61  Marketing Theory and Theory Construction Seminar

Most seminars in this book teach a substantive literature—how the field studies advertising, or branding, or the marketing–finance interface. This one teaches the prior question: how the field builds, frames, argues, and tests theory at all. Its object is not any particular construct but the activity of theorizing—what a marketing scholar means by a “theory,” what separates a theoretical contribution from a clever empirical result, how an abstract construct is turned into something measurable and falsifiable, and how a community decides that a claim has earned the status of knowledge. It is, in short, the philosophy-of-science seminar for marketing, taught not as abstract epistemology but as the working craft a doctoral student must master to write a dissertation that advances theory rather than merely reporting an effect.

The seminar matters now more than it has in decades. Two pressures converge on it. The first is internal: the field’s flagship journals publish a shrinking share of purely conceptual articles, and the discipline has grown anxious that its theoretical output is thinning even as its empirical output explodes (doi:10.1509/jmkg.74.1.1). The second is external: the credibility revolution—false-positive psychology, the replication crisis, the re-examination of what counts as a finding—has forced every empirical science to ask whether its accumulated “knowledge” is real (doi:10.1177/0956797611417632; doi:10.1126/science.aac4716). Theory construction sits at the intersection of both anxieties: a field that cannot say clearly what its theories are cannot say clearly what it would mean to confirm or refute them.

A student who works through this seminar should come away able to do four things: to read a paper and locate precisely where its theoretical contribution lies (or to diagnose that it has none); to take an abstract construct and specify its nomological network so that it becomes measurable and testable; to recognize which mode of theorizing—conceptual framework, analytical model, grounded induction, typology, empirical generalization—a research question demands; and to judge a literature’s credibility by the standards the credibility revolution has made non-negotiable. These skills connect directly to two earlier chapters: the distinction between a latent construct and an observable variable developed in 1 is the measurement-theoretic backbone of this seminar, and the disciplinary history in Chapter 2 supplies the schools-of-thought and philosophy-of-science context against which these debates play out.

This is the book’s conceptual and methodological seminar—the one that complements its empirical seminars. Where the modeling, consumer-behavior, and marketing-strategy seminars teach how to execute research within a tradition, this seminar teaches how to think about what research in any tradition is for, what a contribution to it looks like, and how the community polices the line between a result and a finding.

61.1 Semester arc

The arc moves from the most general question to the most concrete and back. It opens with what theory is—the metatheoretical canon (Sutton & Staw, Whetten, DiMaggio) that every social science now teaches—and immediately localizes it to marketing’s own anxiety about conceptual contribution (MacInnis, Yadav). From there the seminar descends to the measurement layer: constructs, construct validity, and the nomological net (Cronbach & Meehl, Churchill), the machinery that turns an abstract idea into something a study can be about. With that foundation it confronts the field’s philosophy of science directly—the realism-versus-relativism debates of the early 1980s (Anderson, Hunt, Peter & Olson) that still structure how marketing argues about truth.

The middle weeks survey the modes of theorizing the field actually uses, because “theory” in marketing is not one thing. Empirical generalizations and middle-range theory (Bass) coexist with grand substantive theories of competition (Hunt & Morgan), with analytical/game-theoretic modeling as a distinct theoretical practice (Moorthy), with typologies and frameworks as a legitimate theory form (Doty & Glick), and with grounded, case-based, and interpretive induction (Eisenhardt, Gioia). Each mode has its own standards of rigor, and a central pedagogical aim is to make students fluent in matching mode to question rather than defaulting to whatever their training favored.

The seminar closes by folding the credibility revolution back into theory construction. Replication and pre-registration (Simmons et al., the Open Science Collaboration) are reframed not as mere methodological hygiene but as theory testing—the mechanism by which the community decides a theoretical claim survives. Mechanism and process theory (Bullock et al., Pieters) sharpen what it means to explain rather than merely predict. The final weeks turn reflexive—theory about theory building (Corley & Gioia), problematization as a contribution engine (Alvesson & Sandberg)—and end on a live metatheoretical debate (service-dominant logic) and the question of where marketing’s grand theory, if any, now stands. The reading map uses two tags: [F] = Foundational (the metatheoretical canon a theorist is expected to know cold) and [R] = Frontier/Recent (an active debate or a still-executing agenda). Each week pairs at least one foundational anchor with a live edge. DOIs are reproduced as Crossref-verified; any work without a verified DOI is named without a link and flagged.

