reversescored

Reverse-scored items are survey questions designed so that higher agreement indicates lower levels of the measured construct. They are used to reduce response biases such as acquiescence and to improve the construct validity of scales by requiring respondents to process items with opposite wording.

In practice, many surveys use a Likert scale, for example from 1 (strongly disagree) to 5 (strongly agree). A positively worded item might be “I am satisfied with my progress,” where higher scores reflect higher satisfaction. A reverse-scored item might be “I rarely feel satisfied with my progress.” To compute a single score, responses to reverse items are transformed (e.g., for a 1–5 scale, new score = 6 − original score) before averaging or summing across items.

Designers should use reverse items judiciously. They can help detect inattentive responding but can also confuse respondents, introduce interpretive ambiguity, or create method factors that complicate factor analysis. Avoid double negations and ensure reverse items clearly contrast with forward items. Pretesting, careful translation, and cognitive interviewing can help ensure clarity across populations.

In data analysis, reverse scoring should be completed before computing scale scores and reliability estimates. Researchers should examine item-total correlations and consider potential method effects due to item wording. When reporting results, specify which items were reverse-scored and how scores were transformed.

See also: Likert scales, psychometrics, survey methodology.