Marketing Information and Research

Introduction

Entrepreneurial marketing depends on decisions made under uncertainty. An entrepreneur must decide which customers to serve, what problem to solve, how much customers may pay, which competitors matter, and whether a proposed offering can achieve sustainable demand. Intuition, personal experience, and enthusiasm may generate ideas, but they do not reliably establish market opportunity. Marketing information and research provide disciplined ways to reduce uncertainty and convert observations into evidence for action.

Marketing research is the systematic design, collection, analysis, interpretation, and communication of information relevant to a specific marketing decision. It is not limited to surveys. It includes problem definition, desk research, observation, interviews, experiments, customer discovery, sampling, data analysis, validation, and ethical governance. Good research does not promise certainty. Rather, it improves the quality of decisions by making assumptions visible and testing the assumptions that matter most.

For an entrepreneurial venture, research should be proportionate to the decision. A founder exploring an early idea may need ten carefully selected customer conversations and direct observation before investing in a large survey. A firm estimating national demand may require probability sampling, structured measurement, and statistical analysis. The appropriate method depends on the research question, the stage of the venture, available resources, and the consequences of error.

Learning Objectives

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

1. Explain the role of marketing information in entrepreneurial decision-making.
2. Define a research problem, decision problem, and research objective.
3. Distinguish secondary from primary data and qualitative from quantitative research.
4. Design a practical research process involving customer discovery and validation.
5. Select an appropriate sampling approach and recognize sampling limitations.
6. Construct effective questionnaires and interview guides.
7. Interpret basic research findings without overstating what the evidence shows.
8. Apply ethical, legal, and responsible principles to data collection and use.

1. Marketing Information as a Decision Resource

A marketing information system is an organized arrangement of people, processes, technologies, and records used to identify information needs, collect relevant data, analyze it, and distribute useful intelligence to decision-makers. It commonly draws on four sources:

• Internal records: sales, costs, customer service contacts, website activity, returns, and inventory data.
• Marketing intelligence: ongoing observations of competitors, distributors, regulation, technology, and market trends.
• Marketing research: purpose-designed studies addressing a particular decision.
• Analytical tools: methods that transform data into descriptions, forecasts, segments, or recommendations.

Data are recorded observations, such as a transaction date or an interview statement. Information is data organized and interpreted for a purpose. Insight is a defensible understanding of what the information means for a decision. These terms should not be treated as synonyms. A dashboard containing thousands of numbers may provide data without producing insight if the decision and interpretation are unclear.

Research quality has several dimensions:

• Relevance: Does the evidence address the actual decision?
• Validity: Does the method measure the intended concept?
• Reliability: Would a suitable repetition produce reasonably consistent results?
• Timeliness: Is the evidence current enough for the decision?
• Representativeness: Do the observed cases reasonably reflect the target population?
• Actionability: Can a manager use the findings to choose among alternatives?
• Cost-effectiveness: Is the value of better information worth the resources used?

A precise measurement of the wrong population can be less useful than an imperfect but relevant study of the right population. Research should therefore begin with the decision, not with a preferred method such as “conduct a survey.”

2. From Management Problem to Research Problem

A management decision problem asks what the organization should do. For example: “Should the venture launch a subscription meal service for university students?” A research problem asks what must be learned to make that decision: “What unmet meal-planning needs, usage occasions, price perceptions, and adoption barriers characterize the relevant student segments?”

A practical problem-definition sequence is:

1. Identify the decision-maker and the decision to be made.
2. Describe the symptoms or trigger, such as declining repeat purchase or unexpected competitor entry.
3. Separate symptoms from possible causes.
4. State the population, market, or customer group of interest.
5. Identify information gaps and competing explanations.
6. Specify the decisions that the research will inform.
7. Establish boundaries, resources, timing, and ethical constraints.

A useful research brief includes the background, decision statement, research questions, objectives, target population, proposed methods, deliverables, timetable, budget, and limitations. Objectives should use observable verbs such as estimate, compare, explore, identify, measure, or test.