61.2 Week 1 — What theory is (and is not)

Topic. The metatheoretical baseline: what the word “theory” denotes, and what masquerades as theory but is not (data, lists of variables, diagrams, hypotheses, references).

Subtopics. The logic of explanation (the “why” behind relationships); the difference between description and theory; the components of a theoretical argument (constructs, relationships, boundary conditions, underlying logic).

Methods. Close reading of metatheory; argument reconstruction.

Key readings.

  • Sutton & Staw (1995), “What Theory Is Not,” Administrative Science Quarterly. doi:10.2307/2393788 — the field’s sharpest negative definition: five things that are routinely mistaken for theory. [F]
  • Whetten (1989), “What Constitutes a Theoretical Contribution?,” Academy of Management Review. doi:10.2307/258554 — the positive complement: what-how-why-who-where-when as the anatomy of a theoretical argument. [F]
  • DiMaggio (1995), “Comments on ‘What Theory Is Not,’” Administrative Science Quarterly. doi:10.2307/2393790 — three rival views of good theory (covering law, enlightenment, narrative) that defuse the false unity of the term. [F]

Debate. Is there one kind of “good theory,” or are explanation-as-prediction and explanation-as-understanding irreducibly different goals?

61.3 Week 2 — Conceptual contribution in marketing

Topic. Marketing’s own statement of what a conceptual contribution is, and the empirical claim that the field is producing fewer of them.

Subtopics. Types of conceptual contribution (envisioning, explicating, relating, debating); the conceptual/empirical division of labor; the institutional incentives that depress conceptual work.

Methods. Framework-building; content analysis of publication trends.

Key readings.

  • MacInnis (2011), “A Framework for Conceptual Contributions in Marketing,” Journal of Marketing. doi:10.1509/jmkg.75.4.136 — a 2×4 typology (identifying, relating, delineating, debating × kinds of thinking) that gives conceptual work a vocabulary and a quality bar. [F]
  • Yadav (2010), “The Decline of Conceptual Articles and Implications for Knowledge Development,” Journal of Marketing. doi:10.1509/jmkg.74.1.1 — documents the shrinking share of conceptual articles and argues the field’s knowledge engine depends on them. [R]
  • Summers (2001), “Guidelines for Conducting Research and Publishing in Marketing: From Conceptualization Through the Review Process,” JAMS. doi:10.1177/03079450094243 — the craft-level companion: how conceptualization is judged in the review process. [R]

Debate. Is the decline of conceptual articles a knowledge crisis or a healthy migration of theory into empirical papers’ front ends?

61.4 Week 3 — Constructs and the theory of measurement

Topic. How an abstract construct becomes measurable, and why measurement is itself a theoretical act.

Subtopics. Construct validity and the nomological net; the construct→measure mapping; convergent and discriminant validity; reliability; reflective vs. formative measurement.

Methods. Scale development; the multitrait–multimethod logic; confirmatory factor analysis as construct test.

Key readings.

  • Cronbach & Meehl (1955), “Construct Validity in Psychological Tests,” Psychological Bulletin. doi:10.1037/h0040957 — the founding statement that a construct is defined by its place in a nomological network. [F]
  • Churchill (1979), “A Paradigm for Developing Better Measures of Marketing Constructs,” Journal of Marketing Research. doi:10.2307/3150876 — the marketing paradigm: domain specification → item generation → purification → validity. [F]
  • Peter (1981), “Construct Validity: A Review of Basic Issues and Marketing Practices,” Journal of Marketing Research. doi:10.2307/3150948 — sharpens the validity vocabulary and audits how loosely marketing had been using it. [F]
  • Bagozzi & Phillips (1982), “Representing and Testing Organizational Theories: A Holistic Construal,” Administrative Science Quarterly. doi:10.2307/2392322 — theoretical, empirical, and derived concepts knit into a testable structural whole. [F]

Debate. Is construct validity a property of a measure or of a theory? Can a construct ever be validated outside the network of laws it figures in?

This module is developed as the chapter’s worked treatment in Section 61.17.1.

61.5 Week 4 — Philosophy of science in marketing

Topic. The realism–relativism debate that defined marketing’s self-conception as a science.

Subtopics. Logical empiricism and its critics; scientific realism; relativism/constructionism; falsificationism; what “scientific progress” means for an applied field.

Methods. Philosophical analysis; reconstruction of competing programs.

Key readings.

