Research Problems, Questions and Decision Contexts

Introduction

Entrepreneurial decisions are often expressed in action-oriented language: whether to launch, which customer segment to serve, how much to charge, whether to expand, or how to respond to a competitor. Market research begins when these decisions are translated into questions that can be answered with credible evidence. This translation is not a clerical step. It determines what information is collected, from whom, by which method, at what cost, and with what degree of confidence.

A poorly framed research problem can produce technically sophisticated but managerially useless data. For example, a survey may establish that many respondents “like” a proposed product, yet fail to determine whether they will pay for it, whether the target segment is reachable, or whether the venture can profitably serve it. Conversely, a well-framed problem connects a decision maker’s uncertainty to specific information requirements and an appropriate research design.

This chapter develops a disciplined process for moving from a decision context to a management problem, research problem, objectives, questions, hypotheses, and evidence requirements. It also distinguishes exploratory, descriptive, and causal research and shows how the three designs may be combined in entrepreneurial practice.

Learning Objectives

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

1. Distinguish a management decision problem from a research problem.
2. Diagnose the decision context, uncertainty, constraints, and stakeholders surrounding an entrepreneurial decision.
3. Formulate research objectives, research questions, and testable hypotheses.
4. Translate questions into information requirements and measurable constructs.
5. Select among exploratory, descriptive, and causal research designs.
6. Evaluate the suitability, limitations, and ethical implications of a proposed research plan.
7. Apply the framework to a new-venture market-entry decision.

1. The Decision Context: Research Begins Before Data Collection

A decision context is the set of circumstances in which an organization must choose among alternatives. It includes the decision maker, the available options, the consequences of action or inaction, relevant stakeholders, time and budget constraints, and the uncertainty surrounding the outcome.

Entrepreneurial decisions are especially information-intensive because founders often operate with limited history, incomplete customer knowledge, changing markets, and resource constraints. A decision context may involve:

• launching a new product;
• selecting a target market;
• setting or changing a price;
• choosing a distribution channel;
• evaluating brand positioning;
• forecasting demand;
• responding to a competitor;
• deciding whether to pivot or discontinue an offering.

Research does not eliminate uncertainty. Its purpose is to reduce decision-relevant uncertainty sufficiently to improve the quality of a choice. The value of research therefore depends not only on methodological quality but also on whether the information is relevant to a decision that can actually be made.

1.1 Decision alternatives and information value

A research project should identify the alternatives under consideration. “Understand the market” is not an adequate decision statement because it does not identify a possible action. A stronger statement might be: “Should the venture launch the service nationally, begin with two urban districts, or postpone launch until retention improves?”

Research is most useful when different findings would lead to different actions. If management would pursue the same action regardless of the result, collecting additional information may have little practical value. This principle is sometimes described as the value-of-information perspective: information is valuable when it can change a consequential decision and improve expected outcomes, net of research cost and delay.

1.2 A diagnostic map of the decision context

Before writing a research question, the researcher should document:

This map prevents a common error: treating every interesting market fact as equally important. Research should prioritize uncertainties that are both consequential and reducible through evidence.

2. From Management Problem to Research Problem

The management problem concerns what the organization should do. The research problem concerns what must be learned to support that decision.

For example:

• Management problem: Should a meal-delivery start-up introduce a low-cost subscription for university students?
• Research problem: Estimate the size and characteristics of the relevant student segment, identify the factors affecting subscription adoption, assess willingness to pay, and determine whether a low-cost subscription would improve acquisition without producing unacceptable churn or margin loss.

The first statement requires a decision. The second specifies information that research can generate.

2.1 The problem-definition chain

A useful chain is:

Decision context → management problem → research problem → objectives → questions or hypotheses → information requirements → design and measures → analysis → decision.

Each link should be traceable to the preceding one. If a proposed survey item cannot be linked to an objective, it may be unnecessary. If an objective does not help answer the management problem, it may be scope drift.

2.2 Symptoms are not problems

Managers often present symptoms: declining sales, low website conversion, high customer complaints, or weak repeat purchase. A symptom is an observable outcome, not necessarily its cause. Declining sales may result from lower awareness, distribution failure, product dissatisfaction, price, seasonality, competitor entry, or measurement error.

