2. Research Foundations and Evidence-Based Management

Introduction: Why Leadership Knowledge Requires Evidence

Leadership and organizational behavior are often discussed through compelling stories: a chief executive transforms a failing firm, a charismatic manager inspires extraordinary commitment, or a new work arrangement appears to increase productivity. Such accounts may be useful starting points, but they are not, by themselves, reliable knowledge. Doctoral study requires a more demanding question: What evidence supports the claim, under what conditions, through which mechanisms, and with what degree of uncertainty?

Research foundations provide the intellectual and methodological basis for answering that question. They address how knowledge is produced, what counts as evidence, how theories are constructed and tested, how causal claims are justified, and how findings can be translated into organizational decisions. Evidence-based management extends this reasoning into practice. It does not mean applying research mechanically. Rather, it means integrating the best available research evidence with professional expertise, organizational data, stakeholder values, and the contextual realities of a decision.

The subject is especially important in leadership and organizational behavior because many relevant phenomena are complex, relational, historical, and multilevel. Leadership may be enacted by individuals but shaped by teams, organizational systems, institutional rules, and national cultures. Employee attitudes may be influenced simultaneously by supervisors, peers, technology, job design, and labor-market conditions. Consequently, no single method or epistemological position is sufficient for every research problem.

Learning Objectives

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

1. Distinguish major epistemological positions and explain their implications for organizational research.

2. Evaluate theory construction, conceptualization, measurement, and theory testing.

3. Differentiate association, prediction, explanation, and causation.

4. Analyze multilevel relationships in leadership and organizational behavior.

5. Compare qualitative, quantitative, mixed-methods, and review designs.

6. Assess replication, research transparency, validity, reliability, and research ethics.

7. Translate scholarly evidence into responsible managerial decisions.

8. Critically evaluate the limits and practical significance of organizational research.

From Organizational Questions to Researchable Problems

Research begins not with a method but with a problem. A management problem asks what an organization should do; a research problem asks what is not sufficiently understood and how that uncertainty can be investigated. For example, “How can we improve retention?” is a managerial question. A researchable version might be: “Through what mechanisms does perceived supervisor support influence voluntary turnover among early-career professionals, and how does this relationship vary across occupational groups?”

A well-formed research problem normally includes a phenomenon, a population or context, a conceptual tension or gap, and an approach to evidence. Research questions may be descriptive, explanatory, predictive, interpretive, evaluative, or design-oriented. These purposes should not be confused. A descriptive study may establish patterns without explaining them; an explanatory study seeks mechanisms; an evaluative study assesses whether an intervention works in a particular setting.

The doctoral researcher must also distinguish a literature gap from a mere absence of studies. A worthwhile gap may involve contradictory findings, an untested mechanism, boundary conditions, neglected populations, inappropriate assumptions, or a mismatch between theory and method. Novelty is valuable, but contribution also depends on rigor, relevance, and the ability to revise or extend existing knowledge.

Epistemological Positions and Their Consequences

Epistemology concerns the nature, sources, and justification of knowledge. It influences what a researcher regards as a legitimate question, what counts as evidence, and how findings should be interpreted. Ontology concerns assumptions about what exists—for example, whether organizational culture is an objective property, a shared perception, or a socially constructed accomplishment. Methodology refers to the logic connecting research questions, data, methods, and claims.

Positivist and Post-positivist Approaches

A positivist orientation emphasizes observable phenomena, systematic measurement, regularities, and the testing of propositions. In organizational research, it is commonly associated with surveys, experiments, archival data, statistical modeling, and hypothesis testing. A strict positivist view assumes that reality can be observed objectively and represented through general laws.

Post-positivism modifies this position by recognizing that observation is imperfect, measurements are fallible, and knowledge is provisional. Researchers seek systematic evidence and causal explanations, but they acknowledge uncertainty, researcher influence, and the possibility that conclusions may be revised. Confidence intervals, robustness checks, preregistration, and replication are consistent with this more cautious view.

