To reverse or to not reverse Likert-type items: That is the question

Area of Science:

• Psychometrics
• Survey Methodology
• Educational Measurement

Background:

• The use of reversed items in response measures is debated due to potential negative effects on psychometric properties.
• Some researchers advocate for reversed items, while others reject them, leading to ongoing controversy in test construction.

Purpose of the Study:

• To evaluate the efficacy of using reversed items in conjunction with a bias control procedure.
• To compare the psychometric properties of tests with and without reversed items, before and after bias control.

Main Methods:

• Two forms of a test were analyzed: one with direct and reversed items, and another with only direct items.
• A procedure to control response bias effects was applied to both test versions for comparison.

Main Results:

• After controlling for response biases, both test versions demonstrated nearly equivalent factorial structure and reliability.
• Without bias control, the version incorporating reversed items showed significantly poorer psychometric properties.

Conclusions:

• Reversed items should not be used without implementing a procedure to control response bias effects.
• Mitigating bias effects when using reversed items leads to psychometric outcomes comparable to using only direct items, with the added benefit of controlling acquiescence.