The research task is to distinguish the symptom from the underlying problem. A diagnostic framework should ask:

1. What changed, and relative to which baseline?
2. When did the change occur?
3. Which customer, product, channel, or geographic groups are affected?
4. What plausible explanations are consistent with the evidence?
5. Which explanations can be investigated with available methods?

Prematurely assuming a cause can bias the research toward confirming management’s preferred explanation.

2.3 Scope and unit of analysis

A research problem should specify the unit of analysis: an individual customer, household, business account, transaction, store, geographic market, or competitor. Confusing units can produce invalid conclusions. For example, average purchase frequency by city does not establish that every customer in that city purchases frequently. Similarly, competitor-level market share data cannot directly explain individual customer motivations.

The population and boundary of the study should also be stated. “Young consumers” is vague; “residents aged 18–29 who purchased prepared meals at least twice in the past month in the selected metropolitan area” is more operational, although it may still require refinement.

3. Objectives, Questions, and Hypotheses

3.1 Research objectives

A research objective states what the study will accomplish. Good objectives use observable verbs such as estimate, describe, compare, assess, identify, explain, test, or evaluate. They should be specific enough to guide method selection.

Weak objective: “Study customer attitudes.”

Stronger objectives:

• estimate the proportion of target customers who recognize the brand;
• identify the attributes used to evaluate competing services;
• compare stated subscription interest across price levels;
• test whether delivery reliability predicts repeat purchase after controlling for usage frequency.

Objectives may be primary or secondary. A primary objective directly informs the decision; secondary objectives provide supporting diagnosis or segmentation.

3.2 Research questions

A research question is a focused question that can be answered through evidence. It should identify the population, constructs, relationship, comparison, or time frame where relevant.

Common forms include:

• What is happening? What proportion of target users currently use the category?
• Who differs? Which customer segments have the highest unmet need?
• Why or how? Why do prospective users reject the current solution?
• What is related? Is perceived convenience associated with adoption intention?
• What causes change? Does a free-trial offer increase completed first purchases?
• What should be chosen? Which positioning statement produces the strongest qualified response?

“Do customers like the idea?” is usually too broad. It should be decomposed into perceived usefulness, perceived risk, purchase intention, willingness to pay, and barriers to adoption.

3.3 Hypotheses

A hypothesis is a provisional, empirically testable statement about a population or relationship. It is not merely a prediction of a preferred result. A good hypothesis identifies the constructs, direction where justified, and unit of analysis.

Examples:

• H1: Among qualified target customers, higher perceived convenience is positively associated with intention to subscribe.
• H2: Customers exposed to a transparent delivery-time guarantee have a higher completed-purchase rate than comparable customers not exposed to the guarantee.
• H3: The effect of price on purchase intention differs between high-frequency and low-frequency category users.

The corresponding null hypothesis for a two-group mean comparison commonly states that the population means are equal:

Here, μ₁ and μ₂ are the population means for groups 1 and 2. This formulation is applicable when the outcome is measured on a scale for which a mean comparison is meaningful and the research design supports comparison of the groups. A null hypothesis is a statistical testing convention; it does not mean that the researcher believes no difference exists in reality.

Hypotheses should not be forced into exploratory work. When little is known, open-ended inquiry may be more appropriate than testing a narrowly specified relationship.

3.4 Constructs, variables, and operationalization

A construct is an abstract concept such as trust, perceived value, loyalty, or entrepreneurial intention. A variable is a measurable representation of a construct or characteristic. Operationalization specifies how the construct will be observed.

For example, “trust in the platform” could be operationalized through several rating items concerning payment security, reliability, and honesty. A single item may be convenient but can inadequately represent a multidimensional construct. Operational definitions should specify the response scale, reference period, population, and treatment of missing values.

A useful distinction is:

• Independent variable: a presumed predictor or manipulated condition.
• Dependent variable: the outcome to be explained or predicted.
• Mediator: a mechanism through which an effect may operate.
• Moderator: a condition under which the relationship changes.
• Control variable: a measured factor included to reduce confounding or improve precision.

