COMBINING LIKERT ITEMS WITH DIFFERENT NUMBER OF RESPONSE CATEGORIES

Three assumption-free approaches are described to convert and combine scores of Likert items with different response categories to obtain comparable test scores for a single sample. Generated data were continuous, equidistant, normally distributed, avoiding tied scores and thus facilitate ranking the respondents and undertaking analysis under parametric set up. Empirical verification was undertaken involving 5 items in each of 3, 4, 5 and 7-point scale to 100 subjects. Correlations at the level of 0.99 between a pair of approaches were observed with marginal difference in coefficient of variation (CV). Minimum CV and maximum alpha were found when test scores were obtained as sum of standardized equidistant item scores converted to [1, 5]. Thus, the said Approach appears to have advantages. Use of such methods of combining scores of Likert items is recommended for clear theoretical advantages and easiness in calculations. Future studies are suggested.