  • Anderson (1983), “Marketing, Scientific Progress, and Scientific Method,” Journal of Marketing. doi:10.1177/002224298304700403 — the relativist/constructionist challenge to naive empiricism. [F]
  • Peter & Olson (1983), “Is Science Marketing?,” Journal of Marketing. doi:10.1177/002224298304700412 — the provocative companion arguing science is itself a marketed product. [F]
  • Hunt (1990), “Truth in Marketing Theory and Research,” Journal of Marketing. doi:10.1177/002224299005400301 — the scientific-realist rebuttal defending a correspondence notion of truth. [F]

Debate. Is marketing knowledge discovered (realism) or constructed (relativism)? Does the answer change how we should evaluate theories?

61.6 Week 5 — Middle-range theory and empirical generalizations

Topic. Theory between grand systems and raw data: lawlike regularities that replicate across contexts.

Subtopics. Mertonian middle-range theory; empirical generalizations as a knowledge mode; the relationship between regularity and explanation; meta-analysis as generalization machinery.

Methods. Empirical-generalization methodology; meta-analysis; the generalize-then-explain workflow.

Key readings.

  • Bass (1995), “Empirical Generalizations and Marketing Science: A Personal View,” Marketing Science. doi:10.1287/mksc.14.3.g6 — the manifesto for empirical generalizations as the durable, replicable core of marketing knowledge. [F]
  • Hunt & Morgan (1995), “The Comparative Advantage Theory of Competition,” Journal of Marketing. doi:10.1177/002224299505900201 — a worked example of building up from generalizations toward a substantive general theory (revisited in Week 14). [R]

Debate. Is an empirical generalization a theory, a fact to be explained, or both? Can a field accumulate knowledge through regularities without underlying mechanism?

61.7 Week 6 — Substantive grand theory: competition

Topic. What an ambitious, field-spanning substantive theory in marketing looks like, using the theory of competition as the case.

Subtopics. Resource-advantage (R-A) theory; disequilibrium vs. neoclassical equilibrium; theory as an evolutionary process; the criteria for a “general theory.”

Methods. Theory construction; comparative evaluation of research programs.

Key readings.

  • Hunt & Morgan (1995), “The Comparative Advantage Theory of Competition,” Journal of Marketing. doi:10.1177/002224299505900201 — marketing’s most fully developed home-grown general theory; a model of how to assemble a theory from premises. [F]
  • Hunt (1983), “General Theories and the Fundamental Explananda of Marketing,” Journal of Marketing. doi:10.1177/002224298304700402 — what marketing’s grand theory would have to explain; sets the explananda. [F]

Debate. Does marketing need a single general theory, or is the field constitutively a federation of middle-range theories? Are the explananda of Hunt (1983) still the right ones?

61.8 Week 7 — Theorizing in consumer behavior: validity and application

Topic. How a behavioral, experiment-heavy subfield builds theory, and the external-validity debate that shaped it.

Subtopics. Internal vs. external validity; effects-application vs. theory-application research; the role of the experiment in theory testing; generalizing from the lab.

Methods. Experimental design as theory test; the theory-vs-application distinction in research design.

Key readings.

  • Calder, Phillips & Tybout (1981), “Designing Research for Application,” Journal of Consumer Research. doi:10.1086/208856 — the classic split between theory-testing and effects-application research and what each demands of design. [F]
  • Lynch (1982), “On the External Validity of Experiments in Consumer Research,” Journal of Consumer Research. doi:10.1086/208919 — reframes external validity as a function of background-factor interactions, not surface realism. [F]

Debate. When is external validity a genuine threat to a theory test versus a category error? Does “generalizability” mean the same thing for theory and for application?

61.9 Week 8 — Models as theory: analytical theorizing

Topic. Formal/analytical modeling as a distinct theoretical practice with its own warrants.

Subtopics. Assumptions and tractability; the role of mathematical deduction; what a model “proves”; robustness vs. realism; game theory as marketing theory.

Methods. Analytical/game-theoretic modeling; comparative statics; the assumption–implication contract.

Key readings.

  • Moorthy (1993), “Theoretical Modeling in Marketing,” Journal of Marketing. doi:10.2307/1252029 — the definitive statement of what analytical models contribute and how to read them as theory. [F]
  • Sutton & Staw (1995), “What Theory Is Not,” Administrative Science Quarterly. doi:10.2307/2393788 — revisited from the modeling side: a model’s equations are not theory unless the why is articulated. [F]

Debate. Do unrealistic assumptions invalidate a model, or is realism the wrong test (the Friedman “as-if” position)? Is a formal result a theory or a tool?

61.10 Week 9 — Typologies, taxonomies, and frameworks

Topic. Classification as theory: when a typology or framework is a genuine theoretical contribution rather than a descriptive convenience.

Subtopics. Typologies vs. taxonomies vs. classifications; ideal types and configurations; grand theories vs. middle-range frameworks; the testability of typological theories.

Methods. Configurational analysis; the logic of ideal-type construction.

Key readings.

  • Doty & Glick (1994), “Typologies as a Unique Form of Theory Building: Toward Improved Understanding and Modeling,” Academy of Management Review. doi:10.2307/258704 — the canonical defense of typologies as theories with falsifiable, configurational predictions. [F]
  • MacInnis (2011), “A Framework for Conceptual Contributions in Marketing,” Journal of Marketing. doi:10.1509/jmkg.75.4.136 — situates delineating/relating frameworks within the broader space of conceptual contribution (revisited from Week 2). [R]

Debate. When does a 2×2 stop being a heuristic and become a theory? Can a typology be falsified, or only found more or less useful?

61.11 Week 10 — Grounded, case-based, and interpretive theorizing

Topic. Building theory up from qualitative data, and the rigor standards that make induction credible.

Subtopics. Grounded theory; theory-building from cases; the Gioia methodology (first-order/second-order coding, the data structure); analytic generalization; trustworthiness criteria.

Methods. Multiple-case design; constant comparison; systematic qualitative coding.

Key readings.

  • Eisenhardt (1989), “Building Theories from Case Study Research,” Academy of Management Review. doi:10.2307/258557 — the roadmap for inducting testable theory from a small number of cases. [F]
  • Gioia, Corley & Hamilton (2013), “Seeking Qualitative Rigor in Inductive Research: Notes on the Gioia Methodology,” Organizational Research Methods. doi:10.1177/1094428112452151 — the now-standard template for demonstrating rigor in interpretive theory-building. [R]

Debate. Does systematizing qualitative coding buy rigor at the cost of interpretive insight? Is “analytic generalization” a real warrant or a relabeling of small-sample induction?

61.12 Week 11 — Replication and credibility as theory testing

Topic. The credibility revolution recast as the community’s mechanism for deciding which theoretical claims survive.

Subtopics. Researcher degrees of freedom and false positives; direct vs. conceptual replication; pre-registration; the reproducibility of an evidence base; what a failed replication does (and does not) say about a theory.

Methods. Pre-registration; multi-lab replication; meta-analytic reproducibility assessment.

Key readings.

  • Simmons, Nelson & Simonsohn (2011), “False-Positive Psychology,” Psychological Science. doi:10.1177/0956797611417632 — how undisclosed flexibility manufactures significant results, motivating pre-registration. [F]
  • Open Science Collaboration (2015), “Estimating the Reproducibility of Psychological Science,” Science. doi:10.1126/science.aac4716 — the large-scale replication audit that reset the field’s priors. [R]
  • Lynch, Bradlow, Huber & Lehmann (2015), “Reflections on the Replication Corner: In Praise of Conceptual Replications,” IJRM. doi:10.1016/j.ijresmar.2015.09.006 — marketing’s own argument that conceptual replication tests theory, not just results. [R]

Debate. Does a failed direct replication falsify a theory or only an operationalization? Are conceptual replications more theory-relevant but less diagnostic?

61.13 Week 12 — Mechanism and process theory

Topic. The shift from “does X affect Y” to “through what mechanism”—and the inferential cost of taking mediation seriously.

Subtopics. Mechanistic vs. variance explanation; mediation as a theory of process; the endogeneity of mediators; experimental causal-chain designs.

Methods. Mediation and moderated-mediation analysis; experimental manipulation of the mediator; causal-inference critiques.

Key readings.

  • Bullock, Green & Ha (2010), “Yes, But What’s the Mechanism? (Don’t Expect an Easy Answer),” Journal of Personality and Social Psychology. doi:10.1037/a0018933 — why statistical mediation rarely identifies a mechanism without strong, often untestable, assumptions. [F]
  • Pieters (2017), “Meaningful Mediation Analysis: Plausible Causal Inference and Informative Communication,” Journal of Consumer Research. doi:10.1093/jcr/ucx081 — marketing’s constructive response: how to do and report mediation so it bears on theory. [R]
  • Pham (2013), “The Seven Sins of Consumer Psychology,” Journal of Consumer Psychology. doi:10.1016/j.jcps.2013.07.004 — a catalog of theorizing pathologies, including mechanism-by-assertion. [R]

Debate. Is process evidence necessary for a theoretical contribution, or has the field fetishized mediation at the expense of robust effects?