These labels describe an analytical model, not proof of causation. A variable called “independent” in a regression is not necessarily causally independent.

4. Information Requirements and the Research Specification

Information requirements translate objectives into data needs. For each objective, the researcher should specify the construct, indicator, source, time frame, required precision, and intended use.

Information requirements should distinguish facts from perceptions. A respondent’s statement that a product is “too expensive” is a perception; actual price paid, income, alternatives, and contribution margin are additional evidence.

4.1 Measurement quality

Three forms of quality are especially important:

• Reliability: consistency of a measure across items, occasions, or raters.
• Validity: whether the measure represents the intended construct or supports the intended inference.
• Sensitivity: ability to detect meaningful differences or changes.

A reliable measure can still be invalid. A scale that consistently measures brand familiarity rather than trust is dependable but unsuitable for a trust objective.

4.2 Decision thresholds and practical significance

Statistical evidence should be connected to a managerial threshold. A small difference may be statistically detectable in a large sample but commercially irrelevant. Conversely, a meaningful business effect may fail to reach conventional statistical significance in a small exploratory study.

Researchers should therefore predefine, where feasible, what magnitude of change would justify action. For example, management might regard an increase of at least 5 percentage points in qualified conversion as commercially meaningful, subject to profitability and operational capacity. Such thresholds are decision criteria, not universal statistical laws.

5. Three Purposes of Research

The distinction among exploratory, descriptive, and causal research concerns the purpose and logic of inquiry. They are complementary rather than mutually exclusive.

5.1 Exploratory research

Exploratory research investigates poorly understood problems, generates concepts, uncovers language used by customers, and develops possible explanations or hypotheses. It is flexible and generally uses small, purposive, or convenience samples rather than samples designed for population estimation.

Common methods include:

• literature and secondary-data review;
• expert interviews;
• depth interviews;
• focus groups;
• observation and contextual inquiry;
• diary studies;
• social listening and review analysis;
• prototype testing with open-ended feedback.

Its outputs include themes, customer journeys, problem framings, alternative explanations, and candidate survey items. Exploratory findings should not automatically be reported as prevalence estimates. Five interviewees describing a problem does not establish that 5% or 50% of the market shares it.

Exploration is particularly valuable before questionnaire design because it helps the researcher use relevant categories and avoids imposing the venture’s assumptions on respondents.

5.2 Descriptive research

Descriptive research measures the characteristics, behaviors, attitudes, or distribution of a defined population or sample. Typical questions ask who, what, when, where, and how much.

Methods include structured surveys, panels, observation with coding schemes, market databases, transaction analysis, and audits. Descriptive designs require careful definition of the population, sampling frame, measurement period, and weighting procedures where applicable.

Examples include estimating:

• aided and unaided brand awareness;
• category purchase incidence;
• customer satisfaction distribution;
• market share by channel;
• average order value;
• segment size and profile.

A descriptive association does not establish causation. If customers who use a loyalty program purchase more frequently, the difference may reflect self-selection: frequent buyers may be more likely to join.

5.3 Causal research

Causal research examines whether changing one factor produces a change in an outcome. Strong causal inference requires more than correlation. It generally depends on temporal precedence, covariation, and credible elimination of alternative explanations.

The strongest practical design is often a randomized experiment. Eligible units are assigned to treatment and control conditions, and outcomes are compared. For a binary outcome, such as purchase, the difference in conversion rates can be expressed as:

where p_T is the proportion achieving the outcome in the treatment group and p_C is the corresponding proportion in the control group. The estimate is interpretable as an average treatment effect when assignment is randomized, outcomes are measured comparably, interference is limited, and implementation is sufficiently consistent.

For example, if 12% of 1,000 treatment users purchase and 9% of 1,000 control users purchase, then:

The observed difference is 3 percentage points. It should not be described as causal unless the experimental conditions support that inference. The commercial value also depends on incremental revenue, variable cost, customer quality, and possible longer-term effects.

