Conceptual framework in research: meaning and example

A conceptual framework is a map of your own study. It names the main concepts or variables you will study and shows how you expect them to relate to each other. Editage puts it in one line: it is "the map of what variables you will study and how you expect the parts to connect."

conceptual frameworknoun

A diagram or a written explanation, built by the researcher for one study, that sets out the key concepts or variables and the relationships expected between them.

The idea is old. Miles and Huberman, in their 1994 book Qualitative Data Analysis, described a conceptual framework as something that explains, in a diagram or in words, the main things to be studied and the presumed relationships among them. Most frameworks today use both: a diagram of boxes and arrows, and a few paragraphs that explain it.

You do not copy a conceptual framework from anywhere. You assemble it from three things: the theories in your field, the findings of earlier studies, and what you know about your own setting. That is why it usually grows out of your literature review, and why no two studies have quite the same one.

This is where many students get confused, because many papers use the two terms as if they meant the same thing. They do not. A theoretical framework rests on an established theory that others have already tested. Editage calls it "the foundation of theory you are basing your study on." A conceptual framework is built for your study alone, and it shows the particular variables you will measure.

The Technology Acceptance Model, from a 1989 paper by Fred Davis in MIS Quarterly, proposes that people accept a new technology mainly because of two beliefs: perceived usefulness and perceived ease of use. Davis defined perceived ease of use as "the degree to which a person believes that using a particular system would be free of effort." That theory says nothing about Aizawl, about vegetable shops or about mobile signal in a hill town. Esther's conceptual framework is where those details come in.

They are not alternatives. Most studies that investigate relationships between variables need both.

In a quantitative study, the boxes in a conceptual framework are usually variables, and each kind plays a different part. These are the ones you will meet most often:

• Independent variable: the presumed cause or predictor. For Esther, how useful and how easy a shopkeeper finds UPI.
• Dependent variable: the outcome you measure. For Esther, whether the shopkeeper is still taking UPI payments six months later.
• Mediating variable: a step on the path that explains how or why the cause leads to the effect. For Esther, the shopkeeper's intention to keep using UPI.
• Moderating variable: something that changes the strength or the direction of an effect without being caused by it. For Esther, the shopkeeper's age.
• Control variable: something you hold constant or measure so that it does not distort the result, such as the type of shop.

Each variable also needs a way to be measured, such as a set of questionnaire items. A box labelled "engagement" or "awareness" with no clear measure is a common weakness. If your study tests predictions, each arrow usually becomes one hypothesis.

Guides from Editage and the University of Southern California library suggest much the same order. Esther followed it over two weekends.

1. Write down your research question.The question decides which variables matter. Esther's was: what makes women shopkeepers in Bara Bazar continue using UPI?
2. List the key concepts from your literature.Go back through your notes and pull out every factor earlier studies linked to your outcome. Esther found usefulness, ease of use, trust and age again and again.
3. Choose your independent and dependent variables.Put the cause on the left and the outcome on the right, and draw an arrow that starts at the independent variable and points to the dependent one.
4. Add mediators, moderators and controls.Ask how the cause works, for whom the effect might be stronger, and what else could explain the outcome.
5. Justify every arrow.Each arrow should trace back to a theory or a published finding. Editage warns that an arrow without one is "a guess wearing a box."
6. Write the narrative.A few paragraphs that define each concept, cite the sources behind each arrow and state your assumptions.

A literature review matrix makes step two much faster. Esther added a column called "variables measured" and filled it for every paper. Tools such as Fiberarticle's Extract can draft a table like this across many papers, with a short supporting quote for each cell, but she still checked every cell against the paper before drawing an arrow from it.

This is the framework Esther took to her next meeting with Dr. Vanlalruata. The diagram sat on one page. The table below sets out the same thing in words.

Under the diagram she wrote a short narrative. Its opening paragraph read like this:

Drawing on the Technology Acceptance Model, this study proposes that perceived usefulness and perceived ease of use shape a shopkeeper's intention to continue using UPI, which in turn predicts continued use six months later. Two factors from the setting are added: trust in UPI, drawn from earlier studies of digital payments, and network signal, which emerged from pilot visits. The effect of ease of use is expected to be weaker for younger shopkeepers. Type of shop and years in business are treated as control variables.

Notice what the example does. Every box can be measured, every arrow has a named source, and a reader can see the whole study on one page before reading a single result.

A conceptual framework is not only for surveys and experiments. In qualitative research it is looser. It names the ideas you expect to matter, and those ideas shape your interview guide and your first codes, but you do not test it with statistics.

Editage describes the conceptual framework as "a working proposition that is expected to shift," especially in qualitative work. That is a strength, not a weakness. Esther also planned twelve interviews with shopkeepers who had stopped using UPI. Her first framework had no box for fear of being cheated. After four interviews it did, because three women spoke of fake payment screenshots shown by customers.

• Start with a rough map of the ideas you expect, drawn from your literature.
• Use it to plan your interview questions, not to limit them.
• Change it when the data shows something you did not expect, and keep each version.
• In your thesis, show the final framework and explain briefly how it changed.

For how this kind of study differs from a survey, see qualitative vs quantitative research.

There is no single rule for where it goes. Editage notes that examiners usually look for the conceptual framework at the end of the literature review or at the start of the methodology chapter. Rules differ between universities and even between departments: some want a separate section with its own heading, others are happy with a diagram inside the review. Look at recent theses from your own department and ask your guide.

These are the mistakes that committees and reviewers notice most:

• treating the conceptual framework and the theoretical framework as the same thing;
• drawing arrows with no theory or finding behind them;
• using boxes so vague that nobody could measure them;
• showing only the cause and the effect, with no mediators, moderators or controls where the study needs them;
• copying another study's diagram without adapting it to your setting;
• drawing the framework once and never using it again.

That last one matters most. The USC writing guide warns, about theory, "Do not leave the theory hanging out there in the introduction never to be mentioned again." The same is true of a conceptual framework. Your methodology should measure each box, and your discussion should come back to each arrow and say whether the data supported it. If you are still at the proposal stage, how to write a research proposal shows where a first version of the framework fits.

Esther's diagram now hangs above her desk. When the committee met again, the same member looked at it for a minute and asked only one question: why age was a moderator and not a control. Esther pointed to the two studies beside that arrow.