Turn the prompt into a writing decision
A proposal essay needs a recommendation and reasons; a comparison needs a relationship between alternatives; an evaluation needs criteria for judgment. Before drafting, state the question the essay must answer. “Should the library open later one evening?” is narrower than “Write about the library.”
List the evidence available and the format required. In the prepared example, a small survey can support a limited preference claim. It cannot independently establish better grades, universal demand or the cost of extended staffing. Deciding the scope first prevents a polished paragraph from promising more than the evidence supplies.
Build an outline around claims
For the library proposal, an outline might have three jobs: explain the access problem, justify a limited trial using the supplied survey, and address staffing constraints. Attach the evidence and the inference to each job before writing full sentences. A heading such as “Survey results” describes a source; “A trial would test expressed demand” states a paragraph purpose.
If two headings make the same point, combine them or give each a different role. If a claim has no evidence, mark the gap before expanding the outline. Writing Lab can help inspect the structure, while the final priorities and wording remain decisions you make from the assignment.
Keep evidence attached to the source
Record which source supplies each fact, what it actually measured and the context needed to interpret it. Our 12-of-20 survey result is invented practice data, not a published study. In your essay, use the real source and the citation style required by your course.
A number needs a denominator and a population. “60% preferred later hours” is incomplete without saying 60% of which respondents. A preference survey is also different from an attendance count or a grade comparison. Do not let a source about one quantity become evidence for an unrelated outcome.
Diagnose a vague sentence with actual evidence
The original draft says “Everyone needs the library open late.” The available practice data support “12 of the 20 respondents preferred later hours.” This revision narrows the population and replaces an absolute claim with a measured result. The next sentence should explain why that result supports a limited trial, rather than simply repeating the percentage.
The claim that later hours “will improve every student’s grades” needs a separate repair: remove it or obtain appropriate evidence. Changing “will” to “might” does not create evidence. A supported paragraph can instead propose collecting attendance and feedback during a trial to assess whether the access change is useful.
Address a counterargument with a relevant response
A reasonable objection is that later opening needs staff time and resources. A response should acknowledge that constraint and explain why the proposal still merits consideration. For example, a one-evening trial limits the initial commitment and produces information about demand.
The survey alone does not establish that staffing is available or affordable. Keep those questions open in the proposal. A counterargument becomes useful when it changes the recommendation or the conditions under which it works; it is weaker when added as a decorative “some people disagree” sentence.
Revise the reasoning before the polish
Start with the largest problem: a missing argument, an unsupported claim, irrelevant evidence or an absent connection. Then inspect paragraph order and transitions. Sentence-level edits come after the paragraph has a clear job. A shorter, accurate claim can strengthen the draft more than a more impressive adjective.
Request feedback on one layer at a time. “Which claim exceeds my evidence?” gives different work from “Which transition is hard to follow?” Choose the suggestion that addresses the identified problem and record a reason for accepting it. Re-read the revised argument before making further edits.
Read the paragraph without the feedback
A possible prepared revision is: “Twelve of the 20 survey respondents preferred later library hours. This limited result supports testing one later opening evening rather than assuming universal demand. The trial should record attendance and staffing requirements before a longer-term change is considered.”
This version has a fact, an inference and a boundary. It still needs the real survey citation in an actual assignment. Read your own paragraph without the comments visible: can a reader identify the claim, see its supporting evidence and understand what remains uncertain? That is a more useful finish than accepting every suggested edit.