Interpretivism and Constructivism

Interpretivist approaches examine how people create meaning in social contexts. Leadership, identity, trust, and organizational culture may not be adequately understood as variables detached from lived experience. Interviews, observation, documents, and interactional data can reveal how organizational members interpret events and negotiate shared realities.

Constructivism emphasizes that knowledge is co-produced through interaction between researchers and participants. The aim is often not universal prediction but a rich, credible interpretation of situated experience. This does not mean that qualitative research is unstructured or subjective in an uncritical sense. Credibility depends on transparent sampling, reflexivity, careful documentation, triangulation, and a persuasive connection between data and interpretation.

Critical and Pragmatist Perspectives

Critical approaches ask how power, inequality, ideology, and institutional interests shape organizational knowledge and practice. They may investigate whose voices are excluded, how apparently neutral systems reproduce disadvantage, or how managerial language legitimizes particular forms of control.

Pragmatism begins with the research problem and selects methods according to what best addresses it. A pragmatist may combine statistical analysis with interviews because numerical patterns and participant interpretations answer different parts of the question. Pragmatism should not be confused with methodological convenience: integration requires a coherent rationale and explicit treatment of tensions between forms of evidence.

These perspectives are not merely philosophical labels. They affect sampling, the role of theory, the meaning of validity, the relationship between researcher and participant, and the kinds of conclusions that can responsibly be drawn. A doctoral researcher should therefore state epistemological assumptions rather than presenting methods as neutral technical instruments.

Theory Construction: From Concepts to Explanations

A theory is more than a list of variables or a summary of findings. It is an organized explanation of relationships among concepts, including the mechanisms and conditions through which those relationships operate. In leadership research, a theory might propose that empowering leadership increases employee voice because it strengthens psychological safety, particularly when team norms support respectful dissent.

The central building blocks are concepts, constructs, propositions, assumptions, mechanisms, boundary conditions, and levels of analysis. A construct is an abstract concept developed for theoretical or empirical use, such as psychological safety or leader-member exchange. Operationalization specifies how a construct will be observed or measured. The conceptual definition must come before the operational definition; otherwise, researchers risk treating a convenient scale as if it fully represents the phenomenon.

Theory can be developed through deduction, induction, or abduction. Deduction begins with theory and derives testable hypotheses. Induction develops concepts or patterns from empirical material. Abduction moves iteratively between surprising observations and plausible explanations. Strong doctoral research often combines these logics, while clearly identifying when a claim is exploratory and when it is confirmatory.

Mediation explains how or why an antecedent relates to an outcome. Moderation specifies when, for whom, or under what conditions a relationship changes. A multilevel model may propose that individual perceptions are influenced by team climate, while team climate itself is shaped by organizational policies. These distinctions matter because statistical association alone does not establish a mechanism or a boundary condition.

Theory development should also address rival explanations. If employee engagement rises after a leadership program, alternative explanations may include concurrent pay changes, seasonal workload variation, selection into the program, or measurement effects. Theoretical quality is strengthened when a study explains why its proposed account is more credible than plausible alternatives.

Causality, Validity, and the Logic of Evidence

Causal reasoning is central to managerial decisions, yet organizations rarely provide perfect experimental conditions. A causal claim asserts that changing one factor would produce a difference in an outcome, holding relevant alternatives constant. Correlation, temporal sequence, and theoretical plausibility are useful, but none alone is sufficient.

An experiment supports causal inference through manipulation, comparison, and—ideally—random assignment. Randomization helps distribute observed and unobserved confounders across conditions. A field experiment can provide strong practical relevance, although implementation may be difficult and treatment effects may vary across settings.

Quasi-experimental designs are valuable when random assignment is impossible. Difference-in-differences compares changes over time between an intervention group and a comparison group, relying on assumptions such as parallel pre-intervention trends. Regression discontinuity exploits a treatment threshold, while instrumental-variable designs use a variable related to treatment but not directly to the outcome except through treatment. Each design depends on assumptions that must be examined rather than treated as technical footnotes.

