Quick answer: To learn how to write recommendations in a dissertation, start from your findings: each recommendation should name who should act, what they should do, and which finding justifies it.

Group recommendations by audience (practice, policy, and further research), keep them specific and feasible, match their strength to your evidence, and place them in the conclusion chapter or a dedicated recommendations section after your discussion.

Recommendations are where a dissertation shows that its findings matter beyond the page. Examiners read them closely because weak recommendations are easy to spot: they are generic (“more training is needed”), disconnected from the data, or so ambitious that no one could act on them. This guide explains how to write recommendations in a dissertation step by step, with a formula you can apply to each one, illustrative examples across disciplines, and the mistakes that most often draw examiner comments.

Recommendations versus implications versus future research

Students often blur three related parts of the final chapter. Separating them makes each one sharper.

Implications describe meaning; recommendations prescribe action. The site’s guide to writing future research recommendations covers the research strand in depth, so this page focuses on recommendations for practice and policy and on how the strands fit together.

How to write recommendations in a dissertation: the five-part formula

A strong recommendation usually contains five elements, often in one or two sentences:

1. Actor: who should act, as specifically as possible, such as “primary school leadership teams” rather than “schools”.
2. Action: what they should do, using a concrete verb such as introduce, revise, monitor, fund or train.
3. Evidence: the finding that supports it, referred to briefly, such as “given that teachers reported…”.
4. Expected effect: what the action should achieve.
5. Condition or caution: any limit on where the recommendation applies, reflecting your sample and design.

Weak (illustrative): “Schools should improve communication with parents.”

Strong (illustrative): “Primary schools in the study district should replace termly paper newsletters with fortnightly short messages through the existing parent app, because parents in all four focus groups reported missing information in printed letters; this is likely to be most relevant to schools with high proportions of working parents, as in the study sample.”

The strong version names the actor, the action and the evidence, and it limits its reach to the setting the study examined.

Step by step: from findings to recommendations

Step 1: List your key findings

Write each main finding as a single sentence, in the order of your research questions. Recommendations that do not trace back to this list do not belong in the dissertation.

Step 2: Ask who could act on each finding

For each finding, list the people or organisations in a position to respond: practitioners, managers, policymakers, professional bodies, software designers, educators or researchers. Some findings will have no practical audience; that is fine.

Step 3: Draft one recommendation per finding and audience

Apply the five-part formula. Keep each recommendation to one action; if you find yourself writing “and also”, split it.

Step 4: Check strength against evidence

Match your language to your design. A small qualitative study can recommend that an organisation “consider” or “pilot” an approach; a large evaluation with consistent effects can support “should introduce”. Overclaiming is one of the most common examiner criticisms. If you have already written your limitations, the guide on writing the limitations section will help you calibrate.

Step 5: Prioritise and group

Order recommendations by importance or by audience, and group them under subheadings if there are more than four or five. Most dissertations need between three and eight recommendations in total; more than that usually signals that some are generic.

Step 6: Check feasibility

Ask whether the actor could realistically do what you propose with the resources they have. Where cost or training is involved, say so and suggest a first step, such as a pilot.

Illustrative examples across disciplines

The examples below are illustrative, written for fictional studies, to show the formula in different fields.

• Nursing: “Ward managers on units using rotating shifts should schedule protected mentoring time for nurses in their first year, since participants linked their intention to leave to the lack of contact with mentors on opposite shifts.”
• Business: “Small retailers adopting online ordering should assign one staff member to monitor customer messages during opening hours, because response time was the strongest predictor of repeat purchase in the survey.”
• Education: “Teacher education programmes in the region should add a module on assessment design for mixed-ability classes, as newly qualified teachers reported feeling least prepared in this area.”
• Public health: “The local health authority should pilot text-message reminders for second-dose appointments in the two districts with the lowest completion rates before considering wider rollout.”
• Computer science: “Developers of student-facing learning platforms should provide keyboard navigation for all assessment pages, given that usability testing found this to be the most frequent barrier for screen-reader users.”

Revising a weak recommendations section: before and after

The passage below is illustrative, adapted from the kind of draft supervisors often return, followed by a revision.

Before: “Based on the findings, it is recommended that organisations improve employee wellbeing. Managers should communicate better and there should be more training. The government should also do more to support workers. Further research is needed in this area.”

Problems: no finding is named, every actor is vague, “improve” and “do more” are not actions, and the closing sentence about research says nothing about what to study or how.

