DEV Community: SUHI RESEARCH MARKETTING

How to Choose the Right Theory for Your PhD Research (6 Steps)

SUHI RESEARCH MARKETTING — Wed, 07 Oct 2026 09:51:38 +0000

If you're doing a PhD in computer science, information systems, engineering, or any applied field, you've probably been told: "Your study needs a theoretical foundation."

Then you open Google Scholar, and the options are endless. Which theory fits? How do you know you've picked well and not just grabbed the first popular one?

Choosing a theory is a design decision, much like choosing an architecture. Pick well and your hypotheses, methodology, and discussion all line up. Pick badly and you'll fight your own framework for the rest of the project.

Here is a six-step process to make the choice deliberately.

Why This Decision Matters

A theory does three jobs in your thesis:

Explains why your variables should relate to each other.
Guides your hypotheses and research design.
Anchors your discussion, so you can say what your results mean beyond the numbers

Without it, a study is just a pile of observations. Examiners will ask, "Why did you expect this result?" Your theory is the answer.

Step 1: Define Your Research Problem Clearly

Before you search for any theory, be clear about what you are actually trying to understand. Ask yourself what is happening, who it is happening to, and what you think is driving it. Then explain it in two or three plain sentences, as if you were talking to a friend outside your field.

If you struggle to do this, you're not ready to choose a theory yet. A vague problem leads to a vague theory choice, and you will end up bending your study to fit whatever framework you grabbed first. Theories should fit problems, not the other way around.

Step 2: Identify the Type of Question You're Asking

Different questions point to different families of theories, so work out what kind of question yours is before you start searching.

If you are asking why people adopt or reject something, such as a new tool, app, or practice, look at technology acceptance and behavioural intention theories. If your question is about how organisations behave or perform, resource-based, institutional, and contingency perspectives are good places to begin. Questions about why people make certain choices usually lead to behavioural and cognitive models. If you are studying how systems or processes change over time, look into diffusion, change management, or socio-technical theories. And if your focus is on how people learn or build skills, learning and cognitive development theories are the natural neighbourhood.

You don't need to memorise these groupings. You only need to know which area of theory your question belongs to, so your search starts in the right place instead of across the whole literature.

If you'd like, I can rewrite the rest of the article in the same flowing style, without tables or fill-in templates, so the whole piece is consistent.

Step 3: Build a Shortlist of 3 to 5 Candidates

Now search. Good places to start:

The base papers closest to your topic. Check which theories they use and why.
Review articles and meta-analyses on your subject, which often list the dominant theories in the field.
Textbooks and handbooks in your discipline, which summarise major theories in plain language.

Note each candidate with a short summary: who proposed it, what it claims, and which constructs or variables it includes. Resist the urge to commit to the first one you find.

Step 4: Evaluate Each Candidate With a Fit Test

Score every theory on the shortlist against these questions:

1.** Relevance:** Does it address your type of problem and phenomenon?
2.** Variable fit:** Do its constructs map to the variables you want to study?

Context fit: Has it been tested in settings similar to yours, or only in very different ones? 4.** Evidence:** Is it well supported in the literature, or heavily criticised? 5.** Measurability:** Are there validated scales or data sources for its constructs?
Scope: Is it focused enough to be useful, without being so broad that it explains nothing?

A simple 1 to 5 score per question gives you a quick comparison. If two theories tie, move to the next step.

Step 5: Decide Between One Theory, a Combination, or an Extension

Sometimes one theory covers your problem. Often it doesn't. You have three options:

Use a single theory when it fits cleanly. This is the simplest to defend.
Combine two theories when each explains a different part of your problem. You must justify why they belong together.
Extend a theory by adding a variable the original misses. This is common and often becomes your contribution.

A warning: don't stack theories just to look sophisticated. Every added theory must earn its place, or it will weaken your argument.

Step 6: Test the Choice Before You Commit

Before you build your whole study on it, run three quick checks:

Explain it in two sentences to someone outside your field. If you can't, you don't understand it well enough yet.
Sketch the model. Draw your conceptual framework using the theory's constructs. If the diagram feels forced, the fit is poor.
Discuss it with your guide. Show your shortlist, your scoring, and your reasoning. Their feedback now saves months later.

If it passes, you have your theory. If not, go back to your shortlist, which is why you kept more than one.

Common Mistakes to Avoid

Choosing the most popular theory by default. Popular does not mean suitable.
Picking a theory first, then forcing the problem to fit. Start with the problem.
Confusing the theoretical framework with the conceptual framework. The theory is the established foundation, while the conceptual framework is your own model built on it.
Ignoring criticism of the theory. Know its weaknesses, because examiners will.
Using too many theories. More is not better. Justified is better.

