arnold on Sep 16, 2008
Bartlett et al (1960) asked this same question.
This brief article briefly compared the Likert-Type measurement against the forced choice option. The authors argue that the traditional type (Likert-Type) of attitude measurement may be subject to many types of bias, for instance the desire to raise expectation on a post-test to provide positive reinforcement that a change was observed.
The authors hypothesized that the traditional scale (Likert) will yield a significant change in contrast to the force-choice scale which will yield no change.
Results: The two scales had a low level of correlation .22 and did not approach unity. Test-retest reliability was modest for both scales (.71 Likert .61 Forced Choice)
Discussion: The correlation indicates that while the two scales have some common variance they are measuring different things. The present study found a significant change in attitude score for the Likert scale where bias was not adequately controlled…forced choice showed no significant change in attitude.
Takeaway: Traditional methods of attitude measurement are subject to bias. Attitude change may reflect bias and not actual changes in attitude.
Bartlett, Quay, Wrightsman (1960), “A Comparison of Two Methods of Attitude Measurement: Likert-Type and Forced Choice.” Educational and Psychological Measurement, 20, 699 – 704.