Which Theoretical Framework Should a Business or Management Dissertation Use? Fourteen Frameworks Compared (2026)

Supervisors reject business dissertation proposals for one theoretical reason more often than any other: the theory named in Chapter 2 does not operate at the level the study measures. A student surveys 220 employees about engagement and frames the study in the resource-based view, which is a theory about firms, not people. A student compares two firms’ adoption of an ERP system and frames it in social exchange theory, which is about dyads. The framework has to sit at the same level as the data. This guide compares the fourteen theories that appear most often in management, marketing, finance, human resources and information systems dissertations, sets out what each one actually explains, and gives the rule for matching one to your question.

The short version

• Choose the theory by level of analysis first: individual, dyad or team, firm, or field and industry. A theory one level away from your data is the most common fatal flaw in a business Chapter 2.
• A theoretical framework is not a literature review. It is the small set of constructs and propositions that generate your hypotheses; everything else belongs in the review. Our explainer on the difference between a theoretical and a conceptual framework draws the line.
• Most business dissertations need one primary theory and at most one supporting one. Three theories in a master’s dissertation is a warning sign, not a strength.
• Cite the founding source, not a textbook summary of it. Committees check whether Barney (1991) is in your reference list when you claim to use the resource-based view.
• State the boundary conditions. Every theory below has a well-known critique; naming it in your limitations section is worth more marks than pretending it does not exist.

The fourteen frameworks, compared

Strategy questions: resources, capabilities and knowledge

If your dependent variable is firm performance, competitive advantage or growth, you are in the strategy family. The resource-based view holds that resources which are valuable, rare, imperfectly imitable and non-substitutable can sustain advantage, and Barney’s 1991 article in the Journal of Management is the source to cite, with Wernerfelt (1984) as the earlier statement of the idea. Its best-known critique, Priem and Butler’s 2001 exchange in the Academy of Management Review, argues that the core proposition risks being tautological: valuable resources are identified by the advantage they produce. You handle this by defining and measuring the resource independently of the outcome — measure the training investment, not the “human-capital advantage”.

Dynamic capabilities extends the view to changing environments: Teece, Pisano and Shuen (1997) describe the capacity to sense opportunities, seize them and reconfigure the asset base. Choose it when your study is about change — digital transformation, market entry, crisis response — rather than about a static resource stock. The knowledge-based view (Grant, 1996) narrows the resource to knowledge itself and suits studies of knowledge transfer across units or alliances, while absorptive capacity (Cohen and Levinthal, 1990) explains why a firm’s existing knowledge determines what new knowledge it can use, which is the natural frame for R&D, technology transfer and innovation-uptake questions.

Governance questions: contracts, incentives and signals

Finance, accounting and corporate governance dissertations mostly live in the contractual family. Agency theory (Jensen and Meckling, 1976; Eisenhardt, 1989) treats the firm as a nexus of contracts and models the divergence of interest between shareholders and managers; it is the default frame for studies of executive compensation, board composition, ownership concentration and earnings management. Stewardship theory (Davis, Schoorman and Donaldson, 1997) is its deliberate mirror image, assuming managers are motivated to act as stewards, and it is a defensible frame when your sample is family firms, cooperatives or non-profits, where the agency assumption is contested.

Transaction cost economics, from Coase (1937) through Williamson, explains governance choice — make, buy or ally — through asset specificity, uncertainty and frequency. Use it when the outcome is a boundary decision: outsourcing, vertical integration, franchising, supplier relationships. Signaling theory (Spence, 1973; Connelly and colleagues, 2011) handles information asymmetry from the other direction, and it frames studies of what firms disclose voluntarily, how markets react to IPO prospectuses, or why employer branding influences applicant quality.

Field-level questions: legitimacy, conformity and competing logics

When the puzzle is why organizations in a sector all do the same thing — adopt the same sustainability report, the same quality certification, the same governance code — the answer is usually institutional rather than economic. DiMaggio and Powell (1983) identify three mechanisms of isomorphism: coercive pressure from regulation and resource dependence, mimetic imitation of visible peers under uncertainty, and normative pressure through professional training and associations. Naming the specific mechanism your data test is what separates a real institutional framework from a decorative one; a study of ESG disclosure that shows regulatory deadlines drive adoption is testing coercive isomorphism, and should say so.

Institutional logics (Thornton and Ocasio, 1999) is the finer-grained descendant, useful when two belief systems collide inside one organization — the professional logic against the managerial logic in a hospital, the market logic against the community logic in a cooperative bank. It is a qualitative-first framework: it earns its place in a case study or a set of interviews, rarely in a survey.