Longitudinal studies improve temporal reasoning and can distinguish stable differences from change. Panel designs, repeated measures, event histories, and experience-sampling methods are especially relevant to dynamic leadership and employee behavior. Cross-sectional designs may identify associations but are vulnerable to reverse causality and common-method bias.

Validity has several dimensions. Construct validity concerns whether a measure represents the intended construct. Internal validity concerns whether an observed effect can reasonably be attributed to the proposed cause. External validity concerns generalizability across people, settings, times, and tasks. Statistical conclusion validity concerns the appropriateness and precision of inferences from data. These forms of validity can conflict. A tightly controlled laboratory study may have high internal validity but limited organizational realism; a rich field study may offer contextual relevance while facing greater confounding.

Effect size and practical significance should accompany statistical significance. A small effect may be statistically detectable in a large sample but unimportant for managerial action. Conversely, a potentially important effect may be uncertain in a small study. Decision-makers should consider magnitude, uncertainty, cost, feasibility, distributional consequences, and the consequences of being wrong.

The Multilevel Character of Organizations

Organizations are nested systems. Individuals belong to teams, teams operate within departments, and departments exist within organizations and institutional environments. A leadership behavior enacted by one manager may influence an employee directly, shape team norms indirectly, and contribute to organizational routines over time.

Multilevel research distinguishes the level at which a construct is theorized from the level at which it is measured. Individual psychological safety is not automatically the same as team psychological safety. Aggregating individual responses to create a team-level construct requires justification, such as sufficient within-team agreement and meaningful between-team variation. Researchers must avoid ecological fallacies—inferring individual relationships from group-level data—and atomistic fallacies—inferring group processes from individual-level relationships.

Hierarchical linear models and multilevel structural equation models can estimate relationships within and between levels. Cross-level interactions examine whether a higher-level condition changes an individual-level relationship, such as whether team climate moderates the effect of supervisor support on employee voice. However, statistical complexity does not substitute for multilevel theory. The researcher must explain how processes at one level connect to processes at another and whether the proposed direction is plausible.

Choosing Among Research Designs

Quantitative Designs

Quantitative research is appropriate when concepts can be operationalized and the study seeks to estimate prevalence, relationships, effects, or prediction. Surveys provide efficient access to perceptions and attitudes but may suffer from response bias, weak temporal ordering, and common-method variance. Archival and digital trace data can capture behavior at scale, although they may reflect institutional recording practices rather than the phenomenon itself.

Experiments are strongest for testing interventions under controlled conditions. Statistical models such as regression, structural equation modeling, multilevel analysis, and latent growth models allow researchers to estimate complex relationships. Their credibility depends on sound measurement, appropriate model specification, missing-data treatment, transparent reporting, and sensitivity analysis.

Qualitative Designs

Qualitative designs are valuable when the phenomenon is poorly understood, context-dependent, processual, or concerned with meaning and identity. Case study research investigates a bounded system using multiple sources. Ethnography examines culture and practice through sustained engagement. Grounded theory develops a conceptual account through systematic comparison of data. Narrative inquiry studies how individuals construct experience through stories.

Quality depends on purposeful sampling, theoretical sensitivity, reflexive awareness, analytic transparency, and attention to negative cases. Saturation should not be treated as a mechanical sample-size rule; researchers must explain what kind of saturation was sought and how additional data ceased to alter the developing account.

Mixed Methods and Systematic Reviews

Mixed-methods research integrates qualitative and quantitative evidence within a single inquiry. An explanatory sequential design may begin with a survey and use interviews to interpret unexpected results. An exploratory sequential design may develop measures from qualitative findings. Convergent designs collect different forms of evidence in parallel. Integration—not merely placing two studies side by side—is the defining feature.

Systematic reviews use explicit search, selection, appraisal, and synthesis procedures to reduce arbitrary conclusions from a large literature. Meta-analysis statistically combines comparable effect estimates, while qualitative evidence synthesis integrates interpretations across studies. Reviews must consider publication bias, heterogeneity, study quality, measurement differences, and dependence among samples. A precise pooled estimate can be misleading when the underlying studies are systematically biased or address different constructs.

