Your guide has read the draft and written one line in the margin: “Where is your conceptual framework?” You have forty pages of literature review, five objectives and a questionnaire ready to distribute. And now the whole thing is stalled on a diagram you were never taught to draw.
This costs more than a weekend. In most Indian management departments a framework objection means the synopsis goes back to the committee, and committees meet once a cycle. A single missing diagram can push submission by a full semester — while your questionnaire sits undistributed because the constructs it measures have not been fixed yet.
Build your framework and Chapter 2 with Tesify — free to start
What a conceptual framework actually is
A conceptual framework is a diagram plus a paragraph that states which variables your study includes, how you expect them to relate, and which theory says so. That is the entire definition. It is not a summary of your literature review, it is not a list of objectives, and it is not a flowchart of your research process — the “research design” flowchart showing “literature review → questionnaire → data collection → analysis” is a different diagram and its presence is often what makes a guide say the framework is missing.
Two words get confused, and Indian examiners ask about the difference:
- The theoretical framework is the established theory you are borrowing — the Technology Acceptance Model, the Theory of Planned Behaviour, the Resource-Based View, Herzberg’s two-factor theory, SERVQUAL, the Job Demands-Resources model. It exists in the literature with a named author.
- The conceptual framework is your specific application of it: the particular variables you are measuring, in your particular context, with the relationships you are testing. It is yours, and it should not exist anywhere in the literature exactly as you have drawn it.
A dissertation that borrows a published model unchanged and applies it to a new sample is acceptable at Master’s level, but say so honestly: “this study adapts the model proposed by [author], extending it with [your addition] in the Indian [sector] context.”
The six-step build
Step 1: Write your research question as a sentence containing exactly two things
Strip the question down to a cause and an effect. “What is the effect of perceived organisational support on employee retention among IT professionals in Bengaluru?” contains one independent variable, one dependent variable and a population. If you cannot reduce your question to that shape, the framework will not resolve — and neither will the analysis chapter.
Step 2: Name the theory you are standing on
Go back through the literature review you have already written and find the theory the strongest papers keep citing. That is your anchor. State it explicitly: “This study is grounded in social exchange theory, which holds that employees reciprocate the support they perceive from their organisation.”
An examiner’s question here is nearly guaranteed: “why this theory and not that one?” Prepare two sentences naming the alternative you rejected and why it fits your context less well. If your literature review is not yet organised well enough to surface the dominant theory, our guide to writing a literature review that argues rather than lists covers the synthesis matrix that makes this obvious.
Step 3: Classify every variable into one of five roles
This is the step that turns a vague model into a testable one. Each variable in your study is exactly one of:
- Independent variable — the presumed cause you are studying.
- Dependent variable — the outcome.
- Mediator — the mechanism through which the cause produces the effect. It sits between them and is itself caused by the independent variable. Job satisfaction mediating between perceived support and retention, for example.
- Moderator — a variable that changes the strength or direction of the relationship without being caused by it. Tenure, gender, firm size, work arrangement.
- Control variable — something you measure and account for so that it does not confound the result, without theorising about it.
The single most common framework rejection in an MBA dissertation is a variable drawn as a mediator when the theory says it is a moderator, or vice versa. The test is simple: ask whether your independent variable plausibly causes the variable in question. If yes, it is a mediator. If it merely changes the picture — the relationship is stronger for younger employees — it is a moderator.
Step 4: Draw the diagram with directional arrows only
Boxes for constructs, single-headed arrows for hypothesised causal paths, double-headed arrows only for correlations you are not claiming direction on. Moderators attach to the arrow, not to the box. Label every arrow H1, H2, H3 and so on.
Keep it to one page and to no more than about seven constructs. A framework with fourteen boxes is not ambitious, it is a dissertation you cannot finish — each additional construct means another validated scale, more questionnaire items, higher respondent fatigue and a larger sample.
Step 5: Convert every arrow into a hypothesis
Each labelled arrow becomes one numbered hypothesis, stated in the alternative form and in the direction the theory predicts:
H1: Perceived organisational support has a significant positive effect on employee retention.
H2: Job satisfaction mediates the relationship between perceived organisational support and employee retention.
H3: Tenure moderates the relationship between perceived organisational support and employee retention, such that the relationship is stronger for employees with less than three years of service.
Every arrow must have a hypothesis and every hypothesis must have an arrow. When those two lists match exactly, the framework is internally consistent — and that consistency is the first thing a committee checks.
Step 6: Attach a measurable scale to every construct
The framework is not finished until each box has an instrument behind it. For each construct, name the scale you will use, its author and year, and its number of items. Use published, validated scales wherever they exist rather than writing your own items — the reliability evidence already exists and you will need it. Report the Cronbach’s alpha you compute on your own sample, and see our note on what counts as an acceptable alpha for the value to defend.
