Log-in | Contact Jeff | Email Updates

Question 119:



No answer provided yet.Yes, it is a common practice to convert the semantic labels into numeric equivalents. The data should be treated as ordinal unless you have tested the scale and established intervals (See Nunnallyís Psychometric Theory, 2nd Ed page 61 for an example). You can still run parametric tests, such as t-tests and ANOVA on the data, however, when you interpret the results donít make statements like, group A was twice as satisfied as group B (thatís an interval statement). You can say that Group A is more satisfied (ordinal statement). Know that some in the field of psychometrics will even take exception with what Iíve recommended. In practice, it is usually fine: see Hyperstat for a discussion.

Not what you were looking for or need help?

Ask a new Question

Browse All 869 Questions

Search All Questions: