devami/CompositeReliability_1Factor.qmd at main · KODAQS/devami · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
---
title: "Composite Reliability of Unidimensional Model - AVE und Omega"
format: html
author:
    - Melanie Viola Partsch
    - Katharina Groskurth
    - Matthias Blümke
    - Isabelle Schmidt
keywords:
  - Survey Data
  - Measurement
  - Validity
license: "CC BY-NC 4.0"
citation:
    type: "document"
    title: |
        R Code for the Development and Validation of Measurement Instruments in the Social Sciences: Psychometric Analyses (Dimensionality, Reliability, Measurement Invariance)
    container-title: GESIS, Cologne. Data File Version 1.0.0,
    author:
      - Melanie Viola Partsch
      - Katharina Groskurth
      - Matthias Blümke
      - Isabelle Schmidt
    doi: 10.7802/1.1982
    issued: 2020
---

Load the required packages
```{r}
#| label: setup
library(rio)
library(lavaan)
library(semTools)

packageVersion("rio")
packageVersion("lavaan")
packageVersion("semTools")
```

Read the example data.

```{r}
#| label: read_data
example1 <- rio::import("example1.sav")
```

Specify and estimate tau-congeneric measurement model

```{r}
SOC_TC <- "SOC =~ NA*bfi2_1 + bfi2_16R + bfi2_31R + bfi2_46
          SOC ~~ 1*SOC"
SOC_TC.M1 <- cfa(SOC_TC, data = example1, estimator = "MLR")
```

Default for missing data handling is listwise deletion. For the use of FIML additionally specify `missing = "ML"`

Compute omega and AVE (average variance extracted) coefficient using the reliability funtion of the semTools package
```{r}
omega_1F <- semTools::reliability(SOC_TC.M1)
omega_1F
```

omega coefficient to report

```{r}
round(omega_1F[2, 1], digits = 3)
```

AVE coefficient to report

```{r}
round(omega_1F[5, 1], digits = 3)
```

Determine confidence intervals of omega and AVE coefficients

```{r}
SOC.items <- example1[, c("bfi2_1", "bfi2_16R", "bfi2_31R", "bfi2_46")]
N <- nrow(SOC.items)
R <- 1000
omega.boot <- NULL
AVE.boot <- NULL
for (i in 1:R) {
    idx <- sample.int(N, N, replace = TRUE)
    SOC_TC.M1_BS <- cfa(SOC_TC, data = SOC.items[idx, ], estimator = "MLR")
    omega.boot[i] <- round(semTools::reliability(SOC_TC.M1_BS)[2, 1], digits = 3)
    AVE.boot[i] <- round(semTools::reliability(SOC_TC.M1_BS)[5, 1], digits = 3)
}
```

90% confidence interval of omega coefficient
```{r}
round(quantile(omega.boot, c(0.05, 0.95)), digits = 3)
```
95% confidence interval of omega coefficient
```{r}
round(quantile(omega.boot, c(0.025, 0.975)), digits = 3)
```
99% confidence interval of omega coefficient
```{r}
round(quantile(omega.boot, c(0.005, 0.995)), digits = 3)
```

90% confidence interval of AVE coefficient
```{r}
round(quantile(AVE.boot, c(0.05, 0.95)), digits = 3)
```

95% confidence interval of AVE coefficient
```{r}
round(quantile(AVE.boot, c(0.025, 0.975)), digits = 3)
```

99% confidence interval of AVE coefficient
```{r}
round(quantile(AVE.boot, c(0.005, 0.995)), digits = 3)
```