61.14 Week 13 — Theory about theory building

Topic. The reflexive turn: theorizing the act of contribution itself, and how novelty is generated.

Subtopics. What “advancing theory” means; revelatory vs. incremental contribution; problematization vs. gap-spotting as engines of research questions; the rhetoric of contribution.

Methods. Metatheoretical analysis; assumption-challenging research design.

Key readings.

  • Corley & Gioia (2011), “Building Theory About Theory Building: What Constitutes a Theoretical Contribution?,” Academy of Management Review. doi:10.5465/amr.2009.0486 — maps contribution along originality × utility and argues for prescient theorizing. [F]
  • Alvesson & Sandberg (2011), “Generating Research Questions Through Problematization,” Academy of Management Review. doi:10.5465/amr.2009.0188 — replaces gap-spotting with the disciplined challenging of a literature’s assumptions. [R]

Debate. Is “interesting” theory (assumption-violating) more valuable than careful, cumulative theory? Does the contribution imperative reward novelty over truth?

61.15 Week 14 — The future of marketing metatheory

Topic. A live metatheoretical debate and a synthesis: where marketing’s theorizing is, and where it should go.

Subtopics. Service-dominant logic as a candidate integrating metatheory; foundational premises vs. relabeling; the relationship between metatheory and empirical generalization; the agenda for the next generation of theorists.

Methods. Conceptual synthesis; evaluation of a metatheoretical program.

Key readings.

  • Vargo & Lusch (2004), “Evolving to a New Dominant Logic for Marketing,” Journal of Marketing. doi:10.1509/jmkg.68.1.1.24036 — the most-debated recent attempt at a unifying metatheory; a capstone test case for the semester’s criteria. [R]
  • Hunt (1983), “General Theories and the Fundamental Explananda of Marketing,” Journal of Marketing. doi:10.1177/002224298304700402 — the benchmark against which any candidate metatheory’s explanatory scope is judged (revisited from Week 6). [F]

Debate. Does service-dominant logic add explanatory power, or relabel existing constructs? Should the field invest in grand metatheory at all, or in better-tested middle-range theory?

61.16 Foundational vs. frontier at a glance

Foundational core (every theorist must know): Cronbach & Meehl (1955); Churchill (1979); Peter (1981); Bagozzi & Phillips (1982); Anderson (1983); Hunt (1983); Calder, Phillips & Tybout (1981); Lynch (1982); Peter & Olson (1983); Sutton & Staw (1995); Whetten (1989); DiMaggio (1995); Eisenhardt (1989); Doty & Glick (1994); Bass (1995); Hunt (1990); MacInnis (2011); Corley & Gioia (2011); Simmons, Nelson & Simonsohn (2011). These are the canon of how marketing argues about theory and measurement, and the field still builds on each.

Frontier / actively updated (refresh each edition): Yadav (2010); Bullock, Green & Ha (2010); Alvesson & Sandberg (2011); Gioia, Corley & Hamilton (2013); Pham (2013); Vargo & Lusch (2004); Open Science Collaboration (2015); Lynch et al. (2015); Pieters (2017). The split is pedagogical, not chronological: Vargo & Lusch (2004) is “frontier” because the debate it launched is still open, while Churchill (1979) is foundational because the paradigm it set is still executed. Each module pairs a foundational anchor with a live edge so students see both the settled craft and its contested frontier.

61.17 How this chapter expands

The weekly map is a backbone, not a ceiling. The chapter is designed to grow along several axes.

  1. A measurement-theory spine as a parallel track. Weeks 3 and 10 already touch construct validity and qualitative rigor; a future edition should run a measurement companion through every week—formative vs. reflective specification, common-method bias, measurement invariance, and the construct-proliferation problem—so the seminar teaches how marketing operationalizes its theories, not only how it states them. The worked section below models this.
  2. A credibility-revolution module per tradition. Replication (Week 11) is taught generally; the field now needs tradition-specific versions—what pre-registration means for analytical modeling, for structural estimation, and for interpretive work, each of which has a different relationship to falsification.
  3. Emerging modes of theorizing. Computational and simulation-based theory, theory generated with or about machine-learning systems, and abduction/retroduction as an explicit logic deserve their own modules as the field’s toolkit widens; each should follow the template of foundational anchor + frontier paper + a debate about what counts as a contribution.
  4. An internationalized and cross-disciplinary canon. The current map is US/JM-centric and leans on management metatheory; European philosophy-of-marketing work and adjacent debates in economics and psychology would broaden how the seminar weights realism, interpretivism, and formalism.