Replication, Transparency, and the Development of Trustworthy Knowledge

Replication is essential because a single result may reflect sampling variation, measurement choices, selective reporting, or contextual coincidence. Direct replication seeks to reproduce the original procedures as closely as feasible. Conceptual replication tests the same theoretical proposition using different operationalizations, populations, or designs. Both contribute differently: direct replication examines reliability of a finding, while conceptual replication examines robustness of an idea.

Open materials, data where ethically permissible, analytic code, preregistration, registered reports, and clear disclosure of deviations strengthen transparency. Researchers should distinguish exploratory analyses, which generate possibilities, from confirmatory analyses, which test pre-specified expectations. This distinction does not make exploration inferior; it prevents exploratory discoveries from being presented as if they had been predicted in advance.

Research synthesis should also account for null findings and contradictory evidence. Publication bias occurs when the availability or visibility of results depends on their direction or statistical significance. Doctoral researchers should resist p-hacking, HARKing (hypothesizing after results are known), selective outcome reporting, and inflated claims of generalizability. Trust is built through candor about uncertainty, limitations, and competing interpretations.

Research Ethics and Responsible Organizational Inquiry

Organizational research involves power relationships. Employees may feel unable to refuse participation when a manager sponsors the study, and confidentiality can be difficult in small teams. Ethical research therefore requires informed consent, voluntary participation, risk minimization, privacy protection, secure data handling, and fair treatment of participants.

De-identification is not always sufficient when datasets contain job titles, timestamps, locations, or other combinations that permit re-identification. Researchers should collect only necessary data, restrict access, define retention periods, and communicate how findings will be reported. Artificial intelligence and automated analytics increase risks related to surveillance, opaque classification, discriminatory prediction, and secondary use of employee data.

Ethics also concerns interpretation and consequences. A study may be methodologically sound yet harmful if managers use it to label employees, justify excessive monitoring, or remove human judgment from consequential decisions. Researchers should consider stakeholder interests, distributive effects, cultural sensitivity, and whether participants can challenge inaccurate conclusions. Institutional review processes are important, but ethical responsibility remains with the researcher throughout design, analysis, publication, and application.

Evidence-Based Management as Translation, Not Mechanical Application

Evidence-based management is a disciplined decision process. A useful framework involves five activities: asking an answerable question, acquiring relevant evidence, appraising its quality and applicability, aggregating it with organizational and stakeholder knowledge, and acting while monitoring outcomes.

The best available external evidence may come from systematic reviews, well-designed experiments, longitudinal studies, or credible qualitative research, depending on the question. Internal evidence includes turnover data, employee surveys, performance indicators, implementation records, and local experiments. Professional expertise helps interpret feasibility and unintended effects. Stakeholder values matter because management decisions distribute benefits, burdens, opportunities, and risks.

Translation requires attention to boundary conditions. A leadership intervention supported in one industry may not work where authority structures, labor institutions, or skill requirements differ. Managers should identify the mechanism expected to produce change, specify indicators of success, pilot where appropriate, and establish a feedback cycle. A decision should not be justified solely by the phrase “research shows.” The relevant questions are: Which research? How credible? How similar is the context? What would count as failure? Who bears the risk?

Digital tools can support evidence-based management through dashboards, natural-language analysis of employee comments, predictive models, and randomized testing of communication or training interventions. These tools are useful only when data quality, privacy, fairness, interpretability, and human oversight are addressed. Prediction is not causation: a model may identify employees likely to leave without revealing what intervention would retain them or whether intervention would be beneficial.

Illustrative Case: The Leadership Development Initiative at Meridian Systems

Background

Meridian Systems, a fictional mid-sized technology services firm, experienced rising voluntary turnover among project professionals. Senior leaders attributed the problem to inconsistent supervisory leadership and proposed a six-month transformational-leadership program for 40 managers.

Situation and Problem

After the program, employee engagement scores increased by eight percentage points in participating units. The executive team concluded that the program had caused the improvement and planned to extend it across the company. However, the participating managers had volunteered for the program, two major client contracts had begun during the same period, and a company-wide pay adjustment had been introduced.

Analysis

The initial evidence was a before-and-after comparison without a control group. It showed temporal association but could not isolate the program effect. Selection bias was plausible because motivated managers may have volunteered. History effects were also likely because the client contracts and pay adjustment occurred during the same period. In addition, engagement was measured through the same survey instrument before and after the intervention, creating possible response and common-method concerns.

A stronger evaluation would compare participating units with similar nonparticipating units, examine pre-intervention trends, track turnover and behavioral indicators, and collect qualitative evidence about how managers applied the training. A difference-in-differences design could estimate whether engagement changed more in participating units than in comparable units, provided the parallel-trends assumption was credible. Interviews could examine whether psychological safety, feedback quality, or role clarity constituted the proposed mechanism.

Approach and Decision

Meridian established a staged implementation. Managers were assigned to early and later cohorts using a transparent allocation process. The evaluation included repeated measures of engagement, turnover, absenteeism, and team climate, alongside interviews with employees and managers. Data were reported in aggregate, and participation in interviews was voluntary. The organization pre-specified primary outcomes and agreed that the program would be modified or discontinued if improvements were not distinguishable from comparison trends after one year.

Outcome

The evaluation found modest improvement in feedback quality and team psychological safety in early-cohort units, but no clear short-term reduction in turnover. Interviews indicated that the program was more effective where managers had discretion to redesign team routines and where workload was manageable. Leaders therefore retained the training but paired it with workload review and team-level implementation support rather than claiming that training alone solved retention.

Learning Points

The case demonstrates the difference between a promising signal and a causal conclusion. It also shows why mechanisms, context, comparison groups, multiple outcomes, and ethical data practices matter. Evidence-based management does not delay every decision until perfect evidence exists; it improves decisions by making assumptions visible, using proportionate evaluation, and learning from implementation.

Common Errors, Limitations, and Practical Safeguards

Several recurrent errors weaken organizational research. Researchers may select a method before clarifying the question, confuse a measure with a construct, infer causality from cross-sectional data, ignore nested data structures, or overgeneralize from convenience samples. Statistical significance may be emphasized while effect size and uncertainty are neglected. Qualitative accounts may be treated as anecdotal when their sampling and analysis are not explained; quantitative studies may be treated as objective when measurement and modeling choices are hidden.

Practical safeguards include a theory-method alignment table, pilot testing, power and precision planning, transparent inclusion criteria, missing-data analysis, robustness checks, rival explanations, reflexive memos, audit trails, and independent review. Researchers should report limitations as implications for interpretation rather than as ritual disclaimers. Managers should pilot high-risk interventions, monitor unintended consequences, and avoid using research findings outside their evidential scope.

Emerging Directions

Leadership and organizational behavior research is increasingly shaped by remote and hybrid work, algorithmic management, people analytics, ecological and sustainability concerns, and changing expectations about inclusion and employee autonomy. Experience-sampling and digital trace methods permit fine-grained analysis of interaction and work rhythms, but they also intensify privacy concerns. Artificial intelligence may assist literature discovery, coding, forecasting, and decision support, yet researchers must verify outputs, disclose analytic procedures, and remain alert to biased training data and fabricated or unverifiable sources.

Future scholarship is likely to place greater emphasis on open and cumulative science, intervention research conducted with organizations, cross-cultural validity, longitudinal and multilevel designs, and the practical consequences of management knowledge. The central challenge will be to combine methodological sophistication with humility about what evidence can establish.

