A conceptual framework is the diagram that shows, on one page, exactly what your study measures and how those things relate. Philippine research manuals almost always require it, panels always ask about it, and it is the single most efficient page in a thesis: a good one answers four defense questions before they are asked.
This guide takes you from your statement of the problem to a finished diagram in seven steps, with a worked example. Your school’s research manual governs the exact placement and labelling, so check yours before you draw anything.
Step 1: Understand what it is — and what it is not
Expected output: clarity on which of the two frameworks you are being asked for.
Students conflate two different things because manuals often ask for both on adjacent pages.
| Theoretical framework | Conceptual framework | |
|---|---|---|
| What it contains | An existing, named theory by a named author | Your own study’s variables and their relationships |
| Where it comes from | Published literature | Your statement of the problem |
| Form | Usually narrative | Usually a diagram plus a narrative paragraph |
| Who owns it | The theorist | You |
The theoretical framework is borrowed; the conceptual framework is built. A panel that asks “who propounded this theory and in what year?” is asking about the theoretical framework. A panel that asks “why does this arrow point this way?” is asking about the conceptual one.
Step 2: List your variables straight from the statement of the problem
Expected output: a plain list of every variable your problems actually measure, and nothing else.
Do not start by drawing. Start by reading your own statement of the problem and writing down each thing it asks about. If a concept appears in your framework but in none of your problems, it does not belong in the diagram. If a variable appears in your problems but not in your framework, the diagram is incomplete.
Worked example. A study asks: (1) What is the profile of the respondents in terms of sex, year level and weekly study hours? (2) What is the level of use of four study habits? (3) What is the academic performance of the respondents? (4) Is there a significant relationship between study habits and academic performance?
The variable list is therefore: sex, year level, weekly study hours, four named study habits, academic performance. Six items, no more.
Step 3: Classify each variable
Expected output: every item on your list labelled.
- Independent — what you think does the influencing. In the example, the four study habits.
- Dependent — what is being influenced. Academic performance.
- Moderating — something that changes the strength of the relationship. Sex or year level, if your problems actually test them that way.
- Intervening — something that sits between the independent and dependent variable in the causal chain.
- Process — an activity rather than a measured quantity, which matters for the next step.
Be honest here. If your design is correlational, you have variables that are associated, and labelling one “independent” is a convention rather than a causal claim. Say so if asked; the alternative is being told that your data cannot support the arrow you drew.
Step 4: Choose the diagram model
Expected output: a decision between the two models Philippine manuals use, with a reason.
The IPO model — Input, Process, Output. Three boxes in a horizontal flow, sometimes with a feedback arrow returning from Output to Input. Use it when your study describes a process or evaluates something and produces a deliverable: capstone projects, systems development, programme evaluations, module development, training interventions.
Worked IPO example for a capstone. Input: user requirements, existing manual records, hardware and software specifications. Process: system analysis, design, coding, testing, user acceptance evaluation. Output: a deployed records management system with an evaluation report. Feedback: evaluation results returning to refine the design.
The independent-dependent variable model. Two boxes, with an arrow from the independent variables to the dependent variable and moderating variables entering from above or below. Use it when your study measures relationships or differences rather than producing something.
Worked example for the study habits study. Left box: “Study Habits” listing the four named habits. Right box: “Academic Performance” listing the indicator used, such as general weighted average. Arrow from left to right. A box below labelled “Respondent Profile — sex, year level, weekly study hours” with an arrow to the main arrow, if and only if Problem 4 actually tests those as moderators.
The commonest error is choosing IPO by default. IPO is a process model. A purely correlational study forced into IPO boxes produces a diagram that does not match its own statistical treatment, and a panel will see it immediately.
Step 5: Draw the boxes and list the indicators inside them
Expected output: a clean diagram in which every box names its contents.
An empty box labelled “Study Habits” tells a reader nothing. List the indicators inside it — the four named habits — so the diagram shows what will actually be measured. Panels read the diagram as a promise about the instrument, and an indicator that appears in a box must appear as items in the questionnaire.
Three practical rules. Keep it to one page, portrait or landscape, and readable at printed size. Use plain rectangles and straight arrows; decorative shapes and colour gradients add nothing. And label the figure properly — “Figure 1. Conceptual framework of the study” — because manuals check figure numbering.
Step 6: Draw the arrows, and be ready to defend every one
Expected output: arrows you can each justify in one sentence.
