How to Write a Critical Comparative Review of Policy Research Papers
A critical comparative review of two policy research papers judges how each paper’s design, evidence and ethics fit the policy problem, then compares them against suitable quality criteria. Do not summarise the papers side by side. Argue which strategy answers which question better, propose an alternative design and test whether the recommendations follow from the evidence.
You have chosen your two papers, one review and one case study, and now you are staring at a 2,500-word limit and five parts to cover. That is the situation facing many MPA students on SPY00157M Exploring Policy Research at the University of York. The papers are rarely the problem. The structure is.
This guide explains what markers want in each part, how to spend your words, and where marks are lost. It is guidance for planning your own work, and your institution’s rules on AI and integrity apply. This module permits no generative AI.
cta_essay_sample_1
What the brief asks, and what markers reward
The task is to select two papers on the same policy problem, one a review study and the other a case study, and write a critical comparison. The indicative split below is the one most students follow, and it shows where the weight sits.
| Part | Approx. words | What earns marks |
|---|---|---|
| Introduction | 200 | Full references for both papers, purpose, roadmap |
| 1. Problem and questions | 400 | How each paper frames the problem and phrases its questions; role of research in policy |
| 2. Designs and methods | 700 | Critical comparison against fitting quality criteria |
| 3. Ethics | 300 | Reflexivity about power, consent and vulnerability |
| 4. Alternative design | 300 | A justified method that fills a gap |
| 5. Recommendations | 400 | Judging whether conclusions follow from evidence |
| Conclusion | 200 | A judgement, not a recap |
Part 2 carries the most words and the most analysis, so protect it. Many students overspend on describing the papers in Part 1 and arrive at Part 2 with 300 words left.
A step-by-step approach to the critical comparative review
- Confirm that your papers qualify: one review study and one case study, both on the same policy problem.
- Read each paper in full and note its question, design, sources of evidence, sample, limitations and recommendations.
- Build a comparison grid before you write, with one row per criterion and one column per paper.
- Choose quality criteria that suit each design, and decide how you will justify them.
- Draft Part 2 first, because the other parts depend on your judgement there.
- Write Parts 3 to 5 so that each builds on a weakness you identified.
- Write the introduction and conclusion last, then trim to the word limit and declare the count.
Part-by-part guidance
Introduction and Part 1: the problem and the questions
State the policy problem in a sentence, then show how each paper frames it. Two papers on the same problem often define it differently. One might treat homelessness as a public health matter and the other as a service-delivery issue. That difference explains why their questions and designs diverge.
Comment on how the research questions are phrased: causal, descriptive, exploratory or evaluative. Then connect research to policy making. Weiss (1979) separates several meanings of research use, including problem-solving, interactive and enlightenment models. Use one or two of them to say how each paper is likely to influence decisions.
Part 2: comparing designs and methods
This is where the module learning outcomes bite: advantages and disadvantages of techniques, data sources and their limits, and merits of competing designs in a specific context. A short table helps, but the argument must sit in prose beneath it.
Match the criteria to the design. A systematic review can be judged with a tool such as AMSTAR 2 (Shea et al., 2017), looking at search comprehensiveness, risk of bias in included studies and how heterogeneity was handled. A case study can be judged on construct validity, use of multiple sources and a clear chain of evidence (Yin, 2018). Applying the same checklist to both is a common and avoidable error.
Part 3: ethics
Do not list consent forms. Ask whether the authors reflect on their own position, the power gap with participants, and risks to vulnerable groups. A review inherits the ethics of its primary studies, so ask what it can and cannot say about them. For a case study, consider who administered data collection and whether that affects free consent.
Part 4: an alternative design
Propose one design, not three. Say which weakness in your papers it addresses, what it would involve in practice, and what it would cost in rigour or feasibility. A realist evaluation (Pawson and Tilley, 1997), a longitudinal mixed-methods study or a comparative case design could each work, depending on the gap you found.
Part 5 and the conclusion
Take each recommendation and ask whether the evidence supports it, whether it is feasible and who it leaves out. If you suggest your own, keep it tied to a weakness you have already named. End with a judgement about what the two papers together can and cannot tell a policymaker.
cta_essay_sample_2
Marks-maximising checklist
- Both papers appear in full Harvard form in the introduction and reference list.
- Every claim about a paper is checked against the paper itself, not its abstract.
- Quality criteria differ by design and are justified.
- Each strength or weakness is explained in relation to the policy problem, not in the abstract.
