How to write limitations of the study: with examples

The limitations of a study are the weaknesses in its design, data or analysis that affect how far its findings can be trusted or applied. AJE, an author editing service, describes them as "the practical or theoretical shortcomings of a study that are often outside of the researcher's control". Time, money, access to people, the tools you could use and the design you could afford all set limits.

Students often write this section as a confession, kept short so nobody notices. It works the other way. AJE puts it plainly: "limitations boost a study's credibility". A limitation you name yourself shows that you understand your own method. One the examiner has to point out looks like something you missed.

limitationnoun

A constraint on your study that you could not avoid, which narrows what your findings can show or how widely they apply.

Journals take this seriously too. Reporting guidelines, the checklists many journals ask authors to follow, include an item for it. The STROBE checklist for observational studies, for example, asks authors to "discuss limitations of the study, taking into account sources of potential bias or imprecision".

Two words get mixed up in many Indian theses, often under one heading called "scope and limitations of the study". A common way to separate them: limitations are what you could not do, and delimitations are what you chose not to do.

Editage gives a quick test for the difference: if unlimited time and money would have changed the decision, it is a limitation. Hemlata would have tested milk samples for antibiotic residues if she had the money. That makes it a limitation. Choosing cows and not buffaloes was a decision about her question, not a constraint. That is a delimitation.

A delimitation can still cost you something. Hemlata chose villages near the city, which was a delimitation, but what that choice means for her findings belongs among the limitations.

Your scope belongs at the start. Bikaner Technical University's PhD thesis guidelines, for example, ask the introduction chapter to set out the aim, sub-objectives and scope of the work. Check what your own university's format asks for, and where.

Guides sort limitations in different ways. AJE uses five types: theoretical, methodological, empirical, analytical and ethical. For a student writing a thesis, it is easier to think about where in the study the limit came from.

"Small sample size" on its own tells the reader little. Say why it matters: is it too small to detect a difference you cared about, or too narrow to represent the people you want to talk about? Our guide to sampling methods in research explains how the way you chose people shapes what you can claim about them.

In most journal articles, limitations sit inside the discussion section. AJE notes they are placed "either at the beginning of the discussion section" or at its end, and that "it varies depending upon the target journal or publication guidelines". Editage adds that many fields put them just before the concluding paragraph, while engineering and computer science papers often have a separate "Limitations and future work" section.

In a thesis, the usual places are a subsection of the discussion chapter or a section of the final chapter. Some formats ask for scope and limitations in Chapter 1 as well. Whatever your format, do not hide a limitation in the middle of the results. AJE advises keeping the section "prominently positioned and clearly introduced".

Treat these numbers as a rough guide, not a rule. Fewer than three can suggest you have not looked hard. Many more can bury the important ones. If you are writing the discussion around them, our blog on how to write discussion in research paper shows how limitations fit with the other parts.

Two researchers at the University of Michigan Medical School, Ross and Bibler Zaidi, wrote a short paper on this in Perspectives on Medical Education in 2019, called "Limited by our limitations". Their complaint was simple: "Too often, authors use generic descriptions to describe study limitations." They also noticed that "too often, authors only list the potential limitations" without explaining them. Editage gives a similar four step pattern. Put together, each limitation gets four parts:

1. Name it.State the limitation plainly, and say where it came from: the sample, the design, the measurement, the setting or the resources.
2. Say what it does to your findings.Which way might it push your results, and roughly how much? Which claim can you no longer make because of it?
3. Say what you did about it.Describe any step you took to reduce it, or why you still chose this approach over the alternatives.
4. Point to the next study.Name the specific design, measure or sample that would remove this limit. This is where your future scope comes from.

Parts two and four are the ones students skip, and they are the ones examiners read most closely. A limitation without its consequence is just a complaint. A limitation with its consequence is a careful boundary around your claim.

Open the section with a neutral line, such as "Several limitations should be kept in mind when reading these findings." Avoid "unfortunately", "sadly" or "we regret". You are describing your study, not apologising for it.

Hemlata started again with her methods chapter open. She found three real limitations: the answers were self-reported, she had no lab test of the milk, and all six villages were close to the city. Here is how her first draft compared with the rewrite.

Her full paragraph on the first limitation used all four parts:

First, withdrawal practice was self-reported. Farmers were interviewed by a veterinary student, and some may have given the answer they felt was expected, so the share who said they always wait the full withdrawal period (86 of 120, or 72 percent) is more likely an overestimate than an underestimate. To reduce this, questions were asked in Rajasthani by a local livestock assistant, and farmers were told their names would not be recorded with their answers. Testing milk samples for antibiotic residues alongside the survey would show how closely reported practice matches what reaches the milk can.

The paragraph names the limit, says which way it pushes her main number, explains what she did to reduce it, and turns the gap into a specific next study. Her finding still stands as a careful description of what farmers say they do.

These are the mistakes Hemlata's first draft made, and the ones that Editage, AJE and Ross and Bibler Zaidi warn about most:

• Generic lines. "Small sample size" and "limited time" with no detail could belong to any thesis.
• Listing without explaining. A list of weaknesses with no word on what each one does to the findings.
• Too many. Fifteen small worries bury the two that matter.
• Apologising. "Unfortunately, this study suffers from many problems" makes the reader doubt everything.
• Errors called limitations. A wrong formula or a coding slip should be fixed, not confessed.
• New results. A fresh analysis slipped into the limitations looks like an attempt to rescue the study. Put it in the results.
• A different story elsewhere. An abstract or conclusion that claims more than the limitations allow.
• "Further research is needed." Name the study that is needed.

Examiners often ask about limitations at the viva, so this section is also your preparation. If you can explain each one calmly in writing, you can explain it out loud. Our list of PhD viva questions includes the ones examiners ask about the weak points of a study.

Hemlata's new section was about 300 words, three limitations in three short paragraphs. At her viva, the external examiner asked how she knew the farmers were telling the truth. She said she did not know for certain, explained which way the answers were likely to lean and why, and described the residue study that could check them. The examiner nodded and moved on to the next question.