Writing a research gap statement

What a gap statement actually does

In a few lines, a gap statement has three jobs to do. First it tells the reader what is already known, with citations, so they can see you know your field. Then it names one thing which is not known yet, or not settled, or not trustworthy. Last, it explains why somebody should spend three to five years of work to close that gap.

When committees send back gap statements, mostly it is for the same few reasons. Sometimes there is already a paper that fills the “gap”, and the scholar did not find it. Sometimes the gap is real but so small that nobody cares. And many times the gap and the objectives don’t match, so even after the study the gap stays open. You can catch all of these on your own before you submit to the committee.

Six kinds of research gap

Each author has his own way of naming gaps, and you need not worry much about which label is correct. The useful thing to ask is what evidence each kind needs. The six kinds in the table cover almost every gap we see claimed in a PhD proposal. Many good studies mix two of them. For example, a methodological gap can also give rise to a context gap.

A note on the Indian-context gap

In Indian proposals the most common gap claim is “this has been studied abroad but not in India”. The claim can be fair. A lot of models, scales and designs were tested in other countries, and conditions in India really are different. Take the monsoon climate, respondents speaking many languages, informal labour markets, and rules and laws that are not the same. Your claim gets strength only when you say which of these differences can change the result, and how it may change. Many scholars are taking a design from some foreign study and just writing “India” or their state name in it, and that claim stays weak.

Where to look for a real gap

You find gaps only by reading closely. Slowly you start noticing what the papers are still not able to answer. These are the places worth looking first.

• Recent review articles. Most systematic reviews, meta-analyses and narrative reviews from the last three to five years close with a part on limits and future work. Go and check whether later papers have already picked up that idea.
• Limitations sections of the closest primary studies. Authors mostly confess their problems here, like a small sample from one site, follow-up of only few months or a scale nobody validated. When five or six papers are admitting the same limitation, you may have found a methodological gap for the whole field.
• Contradictions. Take the ten to twenty studies nearest to your question and put them in a simple table, with one column each for sample, setting, method, measure and finding. Wherever the findings disagree, go along the rows and see which column is different. Many times that column is your gap, and it can also become your hypothesis.
• Theory that does not fit. The main framework predicts one result, but the data from the field point to something else.
• Practice and policy documents. Government reports, standards and clinical guidelines often say what people working in the field need and are not able to find.

Maintain a log of every search with the database, the date, the search string and how many results you kept. In the RAC somebody is sure to ask how you know “no study has examined…”, and this log is your answer. If your field expects a systematic review, you build it on the same log.

Why “no one has studied X in Y” is not enough

Every gap needs something to be absent, but absence alone does not make a gap. Nobody has studied job satisfaction among left-handed accountants, and that is because there is no reason to expect anything interesting from it. An examiner reading “no study has been done” will be thinking two things in his mind. Are you sure? And so what?

For “are you sure”, show your search log, and also write what the nearest studies actually did. For the second, you need a reason why the missing knowledge would change something. Maybe a theory will need revising, or a method will need replacing. Or some policy or design decision is right now resting on evidence collected somewhere else.

In Organization, Sandberg and Alvesson argue that only filling gaps tends to give small, step-by-step work, and the studies that shift a field more often question its assumptions. If your committee is asking again and again why your work matters, read that paper once.

A four-part pattern for writing the gap

In any subject, most strong gap statements go through four moves in the same order. Write each move as its own sentence first, and join them later.

1. Established: what the literature has already shown, with citations. One or two sentences are sufficient here.
2. Unresolved: the exact thing that is missing, conflicting or unreliable, together with evidence for it. No need to write “this is a methodological gap”, the reader should catch the kind from your wording itself.
3. Consequence: why it matters. What is it that nobody can know, decide or build till this gap is closed?
4. Response: what your study will do, in a single sentence. That same sentence becomes your aim.

Template

“Studies of [topic] have established that [finding] [citations]. However, [specific unresolved issue], because [reason: existing studies used method M / were conducted only in setting S / report conflicting results for reason R]. As a result, [consequence for theory, method or practice]. This study therefore [aim] by [approach, in brief].”

After all four moves are in place, write it again in your own words. Committee members can recognise the bare template immediately.

Research gap statement examples: weak versus strong

All the examples below begin from an imagined reading of the literature. In your own statement, you will need real citations for every such claim.

Engineering: crack detection on concrete bridges

Weak: “No one has applied deep learning to crack detection on bridges in Kerala. This study proposes a CNN-based crack detection model.”

Strong: “On public benchmark datasets, convolutional models for surface crack detection report high accuracy. Suppose, however, that the few studies which tested them on field inspection images all found accuracy dropping sharply, and the authors said the lighting and surface conditions in the field were not there in the benchmarks. None of them looked at wet surfaces, algal staining and the low-contrast light of the monsoon, which are typical for coastal concrete bridges for much of the inspection calendar. Till these conditions are tested, inspection agencies have no way to know if such models are reliable enough to support manual inspection. This study builds a labelled image set from coastal bridge inspections, releases it publicly, and evaluates existing and adapted models on it.”

