I have 8 items of a scale measuring one variable (Social Media Use):
The first item has options 1 to 9, the second item has options 1 to 6, and the remaining six items have options 1 to 5 (likert scale).
How do I standardise the first two items on SPSS (due to their differing ranges)? How do I change the first two items so that all items on the scale have the same range (eg: 1 to 5)?
Agree with Jimmy. Here's the HowTo: When you choose "Analysis" > "Descriptives," there's a checkbox "Standardized values" (don't remember the exact wording, no SPSS on my computer). This will give you the Z scores in a new variable.
Jimmy Y. Zhong Tanja Gabriele Baudson Thank you! So then I can calculate the sum of these Z scores to get the total scale score to then test for regression with another variable?
The answer of Mr. Zhong ("For Z-scores, you need to average them, not sum them") is not correct. For all analyses based on correlations it makes no difference at all - you may sum the z-scores or average them - the results of your analysis (regressions etc.) will be identical.
By the way: the mean of z-scores is M = 0, the standard deviation is SD = 1.
The Z-scores (Z = capital letter) are scores with M = 100 and SD = 10. You can tranform z-scores into Z-Scores or into T-Scores (M = 50, SD = 10) or into any other common score - all these transformations are based on z-scores.
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Social Media