For students with a draft final chapter: Chapter Five of a Nigerian project, or the summary of findings, conclusions and recommendations of a Philippine thesis. The page goes through the chapter part by part and shows what each part has to trace back to. University guides from Nigeria, Kenya and the Philippines are quoted where they set a rule.
From within your field, Esmond points out where your conclusion goes further than your data, contradicts your results, or leaves out findings that don't fit.
Chapter Five adds no new data. It condenses what Chapters One to Four did and found, says what that means for the objectives, and says who should act on it. The thesis format of the graduate school at Jomo Kenyatta University of Agriculture and Technology (JKUAT) in Kenya describes the chain in three lines: “The summary which is just an extended abstract”, then “The conclusions which is derived from the summary”, then “The recommendations which is derived from the conclusion”.
A supervisor can read that chain backward: each recommendation should rest on a conclusion, each conclusion on a finding, and each finding on a table or test in Chapter Four. A missing link anywhere in the chain becomes a question in the margin.
Where the chapter sits depends on the outline. Newgate University Minna's guidelines put the conclusion and recommendations in a sixth chapter, after a separate discussion chapter. The University of Ibadan's Postgraduate College calls the part the conclusion and notes that it may also hold recommendations, depending on the discipline. For conclusions outside this five-chapter model, see what goes in a thesis conclusion?; for other chapters, thesis chapter feedback.
A hypothetical example of a commentYour second recommendation asks the state ministry to fund school libraries, but none of your findings in Chapter Four concerns library access. Which finding does this recommendation come from?
A first comment on one part of your own draft, right away and without an account.
Get a free previewSome outlines split the summary in two. The summary of the study restates the problem, objectives, design, sample, instrument and method of analysis in a few paragraphs. The summary of findings then lists one finding per research question or hypothesis, in the order of Chapter One.
A reader compares each finding with Chapter Four. The numbers must be the same (a mean of 2.81 there can't become 2.9 here), the decision on each hypothesis must be the same, and nothing may appear that Chapter Four didn't report. A summary that adds a result, or drops a hypothesis that wasn't supported, breaks that match.
A conclusion states what the findings mean for each objective, within the population you studied. Emilio Aguinaldo College Cavite's manual puts it in one sentence: “Conclusions should fit together with the findings of the study.” The EMT 417 notes at the Federal University of Agriculture, Abeokuta add a limit: a conclusion “Should be based strictly on the outcome of one’s research and not the work of other people.”
Three kinds of conclusion draw questions. One that claims more than the design allows: a correlational study that concludes one variable causes another. One that reaches past the sample: 150 students in one faculty can't carry a conclusion about all Nigerian undergraduates. And one built on a hypothesis that was not rejected: "gender has no effect" says more than a non-significant test shows. The safer form is that the study found no significant difference in that sample.
The Emilio Aguinaldo College manual also asks for some modesty: it advises against statements that make findings and conclusions look definitive and incontestable, and asks that conclusions stay open to what other researchers may find.
Newgate University Minna's guidelines ask that recommendations be drawn from the key findings and offer practical solutions to the problems identified. A recommendation a reader can trace names the finding it rests on, who should act (a school's management, a ministry, a hospital unit, lecturers) and what they should do. "Government should do more" names none of them.
A recommendation with no finding behind it either needs one or belongs in the section on further studies. A finding that called for action and has no recommendation is a gap of the opposite kind.
Newgate University Minna's guidelines put the emphasis of the final chapter on the findings and their contribution to the existing body of knowledge. The University of Ibadan's manual ties it to the conclusion: the inferences drawn from the findings establish the significance of the research and its contribution to knowledge.
A contribution has to follow from your findings and match the gap named in Chapter Two: evidence on a relationship in a population or setting not studied before, an adapted instrument shown to be reliable in a new context, or a result that settles a disagreement between earlier studies. "This is the first study of its kind" is only as strong as the review behind it.
Limitations name what the design couldn't do and how that affects the conclusions: one institution or local government area, self-reported data, a cross-sectional survey, a low return rate. The limitations section page covers this part in more depth.
Suggestions for further studies are strongest when they follow from those limitations or from a question your findings opened: the same design in another state, a longitudinal follow-up, a variable that your review flagged and your instrument didn't measure. The Emilio Aguinaldo College manual describes recommendations for future research as including suggested improvements in the way the research is conducted and related future research directions. A list of new topics with no tie to your study reads as filler.
In a quantitative project, the summary of findings can run hypothesis by hypothesis, with the decision for each, and the conclusions stay within the sample and the test used. In a qualitative project, the findings are the themes from Chapter Four, the conclusions say what those themes show about the research questions for these participants, and the claims are about transferability to similar settings, not about percentages of a population.
| Part | Traced back to | Question you may hear |
|---|---|---|
| Summary of findings | The tables and tests in Chapter Four | Where in Chapter Four is this finding? |
| Conclusions | The objectives in Chapter One | Which objective does this conclusion answer? |
| Recommendations | A numbered finding | Which finding supports this, and who should act on it? |
| Contribution to knowledge | The gap in Chapter Two | What does this study add that the reviewed studies didn't? |
| Limitations | The design and sample in Chapter Three | How does this limit what you conclude? |
| Suggestions for further studies | The limitations and open questions | Why this study next, and not another? |
These fragments are made up. None of them comes from a real thesis.
Esmond evaluates every assignment before you submit. It never writes it.
Esmond is an AI tool that doesn't write for you. If your course doesn't allow any AI tools, don't use Esmond for that work. If you're unsure, ask your instructor or advisor.
In Chapter Five, Esmond can flag what a supervisor or panel would ask about in your own text: a summarized finding that Chapter Four doesn't contain, a figure that changed between the two chapters, an objective with no conclusion, a conclusion wider than the sample or the design, a recommendation that no finding supports, or a contribution that the literature review contradicts. The sample report shows how such comments are laid out, and how Esmond works explains the steps.
Esmond reads what you upload. If the file holds the whole draft, it can set this chapter against the others; if it holds one chapter, the review stays within that chapter and may raise a point that another chapter already answers.
Esmond doesn't propose recommendations or conclusions for you, and it doesn't check the chapter against your department's outline. What you keep and what you change is your decision.
The summary restates what was done and what was found, with the same figures as Chapter Four. The conclusion says what those findings mean for each objective, within the sample and design used. JKUAT's thesis format describes the conclusions as derived from the summary.
None of the guides quoted here sets a number. A reader checks that each finding calling for action has a recommendation, and that each recommendation names who should act and what they should do.
What your study adds to what the reviewed literature already showed. It has to follow from your findings and match the gap named in Chapter Two.
Read on November 8, 2026.
Get your paper evaluated before you submit