I used "Teacher Sense of Efficacy Scale" designed by Tschannen-Moran and Woolfolk-Hoy (2001) to measure pre-service teachers' self-efficacy. The five-point Likert scale was used in my study with options of (1) not at all to (5) a great extent rated by the participants. I want to categorize the responses of teacher candidates regarding their confidence in teaching into three levels: low, moderate and high.
Please check it up: https://www.wikihow.com/Avoid-Bias-in-Qualitative-Research
Good Luck.
One way to deal with this is to perform a so-called binning procedure in SPSS. Go to Transform > Optimal binning
Good luck!
Do you have a continuous variable that is the sum of several Likert-scored items? If so, why do you want to "throw away information" by converting it into three categories?
You can’t use a three category ordinal variable in a correlation. That requires an interval level variable. You could use the three category variable for an ANOVA.
Hey !
You can do Mean +_ 1/2 SD to classify the groups into high, moderate and low groups. Mean+1/2 SD will be high, Mean-1/2 SD will be low and the in between values will be moderate. Ling Kee Htang
I can suggest two ways:
1. Look back at the definition of the scale. Have a group of experts decide on the cutoffs, in the form of standard setting, guided by the item scoring scheme.
2. Compare the score against external criteria, in the form of known-group validity. These can be other scale, expert assessment etc.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India