There is an evidence that the reliability of the entire scale can be increased when each respondent could express his or her agreement or disagreement with every item using a five-point scale (Jenkins & Taber, 1977; Remmers & Ewart, 1941).

Revilla, Saris & Krosnick (2014) recommended that if researchers want to use agree-disagree, they should offer 5 answer categories rather than 7 or 11 because the latter results in data of lower quality.


References

Jenkins, G.D. & Taber, T. D. (1977). A Monte Carlo Study of Factors Affecting Three Indices of Composite Scale Reliability. Journal of Applied Psychology, 62(4), 392.

Remmers, H. H. & Ewart, E. (1941). Reliability of Multiple-Choice Measuring Instruments as a Function of the Spearman-Brown Prophecy Formula. Journal of Educational Psychology, 32(1), 61.

Revilla, M.A., Saris, W. E. & Krosnick, J.A. (2014). Choosing the Number of Categories in Agree-Disagree Scales. Sociological Methods and Research, 43(1), 73-97.