Quick Recap

State your research problem clearly.
Identify what type of question you are asking.
Build a shortlist of 3 to 5 theories.
Score each one with a fit test.
Decide: single, combined, or extended.
Test and validate with your guide.

Final Thoughts

A good theory makes the rest of your PhD easier. Your hypotheses have a basis, your design has direction, and your discussion has something to say. A few extra days spent choosing carefully will save you months of rework.

If you're stuck between candidates, share your problem statement in the comments. Describing your problem often reveals which theory fits, and other scholars here may have used it in similar work.

Which theory did you end up choosing, and why? I'd like to hear how you decided.

The Paraphrasing Mistake That Turnitin Flags as Plagiarism

SUHI RESEARCH MARKETTING — Mon, 24 Aug 2026 06:57:44 +0000

You did everything "right." You read the source, closed the tab, wrote the idea in your own words, cited the author. Then your Turnitin report comes back at 34% similarity and your guide wants a meeting.

This is one of the most common — and most misunderstood — false alarms in academic writing. Except it's not always a false alarm. Let's break down what's actually happening.

The Mistake: Patchwriting

There's a name for this in academic integrity literature: patchwriting. It's the gap between paraphrasing and copying, and most researchers fall into it without realizing.

Patchwriting looks like this — take the original sentence, swap a few words for synonyms, maybe restructure the clause order slightly, keep the same sentence architecture and logical flow. The result reads like your own writing. Turnitin's similarity algorithm doesn't care. It's not checking whether words are identical — it's checking sentence structure, phrase sequencing, and pattern similarity. Patchwriting trips every one of those checks.

Original:

Climate change has significantly altered precipitation patterns across South Asia, leading to more frequent extreme weather events.

*Patchwritten (still flagged):
*
Climate change has considerably changed rainfall patterns throughout South Asia, resulting in more common extreme weather occurrences.

Different words. Same skeleton. Turnitin flags this correctly — this is not paraphrasing, it's synonym substitution.

Why This Happens (It's Not Carelessness)

Most researchers don't patchwrite because they're being lazy. It happens because of how paraphrasing is actually done under time pressure:

You read the source sentence
The source sentence is still sitting in your working memory, structurally intact
You write "your version" while the original's sentence shape is still the scaffold you're building on

This is a cognitive problem, not an ethics problem, in most cases. But the tool — and your committee — can't distinguish intent from outcome.

The Actual Fix: Read, Close, Recall, Write

The technique that reliably breaks patchwriting is deceptively simple:

Read the source passage fully — don't paraphrase sentence-by-sentence as you go
Close the source. Physically. Don't have it open in a split window "just to check"
Recall the idea from memory and write it down without looking back
Compare afterward — check the source only after you've written your version, to catch accidental structural overlap

This forces your brain to reconstruct the idea, not the sentence. Because you literally cannot see the source while writing, you can't inherit its syntax.

The Threshold Question: What Actually Counts as a Violation

This is where a lot of scholars get confused, and where India-specific context matters. Under UGC's 2018 regulations, plagiarism severity is tiered:

Below 10% — generally acceptable, no penalty (this includes quotes, references, bibliography — most tools exclude these automatically)
10–40% — minor penalty, resubmission usually required within a set timeframe
40–60% — more serious penalty, can affect degree progression
Above 60% — can result in suspension or expulsion, depending on the institution

A 34% similarity score isn't automatically catastrophic — but it's high enough that your committee will want to see where the overlap is concentrated. If it's clustered in a few paraphrased paragraphs from one or two sources, that's a patchwriting pattern, and that's fixable. If it's scattered thinly across dozens of sources, that's usually just common phrasing and citations, and far less concerning.

A Quick Pre-Submission Check

Before you submit anything through your institution's Turnitin account:

Run it through a checker yourself first, if you have access to one, so you're not seeing your score for the first time alongside your guide
Look at where the flagged overlap sits — literature review sections are almost always the highest-risk zone, since that's where most paraphrasing happens
For any paragraph flagged at high similarity, apply the read-close-recall-write method and rewrite from scratch — don't edit the flagged sentence, replace it entirely
Double-check that block quotes are properly formatted (quotation marks + citation) — sometimes what gets flagged is a legitimate quote that's just missing formatting, not a paraphrasing issue at all
The Real Takeaway

A high similarity score isn't a verdict — it's a diagnostic. The mistake worth understanding isn't "I copied something," it's "I paraphrased too close to the source's original structure without realizing it." Once you can see the difference, it's a completely fixable habit, not a character flaw.

If you're prepping a thesis chapter and want a plagiarism check before it reaches your guide's desk, that's something we help with regularly.

reach out