- Implications explain what your results mean; recommendations state what someone should do about them. Keep the 2 ideas separate.
- Implications belong in the discussion; recommendations sit in the conclusion or in a clearly labeled subsection near the end.
- Tie every implication to a named finding, and match the certainty of your wording to the strength of your evidence.
- Build each recommendation from 4 parts: audience, action, context, and justification.
What Are Implications in a Study?
Implications are statements that explain what your findings mean for theory, practice, policy, or future research. They interpret significance instead of restating results.
Your results can be that “mean blood pressure significantly differed between the treatment and control groups at 3, 6, and 12 months post-intervention”. An implication explains why that observation matters and what changes because of it. For example, “this finding indicates that app-based nutritional education can be a viable component of blood pressure management strategies for hypertensive adults”.
Strong implications share 5 features:
- They refer to a specific finding rather than to the study as a whole.
- They identify who is affected: other researchers, clinicians, teachers, managers, or regulators. And if the implication pertains to patients or the general public, they specify who: children, adults, older adults, etc.
- They state a consequence: a model is extended, a care routine is enhanced, a cost is reduced.
- They use modal verbs such as suggests, could, or indicates when the evidence is inconclusive in some way (e.g., from a cross-sectional study).
- They acknowledge the study’s own delimitations: sample, setting, and time frame.
What Are Recommendations in Research?
Recommendations are concrete, actionable proposals that follow from your findings. They tell a defined audience what to do next: adopt, revise, pilot, monitor, fund, or investigate.
A recommendation converts the implications of a finding into an instruction. If the reader cannot picture the action being carried out on a specific Monday morning, the wording is still too abstract to be useful.
Useful recommendations are:
- Specific: they name the action and the setting, not a general direction.
- Feasible: they respect budget, staffing, and regulatory limits.
- Traceable: each item links back to a result reported in the paper.
- Prioritized: high-impact actions appear first.
- Measurable: they indicate how success would be judged.
Example
A study on attrition in elementary school teachers finds a moderate correlation between job satisfaction and number of children in the class with autism. The recommendation cannot be “remove all children with autism from gen-ed classrooms”. Instead, it can be “future studies should explore whether increasing support from school-based behavioral interventionists can improve job satisfaction and reduce attrition in gen-ed teachers”.
Implications vs. Recommendations: Key Differences
Implications describe meaning; recommendations prescribe action. Implications answer the reader’s question “so what?”, while recommendations answer “what next?” The table below summarizes how the 2 sections differ in practice.
| Feature | Implications | Recommendations |
| Core question | What does this finding mean? | What should be done now? |
| Mood | Descriptive and interpretive | Prescriptive and directive |
| Primary audience | Researchers and theorists | Practitioners, policymakers, funders, clinicians, managers |
| Typical verbs | suggests, indicates, extends, challenges | should, adopt, pilot, allocate, evaluate |
| Common failure | Claiming more than the data support | Advice too vague to act on |
Note that depending on your target journal’s guidelines, your implications and recommendations can both go together in the Discussion section, or your recommendations may go into a separate Conclusion section.
Types of Research Implications
Most studies produce more than 1 kind of implication. Sorting them by type prevents you from repeating a single point in 3 different sentences.
| Type | Focus | Typical opening phrase |
| Theoretical | Concepts, models, and assumptions in the literature | “These results extend [model] by showing that …” |
| Practical | Daily decisions made by practitioners | “For clinicians, the findings indicate that …” |
| Policy | Rules, standards, funding, and regulation | “At the policy level, the data support revising …” |
| Methodological | Instruments, measures, and study design | “The low reliability of [scale] suggests that …” |
| Social or economic | Communities, equity, costs, and access | “Wider adoption could reduce [cost] for …” |
| Future research | Unanswered questions and next studies | “Further work is needed to test whether …” |
Theoretical Implications
Theoretical implications explain how your study changes the way a concept is understood or the overall pattern of literature on a topic. Here’s how:
- Confirm: results replicate an established model or what other researchers have found, in a new sample or setting.
- Extend: previous findings or a new model now applies to a group, condition, or period not previously studied.
- Qualify: you’ve found that a theory or finding that was assumed to be general, now holds only under stated conditions that you can specify.
- Challenge: results contradict what existing studies have found or a popular existing theory.
