To write recommendations in a dissertation, turn your key findings into specific, realistic actions for practitioners, policymakers or future researchers, each one clearly linked to your evidence. Good recommendations show that you understand what your results mean beyond the page.
This guide explains where recommendations usually sit, the main types, a simple formula for writing each one, how many to include and the mistakes examiners notice. Conventions vary by discipline and university, so check your handbook for any required format.
The Three Things Every Recommendation Needs
- A clear audience: who should act, such as ward managers, school leaders, regulators or researchers.
- A specific action: what they should do, stated precisely enough to act on.
- A link to evidence: which finding supports it, so the recommendation follows from your study rather than your opinion.
If a recommendation could have been written without doing your research, it is probably too general.
Where Recommendations Go in a Dissertation
Recommendations most often appear in the final chapter, after the summary of findings and before or after the limitations. Some programs place them at the end of the discussion chapter instead, and some applied dissertations give them their own short chapter.
| Dissertation type | Typical placement |
|---|---|
| Five-chapter empirical dissertation | A subsection of Chapter 5 (conclusion or discussion and conclusion) |
| Separate discussion and conclusion chapters | Usually the conclusion chapter |
| Practice doctorate or applied project | Often a dedicated section on implications and recommendations |
| Business or management dissertation | Often a prominent section, sometimes with an implementation plan |
| Humanities dissertation | Often brief, or limited to directions for future research |
Types of Recommendation
Most dissertations include two or three types. Use subheadings so readers can find the ones relevant to them.
- Practice: changes for professionals or organizations, such as new procedures, training or tools.
- Policy: changes to rules, guidelines, funding or regulation at organizational, regional or national level.
- Future research: studies that would address your limitations or extend your findings.
- Theory: refinements to a framework or model that your findings suggest.
- Education: changes to curricula or professional development, common in nursing and teaching dissertations.
Recommendations Versus Implications
Implications explain what your findings mean; recommendations say what someone should do about it. "Night-shift nurses report higher burnout" has implications for workforce planning. "Managers should limit consecutive night shifts to three" is a recommendation. Many dissertations discuss implications first, then make recommendations.
A Simple Formula for Each Recommendation
A reliable recommendation has four parts. Writing each one this way keeps it specific and evidence-based.
- Finding: the result it rests on.
- Action: what should be done.
- Actor: who should do it.
- Rationale or expected benefit: why it should help, and any conditions.
Weak. Hospitals should do more to support new nurses.
Stronger. Because participants described hiding uncertainty in their first months (Theme 1), ward managers should assign every newly qualified nurse a named mentor for at least the first twelve weeks, with protected weekly check-ins. This would give new nurses a safe route to ask questions, which participants linked to faster confidence.
The stronger version names the finding, the action, the actor and the expected benefit in three sentences.
Writing Recommendations for Future Research
Future research recommendations should grow from your limitations and from questions your findings raised. Be specific about population, design and focus.
Generic. More research is needed on this topic.
Specific. This study used a single-site qualitative design. A multi-site study across rural and urban hospitals could test whether the mentoring patterns identified here hold in different staffing contexts, and a longitudinal design could examine whether early mentoring affects retention at two years.
Two to four well-argued research directions are usually more convincing than a long wish list. Where you can, say what design, sample or measure would answer each open question, so another researcher could pick it up and start.
Keep Recommendations Realistic and Proportionate
Examiners check whether your recommendations match the strength of your evidence. A small qualitative study can suggest changes for similar settings, but it cannot justify national policy reform on its own.
- Match the scale of the recommendation to your sample and design.
- Use tentative language where evidence is limited: "could consider", "may benefit from piloting".
- Consider cost, feasibility and who has the authority to act.
- Do not recommend something your study did not examine.
Prioritize
If you have several practice recommendations, order them by importance or by ease of implementation, and say which you would start with. A short table can help in applied dissertations.