What changed: the weak version only puts a method in a place. The strong version claims an empirical and context gap. It gives a mechanism (surface and lighting conditions) and a practical consequence (how reliable inspection is).

Management: hybrid work and employee engagement

Weak: “Only a few studies have looked at employee engagement in Indian IT companies after COVID-19.”

Strong: “Studies on hybrid work and engagement disagree. Some report higher engagement among hybrid workers and others report lower. Suppose recent reviews suggest that the difference may come from the quality of supervision and not from the work arrangement as such, while most studies measure only the arrangement. If so, nobody has checked yet whether the effect on engagement comes from hybrid work itself or from supervisory support. Firms that are designing return-to-office policies are acting as if the first is true. This study tests whether supervisory support moderates the relationship between hybrid work intensity and engagement among employees of Indian IT services firms.”

What changed: here the gap is an evidence gap, built on findings that conflict. It offers a likely explanation, and that explanation turns into the hypothesis. The Indian IT setting is only the place where the test happens, not the full gap.

Life sciences: antimicrobial activity of a medicinal plant

Weak: “Nobody has studied the antibacterial activity of the leaf extract of [plant] against multidrug-resistant strains.”

Strong: “Several studies report that extracts of [plant] show antibacterial activity. Suppose these studies used different solvents and different plant parts, and depended on disc-diffusion zones instead of minimum inhibitory concentrations. Very few of them used standard reference strains, so you cannot compare or pool their results. Nobody can say now how much evidence is there for using the plant. This study determines minimum inhibitory concentrations of standardised extracts against reference and clinical multidrug-resistant isolates, using a recognised susceptibility-testing protocol.”

What changed: this one is a methodological gap. People have already looked at the plant. The trouble is that the way they looked cannot answer the question.

Social sciences: women’s mobile phone use in rural households

Weak: “There is no study on smartphone use among rural women in Madurai district.”

Strong: “Most research on the gender digital divide has measured whether an individual owns a phone. Suppose field studies report that in many rural households one phone is shared, and the family negotiates who uses it. In that case ownership surveys would wrongly classify women who have regular but mediated access, and also women who own a phone but control it very little. Programmes sending welfare and health information by phone depend on getting this right. This study develops and tests a measure of phone access and control in shared-device households of rural Tamil Nadu.”

What changed: the gap here is theoretical or conceptual, about how the key variable gets defined. Tamil Nadu is a good setting to study it. The place comes out of the problem, and not the reverse.

From gap to research questions and objectives

In Indian synopses, the structural fault we see most is a gap going in one direction and objectives going in another. Committee members read the two together, side by side. Every part of the gap should give one research question, and every research question should give one objective that has a method attached.

• Gap: we don’t know yet whether hybrid work is acting on engagement directly or through the support given by supervisors.
• Research question 1: how are hybrid work intensity and engagement related? Objective: to measure this relationship in a sample of IT services employees.
• Research question 2: does supervisory support moderate that relationship? Objective: to test the moderation with validated scales.
• Research question 3: which practices are reported by supervisors of hybrid teams with high engagement? Objective: to find them out through interviews.

You can do two quick tests. Remove one objective and see whether any part of the gap is left open. If nothing is left open, that objective does not belong there. Now imagine every objective is met and the gap is still open. Then either one objective is missing or the gap you claimed is too big.

Some scholars write “to collect data” as an objective, but that is only a task you do on the way. In a synopsis, objectives should be written as things you are going to find out.

How to check a gap statement before you submit

Go through this list before your synopsis reaches the Research Advisory Committee. A member who has prepared well will ask you these same questions.

• You can name the three to five specific studies nearest to your question, and for each one you can say exactly what it did not do.
• You have written the gap as a problem in knowledge. It is not written as the lack of your own method or your own place.
• You searched minimum two databases (for example Scopus or Web of Science, plus one subject database) and you saved the search strings.
• You went through papers and preprints of the last two to three years. In a busy field, gaps get closed very fast.
• There is a “so what” sentence. It says which theory, method, policy or practice will change once the gap is filled.
• Every research question covers some part of the gap, and no part of the gap is left without a question.
• You have enough time, money, data and skills to fill the gap. If you cannot close a gap, however real, it is not your gap.

If your gap fails on more than two points, mostly the fix is to read more and not to reword. Take again the ten studies nearest to your topic and make comparison table we described earlier. Also discuss about it with your guide. If you want somebody outside to check it, our topic and gap selection support reviews your draft gap against the literature you have collected, and we tell you plainly whether it holds or not. For the review chapter that the gap stands on, see literature review guidance.