## 5.13 Reliability Analysis

We need to also assess the reliability of these factors. We will use the Cronbach’s alpha and values greater than 0.70 indicate good reliability estimate.

```
# subset data for only ICT Interest
INTICT.only <- subset(data.us.only, select = INTICT1:INTICT6)
INTICT.alpha <- alpha(INTICT.only)
INTICT.alpha$total[1]
```

```
## raw_alpha
## 0.7848381
```

```
# subset data for only Perceived ICT Competence (COMPICT)
COMPICT.only <- subset(data.us.only, select = COMPICT1:COMPICT5)
# find Cronbach alpha
COMPICT.alpha <- alpha(COMPICT.only)
COMPICT.alpha$total[1]
```

```
## raw_alpha
## 0.8335219
```

```
# subset data for only Perceived Autonomy related to ICT Use (AUTICT)
AUICT.only <- subset(data.us.only, select = AUICT1:AUICT5)
# find Cronbach alpha
AUICT.alpha <- alpha(AUICT.only)
AUICT.alpha$total[1]
```

```
## raw_alpha
## 0.8461312
```

```
# subset data for only ICT as a topic in Social Interaction (SOIAICT)
SOIAICT.only <- subset(data.us.only, select = SOIAICT1:SOIAICT5)
# find Cronbach alpha
SOIAICT.alpha <- alpha(SOIAICT.only)
SOIAICT.alpha$total[1]
```

```
## raw_alpha
## 0.8553583
```

From the results, all the factors have reliability estimate greater than 0.70, which implies that the factors have good internal consistency.

We can use the code below to export the reliability results for all the factors in a `.csv`

file.

```
# Combine result in a row
reliability.factors <- rbind(INTICT = INTICT.alpha$total[1],
COMPICT = COMPICT.alpha$total[1],
AUICT = AUICT.alpha$total[1],
SOIAICT = SOIAICT.alpha$total[1]
)
# Rename column
colnames(reliability.factors) <- "Cronbach's Alpha"
# Round result to three decimal places
reliability.factors <- round(reliability.factors, digits = 3)
reliability.factors
```

```
## Cronbach's Alpha
## INTICT 0.785
## COMPICT 0.834
## AUICT 0.846
## SOIAICT 0.855
```