rm(list=ls())
library(tidyverse)
library(psych)
# load in the dataset
<- read_csv('data/responses.csv')
responses #| label: sumscores
# take all the item scores
<- responses %>%
item_responses select(ends_with("_score"))
12 Exploring alpha
Internal consistency is a measure of the strength of the relationship between questions in the same scale. There are many different measures of internal consistency, with the most common cited being Cronbach’s alpha. However, there are many other measures of internal consistency, including omega, glb, and others.
12.1 From alpha to omega
Read the following article, and note the three main assumptions of alpha. https://tidsskriftet.no/en/2022/08/medicine-and-numbers/internal-consistency-alpha-omega
12.2 Computing variations on alpha
<- psych::alpha(item_responses)
mdl_alpha print(summary(mdl_alpha))
Reliability analysis
raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
0.81 0.81 0.83 0.18 4.3 0.0048 0.6 0.2 0.16
NULL
<- psych::omega(item_responses) mdl_omega
print(summary(mdl_omega))
Omega
psych::omega(m = item_responses)
Alpha: 0.81
G.6: 0.83
Omega Hierarchical: 0.59
Omega H asymptotic: 0.7
Omega Total 0.84
With eigenvalues of:
g F1* F2* F3*
3.0 1.6 1.3 0.0
The degrees of freedom for the model is 133 and the fit was 0.24
The number of observations was 3061 with Chi Square = 726.19 with prob < 0
The root mean square of the residuals is 0.03
The df corrected root mean square of the residuals is 0.04
RMSEA and the 0.9 confidence intervals are 0.038 0.035 0.041
BIC = -341.34Explained Common Variance of the general factor = 0.5
Total, General and Subset omega for each subset
g F1* F2* F3*
Omega total for total scores and subscales 0.84 0.81 0.70 NA
Omega general for total scores and subscales 0.59 0.47 0.36 NA
Omega group for total scores and subscales 0.24 0.33 0.34 NA
NULL
12.3 Extension exercise
The value of alpha is closely related to the length of the test and the number of items. In general, the longer the test, the higher the alpha. Using the test dataset explore the relationship between alpha and the length of the test.