- Ethics goes beyond procedure into reflexivity and power.
- The alternative design names its own limits.
- Word count is declared and the formatting follows the brief: 1.5 spacing, 12-point text, anonymous submission.
If you want to see these habits in action, our worked sample for reference only compares a systematic review and a case study on Housing First. Use it to study the structure, not to copy.
Common mistakes that cost marks
| Mistake | Why it loses marks | Fix |
|---|---|---|
| Summarising each paper in turn | The brief wants comparison and critique | Organise by criterion, not by paper |
| One checklist for both designs | Misjudges the case study or the review | Pick criteria suited to each |
| Generic ethics paragraph | No reflexivity on this policy context | Name the participants, the power gap and the risks |
| Alternative design with no justification | Looks like wishful thinking | Link it to a named gap and admit its costs |
| Recommendations accepted at face value | Part 5 asks you to evaluate them | Trace each back to the evidence |
| Going over the limit | Content beyond the limit may go unmarked | Trim description, not analysis |
What students find hardest
The same few worries come up. Students ask how critical they are allowed to be about a published paper, whether a government evaluation counts as a case study, and what to do when a paper says little about ethics. The answers are usually the same: criticism is expected if it is evidence-based, you should explain why your chosen paper meets the case study definition, and silence on ethics is itself something to discuss.
Another worry is the balance between two kinds of paper that are not obviously comparable. Say so in your introduction. The mismatch is the point, because it lets you show how design shapes what a paper can claim.
If your chosen topic leans on interviews or surveys, brush up on qualitative research methods and quantitative research methodology before you write Part 2. For the comparison itself, our comparative analysis support covers structure and argument, and our social policy assignment support covers the policy context. For feedback on a draft before you submit, there is also academic review help.
On referencing, a count of sources matters less than their quality, but see our guide on how many references suit a 2,500-word essay. If you analyse documents or interview transcripts in your own alternative design, our post on content analysis in qualitative research may help.
We provide model answers, guidance and academic support. You remain responsible for how you use them and for your university’s rules. Your files and identity are kept confidential, and you can read student reviews to judge the service for yourself.
cta_essay_sample_3
Public Administration Expert · Qualification: Bachelor’s degree in Public Administration, University of Manchester · Experience: 8 years as a public administrator and 6 years with Students Assignment Help UK.
Frequently asked questions
What is a critical comparative review in policy research?
It is an evaluative comparison of two research papers on the same policy problem. You assess how their questions, designs, evidence and ethics differ, judge them against fitting quality criteria, and decide what each contributes to policy. It is an argument about the research, not a summary of the findings.
Can I use the same review paper I used for my formative task?
The brief for this module says the review study can be the same as the formative one. Check your module guidance and confirm with your tutor, especially if you plan to change the case study you compare it with.
What counts as a case study for this assignment?
A case study examines a bounded programme, place or organisation in depth, using several sources of evidence (Yin, 2018). Evaluations of services in a given locality often fit. If your paper is borderline, explain in your introduction why you treat it as a case study.
Which quality criteria should I use?
Use criteria that suit each design. For systematic reviews, AMSTAR 2 is a recognised tool (Shea et al., 2017). For case studies, consider construct validity, multiple sources and chain of evidence (Yin, 2018). State why each fits and apply them to the policy problem.
How do I write about ethics if the papers say little about it?
Treat the gap as evidence. Discuss what ethical issues the design raises, whom it affects and who holds power, then judge whether the authors addressed them. Do not invent details the paper does not report. Reflexivity means questioning positions and risks, not listing approvals.
Can I use ChatGPT for this assignment?
The brief sets GenAI at level 1, which means no use of generative AI. Follow that rule and your university’s academic integrity policy. Our samples are for study and reference only and should not be submitted as your own work.
References
Pawson, R. and Tilley, N. (1997) Realistic evaluation. London: SAGE.
Shea, B.J., Reeves, B.C., Wells, G., Thuku, M., Hamel, C., Moran, J., Moher, D., Tugwell, P., Welch, V., Kristjansson, E. and Henry, D.A. (2017) ‘AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both’, BMJ, 358, j4008. doi:10.1136/bmj.j4008.
Weiss, C.H. (1979) ‘The many meanings of research utilization’, Public Administration Review, 39(5), pp. 426-431. doi:10.2307/3109916.
Yin, R.K. (2018) Case study research and applications: design and methods. 6th edn. Thousand Oaks, CA: SAGE.