Causal methods include A/B tests, field experiments, randomized pricing or messaging tests, and quasi-experiments such as difference-in-differences or interrupted time series. Quasi-experimental designs can be useful when randomization is infeasible, but their assumptions must be made explicit.

5.4 Comparison of research purposes

A staged project might begin with interviews, proceed to a representative survey, and then test a message or offer experimentally. This sequence is not mandatory; the appropriate design depends on the decision, prior knowledge, resources, and risk.

6. Research Design as a Logic of Evidence

Research design is the plan connecting questions to observations and conclusions. It includes the time horizon, data sources, sampling, measurement, procedures, analysis, and safeguards against bias.

6.1 Cross-sectional and longitudinal designs

A cross-sectional study observes units at one point or short period. It is efficient for describing current awareness, preferences, or behavior but is weak for establishing change or temporal order.

A longitudinal design observes the same units or comparable populations over time. It can examine retention, attitude change, cohort behavior, and post-purchase development. Attrition, repeated-measurement effects, and changing market conditions must be managed.

6.2 Primary and secondary research

Primary research collects data for the current problem. Secondary research uses existing data collected for another purpose, such as government statistics, industry reports, platform analytics, reviews, or internal sales records.

Secondary data can be faster and less expensive, but the researcher must check relevance, definitions, coverage, timeliness, missingness, and possible commercial or political bias. A market-size estimate based on a different geography or customer definition may be precise yet irrelevant.

6.3 Sampling and generalization

A target population is the complete set of units to which the researcher wishes to generalize. A sampling frame is the operational list or mechanism from which units are selected. Coverage error occurs when the frame excludes relevant members or includes ineligible units.

Probability sampling gives known, non-zero selection probabilities and supports design-based generalization under suitable conditions. Non-probability samples may be practical for exploration or hard-to-reach populations but require caution when estimating population prevalence.

Nonresponse, inaccurate recall, social desirability, and question wording can create bias even in a probability sample. A large sample does not automatically correct systematic error.

For a proportion estimate under a simple random sampling model, a commonly used approximate standard error is:

Here, p̂ is the sample proportion, n is the sample size, and SE(p̂) is the estimated standard error. The formula assumes independent observations and a simple random sample; complex sampling, weighting, clustering, and finite-population effects may require adjustments. It measures sampling variability, not total survey error.

7. From Questions to an Analysis Plan

An analysis plan should be drafted before data collection, especially when the decision is consequential. It should identify:

• the outcome and explanatory variables;
• the unit of analysis;
• data cleaning rules;
• treatment of missing and ineligible responses;
• subgroup comparisons;
• planned models or coding procedures;
• decision thresholds;
• sensitivity checks and limitations.

Pre-specification reduces the temptation to select only favorable analyses after seeing the data. Exploratory analyses remain legitimate, but they should be labeled as exploratory and used to generate subsequent hypotheses rather than presented as confirmatory tests.

Researchers should separate statistical uncertainty from business uncertainty. A precisely estimated conversion rate may still be uncertain in a new market because competitor reaction, capacity, regulation, or supply costs are unknown.

8. Applied Case: FreshRoute Student Subscription

The following is an illustrative fictional case.

Background

FreshRoute is a small meal-delivery venture serving three metropolitan districts. Its current customers order individual meals. Management is considering a discounted monthly subscription aimed at university students, but the venture has limited delivery capacity and cannot afford a failed launch.

Management problem

Should FreshRoute launch the student subscription, and if so, in which districts, at what price, and with what service conditions?

Initial assumptions and risks

The founder believes students value low price above all else. However, interviews with delivery staff suggest that late delivery and inflexible cancellation may be more important barriers. A low price could increase demand beyond capacity and reduce contribution margin.

Research problem

FreshRoute must estimate the size and characteristics of the qualified student segment, identify the attributes and barriers affecting subscription adoption, estimate response to alternative prices and service promises, and determine whether a pilot can generate profitable incremental demand without unacceptable service deterioration.

Objectives

1. Estimate the proportion of students in the three districts who purchase prepared meals at least twice monthly.
2. Identify the relative importance of price, delivery reliability, menu variety, and cancellation flexibility.
3. Compare subscription interest and expected usage across student segments.
4. Test whether a delivery-time guarantee increases completed subscription sign-up in a controlled pilot.
5. Estimate unit economics under realistic usage and capacity assumptions.

