A conceptual framework is a visual or written map of the key concepts or variables in your study and the relationships you expect between them. It is usually a diagram of boxes and arrows, with a short explanation. To build one, start from your research question, list the variables, decide which affects which, add any variables that change or explain those relationships, and support each link with theory or previous research.
What a conceptual framework shows
- the concepts or variables your study examines;
- the relationships you expect between them (the arrows);
- the direction of those relationships;
- where your hypotheses come from.
It differs from a theoretical framework, which describes existing theories. A conceptual framework is your own model for your specific study, usually built from theory and earlier findings.
Types of variables
| Variable | Role | Example |
|---|---|---|
| Independent | The presumed cause | Hours of daily social media use |
| Dependent | The outcome you measure | Sleep quality |
| Mediating | Explains how or why the independent variable affects the dependent variable | Bedtime delay |
| Moderating | Changes the strength or direction of the relationship | Age |
| Control | Held constant or adjusted for, so it does not distort the results | Caffeine intake |
How to build a conceptual framework in 5 steps
1. Start with your research question
How does daily social media use affect sleep quality among university students?
2. Identify the independent and dependent variables
- Independent variable: daily social media use (minutes).
- Dependent variable: sleep quality (questionnaire score).
3. Add mediators and moderators
Ask how the effect might work, and for whom it might be stronger.
- Mediator: social media use may delay bedtime, which then lowers sleep quality.
- Moderator: the effect may be stronger for students who use their phones in bed.
4. Add control variables
List other factors that affect sleep and that you will measure and adjust for, such as caffeine intake, workload and existing sleep disorders.
5. Draw it and explain it
Draw boxes for the variables and arrows for the expected relationships, then explain each arrow in a sentence or two, citing the theory or studies that support it.
Example framework
In words:
Daily social media use → Bedtime delay → Sleep quality
Phone use in bed strengthens the link between social media use and bedtime delay.
Caffeine intake and workload are controlled for.
From this framework come testable hypotheses:
- H1: More daily social media use is associated with poorer sleep quality.
- H2: Bedtime delay mediates the relationship between social media use and sleep quality.
- H3: The relationship between social media use and bedtime delay is stronger among students who use their phones in bed.
See how to write a hypothesis.
Conceptual frameworks in qualitative research
In qualitative studies, the framework is usually less about measurable variables and more about the concepts that guide data collection and analysis, for example "identity", "belonging" and "support networks" in a study of first-generation students. It can change as your analysis develops.
Where it goes in a thesis
At the end of the literature review, or at the start of the methodology chapter, once you have explained the research it is built on.
Common mistakes
- Including every possible variable instead of the ones your study measures.
- Drawing arrows without explaining or citing the reasons for them.
- Confusing mediators (how it works) with moderators (when or for whom it works).
- Building the framework after the analysis.
Turn the framework into hypotheses
The free hypothesis generator turns your variables into testable hypotheses with matching null hypotheses. The research question generator helps if you are still shaping the question.
Frequently asked questions
Is a conceptual framework the same as a theoretical framework?
No. A theoretical framework describes existing theories; a conceptual framework is your own model of the variables in your study.
Does a conceptual framework need a diagram?
It is not required, but a simple diagram makes the relationships much easier to follow.
How many variables should it include?
Only those you measure or analyze. For a bachelor's or master's thesis, two to six variables is common.
Can I change my conceptual framework?
In quantitative research, fix it before collecting data. In qualitative research, it can develop as you analyze.