Theoretical vs conceptual framework: what each is and how to build one

What is a theoretical framework in research?

A theoretical framework is a theory which you did not invent yourself. Somebody proposed it first, other researchers tested it, and now it carries a name, an original paper and constructs that are defined. You apply it to your own problem.

Grant and Osanloo (2014) wrote on this for doctoral students in the Administrative Issues Journal. They call the theory the “blueprint” of the entire dissertation. They mean it practically. An examiner should spot the theory in your research questions and in how the literature chapter is arranged. Your measures come from it, and you read the results through it. A theory named once in Chapter 2 and never again is only decoration.

For studies on how people adopt new technology, many times the Technology Acceptance Model (Fred Davis, MIS Quarterly, 1989) or UTAUT is used. Grant and Osanloo’s own example is a study on school leaders, built on Burns’s transformational leadership theory.

Qualitative researches also use theory. There it acts more like a lens than a source of hypotheses. As per Grant and Osanloo, some qualitative designs keep the framework loose deliberately, so that the researcher does not push fixed ideas onto what the data is showing.

What is a conceptual framework in research?

The conceptual framework is your own, and it is made for one study only.

The definition quoted most often comes from Miles and Huberman’s Qualitative Data Analysis (1994, p. 18). As per them, a conceptual framework “explains, either graphically or in narrative form, the main things to be studied” (the key factors, concepts or variables) “and the presumed relationships among them”. So it can be a diagram or written text. And the links are only presumed. Your study is what checks them.

Look at most Indian theses in management, commerce, education, psychology and the health sciences. There it is one figure of boxes and arrows near the end of the literature review, with its hypotheses right after. It picks a few constructs from a theory and adds one or two variables that your review supports. Anything you won’t measure is left out. So a theory may have ten constructs and your framework only five, and examiners accept it if you write why the other five are left out.

What is the difference between a theoretical and a conceptual framework?

In short, the theoretical framework is borrowed from others and it is general. The conceptual framework you build yourself, and it fits your study only.

Why your supervisor may use the words differently

Scholars get stuck here, mostly after reading three websites that disagree. Grant and Osanloo are firm that the two terms are “neither interchangeable nor synonymous”. In the same paper they admit the terms “can be vague and lead to confusion for students and dissertation committee members alike”. Yamauchi and colleagues (2017, School Community Journal) describe conceptual frameworks as models of relationships that “may or may not imply a particular theoretical perspective”. Usage on the ground also varies. Some supervisors call the figure a “research model” or a “proposed model”. Some qualitative researchers say “conceptual framework” for the full argument of a study. And some DC members want just one section with the title “Theoretical/Conceptual Framework”.

In the RAC meeting there is no point fighting over which word is correct. Ask your guide which heading the department is expecting. After that, open two or three recent theses of your department on Shodhganga and follow them. Examiners want to see that the theory is named and that you have understood it. They also check whether each variable has a reason for being there.

How do you build a theoretical and conceptual framework?

First comes the theory, and from that you make your conceptual framework. Both need a finished literature review. If you are still collecting papers, start with our guide to writing a literature review.

1. 1

   Start from your objectives

   Each box in the framework has to connect with at least one of your research questions or objectives.

2. 2

   Shortlist two or three theories

   See which theories the studies closest to your topic have used. Read the original paper of each one itself.

3. 3

   Choose one and say why

   Take the theory whose constructs match your problem most closely. Write a paragraph on why you kept the others aside, as examiners will ask.

4. 4

   Keep, drop, add

   Keep the constructs you will study and drop the rest with a reason. Then add one or two variables which your review justifies.

5. 5

   Draw it and check it

   Number every arrow. Then check that each box has a measure in your questionnaire, dataset or interview guide.

Worked example: teachers and a learning management system

This study is made up. Suppose your topic is how teachers in affiliated arts and science colleges are using the learning management system of the university. One of your objectives (see research objectives and questions) is to find out what drives their intention to use it, and their actual use.

