Tag | Value |
---|---|
file | Reliability_uu-Cronbachs-alpha-003-en_uu-Cronbachs-alpha-003-en |
name | uu-Cronbachs-alpha-003-en |
section | Reliability/Analysis/Cronbach's alpha |
type | schoice |
solution | FALSE, FALSE, TRUE, FALSE |
Type | Interpreting output |
Program | SPSS |
Language | English |
Level | Statistical Literacy |
A group of researchers want to develop an instrument that measures how traumatized children are after experiencing a car accident. Based on existing trauma literature, they created items belonging to one of the following scales: re-experiencing, avoidance, arousal, and sleep problems. Because the children themselves are too traumatized, they are taking their 36 items from 314 parents. Because the researchers expect four subscales, they conduct a factor analysis with 4 factors. Part of the SPSS output is below.
One of the researchers, Rolf, is specifically interested in the sleep problems parents observe in these children. For this reason, he creates a scale for sleep problems consisting of 6 items. When Rolf performs a reliability analysis, he finds a Cronbach’s alpha of 0.725. Below is part of the SPSS output.
Based on this data, Rolf decides not to remove items from this scale. He gives two arguments for this:
I. Removing items creates the danger of affecting content validity. II. Because the scale consists of only 6 items, the scale with all 6 items will be more reliable than a scale with only 5 or even only 4 items.
Both statements are correct.