In order to calculate CVI and CVR, the 4-points of Likert Scale frequently being adapted to determine the item validation. The indicators of 4-points used degree of relevance (1-not relevent, 2-highly not relevent and etc) to rate the items. My proposed actual questionnaires also will be in terms of 4-point Likert scale but using different degree/descriptions such as 1-disagree, 2-strongle disagree. My concern is, can i used different degree/descriptions in the likert scale to represent the actual meaning of the items? Or can CVI and CVR be calculated if the likert scale description changed from relevance to agree for instance?
Switching from a relevance-based descriptor ("not relevant" to "very relevant") to agreement ("strongly disagree" to "strongly agree") does not inherently invalidate the calculation of CVI and CVR, provided the scale remains ordinal and interpretable regarding item importance or relevance. The key is to ensure that experts understand the scale and its application to content validity.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis