Q.Point out what is wrong with the following questionnaire item, and rewrite it as a better question: "Don't you agree that spending too much time on mobile phones is harmful for students?"

The question is a leading question — its wording already suggests the "correct" answer, which biases responses rather than measuring genuine opinion. …

Step 1 — Identify the flaw. The phrase "Don't you agree that...harmful" frames the issue with a built-in value judgement ("harmful") and phrases the question so that agreeing feels like the natural, expected response. This is a classic leading question — one of the qualities a good questionnaire must specifically avoid (per this chapter's list of qualities of a good questionnaire), because it distorts genuine opinion rather than measuring it.

Step 2 — Explain why this matters for data quality. If enough respondents answer leading questions the way they are nudged to, the collected data will show artificially inflated agreement — a form of non-sampling error introduced at the design stage, before any respondent is even contacted, and one that even a perfect sample and a perfect enumerator cannot fix afterward.

Step 3 — Rewrite the question neutrally. A neutral version should present the issue without suggesting an answer, and ideally offer clear, simple response options for easy tabulation: "In your opinion, does spending a large amount of time on mobile phones affect students' studies? (Yes / No / No effect)." …

Students sometimes only say "the question is too personal" or "too long," missing the specific defect being tested here — leading/suggestive wording. Always check first whether a flawed question is leading, am …