The following section supplies the worked treatment the map points to.

61.17.1 Construct validity as a theory of measurement

The deepest idea in the measurement week is that a construct is not validated against a single criterion but constituted by its position in a network of relationships. Cronbach & Meehl (doi:10.1037/h0040957) call this network the nomological net: the interlocking set of theoretical laws relating the construct to other constructs, and correspondence rules relating constructs to observables. A construct \(C\) becomes scientifically meaningful only when the net specifies enough of these links that some are testable. Construct validation is therefore inseparable from theory testing—you cannot confirm that a measure taps \(C\) without simultaneously assuming the theory in which \(C\) is embedded.

Make the mapping explicit. Let an unobservable construct \(\eta\) generate a set of observable indicators \(x_1,\dots,x_p\) under a reflective measurement model, \[ x_j = \lambda_j\,\eta + \delta_j, \qquad \mathbb{E}[\delta_j]=0,\quad \mathrm{Cov}(\eta,\delta_j)=0, \tag{61.1}\] where \(\lambda_j\) is the loading of indicator \(j\) on the construct and \(\delta_j\) is measurement error. The indicators are effects of the construct, not its definition—reverse the arrow (indicators causing the construct) and the model becomes formative, with entirely different validity logic. Equation Equation 61.1 is the formal content of “the construct → measure mapping”: the theory asserts a latent common cause, and the data either support that structure or do not.

Reliability asks how much of the indicators’ variance is the construct rather than noise. For a composite of \(p\) indicators it is summarized by construct reliability, \[ \rho_c=\frac{\bigl(\sum_{j}\lambda_j\bigr)^2} {\bigl(\sum_{j}\lambda_j\bigr)^2+\sum_{j}\mathrm{Var}(\delta_j)}, \tag{61.2}\] which is high when loadings are large relative to error variances. Reliability is necessary but not sufficient: a perfectly reliable measure can still measure the wrong thing.

That is where validity enters, and Churchill’s paradigm (doi:10.2307/3150876) operationalizes it as a disciplined sequence—specify the construct’s domain, generate items from that domain, purify the measure, then assess validity—so that the eventual scale earns its claim to tap \(\eta\). Two validities do the heavy lifting and have a clean formal reading. Convergent validity requires that indicators of the same construct share substantial variance: the average variance extracted, \[ \mathrm{AVE}=\frac{\sum_{j}\lambda_j^{2}} {\sum_{j}\lambda_j^{2}+\sum_{j}\mathrm{Var}(\delta_j)}, \tag{61.3}\] should exceed \(0.5\), meaning the construct explains more of its indicators’ variance than error does. Discriminant validity requires that distinct constructs be empirically separable: a construct’s AVE should exceed its squared correlation with any other construct, \(\mathrm{AVE}_\eta > r_{\eta\xi}^{2}\), so the two are not merely the same latent variable under two names. Together these formalize the multitrait–multimethod intuition—indicators of one construct cohere, indicators of different constructs diverge—and Peter (1981) (doi:10.2307/3150948) is the audit showing how loosely marketing once used these terms.

The payoff is the seminar’s central claim: a construct becomes testable only inside a nomological network. Suppose theory predicts \(C\) relates positively to an antecedent \(A\) and an outcome \(O\). The construct earns nomological validity when its measure reproduces those predicted relationships, \[ \widehat{\mathrm{Cor}}(\hat\eta, A) > 0 \quad\text{and}\quad \widehat{\mathrm{Cor}}(\hat\eta, O) > 0, \tag{61.4}\] as the embedding theory requires. The logic is unavoidably joint: a failure of Equation 61.4 is ambiguous between a bad measure and a bad theory, and only further links in the net can apportion the blame. Bagozzi & Phillips (1982) (doi:10.2307/2392322) push this furthest, representing theoretical, empirical, and derived concepts as one structural system so that measurement and theory are tested in a single holistic construal. This is why measurement is a theoretical act: to say what a construct is, one must commit to the lawful relationships it stands in—and those commitments are exactly what an empirical study puts at risk.

61.18 Key Takeaways