๐ Guide ยท 6-minute read
How to Draw a Conceptual Framework for Your Thesis
A conceptual framework is the one figure that shows what your study tests: the variables, how you expect them to relate, and the hypotheses behind each arrow. Here is how to build one step by step, how to read moderators and mediators, and three worked examples you can open and change.
What a conceptual framework shows
- Independent variables: what you expect to cause or predict a change (for example service quality).
- Dependent variable: the outcome you explain (for example loyalty).
- Mediators: variables in between, through which the effect passes (satisfaction).
- Moderators: variables that make an effect stronger or weaker (age, gender, experience).
- Hypotheses: one labelled arrow each (H1, H2, ...), matching the hypotheses in your text.
- Control variables, if you use them: kept apart, with a dashed arrow to the outcome.
Step by step
- Start from your research question. Name the outcome you want to explain.
- List the variables from the literature. Keep only those your study measures; give each a short name and an abbreviation.
- Choose a theory if one fits (TAM, UTAUT, Theory of Planned Behaviour, S-O-R). It decides most of the arrows for you.
- Write each hypothesis as one relationship: "H1: Service quality (SQ) โ Customer satisfaction (CS)". One arrow per hypothesis.
- Mark mediators and moderators explicitly: "H4: Customer satisfaction mediates SQ โ LO", "H5: Age moderates CS โ LO".
- Check it against your text: every hypothesis appears exactly once, and nothing in the figure is missing from your hypotheses section.
Write it so it can be drawn for you
Paste a description like this into the Studio, choose Flowchart (or leave the type on Auto), and the framework is drawn with every hypothesis once and nothing added:
Conceptual framework for my thesis. H1: Service Quality (SQ) -> Customer Satisfaction (CS) H2: Price Fairness (PF) -> Customer Satisfaction (CS) H3: Customer Satisfaction (CS) -> Loyalty (LO) H4: Customer Satisfaction mediates SQ -> LO H5: Age moderates CS -> LO Control variables: income and gender.
Plain sentences work too ("A affects B, which affects C"). Ask for colours if your department wants them ("light blue boxes for the independent variables").
Example 1: a model with a mediator (TAM)
Perceived usefulness sits between ease of use and intention, so it is drawn in the middle group. Each hypothesis is one arrow with its label.
Research model based on the Technology Acceptance Model (TAM). H1: Perceived Ease of Use (PEOU) -> Perceived Usefulness (PU) H2: Perceived Ease of Use (PEOU) -> Attitude Towards Using (ATU) H3: Perceived Usefulness (PU) -> Attitude Towards Using (ATU) H4: Perceived Usefulness (PU) -> Behavioural Intention (BI) H5: Attitude Towards Using (ATU) -> Behavioural Intention (BI) H6: Behavioural Intention (BI) -> Actual System Use (ASU)
Example 2: a moderator
Supervisor support does not affect satisfaction directly; it changes how strongly stress affects it. So its arrow points at that path, not at a box.
Conceptual framework with a moderator. H1: Work Stress (WS) -> Job Satisfaction (JS) H2: Job Satisfaction (JS) -> Turnover Intention (TI) H3: Supervisor Support (SS) moderates WS -> JS
Example 3: mediation and a variable with dimensions
Engagement mediates the effect of leadership on performance (H4, shown as a note), and it is measured through three dimensions, shown as its own small group.
Conceptual framework with a mediator. H1: Transformational Leadership (TL) -> Employee Engagement (EE) H2: Employee Engagement (EE) -> Job Performance (JP) H3: Transformational Leadership (TL) -> Job Performance (JP) H4: Employee Engagement (EE) mediates TL -> JP Employee Engagement consists of vigour, dedication and absorption.
Common mistakes
- Arrows in the figure that have no hypothesis in the text, or hypotheses with no arrow.
- Drawing a moderator as an ordinary arrow into the outcome, instead of onto the path it changes.
- Drawing the framework as a process, with steps and yes/no decisions. A framework shows variables and relationships, not a procedure.
- Too many variables. If the figure does not fit on one page, the study is probably testing too much.
Questions
What is the difference between a conceptual and a theoretical framework?
A theoretical framework is an existing theory you build on (for example the Technology Acceptance Model). A conceptual framework is your own study's model: the variables and hypotheses you test, often adapted from that theory.
Where does the conceptual framework go in a thesis?
Usually at the end of the literature review, just before or after the hypotheses, so the reader sees how the hypotheses fit together.
How do I show a moderator in a conceptual framework?
Draw an arrow from the moderator onto the arrow it changes, not into the outcome box, and label it with its hypothesis.
Can I make the diagram for free?
Yes. Opening and changing the templates is free, and drawing your own framework works without an account (3 credits a day) or with a free account. PNG, SVG and PDF downloads are free.