Research questions and hypotheses

The study could ask:

• What proportion of eligible students are active category users?
• Which barriers prevent them from ordering more often?
• Does interest differ by district, budget, living arrangement, or purchase frequency?
• Which subscription configuration produces the strongest qualified preference?

A causal pilot hypothesis might be: students randomly exposed to a clear delivery-time guarantee will have a higher completed sign-up rate than students receiving the standard offer, provided the operational promise can actually be delivered.

Staged design

Stage 1: Exploration. Conduct 15–20 semi-structured interviews across districts and usage levels. Ask participants to reconstruct recent meal decisions rather than merely speculate about an abstract product. Analyze recurring needs, language, barriers, and unacceptable conditions.

Stage 2: Description. Use a structured survey with eligibility screening, recent behavior questions, attribute ratings, subscription configuration tasks, and demographic or situational variables relevant to targeting. If population estimates are required, the sample should be recruited and weighted with a defensible coverage strategy rather than relying solely on the founder’s social-media followers.

Stage 3: Pilot experiment. Randomly assign eligible visitors or recruited users to the standard offer or the delivery-guarantee offer. Keep price, landing-page design, and eligibility rules constant. Track sign-up, first order, cancellation, contribution margin, delivery performance, and repeat behavior. Monitor adverse operational effects.

Information-to-decision matrix

Analysis and interpretation

Suppose the pilot produces higher sign-up under the guarantee but also increases delivery costs. The decision should not be based on sign-up alone. FreshRoute should examine incremental contribution after variable delivery cost, cancellation, refunds, and repeat purchase. It should also assess whether the promise is sustainable at larger volume.

If interviews emphasize reliability but the survey asks only about price, the research has failed to reflect the problem context. If the survey finds high stated interest but the pilot produces weak completed orders, the discrepancy is informative: hypothetical intention may not translate into behavior because of price, effort, trust, or competing alternatives.

Outcome and learning points

The appropriate outcome is not necessarily a national launch. A disciplined result may be a limited pilot, a revised service promise, a different target segment, or a decision to postpone. The case illustrates five principles:

1. The decision must be explicit before methods are selected.
2. Exploration can challenge managerial assumptions.
3. Description estimates the market but does not prove what causes adoption.
4. Experiments are useful for offer or message decisions when implementation is feasible.
5. Operational and financial evidence must accompany customer preference data.

9. Common Errors and Critical Limitations

9.1 Starting with a method

Choosing “a survey” or “focus group” before defining the decision often produces data that are easy to collect but difficult to use. The question should determine the method, not the reverse.

9.2 Confusing intention with behavior

Intentions can be useful predictors but are affected by hypothetical bias, social desirability, and lack of commitment. Behavioral evidence, transaction data, experiments, or observed trade-offs generally provide stronger decision evidence where feasible.

9.3 Overgeneralizing qualitative findings

Qualitative research provides depth and meaning, not automatic population estimates. The researcher should report sampling logic, context, contradictory cases, and interpretive limits.

9.4 Treating correlation as causation

Associations may reflect reverse causality, omitted variables, selection, or common causes. Statistical adjustment can improve comparison but does not guarantee causal identification.

9.5 Ignoring competing explanations

A good design actively searches for evidence that could disconfirm the preferred explanation. Rival hypotheses, placebo or control conditions, pre-launch baselines, and sensitivity analyses improve credibility.

9.6 Scope expansion

Research teams may add interesting questions until the project becomes too long, expensive, and unfocused. Prioritize questions by decision relevance, uncertainty, feasibility, and consequence of error.

10. Ethics, Governance, and Responsible Use of Data

Market research involves people, personal information, and potentially consequential decisions. Researchers should obtain informed consent where required, explain the purpose and use of data, minimize collection, protect confidentiality, and retain data only as long as justified. Sensitive characteristics should not be collected merely because they are easy to ask for.