From your review you get two candidate theories. TAM (Davis, 1989) explains use with perceived usefulness and perceived ease of use. The other one is UTAUT, the Unified Theory of Acceptance and Use of Technology. Viswanath Venkatesh, Michael Morris, Gordon Davis and Fred Davis proposed it in MIS Quarterly in 2003, after comparing eight earlier acceptance models. UTAUT has four core constructs (performance expectancy, effort expectancy, social influence, facilitating conditions) and four moderators (gender, age, experience, voluntariness of use).

You go with UTAUT. In a college, the principal and the colleagues clearly have a say in whether teachers log in or not. Campus Wi-Fi and training also differ a lot from college to college. TAM does not have a separate place for either of these. (UTAUT2 by Venkatesh, Thong and Xu in 2012 was made for consumers, so you mention it and keep it aside.)

Next you adapt the theory to your setting. Suppose your review has many Indian studies where workload is coming up again and again. Teachers already have heavy timetables, they don’t want one more step. So you add “perceived workload” as one extension and cite those studies for it. The university has made the system compulsory, so voluntariness is same for everyone in your sample. You drop that moderator and write the reason.

How do you draw a conceptual framework?

Word or PowerPoint shapes are enough. The usual rules are these.

• Independent variables on the left, the dependent variable on the right, and any mediator in between.
• For every relationship you are going to test, draw one arrow with a single head and write the hypothesis number on it.
• Moderators go in a box below, with an arrow to the middle of the path they affect. (If the difference between a mediator and a moderator is unclear, see mediation and moderation analysis.)
• The variables you added yourself get a dashed border. Then the examiner knows which boxes are yours.
• A figure number, a title and a source line are needed, for example “Adapted from Venkatesh et al. (2003)”.

In our example the figure is read from left to right. On the left are five boxes for performance expectancy, effort expectancy, social influence, perceived workload (with a dashed border) and facilitating conditions. Behavioural intention is in the middle and use behaviour is on the right. The first four boxes have arrows to intention, and intention has an arrow to use. Facilitating conditions gets its own arrow going directly to use, same as in the original UTAUT. Age, gender and experience sit in one box below. Written as a list of paths, it looks like this.

Every row turns into a hypothesis written in words. For example, “H2: Effort expectancy has a positive effect on teachers’ behavioural intention to use the learning management system.” For the wording of these, read how to write a research hypothesis. At the testing stage, this same figure becomes your structural model in the SEM software. If you are still deciding on a tool, see AMOS vs SmartPLS.

For a qualitative study you don’t need arrows or hypotheses at all. Mostly a concept map or one page of narrative is given, marked clearly as provisional.

Mistakes that DC members pick up

• Too many arrows. If you cannot name the study that supports an arrow, take that arrow out.
• The original authors’ figure pasted as it is, unchanged and without credit.
• Twelve boxes with only 150 respondents. With each path you add, the sample you need goes up, so discuss about this with whoever is advising you on the analysis, before the questionnaire is sent out.
• Boxes for which the questionnaire has no matching items.

Where does the framework go in the thesis and synopsis?

University ordinances mostly don’t say anything on this. So it depends on the custom of the department and on your guide. In Indian theses the pattern usually goes like this.

• Synopsis: a short subsection naming the theory with its original citation, then the framework figure and the hypotheses or questions coming from it. Half a page to one page.
• Chapter 1: only one paragraph. Name the theory there and tell that details are in Chapter 2. Most departments are not expecting the full framework here.
• Chapter 2 (review of literature): this is where it usually goes. Put the theory early in the chapter. Keep the conceptual framework and the hypotheses for the end, after the review has given a reason for every arrow.
• A separate chapter: in some management and social science theses there is a chapter called “Theoretical framework and hypotheses development”. If your department does it this way, follow the same.

Later the framework appears again briefly in two places. In the methodology chapter, you link each construct to the scale that measures it. In the discussion chapter, you write which paths held and what that means for the theory.

In the synopsis, check that figure, objectives and hypotheses all match with each other. Committees do notice when there are five objectives and the figure has only three variables. The rest of the synopsis is in our synopsis format guide. The same section for admission or fellowship proposals is in the research proposal guide. The full chapter plan is in PhD thesis structure, and sample layouts for the review chapter are in literature review examples and format.