My participants have rated themselves 1-10 on a series of 19 statements. Those statements are supposed to represent 5 different constructs. So for example, questions 2, 6 and 11 are all about relationships. How can I combine one participant's three responses into a number that represents their score for relationships? I have a feeling that it's not as simple as just adding their scores up and dividing by 3. Any advice much appreciated.
Right. Without having standardized the test on a large population, you'd be flying blind, not to mention that the use of only three numeric responses greatly oversimplifies a participant's personality. You're trying to make tea from blackberry leaves. The only thing I can recommend as an alternative to the mean is a sort of Occam's Bludgeon approach: the most interesting cases often lie at one end of the spectrum or the other; consider using the extreme value of the three responses, highest or lowest, instead.
If you want the mean of their score for the 3 questions than yes add and then divide. However, the first question might be what are you trying to achieve by finding the mean? The second question is; Are you going to subject the mean of the three questions for all of your participants to a more robust statistical test? and then finally; what are you trying to discover by calculating the mean?
- Badges
- Science topic
- Similar topics
- Psychology
- Emotion