Every arrow is a claim. A single-headed arrow claims direction; a double-headed arrow claims association without direction; a feedback loop claims the output changes the input. Draw the one you can support.
The justification comes from two places: your literature and your design. “The arrow runs from study habits to academic performance because Santos and Cruz (2024) reported that direction, and because our data collection measured habits before the grading period closed.” That sentence is what turns a diagram into an argument.
If your design cannot support the direction, use a double-headed arrow and describe it as a relationship rather than an effect. Nobody has ever failed for being accurate about what their design can show. The place to establish which studies support which direction is your review of related literature, which is why the framework is drawn after the review and not before it.
Step 7: Write the narrative paragraph
Expected output: one or two paragraphs explaining the figure, placed immediately after it.
Explain the diagram box by box, then arrow by arrow. Do not simply restate the labels; say why each element is there.
Template: “Figure 1 presents the conceptual framework of the study. The first box contains the four study habits measured by the researcher-made instrument: time management, note-taking, reading comprehension and examination preparation. The second box contains academic performance, operationalised as the general weighted average for the first semester. The single-headed arrow indicates the hypothesised direction of influence from study habits to academic performance, consistent with the findings of the studies reviewed in Chapter 2. The lower box contains the respondent profile variables, which are examined as possible moderators under Problem 4.”
The reconciliation check that makes the framework defensible
Run these four before you submit. Each corresponds to a question panels actually ask.
- Every variable in the diagram appears in your statement of the problem. If it does not, delete it from the diagram or add the problem.
- Every variable in the diagram appears in your instrument. A box promising four study habits needs items measuring all four.
- Every relationship in the diagram has a statistical tool. An arrow claiming a relationship needs a correlation or a test behind it — which tool answers which kind of problem is set out in the guide to the statistical treatment of data.
- The diagram matches your design. IPO for process studies, independent-dependent for relationship studies, and no causal arrows on a purely correlational design. If you have not settled the design yet, settle it first using the comparison of quantitative and qualitative research.
The operational definitions inside your boxes must also match the ones you set in Chapter 1, and the instrument that measures them is described in Chapter 3. Where each of these sits across the five chapters is summarised in the overview of Chapters 1 to 5.
Keep the framework and the chapters in agreement
Tesify drafts your chapters in the structure your research manual expects and keeps your variables, instrument and analysis consistent across them, so the diagram you defend is the one your data actually supports. Over 9,000 students have used it to write more than 15,000 chapters, and every word is still written by you.
Frequently asked questions
What is the difference between a theoretical and a conceptual framework?
The theoretical framework is an existing named theory by a named author, taken from the literature. The conceptual framework is your own study’s variables and their relationships, built from your statement of the problem.
Should I use the IPO model?
Use it if your study describes a process or produces a deliverable, such as a capstone system, a developed module or a programme evaluation. Use the independent-dependent model if your study measures relationships or differences.
Where does the conceptual framework go?
Most Philippine manuals place it at the end of Chapter 2, after the review of related literature and the theoretical framework, but some place it in Chapter 1. Follow your own manual.
Can I use a framework from another thesis?
You can adapt one, and you must cite whose it was and state what you changed. What you cannot do is copy a diagram whose variables do not match your own problems, which is exactly what a panel checks first.
How many variables should the diagram have?
As many as your statement of the problem contains, and no more. Undergraduate studies typically have two to five. A crowded diagram usually signals a study trying to do too much.
Do qualitative studies need a conceptual framework?
Many manuals still require one, though it looks different — often concepts and their relationships rather than measured variables, and sometimes presented as emerging rather than fixed. Ask your adviser how your department expects it to be handled.
Can the arrows be double-headed?
Yes, and a double-headed arrow is the honest choice when your design establishes association rather than direction. It is safer than a single-headed arrow you cannot defend.
What software should I use to draw it?
Anything that produces clean rectangles and straight arrows. Word’s shape tools are entirely adequate. Panels care about the logic, not the rendering.
Does the framework change after data collection?
It can, if your findings contradict a hypothesised direction. Discuss that in Chapter 5 rather than silently redrawing the figure, because an unexplained change between the proposal and the final manuscript invites a question.
What is the most common mistake?
Choosing IPO by default for a correlational study. The diagram then promises a process the statistical treatment never analyses, and the mismatch is visible on one page.