After: “Three recommendations follow from the findings. First, line managers in the participating firms should hold a monthly one-to-one check-in with remote employees, because survey respondents without regular check-ins reported significantly higher isolation scores. Second, the firms’ HR teams should add workload planning to existing manager training, since interviewees most often linked stress to unplanned work arriving late in the week. Third, because the study covered three firms in one sector, future research should test whether the association between check-ins and isolation holds in other sectors, ideally with a longitudinal design that can show whether check-ins precede changes in isolation.”

The revision is longer, but every sentence now does work: each recommendation names an actor, an action and the finding behind it, and the research recommendation explains what the current study could not show.

Recommendations in professional doctorates and applied programmes

In applied degrees, recommendations often carry more weight than in research-focused dissertations. Doctor of Nursing Practice projects, Doctor of Education dissertations, MBA consultancy projects and public administration capstones are frequently written for a partner organisation, and examiners assess whether the recommendations are usable in that setting. In these programmes, it is common to add an implementation note to each recommendation: who would lead it, what resources it would need, a realistic timeline and how success would be measured. Some programmes also ask for a dissemination plan describing how the recommendations will be shared with the partner organisation. If your programme uses a template for this section, follow it exactly; if not, a short table with one row per recommendation and columns for action, lead, resources and measure is a clear way to present the detail.

Writing recommendations for policy

Policy recommendations need particular care, because the actors are large and the evidence from a single dissertation is usually limited. Name the level of government or the specific agency that holds the relevant power, refer to the existing policy or programme your recommendation would change, and be realistic about what one study can support. A recommendation that a regulator “review” or “consider” a guideline in the light of your findings is often more defensible than one that it “mandate” a change. Where your study is local, frame policy recommendations as questions for decision-makers to examine rather than as conclusions.

Where recommendations go in the dissertation

Conventions vary by institution and discipline, so check your handbook. Common structures are:

• Within the conclusion chapter, after a summary of findings and before closing remarks. This is common in UK undergraduate and master’s dissertations.
• A dedicated “Recommendations” section or chapter, often required in professional programmes such as education, nursing, business and public administration, and in applied doctorates such as the DNP or EdD.
• At the end of the discussion chapter, in some article-style or scientific theses, followed by a short conclusion.

Wherever they sit, recommendations should come after the discussion of findings, so the reader has seen the evidence before being told what to do with it. If you are unsure how the discussion and conclusion divide the work, the comparison of discussion versus conclusion chapters explains the split.

Common mistakes examiners flag

1. Generic recommendations that could follow from any study, such as “more awareness is needed”.
2. Recommendations with no link to findings, often based on the literature review or personal opinion rather than the data.
3. Overreach: national policy changes recommended from a single-site study of twelve participants.
4. Vague actors such as “the government” or “stakeholders”.
5. Mixing future research into practice recommendations without signposting.
6. New findings or new literature introduced for the first time in the recommendations.

A recommendations checklist

• Every recommendation traces to a stated finding.
• Each names a specific actor and a concrete action.
• The language matches the strength of the evidence and the scope of the sample.
• Practice, policy and research recommendations are clearly separated.
• There are few enough recommendations that each one matters.
• The placement follows your programme’s handbook.

For related guides, see the difference between results and discussion and how to write a qualitative dissertation.

Where Tesify fits

Recommendations are only as strong as the chain from research question to finding. Start structuring your thesis in Tesify. It helps you structure and organize every chapter so that chain stays visible, and 9,000+ students have written 15,000+ chapters with Tesify. Every word stays 100% written by you.

Frequently asked questions (FAQ)

How many recommendations should a dissertation have?

There is no fixed number, but most dissertations make between three and eight focused recommendations. Fewer specific recommendations are stronger than many generic ones.

Where do recommendations go in a dissertation?

Usually in the conclusion chapter or a dedicated recommendations section after the discussion. Check your programme handbook, since some professional degrees require a separate chapter.

What is the difference between recommendations and implications?

Implications explain what the findings mean for theory or practice, while recommendations state what a specific actor should do because of those findings.

Can I make recommendations from a qualitative study?

Yes, but phrase them to match the evidence, for example suggesting that an organisation consider or pilot an approach, and limit them to settings similar to the one you studied.

Should recommendations include future research?

Many dissertations include recommendations for future research, but present them separately from recommendations for practice or policy so readers can tell them apart.

How do I make recommendations specific?

Name the actor, the action, the finding that supports it, the expected effect and any condition on where it applies. Replace general verbs such as improve with concrete ones such as introduce or monitor.

Can I cite literature in my recommendations?

You can refer to supporting literature briefly, but every recommendation should rest mainly on your own findings. Avoid introducing new sources for the first time in this section.

What tense should recommendations be written in?

Use the present tense with modal verbs such as should, could or may, matched to the strength of your evidence.