For example, a weak objective is “Understand students.” A stronger set is:

• Explore how students currently obtain affordable evening meals.
• Identify the attributes that influence selection among alternatives.
• Estimate the proportion of a defined student population that would consider a weekly subscription.
• Test whether a prototype improves perceived convenience without reducing trust.

Research questions may be exploratory, descriptive, or causal. Exploratory questions investigate meanings and possibilities. Descriptive questions estimate characteristics, frequencies, or distributions. Causal questions examine whether changing one factor produces a change in another. Causal claims require stronger designs than simple observation.

3. The Research Process

A flexible research process contains the following stages:

1. Define the decision and problem.
2. Conduct preliminary secondary research.
3. Formulate objectives, research questions, and, where suitable, hypotheses.
4. Choose a research design and sources of evidence.
5. Define the target population and sampling plan.
6. Design instruments such as interview guides, observation protocols, or questionnaires.
7. Pilot-test the instruments and procedures.
8. Collect data ethically and consistently.
9. Clean, code, analyze, and interpret the evidence.
10. Communicate findings, limitations, and recommendations.
11. Validate the recommended action through a market test or experiment where feasible.

The stages are iterative rather than strictly linear. Early interviews may reveal that the original problem is incorrectly defined. A pilot survey may expose ambiguous wording. Analysis may show that an important segment was omitted, requiring additional research.

A research proposal should distinguish method from evidence standard. The method describes how information will be collected; the evidence standard describes what result would be persuasive enough to support a decision. Before collecting data, an entrepreneur should ask: “What finding would cause us to continue, modify, or stop?” This reduces confirmation bias.

4. Secondary Data and Desk Research

Secondary data were collected previously for a purpose other than the current study. Internal sales records, government statistics, academic articles, trade reports, competitor websites, app-store reviews, and public regulatory documents are examples. Primary data are collected specifically for the current research problem.

Secondary research is often the first step because it is relatively fast and inexpensive. It can clarify market size, terminology, competitors, regulations, demographic patterns, and previous findings. It may also prevent duplication of existing work.

Sources should be evaluated for:

• Authority: Who produced the information and with what expertise?
• Purpose: Was it produced for public knowledge, advocacy, promotion, or sales?
• Method: How were data defined, collected, measured, and analyzed?
• Coverage: Which geography, period, population, and channels are included?
• Currency: When were the data collected and updated?
• Comparability: Are definitions and units consistent across sources?
• Bias and conflicts: Could incentives have influenced the findings?

Search-engine rankings are not evidence of source quality. A commercial report may be useful, but its sample, definitions, and sponsorship should be examined. Online reviews are valuable for discovering complaints and language used by customers, but they are usually self-selected and should not be treated as representative estimates of all customers.

A market-size estimate should state its basis. Top-down estimation begins with a broad published market and applies defensible filters. Bottom-up estimation builds from operating assumptions, such as potential outlets × expected customers per outlet × purchase frequency × price. Both are estimates, not facts. Assumptions should be displayed and tested with sensitivity analysis.

5. Primary Research: Qualitative and Quantitative Approaches

Qualitative research

Qualitative research seeks depth, meaning, context, language, and process. Common methods include semi-structured interviews, focus groups, ethnographic observation, diary studies, usability tests, and analysis of naturally occurring conversations. It is particularly useful when the problem is poorly understood, when motivations are difficult to measure directly, or when researchers need to develop concepts and questionnaire items.

A strong qualitative interview uses open prompts, probes for concrete episodes, and avoids leading the participant. “Tell me about the last time you solved this problem” is generally more informative than “Would you like a faster solution?” Researchers should distinguish what participants say they might do from what they have actually done. Interviews can reveal perceived needs, but behavior, payment, and repeated use provide stronger validation.

Qualitative samples are commonly purposive rather than statistically representative. Participants may be selected because they have relevant experience, belong to a particular segment, or can illuminate contrasting cases. Sample adequacy depends on the research purpose and information richness, not on a universal numerical rule. Researchers may stop when additional interviews produce little new insight, while remaining alert to overlooked segments and minority experiences.

Analysis normally involves transcription or detailed notes, coding, comparison across cases, theme development, and interpretation. Researchers should preserve contradictory evidence rather than selecting only quotations that support the preferred idea.

Quantitative research

Quantitative research measures variables numerically and uses structured procedures to describe patterns, estimate proportions, compare groups, or test relationships. Surveys, transaction analysis, controlled experiments, and structured observation are common approaches.

Descriptive statistics may include counts, percentages, means, medians, and measures of variation. A percentage is calculated as:

The part and total must be measured on compatible bases, and the denominator must be stated. For example, if 84 of 240 eligible respondents select an option, the percentage is (84 ÷ 240) × 100 = 35%. This result describes the respondents or sample unless the sampling design supports inference to a wider population.

The arithmetic mean for n observations is:

Here, xᵢ is observation i, n is the number of observations, and x̄ is the sample mean. The mean is sensitive to extreme values; the median may better represent a skewed distribution such as spending.

Quantitative research can estimate associations, but association alone does not establish causation. A higher purchase rate among people exposed to an advertisement may reflect pre-existing differences rather than the advertisement’s effect. Stronger causal evidence requires controls such as random assignment, a comparison group, consistent exposure, and a clearly defined outcome.

Mixed methods

Mixed-methods research combines qualitative and quantitative evidence. A sequential design may begin with interviews to discover relevant language and attributes, followed by a survey to measure their prevalence. Alternatively, a survey may identify an unexpected pattern that is explored through follow-up interviews. Integration should occur during interpretation: researchers should explain how the findings agree, differ, or answer different parts of the problem.

6. Customer Discovery and Validation

Customer discovery is an iterative process for investigating whether a defined customer group experiences a meaningful problem and how it currently handles that problem. It is not primarily a sales presentation. The researcher seeks evidence about actual behavior, context, alternatives, costs, frustrations, and decision authority.

Useful discovery questions include:

• When did you last encounter this problem?
• What did you do, step by step?
• What was difficult, costly, or risky?
• What alternatives did you consider?
• Who else influences the decision?
• Have you paid for a solution or invested time to create one?
• What would prevent you from changing your current approach?

Avoid asking hypothetical questions that invite politeness, such as “Would you buy this?” A stated intention may be useful as one indicator but is weaker than evidence of recent behavior, payment, repeated use, referral, reservation, or willingness to join a realistic trial.

Validation converts a proposition into testable assumptions. A venture may need to test:

• Desirability: Do target customers recognize the problem and value the proposed outcome?
• Feasibility: Can the organization deliver the solution reliably?
• Viability: Can revenue, cost, and retention support a sustainable model?
• Compliance and trust: Can the offering meet legal, safety, privacy, and credibility requirements?

A minimum viable product or service should test the riskiest assumption with the smallest credible experiment. Examples include a landing-page test, a manual service delivered before automation, a prototype usability session, a limited geographic pilot, or a paid preorder. A successful click or sign-up is not automatically proof of a viable business; the test must define the action, audience, conditions, and threshold in advance.

7. Sampling: Choosing Whom to Study

The population is the complete set of units relevant to the research question. A sampling frame is the operational list or procedure from which units are selected. A sample is the subset actually studied. The unit of analysis is what is analyzed, such as an individual, household, business, transaction, or store.

Probability sampling gives units a known, non-zero chance of selection. Its main forms are:

• Simple random sampling: units are selected randomly from a complete frame.
• Systematic sampling: every kth unit is selected after a random starting point.
• Stratified sampling: the population is divided into relevant strata, and samples are drawn within them.
• Cluster sampling: naturally occurring groups are selected, often to reduce fieldwork costs.

Non-probability sampling does not provide known selection probabilities. Convenience sampling uses accessible participants; purposive sampling selects information-rich cases; quota sampling fills specified category totals; and snowball sampling uses participant referrals. These approaches can be practical for discovery but may not support population estimates without strong assumptions.

Sample size is not determined by population size alone. It depends on the desired precision, confidence level, variability, design, budget, and expected response rate. For a simple random sample estimating a proportion, a commonly used planning expression is:

Here, n₀ is the initial required sample size, z is the standard-normal critical value associated with the chosen confidence level, p is the anticipated population proportion, and e is the desired maximum margin of error expressed as a proportion. This formula assumes a probability sample, a sufficiently large population, independent observations, and a proportion estimate under the simple-random-sampling model. If no credible prior estimate of p exists, p = 0.5 is conservative because it produces the largest product p(1−p). Nonresponse and complex sampling require further adjustment.

For illustration, with 95% confidence, z ≈ 1.96, p = 0.5, and e = 0.05:

n₀ ≈ (1.96² × 0.5 × 0.5) ÷ 0.05² ≈ 384.16.

The result is normally rounded up to 385 completed responses under the stated assumptions. This is not a guarantee that the survey is unbiased: coverage error, nonresponse, poor wording, and dishonest or inattentive responses can remain.

Researchers should report who was eligible, how participants were recruited, how many were contacted, how many completed the study, and which groups may be missing. A large convenience sample is not automatically more representative than a smaller probability sample.

8. Questionnaire and Instrument Design

A questionnaire translates concepts into observable measures. Begin with the information needed, not with interesting questions. Each item should have a clear purpose and should contribute to a decision or analysis.

Good questionnaire practice includes:

• Use simple, specific language.
• Ask one thing at a time; avoid double-barrelled questions.
• Define the time period, population, and terms.
• Provide balanced and mutually exclusive response options where appropriate.
• Include “not applicable,” “don’t know,” or “prefer not to answer” when meaningful.
• Avoid leading, loaded, prestige, and emotionally pressuring wording.
• Place screening and eligibility questions early.
• Use a logical sequence: introduction, screening, behavior, evaluation, classification, and sensitive questions.
• Keep the instrument as short as the objective permits.
• Randomize response options when order effects are plausible, while retaining logical scales.
• Pilot-test comprehension, completion time, routing, and data coding.

A question such as “How satisfied are you with our affordable and convenient service?” contains at least two attributes and cannot show which one drives the response. It should be separated into distinct items. Frequency questions should offer a defined period, such as “during the past 30 days.” Rating scales should be balanced and consistently ordered. Researchers must not imply that a scale is an interval measurement if the response categories merely represent ordered judgments.

Interview guides should similarly include a purpose, opening script, consent language, core questions, probes, closing questions, and a recording plan. The interviewer should avoid correcting participants or signaling the desired answer.

9. Data Quality, Analysis, and Interpretation

Before analysis, researchers should check completeness, duplicate records, impossible values, inconsistent coding, straight-line responses, and unusual completion times. Cleaning rules should be documented rather than silently changing inconvenient observations.

A basic analysis plan specifies variables, coding, subgroup comparisons, and decision criteria before examining results where feasible. Cross-tabulation can compare responses across segments, but small subgroup counts may make percentages unstable or disclose individuals. Missing data should be reported and handled consistently; deleting cases can bias results if missingness is systematic.

Interpretation should distinguish:

• What was directly observed.
• What is inferred from the evidence.
• What remains uncertain.
• What action is recommended and why.

Researchers should not claim that a sample “proves” universal demand, that correlation proves causation, or that positive intentions equal purchase. Confidence intervals, significance tests, and predictive models can be useful in appropriate studies, but statistical sophistication cannot correct a biased frame or a poorly defined construct. Practical significance also matters: a small, statistically detectable difference may not justify a costly managerial action.

A decision report should include an executive conclusion, method, sample, principal findings, segment or behavioral differences, limitations, implications, and recommended next steps. Charts should label denominators, units, dates, and filters. Selective quotation and attractive but misleading graphics undermine trust.

