Conceptual & Theoretical Framework

Theoretical or conceptual

The terms are used loosely and sometimes interchangeably, which is why so many students find this confusing. In most doctoral work the distinction is worth holding.

A theoretical framework is borrowed. It is an established theory from your field — a named theory with a literature behind it — that you adopt as the lens through which your problem is understood. It explains why things happen in general, and you are applying that explanation to your case, extending it, or testing whether it holds.

A conceptual framework is yours. It sets out the specific concepts in your study and the relationships you expect between them, usually drawing on more than one theory plus the empirical literature. It explains what you think is going on in this study.

Many theses have both: a theory that grounds the work, and a conceptual model showing how the particular variables or concepts relate in this context. Some fields expect only one. Recent theses passed in your own department will tell you what is expected far more reliably than any general account.

What is not acceptable in either case is a diagram of boxes and arrows with no argument behind it. Arrows are claims. Each one says that something relates to something else, in a direction, and each needs a reason drawn from theory or evidence.

Why most frameworks do no work

The recurring failure is not a wrong framework but an inert one. It is named in the proposal, drawn in chapter two, and then plays no part in anything that follows.

You can see it in a finished thesis. The framework section ends, the methodology begins, and nothing in the design was determined by the framework. The interview guide asks about whatever seemed interesting. The variables were chosen because previous studies used them. The discussion interprets findings against the literature generally rather than against the framework specifically.

A framework that is working shows up repeatedly:

• In the questions. The framework identifies what is worth asking, so the research questions follow from it rather than sitting beside it.
• In the design. It determines what has to be measured or explored — if a construct is in your framework and nothing in your study touches it, one of the two is wrong.
• In the instrument. Interview questions and survey items map back to framework concepts, and you should be able to show that mapping.
• In the analysis. Especially in deductive work, where the framework says what relationships to test.
• In the discussion. Findings are interpreted against what the framework predicted, including where it did not hold — which is often the most interesting result.

The practical test: delete the framework section from your thesis. Does anything else stop making sense? If not, it was not doing any work.

Building one

Frameworks are usually built backwards from the question rather than chosen first and applied.

1. State what you think is going on. In plain language, before any theory: what do you believe explains the phenomenon, and what would you expect to see if you were right?
2. Find who has thought about this. Which theories address this kind of process? Read enough of each to know what it actually claims, rather than adopting one from a summary.
3. Choose, and be able to say why not the others. This is the question you will be asked. Knowing which competing theory you rejected, and on what grounds, is what separates a chosen framework from an inherited one.
4. Identify the concepts that matter for your study. Not everything in the theory — the parts your question actually engages.
5. Specify the relationships. What affects what, in which direction, and why. Every arrow needs a justification you can point to.
6. Check it against your design. Can you actually observe or measure each concept? A framework containing something you have no way of accessing is a framework that will embarrass you in the viva.

On adapting existing frameworks: it is entirely legitimate and often stronger than inventing one. What must be explicit is what you changed and why — adding a construct, removing one that does not apply in your context, or altering a relationship. That reasoning is a contribution in itself.

Diagrams, and their traps

Most conceptual frameworks are presented as a diagram, and diagrams are useful precisely because they force you to be explicit. They also invite three specific problems.

Arrows that mean different things. An arrow can indicate a causal effect, a correlation, a temporal sequence, or that one thing is a component of another. Using the same arrow for all of them hides distinctions that matter. Use a key, and say what each type of line means.

Boxes containing things of different kinds. A framework mixing measurable variables, abstract constructs and contextual factors in identical boxes suggests they will be handled the same way — and they cannot be.

A diagram the text does not explain. The figure should follow the argument, not replace it. If the reader has to work out the reasoning from the picture, the reasoning was not written.

A useful discipline: write the framework as prose first, then draw it. If it cannot be written as a paragraph of connected claims, the diagram is hiding the fact that the argument is not there yet.

What examiners ask about frameworks

• Why this theory and not [obvious alternative]? The most common. Have the rejected option and your reason ready.
• Where does this concept come from? Pointing at a specific source is much stronger than “the literature”.
• Why does this arrow point this way? Particularly where your design cannot establish direction — in which case say the ordering is theoretically specified rather than empirically demonstrated.
• How did the framework shape what you collected? The question that exposes an inert framework immediately.
• What would have falsified it? If no plausible finding would have troubled the framework, it was not doing explanatory work.
• Did your findings support it? Including where they did not, which is a legitimate and often valuable outcome — a framework that failed in a specified way is a contribution, provided you engage with it rather than glossing over it.