Response Interpolation and Scale Sensitivity: Evidence Against 5-Point Scales
Journal of Usability Studies, Volume 5, Issue 3, May 2010, pp. 104 - 110
Abstract
A series of usability tests was run on two enterprise software applications, followed by verbal administration of the System Usability Scale. The original instrument with its 5-point Likert items was presented, as well as an alternate version modified with 7-point Likert items. Participants in the 5-point scale condition were more likely than those presented with the 7-point scale to interpolate, i.e., attempt a response between two discrete values presented to them. In an applied setting, this implied that electronic radio-button style survey tools using 5-point items might not be accurately measuring participant responses. This finding supported the conclusion that 7-point Likert items provide a more accurate measure of a participant’s true evaluation and are more appropriate for electronically-distributed and otherwise unsupervised usability questionnaires.
Practitioner’s Take Away
The following are the main findings of this study:
- Five-point Likert scales are more likely than 7-point scales to elicit interpolations in usability inventories.
- Interpolations are problematic because they cannot be mitigated within an electronic survey medium and require interpretation with facilitated surveys.
- Interpolations provide evidence that 5-point Likert scales may not be sensitive enough to record a usability test participant’s true evaluation of a system.
- Seven-point Likert scales appear to be sensitive enough to record a more accurate evaluation of an interface while remaining relatively compact.
- Seven-point Likert scales appear to be more suited to electronic distribution of usability inventories.
- Practitioners can quickly test Likert items through verbal protocols by using interpolations as a metric.
Article Contents
Download Article
Response Interpolation and Scale Sensitivity: Evidence Against 5-Point Scales