Research Design and the Market Research Process

Introduction

Market research is not simply the collection of customer opinions. It is a disciplined process for converting a managerial or entrepreneurial uncertainty into credible evidence that can support a decision. Entrepreneurs use research to assess whether a problem is important, identify attractive customer segments, estimate demand, understand competitors, test value propositions, and reduce the risk of allocating scarce resources to an unattractive opportunity.

A research plan connects the decision to be made with the evidence required to make it. It specifies the research problem, scope, methodology, sources of data, sampling approach, instruments, fieldwork schedule, budget, analytical procedures, quality controls, ethical safeguards, and limitations. A strong plan is neither the longest nor the most expensive design. It is the design that produces sufficiently trustworthy evidence for the decision at hand within realistic constraints.

In entrepreneurship, research is usually conducted under conditions of uncertainty, limited budgets, incomplete information, and rapidly changing markets. Consequently, the researcher must balance methodological rigor with speed and practicality. This chapter develops that capability by examining the market research process, research paradigms, design choices, validity, reliability, triangulation, project management, and limitations.

Learning Objectives

By the end of this chapter, you should be able to:

1. Translate a managerial decision into a clear research problem and research objectives.
2. Distinguish exploratory, descriptive, and causal research designs.
3. Select appropriate qualitative, quantitative, and mixed-method approaches.
4. Develop a coherent sampling, data-collection, analysis, timing, and budgeting plan.
5. Evaluate validity, reliability, credibility, and triangulation in market research.
6. Identify ethical, technological, practical, and interpretive limitations.
7. Communicate research findings in a form that supports entrepreneurial decisions.

1. From Managerial Problem to Research Problem

The market research process begins with a decision problem, not with a questionnaire. A decision problem describes an action that management may need to take, such as whether to launch a product, enter a segment, change a price, or select a distribution channel. A research problem identifies what must be learned to inform that action.

For example, “Should the firm launch a subscription version of its product?” is a decision problem. A corresponding research problem might be: “What customer segments experience sufficient recurring value to adopt a subscription model, what price structures are acceptable, and what alternatives do they currently use?” The second formulation is researchable because it identifies information requirements.

A useful research brief contains:

• Decision context: the business situation and decision deadline.
• Decision alternatives: the realistic actions under consideration.
• Information gaps: what is not known or is disputed.
• Research objectives: specific knowledge outcomes.
• Research questions: questions that organize investigation.
• Success criteria: the evidence that would make a decision more defensible.
• Scope boundaries: population, geography, product category, time period, and competitors included.
• Constraints: budget, access, legal requirements, skills, and timing.

Research objectives should be specific and observable. “Understand the market” is too broad. “Estimate the proportion of target users who would consider adoption within six months” is more precise, although it still requires definitions of target users, consideration, and the six-month period.

A research question should not presuppose its answer. “Why do customers prefer our superior service?” is biased. “Which service attributes influence customers’ provider choice, and how do customers evaluate competing providers?” is more neutral.

2. Research Paradigms and Design Logic

A research design is the overall plan for obtaining and interpreting evidence. It reflects assumptions about reality, knowledge, and the role of the researcher.

A positivist or post-positivist orientation emphasizes observable indicators, measurement, systematic comparison, and probabilistic explanation. Surveys, experiments, transaction data, and structured competitor audits are commonly associated with this orientation. Post-positivism recognizes that measurements are imperfect and that conclusions are provisional rather than absolutely certain.

An interpretivist orientation emphasizes how people construct meanings and how behavior is shaped by context. Interviews, observation, diary studies, and ethnographic techniques are useful when the researcher needs to understand motivations, language, practices, or unarticulated problems.

A pragmatic orientation selects methods according to the decision and the nature of the question. It allows qualitative and quantitative evidence to be combined when each contributes distinct value. Entrepreneurship research frequently benefits from pragmatism because opportunity assessment involves both meaning—what problem customers experience—and measurement—how widespread, urgent, or economically valuable that problem is.

These orientations are not merely philosophical labels. They influence the role of the researcher, the type of evidence considered persuasive, the meaning of rigor, and the claims that can responsibly be made.

3. Major Research Designs

Exploratory Research

Exploratory research is appropriate when the problem is poorly defined, customer language is unknown, or important variables have not yet been identified. Its purpose is discovery rather than statistical estimation. Common methods include expert interviews, customer interviews, focus groups, observation, social listening, secondary research, and analysis of support complaints.

Exploration is especially valuable at the beginning of an entrepreneurial project. It may reveal that the assumed problem is not the real problem, that different segments describe the same need differently, or that a proposed solution conflicts with existing routines. Exploratory samples are usually purposive rather than statistically representative. The researcher seeks information-rich participants, not necessarily a miniature version of the whole market.

Descriptive Research

Descriptive research answers questions such as who, what, where, when, and how much. It can estimate market characteristics, segment profiles, awareness, usage frequency, satisfaction, purchase intentions, or competitor consideration. Surveys, structured observation, panels, and secondary datasets are common sources.

Descriptive research requires clear operational definitions and careful sampling. A large sample cannot correct a biased sampling frame or a poorly worded question. If a survey reaches only existing followers of a brand, it may describe followers accurately while saying little about the total target market.

Causal Research

Causal research examines whether changes in one factor produce changes in another. Experiments are the strongest design for causal inference because the researcher manipulates an independent variable and attempts to control alternative explanations. Examples include randomized tests of price displays, advertising messages, onboarding processes, or landing-page designs.

A basic causal claim requires more than correlation. The presumed cause must precede the outcome, the variables must covary, and plausible alternative explanations should be controlled or examined. In field settings, perfect control is rarely possible; therefore, conclusions should match the strength of the design.

Diagnostic and Evaluative Research

Diagnostic research investigates why a condition exists. For example, a decline in conversion may be examined through funnel analytics, usability testing, interviews, and competitor comparison. Evaluative research assesses whether a program, product, campaign, or intervention achieved its intended results. These designs are often combined with exploratory, descriptive, or causal methods.

4. Qualitative, Quantitative, and Mixed Methods

Qualitative Methods

Qualitative methods generate rich, contextual evidence. Interviews can explore individual experiences; focus groups can reveal shared language and disagreement; observation can expose behavior that participants do not report; and diary studies can capture experiences over time.

A high-quality interview guide moves from broad experience to focused evaluation. It should use probing questions rather than leading questions. Asking “Tell me about the last time you solved this problem” generally produces better evidence than asking “Would you use our proposed solution?” The former elicits behavior; the latter often elicits politeness or hypothetical intention.

Qualitative analysis commonly involves transcription, coding, category development, comparison across cases, and interpretation. Researchers should distinguish participants’ direct statements from the researcher’s interpretation. Saturation is a practical judgment that additional data are producing few substantially new themes; it is not a mechanical guarantee that all perspectives have been discovered.

Quantitative Methods

Quantitative methods measure variables numerically and examine distributions, associations, differences, or estimated effects. They are suitable when concepts can be operationalized and when the researcher needs estimates that can be compared across segments or over time.

A survey item should measure one concept at a time, use language familiar to respondents, provide balanced response options, and avoid unnecessary complexity. Scales should be selected deliberately. For example, agreement scales measure perceived agreement, whereas frequency scales measure behavior frequency. These are not interchangeable.

For a proportion estimated from a simple random sample, an approximate margin of error at confidence level associated with critical value z is:

Here, MOE is the margin of error, z is the critical value for the chosen confidence level, p is the estimated population proportion, and n is the sample size. The formula assumes a probability sample, independent observations, and a sufficiently large sample for the normal approximation. It does not correct for nonresponse bias, coverage error, measurement error, or a complex sample design. If no prior estimate of p is available, p = 0.5 gives the largest variance and therefore a conservative planning value.

For example, with p = 0.50, n = 400, and z ≈ 1.96 for a 95% confidence level:

The approximate margin of error is 4.9 percentage points under the stated assumptions. It should not be reported as though it describes a non-probability online sample.

Mixed-Method Designs

Mixed methods integrate qualitative and quantitative evidence in a planned sequence or structure. In an exploratory sequential design, qualitative work identifies concepts and language, followed by a survey that measures their prevalence. In an explanatory sequential design, quantitative results are collected first, followed by qualitative work to explain surprising or important patterns. In a convergent design, qualitative and quantitative evidence are collected in parallel and compared.

Integration must occur at the level of research questions, sampling, analysis, or interpretation. Simply placing interview quotations beside survey percentages does not create a mixed-method study. The researcher should state how one evidence stream informs the other and how disagreements will be handled.

5. Scope, Population, Sampling, and Measurement

Scope protects a project from uncontrolled expansion. A market study should specify the product category, customer population, geography, relevant competitors, decision horizon, and assumptions. Scope may be divided into:

• Inclusions: issues and populations directly addressed.
• Exclusions: issues intentionally omitted.
• Dependencies: information required from another source or project.
• Change-control rules: how new requests will be evaluated.

The target population is the full group to which the study intends to apply. The sampling frame is the operational list or mechanism from which participants are selected. These may differ substantially. For example, a database of registered users is not a frame for all potential category users.

Probability sampling includes simple random, systematic, stratified, and cluster sampling. It supports statistical generalization when the frame and implementation are adequate. Non-probability approaches include convenience, purposive, quota, snowball, and judgment sampling. These can be highly practical and appropriate for exploratory research, but population estimates should be presented cautiously.

Sample size is not determined by a universal rule. It depends on the decision risk, desired precision, population variability, subgroup analysis, expected response, design effect, and available resources. Qualitative sample adequacy depends on relevance, diversity of perspectives, depth, and saturation rather than a fixed numerical threshold.

Measurement requires operationalization. A construct such as trust may be represented through several indicators, such as perceived honesty, competence, security, and consistency. Researchers should pretest instruments to identify ambiguity, excessive burden, response-order effects, and technical problems.

6. The Market Research Process

A coherent process is iterative but can be represented as a sequence:

1. Define the decision and research problem.
2. Review existing internal and external evidence.
3. Specify objectives, questions, scope, and deliverables.
4. Select a research design and method combination.
5. Define the population, sampling approach, and measures.
6. Develop instruments and fieldwork protocols.
7. Pilot-test the design.
8. Collect and monitor data.
9. Clean, code, analyze, and integrate evidence.
10. Assess quality, limitations, and alternative explanations.
11. Communicate findings and implications.
12. Record decisions, data, and lessons for future research.

Secondary research should precede expensive primary research where possible. Useful sources include company records, public statistics, trade publications, regulatory filings, competitor websites, app-store reviews, academic studies, search trends, and industry reports. Secondary evidence must be assessed for purpose, date, definitions, sampling, sponsorship, and comparability.

A competitor analysis should examine more than visible features. Relevant dimensions may include target segment, customer problem, positioning, pricing logic, distribution, switching costs, evidence of traction, partnerships, capabilities, regulatory exposure, and likely strategic response. Public information is useful but incomplete; competitor claims should be distinguished from independently verified evidence.

7. Validity, Reliability, Credibility, and Triangulation

Validity concerns whether a study supports the interpretations and uses claimed by the researcher. Important forms include:

• Face validity: whether a measure appears relevant on inspection. It is useful but weak evidence.
• Content validity: whether the measure adequately covers the construct’s important dimensions.
• Construct validity: whether the measure behaves as theory predicts in relation to other variables.
• Criterion-related validity: whether a measure relates to a relevant external criterion.
• Internal validity: whether an observed effect can reasonably be attributed to the proposed cause rather than confounding factors.
• External validity: whether findings can be transferred or generalized to other populations, settings, or times.

Reliability concerns consistency of measurement. Test–retest reliability examines stability over time; inter-rater reliability examines agreement among coders or observers; and internal consistency examines whether items intended to measure the same construct produce coherent responses. Reliability does not guarantee validity. A consistently biased scale can be reliable but invalid.

In qualitative research, parallel concepts include credibility, dependability, confirmability, and transferability. Credibility may be strengthened through prolonged engagement, participant feedback, negative-case analysis, peer debriefing, and transparent documentation. Reflexivity requires the researcher to examine how their assumptions, relationships, and position may influence the study.

Triangulation compares evidence from different sources, methods, researchers, theories, or settings. For example, a claim about demand might be examined through interviews, search behavior, competitor sales proxies, and a survey. Convergence increases confidence, but disagreement is also informative. Different methods may reveal different aspects of a phenomenon rather than one method simply being wrong.

Triangulation should not be treated as a guarantee of truth. If all sources depend on the same flawed assumption, apparent agreement may be illusory. The researcher should document the logic of comparison and explain why evidence is considered convergent, complementary, or contradictory.

8. Timing, Budgeting, and Project Governance

A research schedule should identify dependencies rather than merely list activities. A typical plan includes briefing, secondary research, instrument development, recruitment, pilot testing, fieldwork, data processing, analysis, reporting, and decision review. The schedule should include contingency time for recruitment shortfalls, instrument revisions, poor data quality, and stakeholder review.

A budget can be organized into:

Budget decisions should be linked to decision value. A larger sample is not automatically better if the questionnaire is invalid or the frame is biased. Conversely, a low-cost study can be expensive in strategic terms if it creates false confidence and leads to a poor investment.

Governance should define who owns the data, who can access identifiable information, who approves changes, how quality is monitored, and what happens if results contradict sponsor expectations. A preregistered analysis plan is useful for confirmatory studies because it distinguishes planned tests from post hoc exploration.

9. Ethics, Privacy, and Responsible Use of Technology

Market researchers have obligations to participants, clients, organizations, and the public. Ethical practice includes informed consent, voluntary participation, the right to withdraw, minimization of harm, confidentiality, secure data handling, and honest reporting. Participants should understand the purpose of the study, expected duration, use of recordings, incentives, and relevant risks.

Researchers should collect only data necessary for the stated purpose. Personally identifiable information should be separated from analytic data where feasible, retained only as long as necessary, and protected through access controls. Applicable privacy and consumer-protection laws vary by jurisdiction; legal review is appropriate when collecting sensitive data, tracking online behavior, or transferring data across borders.

Artificial intelligence can assist with transcription, coding suggestions, survey drafting, synthetic data generation, sentiment classification, and competitor monitoring. However, AI outputs may contain fabricated information, biased classifications, privacy risks, and opaque decision rules. Human review, documentation of prompts and models, validation against known examples, and restrictions on uploading confidential data are essential. Synthetic respondents cannot substitute for evidence from real customers unless their limitations are explicit and the use is strictly exploratory.

Researchers must not manipulate participants, conceal sponsorship where disclosure is required, selectively report favorable findings, or present speculative competitor intelligence as fact.

10. Interpreting Findings and Communicating Uncertainty

A research report should distinguish observations, interpretations, assumptions, and recommendations. It should explain what the data show, what they do not show, and how uncertainty affects the decision. Percentages without denominators, charts without base definitions, and averages that conceal important segments can mislead decision makers.

A useful reporting structure is:

1. Decision and research objectives.
2. Method and sample, including exclusions.
3. Principal findings linked to evidence.
4. Segment or competitor differences.
5. Alternative explanations and limitations.
6. Implications for action.
7. Recommended next experiment or information-gathering step.

Entrepreneurs should avoid treating stated purchase intention as equivalent to future sales. Intentions may be affected by social desirability, lack of price realism, unfamiliarity with the product, and the difference between interest and willingness to switch. Behavioral evidence, observed choices, payment tests, retention, or conversion experiments may provide stronger evidence when feasible.

Illustrative Case: GreenRoute Refill Service

Background

GreenRoute, a fictional early-stage venture, is considering a household-refill delivery service in a metropolitan region. The founders believe environmentally concerned households will pay a monthly fee for reusable containers and scheduled replenishment.

Problem

The team initially proposes a large survey asking residents whether they would use the service. An adviser identifies three unresolved issues: the size of the genuinely reachable segment, the inconvenience customers will tolerate, and whether the economics can compete with supermarkets and existing delivery services.

Analysis and Approach

The team develops an exploratory sequential design. First, researchers review public waste data, competitor prices, delivery coverage, customer reviews, and internal landing-page analytics. They conduct 24 semi-structured interviews across apartment residents, families, and environmentally motivated professionals. Participants are asked about their most recent purchase and disposal experiences rather than being shown the concept immediately.

The interviews reveal that “sustainability” attracts attention, but convenience and product availability determine repeat behavior. Participants also distinguish between occasional bulk purchasing and a recurring service. These findings inform a survey measuring current purchasing frequency, acceptable delivery windows, container-return willingness, category priorities, price sensitivity, and competing alternatives.

The survey uses quotas for location and household type to avoid an overconcentration of environmentally engaged respondents recruited from sustainability groups. Because the sample is non-probability, the team avoids claiming a population margin of error. A prototype is then tested through a small randomized landing-page experiment: visitors see either an environmental-benefit message or a convenience-and-price message, with the same sign-up option. The experiment is interpreted as evidence about message response among visitors, not as proof of market-wide demand.

Outcome

The evidence suggests a narrower initial segment than the founders expected: households in dense neighborhoods with predictable consumption of a limited number of products. The convenience message generates more qualified sign-ups than the environmental message, while interviews explain why: customers support sustainability but will not tolerate missed delivery windows or complicated container returns. The team narrows the launch area, tests a smaller product assortment, and measures paid preorders and repeat orders before expanding.

Learning Points

The case illustrates four principles. First, the decision problem must be clarified before selecting a method. Second, qualitative evidence can improve the validity of a quantitative instrument. Third, a field experiment can examine a specific causal question without proving total market demand. Fourth, triangulation does not eliminate uncertainty; it makes assumptions and disagreements more visible and actionable.

Challenges and Limitations

Market research faces structural limitations. Respondents may forget, misunderstand, exaggerate, or answer strategically. Nonresponse can systematically exclude people with different views. Rapidly changing markets can make findings obsolete. Competitor information may be incomplete or intentionally promotional. Small entrepreneurial studies may lack resources for probability sampling or independent replication.

Researcher discretion also creates risks. Coding decisions can privilege expected themes, stakeholder pressure can influence questionnaire wording, and attractive visualizations can imply more precision than the design supports. These risks are reduced—not eliminated—by piloting, audit trails, independent review, transparent limitations, and explicit decision rules.

Research should therefore be treated as an input to judgment, not a replacement for judgment. The appropriate goal is not certainty but a better-informed, more reversible, and more testable decision.

Best-Practice Checklist

Before approving a market research plan, ask:

• Is the decision clearly stated?
• Are the research questions neutral and answerable?
• Is the target population defined?
• Does the sampling approach match the intended claim?
• Are measures operationalized and pretested?
• Do methods fit the stage and uncertainty of the venture?
• Are timing, dependencies, budget, and contingencies realistic?
• Are validity, reliability, and alternative explanations addressed?
• Is triangulation designed rather than merely asserted?
• Are privacy, consent, data security, and AI risks controlled?
• Are limitations and stopping or decision rules documented?
• Does the final deliverable connect evidence to an actionable decision?

Chapter Summary

Market research is a structured process for producing evidence relevant to an entrepreneurial decision. It begins by separating the managerial decision problem from the research problem and then defining objectives, questions, scope, and success criteria. Exploratory research develops understanding, descriptive research estimates characteristics, causal research examines effects, and diagnostic or evaluative research explains or assesses conditions.

Qualitative methods provide contextual depth, quantitative methods support measurement and comparison, and mixed methods integrate distinct forms of evidence. Sampling, operationalization, piloting, timing, budgeting, and governance determine whether a design is feasible and credible. Validity concerns the appropriateness of interpretations, while reliability concerns consistency. Qualitative quality is often discussed through credibility, dependability, confirmability, and transferability. Triangulation can strengthen interpretation but does not guarantee truth.

Ethical research protects participants, limits data collection, secures information, reports findings honestly, and uses AI with human oversight. The most useful research reports communicate uncertainty, limitations, alternative explanations, and practical implications. For entrepreneurs, rigorous research does not remove uncertainty; it converts vague uncertainty into explicit assumptions that can be tested through progressively stronger evidence.

Key Terms

Causal research: Research designed to assess whether one factor produces a change in another.

Construct validity: The extent to which a measure represents the theoretical concept it is intended to measure.

Decision problem: A managerial choice or action requiring evidence.

External validity: The extent to which findings apply beyond the study conditions.

Exploratory research: Research used to clarify poorly understood problems and discover relevant concepts.

Internal validity: The extent to which an observed effect can be attributed to the proposed cause.

Mixed methods: Planned integration of qualitative and quantitative approaches.

Nonresponse bias: Bias arising when people who do not participate differ systematically from participants.

Operationalization: Converting an abstract construct into observable indicators or measures.

Population: The full group to which a study intends its claims to apply.

Post-positivism: An approach emphasizing systematic evidence while recognizing that knowledge and measurement are fallible.

Pragmatism: An approach that selects methods according to the practical research question and decision.

Qualitative research: Research emphasizing meaning, context, experience, and interpretation.

Quantitative research: Research using numerical measurement and structured analysis.

Reliability: The consistency or stability of a measurement process.

Research design: The integrated plan for collecting and interpreting evidence.

Sampling frame: The operational list or mechanism from which a sample is selected.

