flowchart LR S[Specify info<br/>needed] --> M[Mode] M --> C[Content] C --> Q[Question type<br/>open / closed] Q --> W[Wording] W --> O[Order] O --> L[Layout] L --> P[Pre-test] P --> F[Finalise] style S fill:#E3F2FD,stroke:#1565C0 style F fill:#E8F5E9,stroke:#2E7D32
70 Data Collection and Questionnaire Design
70.1 What is Data Collection?
Data collection is the systematic process of gathering information for analysis and decision making. It is the raw material of every statistical and managerial study. Naresh Malhotra’s standard text frames it as “the process by which the researcher collects information from the chosen units of inquiry” (malhotra2019?).
TipTwo Sources of Data
| Source | Description | Examples |
|---|---|---|
| Primary | Collected first-hand for the study | Surveys, observation, experiments |
| Secondary | Already collected for some other purpose | Government reports, journals, company records |
TipComparing Primary and Secondary Data
| Feature | Primary | Secondary |
|---|---|---|
| Cost | High | Low |
| Time | Long | Short |
| Customisation | High | Low |
| Currency | Fresh | Possibly outdated |
| Availability | Always (collect it) | Sometimes |
70.2 Methods of Primary Data Collection
TipSix Common Methods of Primary Data Collection
| Method | What it does | Use |
|---|---|---|
| Observation | Watch behaviour without interaction | Retail traffic, ethnography |
| Survey | Standardised questions to many respondents | Most common method |
| Interview | One-on-one (structured / semi / unstructured) | Depth |
| Focus group | Small moderated discussion | Qualitative insight |
| Experiment | Manipulate variables, observe outcomes | Cause-effect |
| Projective techniques | Indirect — word association, sentence completion | Subconscious motives |
70.2.1 Survey Modes
TipFive Survey Modes
| Mode | Strengths | Weaknesses |
|---|---|---|
| Personal interview | Depth, completion rate | Costly, interviewer bias |
| Telephone | Speed, moderate cost | Sample bias (non-mobile) |
| Wide reach, low cost | Low response rate | |
| Online / web | Low cost, fast, large reach | Sample bias (digitally connected) |
| Mobile / SMS | Reach low-tech areas, fast | Limited length |
70.3 The Questionnaire — Steps in Design
The questionnaire is the central instrument of survey research. Malhotra’s classical 10-step design process (malhotra2019?):
TipTen Steps in Questionnaire Design
| # | Step |
|---|---|
| 1 | Specify the information needed |
| 2 | Choose the type of interview / mode |
| 3 | Determine the content of individual questions |
| 4 | Design the question to overcome inability and unwillingness |
| 5 | Decide the question structure (open / closed) |
| 6 | Determine question wording |
| 7 | Arrange the questions in proper order |
| 8 | Design the form and layout |
| 9 | Reproduce the questionnaire |
| 10 | Pre-test, revise, finalise |
70.4 Question Types
TipTwo Main Types of Questions
| Type | Description | Strength |
|---|---|---|
| Open-ended | No fixed response options | Rich depth, qualitative |
| Closed-ended | Fixed response options | Easier to code and analyse |
TipClosed-ended Question Formats
| Format | Description |
|---|---|
| Dichotomous | Yes/No |
| Multiple choice | Pick one or more |
| Likert scale | Strongly Agree → Strongly Disagree (typically 5 or 7 points) |
| Semantic differential | Bipolar scale (cold ↔︎ warm) |
| Stapel scale | Single adjective with +5 to −5 |
| Constant sum | Allocate fixed total across options |
| Rank order | Rank items in order of preference |
70.5 Common Errors in Questionnaires
TipCommon Errors in Question Design
| Error | Description |
|---|---|
| Leading question | Suggests the desired answer |
| Loaded question | Contains emotionally charged language |
| Double-barrelled | Asks two things in one question |
| Ambiguous | Vague or unclear wording |
| Negative wording | Confuses respondents |
| Long-winded | Too long; respondent gives up |
| Sequence bias | Earlier questions affect later answers |
| Social desirability | Pressure to give the “acceptable” answer |
70.6 Reliability and Validity
Two essential tests of any measurement instrument:
TipReliability vs Validity
| Concept | What it captures | Measures |
|---|---|---|
| Reliability | Consistency — repeats give same result | Test-retest, split-half, Cronbach’s α |
| Validity | Accuracy — measures what it claims | Content, criterion (concurrent / predictive), construct |
A test can be reliable but not valid; a valid test must be reliable. Cronbach’s alpha ≥ 0.7 is the typical reliability threshold.
70.7 Pre-testing the Questionnaire
A pre-test is the trial run of the questionnaire on a small sample to identify problems before full deployment. Pre-testing checks:
- Length and respondent fatigue.
- Clarity of instructions.
- Wording problems.
- Skip patterns and routing.
- Recording and coding ease.
70.8 Practice Questions
Q 01
Sources
Easy
Government statistical reports used in a marketing study are an example of:
View solution
Correct Option: B
Already collected for some other purpose = secondary data.
Q 02
Likert
Easy
A 5-point "Strongly Agree to Strongly Disagree" scale is a:
View solution
Correct Option: A
Likert scale (Rensis Likert, 1932) — agreement scale, typically 5 or 7 points.
Q 03
Errors
Medium
"Do you find this product good and worth its price?" is an example of:
View solution
Correct Option: C
The question asks two things at once — quality AND value. Double-barrelled.
Q 04
Reliability
Medium
A measure that gives consistent results when repeated under similar conditions is said to be:
View solution
Correct Option: B
Consistency = reliability. Validity = accuracy. A test can be reliable but invalid; a valid test must be reliable.
Q 05
Cronbach
Medium
Cronbach's alpha is used to assess:
View solution
Correct Option: A
Cronbach's α (1951) — most-used measure of internal consistency. Threshold typically α ≥ 0.7.
Q 06
Methods
Medium
Word association and sentence completion are examples of:
View solution
Correct Option: B
Projective techniques — indirect methods to surface subconscious motives.
Q 07
Survey Mode
Medium
An advantage of online surveys over personal interviews is:
View solution
Correct Option: A
Online surveys are cheap and fast. They suffer from sample bias toward the digitally connected.
Q 08
Pre-test
Easy
The purpose of pre-testing a questionnaire is to:
View solution
Correct Option: B
Pre-test = trial run to identify wording, length, routing and coding problems.
ImportantQuick recall
- Data collection — primary (first-hand, customised, costly) vs secondary (already collected, cheap, fast).
- Six methods of primary data: observation, survey, interview, focus group, experiment, projective.
- Survey modes: personal, telephone, mail, online, mobile.
- Malhotra’s 10 steps in questionnaire design.
- Question types: open vs closed. Scales: Likert, Semantic Differential, Stapel, Constant Sum, Rank Order.
- Common errors: leading, loaded, double-barrelled, ambiguous, negative wording, long-winded, sequence bias, social desirability.
- Reliability (consistency) vs Validity (accuracy). Cronbach’s α ≥ 0.7.
- Pre-testing is essential before full deployment.