People questions: employees, teams and top managers

Human resources and organizational behavior dissertations survey individuals, so the framework must be an individual-level theory. The job demands-resources model (Bakker and Demerouti, 2007) is the workhorse: job demands drive a health-impairment path towards burnout, job resources drive a motivational path towards engagement, and resources buffer demands. It is a good fit because it generates testable mediation and moderation hypotheses rather than a list of factors. Self-determination theory (Deci and Ryan, 2000) suits questions about intrinsic motivation, autonomy and the effect of incentive design, and social exchange theory (Blau, 1964) frames reciprocity questions such as perceived organizational support and its return in discretionary effort.

Upper echelons theory (Hambrick and Mason, 1984) is the exception in this group: it links the observable characteristics of the top management team — tenure, functional background, education, diversity — to strategic choices and firm outcomes. It sits between the individual and the firm and is the standard frame for board- and CEO-level archival studies.

Technology and adoption questions

Information systems and digital marketing dissertations usually need an acceptance model. The technology acceptance model (Davis, 1989) explains intention to use through perceived usefulness and perceived ease of use; UTAUT (Venkatesh, Morris, Davis and Davis, 2003) consolidates eight earlier models into performance expectancy, effort expectancy, social influence and facilitating conditions, with moderators for age, gender, experience and voluntariness. Both are individual-level. If your unit is the firm adopting the technology rather than the person using it, the technology-organization-environment framework is the correct level, and diffusion of innovations (Rogers) covers spread across a population over time. In services and retail, the SERVQUAL dimensions of Parasuraman, Zeithaml and Berry (1988) remain the standard frame for perceived service quality.

Choosing yours, in four questions

1. What is the unit in your dataset? One row per employee means an individual-level theory. One row per firm-year means a firm-level theory. One row per transaction or contract means transaction cost economics or agency theory. Get this right and half the candidates disappear.
2. Is the outcome an explanation or a prediction? Institutional theory and institutional logics explain patterns and suit qualitative designs; acceptance models and job demands-resources predict variance and suit surveys with structural models. Our comparison of covariance-based and partial least squares SEM covers the estimation choice that follows.
3. Can you measure the constructs the theory names? A framework you cannot operationalize is decoration. If you cannot measure asset specificity in your sample, transaction cost economics is not your theory, whatever the topic.
4. What does the theory forbid? A framework that is consistent with any result is not doing work. Write the prediction the theory rules out, and design the study so that outcome is observable.

Once the framework is chosen, the chapter itself follows a standard shape: define the constructs, state the propositions, derive your hypotheses from them, and show the model in a diagram whose arrows match the hypotheses exactly. Our step-by-step on writing the theoretical framework covers that mechanics, the pillar guide to writing a business management dissertation places the chapter in the whole document, and the same exercise in a different discipline — our guide to choosing a theoretical framework for an education dissertation — shows how differently the theory catalogue looks one faculty away. For a practice-oriented program, our step-by-step MBA dissertation guide shows how much theory an applied brief is expected to carry.

Frequently asked questions

How many theories should a business dissertation use?

One primary framework, and a second only if it does work the first cannot. A common defensible pairing is a firm-level theory for the context and an individual-level theory for the mechanism, with an explicit statement of which hypotheses come from which. Three or more theories in a master’s dissertation usually signals that none of them is being tested.

Can I use the resource-based view for a survey of employees?

Not as the primary framework. The resource-based view explains differences between firms, and a sample of employees inside one or two firms cannot test it. Either move to a firm-level design with multiple organizations, or choose an individual-level theory such as the job demands-resources model or social exchange theory and treat the resource-based view as background.

What is the difference between a theoretical and a conceptual framework in business research?

The theoretical framework is an existing theory you adopt, with its established constructs and propositions. The conceptual framework is the specific model you build for your study, often combining constructs from more than one theory and shown as a diagram of your hypotheses. Most business dissertations need both, in that order.

Is institutional theory suitable for a quantitative dissertation?

Yes, if you measure a mechanism rather than the concept in general. Studies operationalize coercive pressure through regulatory events or deadlines, mimetic pressure through the prior adoption rate among visible peers, and normative pressure through professional membership or accreditation. A study that only asserts that institutional pressures exist is not testing the theory.

Do I need a theoretical framework for a case-study dissertation?

Yes, but its role is different. In a case study the framework provides the sensitizing concepts that guide data collection and the initial coding structure, and the analysis may extend or challenge it. Say up front whether the case is a test of the theory or a source of theory-building; the two require different write-ups.

Which framework should a marketing dissertation use?

It depends on the outcome. Consumer-level intention or behavior studies use the theory of planned behavior or the technology acceptance family; service quality and satisfaction studies use the SERVQUAL dimensions and expectancy-disconfirmation; brand and firm-level studies use the resource-based view or signaling theory. Choose by the unit measured in your data, not by the topic label.