Participants must not be deceived or exposed to unreasonable risk. Experimental offers should not create unfair exclusion, hidden costs, or unsafe service conditions. Incentives should compensate time without becoming coercive.

Digital trace data and artificial intelligence create additional concerns. Scraped reviews, social posts, location data, and platform behavior may contain personal or inferred sensitive information. Public availability does not automatically mean unrestricted ethical use. Researchers should assess legality, consent expectations, re-identification risk, data provenance, and algorithmic bias.

AI tools can assist with interview transcription, coding suggestions, questionnaire critique, segmentation, and synthesis. They should not replace human judgment about consent, context, validity, or causal interpretation. Researchers must verify generated summaries against source material, avoid uploading confidential data to unauthorized systems, document model use, and disclose material limitations.

11. Best-Practice Workflow

A robust workflow is iterative but disciplined:

1. Define the decision, alternatives, timing, and consequences.
2. Audit existing evidence and identify what remains uncertain.
3. Separate symptoms from possible underlying causes.
4. State the management problem and research problem separately.
5. Define the population, unit of analysis, and scope.
6. Write primary and secondary objectives.
7. Convert objectives into answerable questions and, where appropriate, hypotheses.
8. Specify constructs, indicators, sources, and decision thresholds.
9. Select exploratory, descriptive, causal, or mixed designs based on the questions.
10. Prepare a sampling, measurement, analysis, ethics, and governance plan.
11. Pilot instruments and procedures.
12. Report findings with uncertainty, limitations, rival explanations, and implications for action.

The final report should distinguish evidence from interpretation and recommendation. A recommendation is a reasoned managerial judgment based on evidence and constraints; it is not itself a research finding.

Chapter Summary

Entrepreneurial market research is a decision-support process. It begins with a clearly defined decision context rather than a preferred method. The management problem asks what the organization should do; the research problem specifies what must be learned to support that choice. Objectives then define the intended accomplishments, while research questions and hypotheses express the specific relationships, estimates, comparisons, or explanations required.

Information requirements connect abstract objectives to measurable constructs, data sources, populations, and decision thresholds. Exploratory research clarifies poorly understood problems and generates hypotheses. Descriptive research measures characteristics and distributions. Causal research estimates the effects of interventions when design conditions support causal inference. These purposes can be combined in a staged research program.

Sound research requires attention to sampling, measurement, validity, reliability, alternative explanations, practical significance, ethics, and data governance. The FreshRoute case demonstrates how interviews, structured measurement, and a controlled pilot can jointly inform a market-entry decision. The ultimate test of a research plan is whether credible evidence can change and improve an important decision.

Key Terms

Decision context: The circumstances, alternatives, constraints, stakeholders, and uncertainty surrounding a decision.

Management problem: The action-oriented problem concerning what an organization should do.

Research problem: The information problem that must be addressed to support a management decision.

Research objective: A specific statement of what a study intends to accomplish.

Research question: A focused, answerable question addressed through evidence.

Hypothesis: A provisional, empirically testable statement about a population or relationship.

Construct: An abstract concept such as trust, value, or loyalty.

Operationalization: The specification of how a construct will be measured or observed.

Independent variable: A presumed predictor or manipulated condition in an analytical model.

Dependent variable: An outcome to be explained, predicted, or compared.

Mediator: A mechanism through which an association or effect may operate.

Moderator: A condition under which a relationship changes in magnitude or direction.

Exploratory research: Flexible inquiry used to clarify problems, discover themes, and generate hypotheses.

Descriptive research: Structured research used to measure characteristics, behaviors, or distributions.

Causal research: Research designed to estimate whether changing one factor produces a change in an outcome.

Target population: The complete set of units to which the researcher intends to generalize.

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

Reliability: The consistency of a measurement procedure.

Validity: The degree to which evidence supports the intended interpretation or use of a measure.

Information requirement: A defined item of evidence needed to answer a research question or support a decision.

Review Questions

Short Answer Questions

1. How does a management problem differ from a research problem?
2. Why should symptoms be distinguished from underlying problems?
3. What is the purpose of operationalizing a construct?
4. State one principal use of exploratory research and one limitation.
5. Why does correlation alone not establish causation?

