What is recommended as optimal when creating pre and post survey data about self confidence in managing particular clinical scenarios after role playing of the scenarios out?
Does a likert scale anchored at 0=no confidence up to 4=completely confident work better than using a 5 point Liker scale with anchor at -2 with 0 in the center and +2 having completely confidence ?
Thanks for all input !
I would use:
0=no confidence
1=slight confidence
2= moderate confidence
3=high confidence
first one is better choice as it is easy for respondents to respond in a natural way.
Regards
Rafique
I think Joan's version (which should go from 0 to 4 to have 5 levels) is more straightforward for use with a general population. If, however, you are working with a more educated population, the minus through plus version would be just fine.
Wondering if it might be better to have an even number of categories - 4 or 6 rather than 5. With an odd number of categories, respondents may be pulled psychologically to the middle category, which may yield less useful information,
I find your statement very intuitively correct as a concern. Any published data to confirm the even # rather than odd ? Something Kahneman and Taversky were likely to do !
I haven't researched published data to confirm the even # hypothesis - it is more intuition as you say. My wife is a psychologist and experienced researcher and has been helping me launch a small research study of my own which includes a survey with a Likert scale - she recommended the even rather than odd number of categories and I am inclined to go with it.
This would requires a 7 point semantic differential scale (polar opposites) rather than 5 point likert. I've been studying for my dissertation and read a true likert is a scale from agree to disagree.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publications