Practical and Policy Implications
These implications translate results for non-researchers, but people who are involved in the field in some way, e.g., clinicians, managers, teachers, practitioners. Name the decision maker and the decision, then state how your evidence should shift it.
- Identify the setting: hospital ward, classroom, factory floor.
- State the change in behavior, allocation, standard, or workflow.
- Note the conditions under which the change is justified.
- Flag cost, training, or infrastructure requirements that adoption would create.
Methodological Implications
Methodological implications pertain to how future work should be designed or measured. They carry real value when your instruments behaved unexpectedly.
- Report measures that performed poorly and explain what that means for interpretation.
- Suggest sampling improvements: larger samples, stratification, or different recruitment channels.
- Recommend designs that would strengthen causal claims, such as randomized or longitudinal work.
How Do You Write Implications?
Start from a specific finding, state what it changes, name who is affected, then calibrate your certainty to your evidence. The 5 steps below make the process repeatable.
- Step 1: Identify the finding. Restate the result in plain language, including direction and size of the effect.
- Step 2: Ask “so what?” 3 times. Push past the obvious answer until you reach a consequence a reader outside your subfield would care about.
- Step 3: Assign the audience. Decide whether the consequence matters to scientists, practitioners, policymakers, the public, or all 3.
- Step 4: Choose appropriate verbs to avoid overstating. Match verbs to design: use “demonstrates” for experimental evidence and “suggests” for correlational evidence.
- Step 5: Specify boundaries. Add the sample, setting, or time frame that limits generalization. For example, “Our findings on bone turnover markers show that low-impact exercise can be a valuable preventive measure for peri-menopausal women with a family history of osteoporosis”.
Worked Example: Turning a Finding Into an Implication
The table below shows the same result at 4 stages of drafting. Each revision removes vagueness and adds precision.
| Version | Text | Comment |
| Finding | Students using spaced review scored 12% higher on the delayed test. | Neutral statement of the result. |
| Weak implication | This finding has important implications for education. | Names no consequence and no audience. |
| Better implication | The result suggests that spacing improves retention beyond immediate performance. | Adds a consequence but still no audience. |
| Strong implication | For course designers, the 12% gain suggests that review should be spread across weeks rather than massed before the exam, at least in introductory science courses. | Names audience, consequence, and boundary. |
How Do You Write Recommendations?
Name the audience, specify the action, set the context, and justify it with a result. Recommendations that drop any of these 4 elements read as generic advice.
Components of an Actionable Recommendation
| Element | Question it answers | Example |
| Audience | Who acts? | District curriculum committees |
| Action | What exactly should be done? | should schedule 3 spaced review sessions per unit |
| Context | Where, and under what conditions? | in introductory science courses with more than 60 students |
| Justification | Why is this warranted? | because spaced review raised delayed test scores by 12% |
Joined together, the 4 elements produce a single sentence that a reader can act on without returning to your results section.
Ordering and Prioritizing Recommendations
- Lead with the recommendation supported by your strongest evidence.
- Group items by audience so each reader finds a relevant cluster quickly.
- Separate short-term actions from long-term ones.
- Limit the list to 3-6 items; longer lists dilute attention.
- In theory-driven papers, close with research recommendations rather than practice recommendations.
Examples of Implications and Recommendations
Example 1: Health Sciences
- Finding: Nurse-led follow-up calls reduced 30-day readmissions by 18% across 2 urban hospitals.
- Theoretical implication: The result supports continuity-of-care models that emphasize post-discharge contact rather than in-hospital education alone.
- Practical implication: Discharge planning can be strengthened through low-cost telephone contact, without adding clinic visits.
- Recommendation: Hospital administrators should pilot nurse-led calls within 72 hours of discharge for patients with heart failure, then review readmission rates after 6 months.
Example 2: Education
- Finding: Teachers who completed 8 hours of feedback training raised student writing scores by 0.4 standard deviations.
- Theoretical implication: Findings extend feedback-intervention theory by showing that gains depend on training in specificity, not on feedback frequency.
- Practical implication: Short, targeted professional development can outperform generic annual workshops.
- Recommendation: School leaders should replace 1 generic workshop each year with an 8-hour feedback module, and should track writing scores every term.