Long Answer Questions

1. Explain the chain from decision context to management problem, research problem, objectives, questions, and information requirements.
2. Compare exploratory, descriptive, and causal research in terms of purpose, methods, outputs, and limitations.
3. Discuss how sampling, measurement, and nonresponse can affect the credibility of a market-research conclusion.
4. Explain why a research project should include decision thresholds and practical significance, not only statistical results.
5. Discuss the ethical and governance issues involved in using digital trace data and AI in entrepreneurial market research.

Analytical / Application Questions

1. A start-up observes falling sales and immediately commissions a customer-satisfaction survey. Diagnose the weaknesses in this approach and propose a staged research design.
2. Design a research problem, three objectives, two research questions, and one hypothesis for a venture considering a new pricing plan.
3. Using the FreshRoute case, specify which decisions require exploratory, descriptive, and causal evidence, and explain how the findings should be integrated before scaling.

Multiple Choice Questions with Answers

Question 1

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

A. A management problem asks what action should be taken; a research problem asks what must be learned to support that action.

B. A management problem is always quantitative; a research problem is always qualitative.

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

D. A management problem is solved before research begins; a research problem requires no decision context.

Correct Answer: A. A management problem asks what action should be taken; a research problem asks what must be learned to support that action.

Question 2

Which objective is most appropriately written as a research objective?

A. Understand the market completely.

B. Make customers happier.

C. Estimate the proportion of qualified users who would renew within 30 days.

D. Launch the product successfully.

Correct Answer: C. Estimate the proportion of qualified users who would renew within 30 days.

Question 3

What is the primary purpose of exploratory research?

A. To prove that a managerial assumption is correct

B. To discover themes, clarify problems, and generate hypotheses

C. To estimate population prevalence with known sampling error

D. To establish causation through random assignment

Correct Answer: B. To discover themes, clarify problems, and generate hypotheses

Question 4

Which research purpose is most directly concerned with measuring who, what, when, where, and how much?

A. Exploratory

B. Descriptive

C. Causal

D. Experimental-only

Correct Answer: B. Descriptive

Question 5

In a randomized pilot, the treatment conversion rate is 0.12 and the control conversion rate is 0.09. What is the observed difference in conversion rates?

A. 0.03, or 3 percentage points

B. 0.21, or 21 percentage points

C. 1.33, or 133 percentage points

D. 0.75, or 75 percentage points

Correct Answer: A. 0.03, or 3 percentage points

Question 6

Which condition most strongly supports a causal interpretation of a treatment-control difference?

A. The treatment group is larger than the control group

B. The treatment was randomly assigned and outcomes were measured comparably

C. The treatment group reported greater satisfaction before exposure

D. The result matches the founder’s expectation

Correct Answer: B. The treatment was randomly assigned and outcomes were measured comparably

Question 7

Which statement about qualitative interviews is most accurate?

A. A small interview sample automatically estimates market prevalence.

B. Interview findings can reveal themes and language but may not generalize as population proportions.

C. Interviews are inherently causal experiments.

D. Interview data do not require an explicit research objective.

Correct Answer: B. Interview findings can reveal themes and language but may not generalize as population proportions.

Question 8

What is a moderator in a research model?

A. A variable through which an effect operates

B. A variable that changes the strength or direction of a relationship

C. The final business recommendation

D. A source of secondary data

Correct Answer: B. A variable that changes the strength or direction of a relationship

Question 9

Which issue is a potential sampling-frame problem?

A. A questionnaire uses clear wording

B. A customer list excludes users who purchase through an independent channel

C. A researcher pre-specifies an analysis plan

D. A pilot test identifies confusing instructions

Correct Answer: B. A customer list excludes users who purchase through an independent channel

Question 10

Why should AI-generated research summaries be checked against source material?

A. AI tools cannot produce any text

B. Generated summaries may omit context, reproduce bias, or misrepresent the original evidence

C. Verification makes sampling unnecessary

D. AI summaries are automatically legally confidential

Correct Answer: B. Generated summaries may omit context, reproduce bias, or misrepresent the original evidence