10. Ethics, Privacy, and Responsible Data Collection

Ethical research respects autonomy, privacy, fairness, and participant welfare. Core practices include informed consent, voluntary participation, the right to withdraw where feasible, honest purpose statements, proportionate incentives, confidentiality, secure storage, restricted access, and responsible disposal.

Participants should know what information is collected, why it is needed, how it will be used, whether it will be shared, and whom to contact with concerns. Sensitive data require stronger justification and safeguards. Researchers should collect only what is necessary, avoid unnecessary identifiers, and separate contact details from research responses when possible.

Legal requirements vary by jurisdiction, but responsible practice commonly includes a lawful basis for processing personal data, transparency notices, access and correction rights, retention limits, security controls, and special protections for children and vulnerable groups. Publicly accessible information is not automatically ethical to repurpose in a new context. Scraping, facial recognition, location tracking, and profiling can create significant privacy and fairness risks.

Entrepreneurs must also avoid deceptive research. A disguised sales pitch presented as neutral research compromises consent and data quality. Incentives should compensate time without coercing participation. Conflicts of interest, sponsorship, limitations, and adverse findings should be disclosed.

11. Technology and Artificial Intelligence in Marketing Research

Digital analytics, social listening, online experiments, automated transcription, natural-language processing, and AI-assisted coding can reduce cost and increase speed. They can help identify themes, summarize large text collections, personalize recruitment, and detect patterns in customer behavior.

However, automation introduces risks. Models may reproduce historical bias, infer sensitive attributes, hallucinate summaries, misclassify sarcasm, or obscure how conclusions were produced. Data obtained from platforms may not represent customers who are offline or less active. AI-generated research participants, synthetic responses, or fabricated quotations must never be presented as human evidence. Researchers should preserve source records, validate automated coding against human review, document prompts and transformations, and disclose material use of AI.

Technology should support, not replace, problem definition and judgment. A fast analysis of irrelevant or biased data remains irrelevant or biased.

12. Illustrative Case: GreenBite Campus Meals

Background

GreenBite, a fictional early-stage venture, proposed a weekly subscription supplying affordable prepared meals to university students. The founders believed that students wanted healthier food and planned to launch across three campuses.

Problem

The founders initially asked, “How can we convince students to subscribe?” This framed the issue as promotion before establishing whether a sufficiently important problem existed. A revised decision problem was: “Should GreenBite pilot a subscription service, and for which student segment and usage occasion?”

Analysis and approach

The team first reviewed campus dining reports, local competitors, delivery prices, and food-safety requirements. It then conducted 18 semi-structured interviews selected to include commuters, residential students, students with dietary restrictions, and students who regularly purchased prepared meals. Interviewers asked about the last three occasions on which participants obtained evening food, rather than asking whether they liked the proposed service.

The interviews revealed three distinct situations: students needing a meal between classes, students seeking late-night convenience, and students planning several meals in advance. The subscription concept appealed mainly to the third group. Several participants described previous subscription failures caused by inflexible pickup times and unused meals.

The team created a short survey using language from the interviews. Among 240 eligible respondents recruited through a stratified campus intercept plan, 84 indicated that they had purchased prepared evening meals at least four times in the previous month. Of these frequent purchasers, 35% reported that a flexible multi-meal option would be worth trying. The denominator was reported explicitly; the result was not presented as the percentage of all students.

GreenBite then ran a two-week paid pilot with 30 volunteers from the relevant segment. Instead of building an app, staff used a simple ordering form and offered two pickup windows. The pilot measured paid orders, pickup completion, repeat ordering, complaints, and food waste. Participants also completed a brief follow-up interview.

Outcome

The pilot showed that demand existed, but the original weekly subscription was too rigid. Customers preferred purchasing three meals with the ability to skip a week. The team revised the offer, limited the first launch to one campus, and negotiated pickup locations before investing in software.

Learning points

The case illustrates that customer discovery should investigate actual behavior, that survey percentages require clear denominators, and that payment and repeated use provide stronger validation than general enthusiasm. It also shows the value of a staged research design: secondary research reduced avoidable uncertainty, interviews clarified the problem, a survey measured its prevalence, and a paid pilot tested behavior under realistic conditions.

Challenges and Limitations

Marketing research cannot eliminate uncertainty. Markets change, competitors respond, and customers may behave differently when money, time, or reputation is at stake. Research participants may forget, misunderstand, exaggerate, or provide socially desirable answers. Samples may exclude non-users, silent customers, low-connectivity groups, or people who distrust researchers. Small entrepreneurial studies often have limited statistical power and cannot support broad generalization.

Researchers should therefore use triangulation, transparent assumptions, pilots, and repeated learning cycles. They should be especially cautious when evidence is collected from friends, followers, existing customers, or incentivized online panels. The appropriate conclusion may be “the proposition merits a controlled pilot,” not “the market is proven.”

Best-Practice Checklist

Before approving a marketing research project, ask:

1. What decision will the research inform?
2. What assumptions are most dangerous if wrong?
3. Is the target population clearly defined?
4. Can existing evidence answer part of the question?
5. Does the method match the type of evidence required?
6. Are recruitment and sampling limitations explicit?
7. Have instruments been piloted?
8. Are behavioral measures included alongside opinions?
9. What result would change the decision?
10. Are consent, privacy, security, and disclosure requirements satisfied?
11. Are findings reported with denominators and limitations?
12. What is the next validation step?

Chapter Summary

Marketing information supports entrepreneurial decisions by turning uncertain assumptions into structured evidence. Effective research begins with a clearly defined management decision and research problem. Secondary data provide efficient context, while primary qualitative research explores meanings and problems and quantitative research measures patterns and differences. Mixed methods can connect discovery with estimation.

Customer discovery should focus on recent behavior, existing alternatives, costs, and barriers rather than flattering hypothetical intentions. Validation tests desirability, feasibility, viability, and trust through realistic experiments such as prototypes, paid pilots, or preorders. Sampling determines whose evidence is observed and whether conclusions can reasonably extend beyond participants. Questionnaire quality depends on clear wording, appropriate response options, logical sequencing, and pilot testing.

Analysis must distinguish observation from inference and association from causation. Ethical research requires informed participation, privacy protection, data minimization, security, transparency, and fair treatment. Technology and AI can improve efficiency but cannot substitute for sound problem definition, representative evidence, human judgment, and accountability.

Key Terms

Marketing research: Systematic collection and analysis of information for a marketing decision.

Marketing information system: People, processes, technologies, and records that produce useful marketing information.

Decision problem: The managerial choice the organization must make.

Research problem: The information issue that must be addressed to support the decision.

Secondary data: Existing data collected previously for another purpose.

Primary data: Data collected specifically for the current research problem.

Qualitative research: Research focused on meanings, experiences, context, and processes.

Quantitative research: Research using structured numerical measurement and analysis.

Mixed methods: Combination of qualitative and quantitative approaches.

Customer discovery: Investigation of customer problems, behavior, alternatives, and context.

Validation: Evidence-based testing of assumptions about desirability, feasibility, and viability.

Population: Complete set of units relevant to a study.

Sampling frame: Operational list or procedure from which a sample is selected.

Probability sampling: Sampling in which units have known, non-zero selection probabilities.

Non-probability sampling: Sampling without known selection probabilities.

Validity: Degree to which a method measures what it is intended to measure.

Reliability: Consistency of a measurement or procedure.

Margin of error: Planned maximum sampling error around an estimate under stated assumptions.

Informed consent: Voluntary agreement based on adequate information about participation.

Triangulation: Use of multiple methods or sources to examine a research issue.

Review Questions

Short Answer Questions

1. Distinguish a management decision problem from a research problem.
2. Give two examples of secondary data relevant to an entrepreneurial venture.
3. Why are recent behavioral questions often more useful than hypothetical purchase questions?
4. What is the difference between a population and a sampling frame?
5. State three ethical requirements for collecting personal data.

Long Answer Questions

1. Explain the stages of a practical marketing research process and show how the stages may be iterative.
2. Compare qualitative and quantitative research in terms of purpose, methods, sampling, and limitations.
3. Explain how customer discovery and validation contribute to entrepreneurial marketing decisions.
4. Discuss the main sources of sampling error and non-sampling error in a marketing survey.
5. Evaluate how AI and digital analytics can improve marketing research while creating ethical and methodological risks.

Analytical / Application Questions

1. A startup wants to introduce a reusable packaging service but has only asked friends whether they “like the idea.” Design a staged research plan that includes secondary research, customer discovery, sampling, and a validation experiment.
2. A survey reports that 62% of respondents would subscribe to a new service, but the sample was recruited from the venture’s social-media followers. Analyze what can and cannot be concluded, and recommend additional evidence.
3. Using the GreenBite case, propose two changes to the research design that would improve confidence in the launch decision while maintaining a limited budget.

Multiple Choice Questions with Answers

Question 1

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

A. A decision problem asks what should be done; a research problem asks what must be learned.

B. A decision problem uses qualitative data; a research problem uses quantitative data.

C. A decision problem concerns competitors; a research problem concerns customers only.

D. A decision problem is always broader than the market; a research problem is always narrower.

Correct Answer: A. A decision problem asks what should be done; a research problem asks what must be learned.

Question 2

Which is an example of secondary data?

A. A newly conducted interview with a target customer

B. A government report on household spending

C. A prototype usability test conducted by the venture

D. A survey designed for the current launch decision

Correct Answer: B. A government report on household spending

Question 3

What is a major purpose of customer discovery?

A. To persuade every interviewee to purchase immediately

B. To investigate actual problems, behaviors, and existing alternatives

C. To replace all quantitative research

D. To guarantee that a product will be profitable

Correct Answer: B. To investigate actual problems, behaviors, and existing alternatives

Question 4

Which sampling method gives eligible units known, non-zero probabilities of selection?

A. Convenience sampling

B. Snowball sampling

C. Probability sampling

D. Judgment sampling

Correct Answer: C. Probability sampling

Question 5

A survey obtains 84 relevant responses from 240 respondents. What percentage selected the relevant response?

A. 28%

B. 30%

C. 35%

D. 42%

Correct Answer: C. 35%

Question 6

Which question is double-barrelled?

A. How many times did you purchase the product in the past 30 days?

B. Which of these brands did you purchase most recently?

C. How satisfied are you with the service’s speed and reliability?

D. Who usually makes the purchase decision?

Correct Answer: C. How satisfied are you with the service’s speed and reliability?

Question 7

Why is a paid pilot generally stronger validation than a question asking whether customers would buy?

A. It eliminates all sampling error.

B. It observes behavior under more realistic conditions.

C. It guarantees repeat purchase.

D. It requires no ethical safeguards.

Correct Answer: B. It observes behavior under more realistic conditions.

Question 8

Which statement about correlation is correct?

A. Correlation always proves causation.

B. Correlation can indicate association but does not alone establish causation.

C. Correlation is relevant only to qualitative research.

D. Correlation cannot be calculated from numerical data.

Correct Answer: B. Correlation can indicate association but does not alone establish causation.

Question 9

If no credible prior estimate of a population proportion exists, why is p = 0.5 often used in planning a simple random sample?

A. It produces the smallest possible sample.

B. It assumes that exactly half the population will respond.

C. It conservatively maximizes p(1 − p) for the proportion formula.

D. It removes the need to define a population.

Correct Answer: C. It conservatively maximizes p(1 − p) for the proportion formula.

Question 10

Which practice best reflects ethical data collection?

A. Collecting every available identifier in case it becomes useful

B. Presenting a sales pitch as anonymous research

C. Using informed consent, data minimization, and secure storage

D. Removing adverse findings from the final report

Correct Answer: C. Using informed consent, data minimization, and secure storage