A Recommendations Table for Applied Dissertations
Business, management, nursing and education dissertations sometimes present recommendations in a table that links each one to its evidence and implementation details.
| Recommendation | Based on | Responsible | Timeframe |
|---|---|---|---|
| Assign named mentors to all new nurses for 12 weeks | Theme 1: performing competence | Ward managers | Next intake |
| Add an uncertainty-sharing session to induction | Theme 2: permission to not know | Practice education team | Within six months |
| Track mentoring uptake in staff surveys | Limitation: no retention data | Workforce team | Annual survey cycle |
Follow the table with a short paragraph for each recommendation, since examiners expect reasoning, not just a list.
Worked Example: A Full Set of Recommendations
Here is how one master's dissertation turned three findings into a short, well-organized recommendations section. The study surveyed 180 employees of a mid-sized retailer about hybrid working and interviewed twelve team leaders.
Findings. (1) Employees with a fixed weekly office day reported stronger team belonging than those with no fixed pattern. (2) Team leaders felt untrained in managing hybrid teams. (3) Belonging was not linked to self-rated productivity.
Practice. Team leaders should agree one fixed office day per week with each team, since a fixed pattern was associated with stronger belonging (Finding 1). The HR department should offer short training in hybrid team management, because every interviewed leader described learning by trial and error (Finding 2).
Future research. Because the survey was cross-sectional, a longitudinal study could test whether fixed office days build belonging over time or simply attract employees who already feel connected. Objective productivity measures could also clarify Finding 3, which relied on self-report.
Each recommendation names a finding, an action and an actor, and the research directions come straight from the study's limitations. The section is short, but every sentence earns its place.
Common Mistakes in Dissertation Recommendations
- Too general: "improve communication" says nothing actionable.
- Not linked to findings: each recommendation must trace to a result.
- Overreaching: sweeping policy claims from a small sample.
- New evidence: introducing sources or data that never appeared earlier.
- Too many: a dozen thin recommendations dilute your strongest points.
- Repeating limitations: research recommendations should propose solutions, not restate problems.
How Long Should the Recommendations Section Be?
There is no set length. In a master's dissertation, recommendations often take a few hundred words to a page or two; in a doctoral dissertation they may run longer. What matters is that each recommendation is justified.
Proportion matters more than word count. If your recommendations section is longer than your summary of findings, it has probably drifted into opinion. As a rough guide, three to five practice or policy recommendations and two to four research recommendations are common, but follow your program's expectations.
How DissertationHorse Helps with Recommendations
Send your findings, discussion chapter, conclusion draft and program guidelines. A writer at your level prepares a custom recommendations section, with practice, policy and research recommendations each tied to your own evidence, for you to study and use as a reference.
We work only from your actual findings and never invent results. Free revisions within the scope of your original request are included.
Want Recommendations Your Examiners Will Respect?
Send your findings, discussion and conclusion draft. Your writer prepares a custom recommendations section, each one tied to your evidence, for you to study and adapt.
Order Your RecommendationsFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Usually in the final chapter, after the summary of findings. Some programs place them in the discussion chapter or a separate section, so check your template.
There is no fixed number. Many dissertations include three to five practice or policy recommendations and two to four for future research.
Implications explain what your findings mean for theory, practice or policy. Recommendations state specific actions someone should take as a result.
Yes, as long as they are proportionate. Frame them for similar settings and use tentative language, since qualitative findings support transferability rather than generalization.
Mostly they draw on your own findings. You can cite literature briefly to support feasibility, but do not introduce major new sources here.
Yes. Non-significant or unexpected results still suggest directions, especially for future research design, measurement or sample size.
A short list or table can help, but each recommendation still needs a sentence or two of justification. Many examiners prefer paragraphs with clear subheadings.
Only if you have a sound basis for it, such as figures from your own data or a published source. Otherwise, note that costs should be assessed before implementation rather than guessing.
Recommendations usually use "should", "could" or "would", while references to your findings stay in the past tense.