• Probes should be effectively used, keeping a focus on the individual and not just the topic
• Utilize think back questions
• Utilize ‘Consider the statement’ questions
• Avoid asking why, Instead of “Why do you prefer that type of program,” ask “What are the major reasons you prefer that type
of program? What do you like about it?”
• Keep self-disclosure to a minimum, disclose only what is necessary
• Move from aided to unaided questioning
Data Collection – Questionnaire
• A questionnaire consists of a number of questions printed or typed in a definite order on a form or set of forms. The questionnaire is mailed to respondents who are expected to read and
understand the questions and write down the reply in the space meant for the purpose in the questionnaire itself.
• General Form:
• Structured: Pre-determined Likert, True/ False, Yes/ No, Free Response
• Unstructured: Guideline-based/ Thematic
• Question Sequence:
• Initial questions need to be as less invasive as possible to ensure that the individual doesn’t take up a defensive stance.
• Questions that require considerable mental effort, personal character, money
• The goal is to ensure that there is adequate rapport or familiarity before intrusive questions are asked

• Question wording:
• should be easily understood;
• should be simple i.e., should convey only one thought at a time;
• should be concrete
• the size of the questionnaire should be kept to the minimum.
• Questions should proceed in logical sequence moving from easy to more difficult questions
• Technical terms and vague expressions capable of different interpretations should be avoided in a questionnaire
• There should be some control questions in the questionnaire which indicate the reliability of the respondent.
• Adequate space for answers should be provided in the questionnaire to help editing and tabulation.
• Brief directions with regard to filling up the questionnaire should invariably be given in the questionnaire itself.
• Should look interesting
• When choosing between MCQs and Open-ended questions, a key deciding factor is the nature of data you wish to collect, and also your research question (What, Why, How).

• Questionnaires vs Schedules – Role of enumerators

Data Collection – Consumer Panels
• Where a set of consumers are arranged to come to an understanding to maintain detailed daily records of
their consumption and the same is made available to investigator on demands. Such observations are usually
conducted over longer periods of time to ensure a clear understanding of consumer preferences.
• Transitory Consumer Panel: is set up to measure the effect of a particular phenomenon. Usually such a panel
is conducted on a before-and-after-basis. Initial interviews are conducted before the phenomenon takes place
to record the attitude of the consumer. A second set of interviews is carried out after the phenomenon has
taken place to find out the consequent changes that might have occurred in the consumer’s attitude.
• Continuing Consumer Panel: often set up for an indefinite period with a view to collect data on a particular
aspect of consumer behaviour over time, generally at periodic intervals or may be meant to serve as a general
purpose panel for researchers on a variety of subjects. Such panels have been used in the area of consumer
expenditure, public opinion and radio and TV viewership. The representativeness of the panel relative to the
population and the effect of panel membership on the information obtained are the two major problems
associated with the use of this method
Data Collection – Projective Techniques
• Word Association Tests: To understand what kind of associations certain words have, eg., when trying to come
up with a new slogan, or to understand how the consumer interprets a certain brand
• Sentence Completion Tests: Similar to word association, but more specific, therefore allowing the researcher
to investigate contextual associations as well
• Verbal Projection Tests:
• Thematic Apperception Test: Picture cards are shown to the participant and they are required to come
up with a story that has a beginning, a middle and an end.
• Rosenzweig Test: Also called the Picture Frustration Test. Individuals are exposed to comic based
situations where they are required to express how a certain character would react
• Rorschach Test: 10 printed inkblots are shown to the client and the client is asked to express what they
perceive. Validity reliability questionable.
• Holtzman Inkblot Test: Modified Rorschach. 45 inkblots. Utilizes color, shading, etc. Is able to get much
more information due to higher number of items but is yet to be validated for occupational decision
making.