Best Practices for Doctoral Researchers and Evidence-Oriented Managers

A disciplined approach can be summarized as follows:

• Begin with a significant problem and a precise research question.
• Make assumptions about reality, knowledge, and values explicit.
• Define constructs conceptually before selecting measures.
• Match design and method to the claim being made.
• Separate description, prediction, explanation, and causal inference.
• Address mechanisms, boundary conditions, rival explanations, and levels of analysis.
• Use transparent procedures for sampling, analysis, and reporting.
• Treat qualitative and quantitative evidence as different forms of disciplined inquiry, not as a hierarchy of sophistication.
• Examine practical significance, uncertainty, feasibility, and distributional effects.
• Protect participants and scrutinize the consequences of data-driven decisions.
• Translate evidence through local diagnosis, cautious implementation, and evaluation.

Chapter Summary

Research foundations determine how leadership and organizational behavior knowledge is generated, evaluated, and applied. Epistemological positions shape questions, methods, standards of validity, and interpretations. Theory links constructs through mechanisms and boundary conditions; sound research distinguishes conceptual definitions from operational measures and considers rival explanations. Causal claims require more than correlation and should be supported by appropriate designs and explicit assumptions.

Organizations are multilevel systems, so researchers must align theory, measurement, and analysis across individuals, teams, organizations, and institutional contexts. Quantitative, qualitative, mixed-methods, and systematic review designs answer different kinds of questions. Replication, transparency, and responsible reporting strengthen the cumulative character of scholarship. Research ethics includes consent and confidentiality but also extends to privacy, fairness, surveillance, and the consequences of managerial use.

Evidence-based management integrates external research with internal data, professional expertise, stakeholder values, and contextual judgment. Its purpose is not to produce certainty, but to improve the quality, transparency, and learning capacity of organizational decisions.

Key Terms

Epistemology: The study of how knowledge is produced, justified, and evaluated.

Ontology: Assumptions about the nature of reality and what exists.

Methodology: The logic connecting research questions, assumptions, methods, data, and claims.

Construct: An abstract concept developed for theoretical or empirical investigation.

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

Theory: An organized explanation of relationships, mechanisms, assumptions, and conditions.

Mechanism: The process through which a cause is expected to produce an outcome.

Mediation: An explanation of how or why one variable influences another through an intervening process.

Moderation: A condition that changes the strength or direction of a relationship.

Causality: A relationship in which changing one factor produces a change in another under specified conditions.

Internal validity: The credibility of an inference that an observed effect was caused by the proposed factor.

External validity: The extent to which findings apply across contexts, populations, times, or tasks.

Construct validity: The degree to which an operational measure represents the intended construct.

Multilevel research: Research that theorizes or analyzes phenomena across nested levels such as individuals, teams, and organizations.

Replication: A study designed to examine whether a prior finding or theoretical proposition is reproducible.

Systematic review: A transparent, structured synthesis of research using explicit search, selection, and appraisal procedures.

Evidence-based management: The integration of best available research, organizational data, expertise, and stakeholder values in managerial decisions.

Reflexivity: Critical examination of how the researcher’s position, assumptions, and relationships influence inquiry.

Common-method bias: Distortion that arises when predictor and outcome data are collected through similar methods or sources.

Research ethics: Principles governing responsible treatment of participants, data, communities, and the consequences of research.

Review Questions

Short Answer Questions

1. How do epistemological assumptions influence the design of organizational research?

2. Distinguish a construct from its operationalization and provide an example from leadership research.

3. What is the difference between mediation and moderation?

4. Why can a cross-sectional correlation not, by itself, establish causality?

5. What is the principal purpose of replication in organizational research?

Long Answer Questions

1. Compare positivist, interpretivist, critical, and pragmatist approaches to leadership and organizational behavior research.

2. Explain how theory construction links concepts, mechanisms, boundary conditions, and empirical propositions.

3. Discuss the major forms of validity and the tensions that may arise among them in field research.

4. Evaluate the contributions and limitations of quantitative, qualitative, mixed-methods, and systematic review designs.

5. Explain how ethical responsibilities extend beyond informed consent when organizations use employee analytics and artificial intelligence.

Analytical / Application Questions

1. A company reports that employee engagement increased after a leadership training program. Design an evaluation that would distinguish the program effect from selection, history, and measurement explanations.