Triangulation: Comparison of evidence across sources, methods, researchers, theories, or settings.

Validity: The degree to which evidence supports the interpretations and uses claimed.

Scope: The defined boundaries of a research project.

Review Questions

Short Answer Questions

1. What is the difference between a managerial decision problem and a research problem?
2. Distinguish exploratory, descriptive, and causal research.
3. Why does a large sample not automatically produce valid findings?
4. What is triangulation, and why can disagreement between sources be useful?
5. State two ethical responsibilities of a market researcher.

Long Answer Questions

1. Explain how a pragmatic research orientation can support entrepreneurial market research. Include circumstances in which qualitative and quantitative methods should be combined.
2. Discuss the relationship among validity, reliability, credibility, and generalizability in market research.
3. Design a research plan for evaluating a new digital service, including objectives, scope, sampling, methods, timing, and budget categories.
4. Explain why stated purchase intention may be weaker evidence than observed behavior. What methods could strengthen the evidence?
5. Evaluate the role of AI in market research, including its potential benefits, risks, and governance requirements.

Analytical / Application Questions

1. A founder wants to survey only social-media followers and use the results to estimate demand in the entire city. Critically assess the design and propose improvements.
2. Interviews suggest that customers value sustainability, but a randomized message test shows stronger response to convenience claims. How should the researcher interpret and investigate this apparent contradiction?
3. Prepare a concise mixed-method research plan for deciding whether an entrepreneur should launch a subscription pricing model. Specify how each method contributes to the decision and identify at least three limitations.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes a research problem from a decision problem?

A. A research problem identifies what must be learned to inform an action.

B. A research problem is always answered through a survey.

C. A research problem specifies the company’s preferred action.

D. A research problem eliminates managerial judgment.

Correct Answer: A. A research problem identifies what must be learned to inform an action.

Question 2

Which design is most appropriate when a venture does not yet understand the language customers use to describe a problem?

A. Causal research

B. Exploratory research

C. Descriptive research

D. Financial auditing

Correct Answer: B. Exploratory research

Question 3

Using MOE ≈ 1.96 × √(0.50 × 0.50 / 400), what is the approximate margin of error under the stated simple-random-sample assumptions?

A. 0.25 percentage points

B. 2.5 percentage points

C. 4.9 percentage points

D. 9.8 percentage points

Correct Answer: C. 4.9 percentage points

Question 4

Which statement about reliability and validity is correct?

A. Reliability guarantees validity.

B. Validity concerns consistency only.

C. A measure can be reliable but invalid.

D. Neither concept applies to qualitative research.

Correct Answer: C. A measure can be reliable but invalid.

Question 5

What is the principal purpose of a sampling frame?

A. To provide the operational basis from which participants are selected

B. To determine the research budget automatically

C. To guarantee that all respondents answer honestly

D. To replace the need for defining a population

Correct Answer: A. To provide the operational basis from which participants are selected

Question 6

In an exploratory sequential mixed-method design, what normally occurs first?

A. A randomized field experiment

B. Qualitative exploration

C. A final population estimate

D. Competitor acquisition

Correct Answer: B. Qualitative exploration

Question 7

Which practice most directly strengthens causal inference?

A. Using only enthusiastic volunteers

B. Manipulating an independent variable while controlling alternative explanations

C. Increasing the number of interview questions

D. Reporting only statistically significant findings

Correct Answer: B. Manipulating an independent variable while controlling alternative explanations

Question 8

Why should a non-probability online sample generally not be reported with a conventional population margin of error?

A. Online samples are always too small.

B. Non-probability selection does not support the assumptions behind that estimate.

C. Margins of error apply only to qualitative research.

D. The calculation requires competitor data.

Correct Answer: B. Non-probability selection does not support the assumptions behind that estimate.

Question 9

Which is an example of triangulation?

A. Repeating the same leading question three times

B. Comparing interviews, behavioral data, and survey results about the same issue

C. Increasing a budget after data collection

D. Removing findings that disagree with the sponsor

Correct Answer: B. Comparing interviews, behavioral data, and survey results about the same issue

Question 10

Which action best reflects responsible use of AI in market research?

A. Uploading confidential participant data to any public tool

B. Treating generated summaries as verified evidence

C. Using human review and validating AI outputs against known examples

D. Replacing all customer research with synthetic respondents

Correct Answer: C. Using human review and validating AI outputs against known examples