Chapter 19. Limitations, Recommendations, Contribution and Conclusion
19.1 Limitations
Limitations identify features of the study that constrain interpretation. They are not ritual apologies. Each limitation should state what the issue is, why it matters, the direction or nature of possible bias where reasonably known, and what was done to mitigate it. Be specific about sampling, measurement, design, missingness, confounding, self-report, access, time, model assumptions and external validity.
| Limitation | Why it matters | Appropriate response |
|---|---|---|
| Convenience sample | Selection probabilities are unknown; sample may differ from target population | Limit generalisation; describe recruitment and sample; recommend probability sampling where feasible |
| Cross-sectional design | Temporal order and reverse direction are uncertain | Use association language; recommend longitudinal or experimental follow-up |
| Self-report measures | Recall, social desirability and common method effects may distort responses | Use validated scales and confidentiality; triangulate with independent data in future work |
| Single setting | Context may limit transferability | Describe context precisely; replicate across settings |
| Nonresponse/missing data | Respondents may differ from nonrespondents; estimates may be biased | Report patterns; use suitable missing-data methods and sensitivity checks |
| Measurement adaptation | Altered wording may change scale properties | Pilot, document adaptation and assess measurement evidence |
| Unmeasured confounding | Observed association may partly reflect omitted variables | Use theory-driven covariates and avoid causal claims |
19.2 Recommendations
Recommendations should follow from findings and be directed to an identified audience: managers, policymakers, educators, practitioners, researchers or institutions. Separate practical recommendations from future-research recommendations. Each should be feasible, proportionate to the evidence, and linked to a specific result. Avoid generic statements such as “organisations should improve communication” unless the study measured communication and supports that action.
A useful recommendation format is: finding → implication → action → responsible actor → conditions/resources → indicator for evaluation. For example, if a robust association is observed between supervisor support and engagement, the study might recommend piloting manager-support practices and evaluating them prospectively, rather than asserting that a training programme will necessarily cause engagement to rise.
19.3 Contribution to knowledge
A contribution can be theoretical, empirical, methodological, contextual, measurement-related or practical. It must be stated specifically. “This study contributes to knowledge” is not enough. Identify what was previously uncertain, what the study adds, how it differs from prior work, and the boundary within which the contribution holds.
| Contribution type | Questions to answer | Evidence expected |
|---|---|---|
| Theoretical | Does the study refine a mechanism, boundary condition or relationship? | Clear theory-based argument and findings relevant to it |
| Empirical | Does it provide credible new estimates or test an unresolved pattern? | Transparent data, design, estimates and uncertainty |
| Methodological | Does it improve measurement, design or analytic practice? | Demonstrated advantage or documented validation |
| Contextual | Does context test a meaningful boundary condition? | Reasoned context choice and comparison with prior settings |
| Practical | Does it inform a decision or intervention? | Actionable implications proportionate to evidence |
19.4 Conclusion
The conclusion gives a concise, direct answer to the overall research aim. It summarises the principal findings and their significance, states the defensible contribution, notes the main boundary of interpretation, and closes without introducing new evidence. It should not simply copy the abstract, repeat the whole discussion, or make stronger claims than the results support.
19.5 Suggested final chapter structure
- Introduction and reminder of the study aim.
- Summary of key findings by objective.
- Overall conclusion and answer to the research question.
- Contribution to theory, evidence, method or practice.
- Practical and policy implications.
- Limitations and boundaries.
- Recommendations for practice and future research.
- Final concluding paragraph.