Decode the command before you write
Compare asks for similarities and differences; evaluate also asks you to judge the strength of evidence. Underline the object of comparison and the criterion for judgment. A long definition of each term does not answer an evaluative question.
Use the first three minutes of a 25-minute practice block to list your claim, two contrasts and one limitation. Reserve the final three minutes to check whether each paragraph responds to the question.
Build a claim–evidence–limit outline
Start with a one-sentence thesis. In the body, pair each distinction with a concrete implication, then qualify what you can infer. A useful paragraph sequence is claim, evidence or example, interpretation, limitation.
For a methods exam, distinguish association from causal identification. Random assignment can reduce systematic baseline confounding, but measurement error, attrition and noncompliance still matter.
Worked essay: sleep and exam performance
Prompt: Compare an observational study relating sleep duration to exam scores with a randomized intervention assigning students sleep-hygiene guidance. Which design better supports a causal claim and why?
Model response: An observational study can show whether students who report longer sleep also tend to earn higher scores, but study habits, workload or prior attainment may influence both variables. The association alone does not establish that additional sleep caused the score difference. In a randomized trial, assignment to guidance is less likely to be systematically related to those baseline characteristics. If groups differ in scores, the design more strongly supports a causal effect of being offered guidance, subject to implementation and follow-up. It does not automatically identify the causal effect of each extra hour of sleep: students may ignore the advice, sleep may be measured imperfectly and the guidance may change other habits. Thus random assignment strengthens causal inference about the intervention, not every proposed mechanism.
This answer directly compares designs, makes a judgment and states a specific limitation. It avoids treating randomization as proof of perfect compliance.
Mark your own response with an 8-point rubric
Award 0–2 points in each category: thesis (clear and responsive), comparison (accurate distinctions), evidence/reasoning (specific causal logic), and limits (credible qualification). A 2 is explicit and correct, 1 is partial, 0 is missing or misleading. Maximum 8 points; this is a practice rubric, not an official grade.
Example scoring: a response with a clear thesis (2), accurate design comparison (2), one relevant confounder but no explanation (1), and a meaningful noncompliance caveat (2) scores 7/8. Rewrite the weakest category before attempting a fresh prompt.
Turn course material into active practice
Upload your own PDF or DOCX to Evaluo to review concepts and practise by topic. Then write your essay on paper or in your editor and apply the rubric manually: Evaluo is not represented here as an automatic essay grader.
Pair this guide with the short-answer exam guide when your course mixes question formats. Return to the original lecture notes to verify every substantive claim, and practise another question without looking.
Try it yourself
25-minute essay practice
Compare observational and randomized approaches to studying revision habits and midterm outcomes.
- 1.3 min: underline compare and evaluate; write one claim and two contrasts.
- 2.17 min: draft a thesis, two evidence-based paragraphs and a limitation.
- 3.3 min: assess thesis, comparison, reasoning and limits (0–2 each).
- 4.2 min: rewrite one weak sentence and verify your causal claim against course notes.
How to check your work
Do not conclude that correlation proves causation; assignment to advice identifies an intervention effect more directly than the effect of a particular behavior.
Have your notes at hand?
Evaluo builds topic-by-topic questions from your PDF or Word file and shows what to review. Free plan: 1 study material per month.
Common mistakes
Defining instead of evaluating
Two accurate definitions still miss the requested judgment.
Better: State which design supports the claim more strongly and why.
Treating association as causation
Confounders can explain part or all of an observed relationship.
Better: Name a plausible confounder and distinguish it from random assignment.
Assuming randomized means flawless
Attrition, adherence and measurement can weaken conclusions.
Better: Specify which causal effect the experiment can estimate.
Frequently asked questions
How long should I plan an essay before writing?
In a 25-minute practice, spend about three minutes outlining and three minutes checking. Adjust to your exam.
Can Evaluo grade my written essay automatically?
This guide uses a manual eight-point rubric. Evaluo helps you study uploaded material and practise topics; do not assume automatic essay scoring.
About this guide
Written by Gonzalo Olmos, founder of Evaluo. Published November 17, 2026.
Sources
References on learning strategies. The original activities in this guide are educational examples and should be checked against your course materials.