Medical educators are increasingly relying on data derived from Likert-type scales to assess trainee performance, evaluate faculty and measure the impact of educational interventions, according to a recent analysis published in the Journal of Graduate Medical Education.
Developed in 1932 by American social psychologist Rensis Likert, the Likert scale is a psychometric tool used to represent people’s attitudes toward a particular question or statement. The typical scale presents respondents with a series of statements and asks them to indicate their level of agreement or disagreement, usually on a 5- or 7-point scale. The scale is ordinal, meaning responses can be ranked, but the intervals between responses are not necessarily equal.
The growing use of Likert scales in medical education research has prompted a call for greater understanding of how to properly analyze and interpret the data they generate. The Journal of Graduate Medical Education article, authored by Gail M. Sullivan and Anthony R. Artino Jr., aims to provide medical educators without extensive statistical backgrounds with the foundational knowledge needed to assess outcomes measured using these scales.
The authors note that while Likert scales are frequently employed, a common misunderstanding exists regarding the nature of the data they produce. Because the scale is ordinal, standard statistical methods designed for interval or ratio data may not be appropriate. The difference between “always,” “often,” and “sometimes” on a frequency response Likert scale, for example, is not necessarily equivalent.
The analysis emphasizes the importance of selecting appropriate statistical methods based on the characteristics of the data. Researchers must consider the limitations of ordinal scales when drawing conclusions from their findings. The article does not detail specific statistical methods, but rather focuses on establishing a conceptual understanding of the scale’s properties.
The Accreditation Council for Graduate Medical Education (ACGME) has not issued specific guidance on the use of Likert scales, but the increasing prevalence of their use in submissions to the Journal of Graduate Medical Education suggests a growing reliance on this method of data collection within the field. The ACGME did not respond to requests for comment regarding the use of Likert scales in program evaluation.