If your framework contains a construct for which you cannot find a scale, you have a problem to solve now, before data collection — not in Chapter 4.
How the framework determines your analysis chapter
A properly built framework tells you exactly what to run, which is why building it late causes so much damage:
| Framework shape | Analysis it implies |
|---|---|
| One IV, one DV | Correlation and simple linear regression |
| Several IVs, one DV | Multiple regression, with collinearity diagnostics |
| IV → mediator → DV | Mediation analysis with bootstrapped indirect effects |
| IV → DV, moderated by M | Moderated regression with an interaction term and simple slopes |
| Multiple latent constructs with several paths | Structural equation modelling, preceded by confirmatory factor analysis |
| Group differences on one outcome | t-test or one-way ANOVA with post-hoc comparisons |
If you are unsure which test your design implies, our decision table for choosing a statistical test works from the question type down to the method. And if your multiple regression returns a significant model with no significant predictors, that is a collinearity symptom rather than a framework failure — the diagnosis and the four legitimate remedies are set out in our guide to handling multicollinearity with VIF.
The four objections that send a framework back
- “This is a process flowchart, not a framework.” You have drawn your research steps rather than your variables. Redraw with constructs in the boxes.
- “Where is the theory?” The diagram exists but no named theory justifies the arrows. Add the anchoring theory and cite it.
- “Too many variables.” Cut to the constructs your objectives actually require. Objectives and boxes should correspond.
- “How will you measure this?” A construct with no instrument behind it. Name the scale in the framework section itself, not only in the methodology.
If the committee has already returned your synopsis, resist the urge to resubmit the same document with a diagram pasted in. Our guide to repairing a synopsis rejected by the doctoral committee covers how to read what the objection really was, because a framework comment often signals a scope problem underneath it.
Get the framework and Chapter 2 written this week
The framework itself is a diagram and three paragraphs. What takes the fortnight is everything around it — the theory justification, the hypothesis development section where each arrow is argued from cited evidence, the operationalisation table listing every construct against its scale, and the references formatted to your department’s style.
Tesify drafts that structure from your own literature notes and your own variable list. You decide the theory and the variables; it produces the hypothesis development section, the operationalisation table and a consistently formatted bibliography, in the format your university ordinance requires. It supports honest writing — the research stays yours, and every claim traces to a source you supplied.
Start free with Tesify and get your framework chapter drafted
Frequently asked questions
Does Tesify cost anything to try?
There is a free tier, so you can create a project and see the drafted structure before paying anything. Paid plans are billed in rupees for Indian users; check the current pricing page for the exact figure, since plans change. The point of the free tier is that you can judge the output against your own department’s expectations before spending a scholar’s money.
Is using an AI writing assistant allowed for an Indian MBA dissertation?
Indian institutions increasingly permit AI assistance for drafting, structuring and language, while requiring that the research, the data and the analysis are your own, and many now require a declaration of what assistance was used. The rule that matters is your own university’s policy — read it before you start, and disclose if disclosure is required. Passing off generated text as unaided work is the misconduct, not using a tool to draft.
What happens to my data and my draft?
Your project content belongs to you and is visible in your own account. Before uploading anything, check your institution’s rules on unpublished data — if your dissertation involves confidential company information from your internship organisation, the non-disclosure terms you signed apply to every tool you use, including reference managers and cloud storage.
Can a conceptual framework have more than one dependent variable?
Yes, but each additional outcome multiplies the analysis and the sample requirement, and at Master’s level it usually signals a question that has not been narrowed enough. If you genuinely have two outcomes, be sure your objectives state both and that your chosen method can handle them — otherwise run the analysis separately for each and say so.
Where does the conceptual framework go in the dissertation?
Conventionally at the end of the literature review chapter, after the theory and the research gap have been established, immediately followed by the hypothesis development section. Some departments place it in the methodology chapter instead. It also appears in condensed form in the synopsis. Follow your department’s approved format.
Can I use a framework from a published paper as it is?
You may adapt one with full citation, and replication in a new context is a legitimate Master’s contribution. Reproducing the diagram itself may require permission from the publisher, so redraw it and cite the source as “adapted from”. What you cannot do is present another author’s model as your own construction.
How many hypotheses should an MBA dissertation have?
However many arrows your framework contains — commonly three to eight in a one-year dissertation. There is no prescribed number in any UGC or university regulation. What is judged is whether each hypothesis follows from cited evidence and whether your data and method can actually test it.
My study is qualitative. Do I still need a conceptual framework?
You need a conceptual framework but not hypotheses. In qualitative work it maps the concepts you are exploring and their expected relationships, and it functions as a starting orientation that your findings may revise. Say explicitly that the framework is provisional and describe how the data was allowed to challenge it.