2. A researcher finds that empowering leadership predicts employee voice more strongly in some teams than others. Develop a multilevel explanation and identify an appropriate analytical strategy.

3. A senior manager wants to deploy an algorithm that predicts employee turnover. Using the evidence-based management framework, identify the evidence, ethical safeguards, contextual questions, and monitoring procedures required before implementation.

22. Multiple Choice Questions with Answers

Question 1

Which statement best reflects a post-positivist approach to organizational research?

A. Knowledge is entirely subjective and cannot be evaluated systematically.

B. Research can produce useful explanations, but observations and measurements remain fallible.

C. Only findings generated through randomized experiments are legitimate.

D. Organizational reality consists solely of participants’ narratives.

Correct Answer: B. Research can produce useful explanations, but observations and measurements remain fallible.

Question 2

A researcher argues that empowering leadership increases employee voice because it strengthens psychological safety. Psychological safety functions primarily as what in this explanation?

A. A moderator

B. A sampling frame

C. A mediator

D. A control group

Correct Answer: C. A mediator

Question 3

Which design feature most directly strengthens causal inference in a field experiment?

A. Using a larger descriptive sample without comparison groups

B. Random assignment to treatment and comparison conditions

C. Collecting all variables in one survey

D. Selecting only participants who favor the intervention

Correct Answer: B. Random assignment to treatment and comparison conditions

Question 4

A study aggregates employee perceptions to represent team climate. What must the researcher establish before treating the construct as team-level?

A. That every employee gave an identical response

B. That the sample includes only senior employees

C. That there is sufficient within-team agreement and between-team variation

D. That team climate has no relationship with individual attitudes

Correct Answer: C. That there is sufficient within-team agreement and between-team variation

Question 5

Which example illustrates an ecological fallacy?

A. Using interviews to understand how employees interpret leadership.

B. Inferring individual employee behavior from a relationship observed only across departments.

C. Comparing treatment and control groups in a randomized study.

D. Using repeated measures to examine change over time.

Correct Answer: B. Inferring individual employee behavior from a relationship observed only across departments.

Question 6

What is a defining feature of a systematic review?

A. It relies exclusively on the most frequently cited article.

B. It uses explicit procedures for searching, selecting, appraising, and synthesizing studies.

C. It excludes studies with nonsignificant findings automatically.

D. It replaces assessment of study quality with expert opinion.

Correct Answer: B. It uses explicit procedures for searching, selecting, appraising, and synthesizing studies.

Question 7

Which practice best distinguishes exploratory analysis from confirmatory analysis?

A. Reporting exploratory findings as though they were predicted in advance

B. Pre-specifying confirmatory hypotheses and clearly labeling later discoveries as exploratory

C. Avoiding all analyses not supported by the initial literature review

D. Conducting statistical tests until a significant result appears

Correct Answer: B. Pre-specifying confirmatory hypotheses and clearly labeling later discoveries as exploratory

Question 8

In the Meridian Systems case, why was the initial increase in engagement insufficient to establish that training caused the improvement?

A. Engagement can never be measured quantitatively.

B. The study lacked any evidence from employees.

C. Other events and self-selection could also explain the observed change.

D. Training programs cannot affect leadership behavior.

Correct Answer: C. Other events and self-selection could also explain the observed change.

Question 9

Which statement is most consistent with evidence-based management?

A. Managers should apply findings from any study that reports statistical significance.

B. External research should replace professional judgment and stakeholder values.

C. Decisions should integrate credible research with local data, expertise, values, and context.

D. Action should be postponed until research provides complete certainty.

Correct Answer: C. Decisions should integrate credible research with local data, expertise, values, and context.

Question 10

Why is a turnover prediction model not automatically a causal intervention?

A. Prediction identifies likely outcomes but does not necessarily show what action would change them.

B. Prediction models cannot use organizational data.

C. Causal research never uses statistical models.

D. Turnover is not an organizational behavior outcome.

Correct Answer: A. Prediction identifies likely outcomes but does not necessarily show what action would change them.