Colleagues,
The literature is not definitively clear about the use of mid-point categories on self-report Likert scales. For example, some say that mid-point choices increase scale reliability (Courtenay & Weidemann, 1985; Madden & Klopfer, 1978), while detractors claim that mid-pointing could result in satisficing, (Krosnick, Narayan, & Smith. 1996), which could cluster responses at the mid-point. I am in the process of designing a scale and would appreciate your thoughts on the matter.
Best,
Ruth
References
Courtenay, B. C., & Weidemann, C. ( 1985). The effects of a don't know response on Palmore's facts on aging quizzes. The Gerontologist, 25, 177-181.
Krosnick, J. A., Narayan, S., & Smith, W. R. ( 1996). Sati sf icing in surveys: Initial evidence. In M. T. Braverman & J. K. Slater (Eds.), Ad1•ances in survey research (pp. 29-44). San Francisco: Sage.
Madden, T. M., & Klopfer. F. J. ( 1978). The "cannot decide" option in Thurstone-typc attitude scales. Educational and Psychological Measurement, 38, 259-264.