Sentence Templates You Can Adapt
Templates speed up a first draft, but every bracketed item must be replaced with study-specific detail. Otherwise, you sound too generic to peer reviewers.
| Purpose | Template |
| Theoretical implication | “These results extend [theory] by demonstrating that [relationship] also holds for [population].” |
| Contradictory finding | “Contrary to [prior study], we observed [result]; this suggests that [condition] moderates the effect.” |
| Practical implication | “For [practitioner group], the findings indicate that [practice] may [outcome] under [conditions].” |
| Policy implication | “At the policy level, the association between [X] and [Y] supports revising [standard] to [new value].” |
| Practice recommendation | “[Audience] should [action] in [context] in order to [outcome].” |
| Research recommendation | “Future studies should test [relationship] using [design] in [population] to establish [claim].” |
Common Mistakes in Writing Implications and Recommendations
Overclaiming, restating results, and offering advice nobody can act on account for most weak sections. Check your draft against the 7 recurring errors listed below.
| Mistake | Why it weakens the paper | Quick fix |
| Restating results | Adds length without adding meaning | Cut numbers already reported; keep the consequence |
| Overclaiming causality | Correlational data cannot support “causes” | Use “is associated with” or “predicts” |
| Ignoring limitations | Readers distrust unbounded claims | Attach sample, setting, or time frame to each claim |
| Generic advice | “More research is needed” guides nobody | Name the design, population, and question |
| No named audience | Readers cannot tell who should act | Open each recommendation with the actor |
| Unsupported recommendations | Introduces opinions the study never tested | Cite the specific result behind each item |
| Overloading the list | 15 items dilute the 3 that matter | Prioritize and cut to 3-6 recommendations |
Frequently Asked Questions
What is the difference between implications and recommendations in research?
Implications explain what your findings mean while recommendations state what should be done as a result. Implications are interpretive and belong to the discussion, while recommendations are prescriptive and can go in a separate conclusion section if the journal allows. Implications are nearly always presented before recommendations.
How do you write the implications of a study in a thesis?
Create a dedicated subsection in the discussion chapter, then organize it by type: theoretical, practical, policy, and methodological. Open each paragraph with the finding, follow with the consequence, and close with the boundary condition. Most examiners expect 2-4 pages that address every research question at least once.
Where should recommendations be placed in a research paper?
Place them at the end: either as the final part of the conclusion or as a labeled subsection immediately before it. Journals with tight word limits often expect only 3-5 sentences.
What are examples of theoretical and practical implications?
A theoretical implication might state that findings extend self-determination theory to fully remote workers. A practical implication might state that managers can raise retention by granting schedule control. The first changes how a concept is understood while the second changes what someone does on the job.
Can you write recommendations for future research without a limitations section?
You can, but the paper is weaker for it. Limitations identify the gaps that future studies should close, so the 2 sections reinforce each other. If your journal template omits a limitations heading, includes limitations in the Discussion section anyway and link them logically to your recommendations.
How do you write implications for future research in a qualitative study?
Ground them in transferability rather than statistical generalization. State the contexts where your themes may apply, identify participant groups that were not represented, and propose designs that would test the themes further, such as multi-site case studies or longitudinal interviews with 20-30 participants.
What tense should be used for implications and recommendations?
Use present tense for what the findings mean (“the results suggest”), past tense for what you did (“participants reported”), and modal verbs for recommendations (“clinicians should pilot”). Avoid future tense in implications, since it implies a level of certainty that most designs cannot support.
What do I do if my study has weak or small effect sizes?
A small effect is still a finding. Report them honestly and scale your implications to match. To do so, compare the effect you found against benchmarks in your field, and note what acting on it would cost.
You can also frame weak or null results as evidence about boundary conditions rather than as a failed study. For example, a weight loss drug produces negligible weight loss in a sample of older adults. If the study had adequate power and was well designed, it provides valuable evidence that this drug is not effective for older people, and can prevent doctors from unnecessarily prescribing that drug to older patients.
Practical steps while writing about weak or small effects:
- Swap strong verbs for hedged ones: “is associated with a modest increase” rather than “improves.”
- Report confidence intervals so readers can judge precision alongside magnitude.
- Ask whether the intervention or phenomenon you’re studying is cheap to implement or accumulates across a large population. Small effects can still justify action on either ground.
- Consider whether low power, attrition issues, or unreliable measures have affected your results, then recommend the design that would test that explanation.
Comment