Skip to contents

Test equality of coefficient alpha across fits

Source: R/agreement_test.R
alpha_test.Rd

Wald test that coefficient alpha is equal across two or more [alpha()] fits (or, with a single fit, equal to `value`). See [kappa_test()] for the `paired` semantics; `alpha` is the only coefficient, so there is no `coef` argument.

Usage

alpha_test(..., paired = TRUE, value = 0)

Arguments

...

Two or more `misskappa_estimate` objects from [alpha()]. Names become labels in the output.

paired

`TRUE` (default) for same-subject fits, `FALSE` for independent samples.

value

Null value for the difference (or, with one fit, for alpha). Default `0`.

Value

An `htest` object.

Examples

# Do the three Holzinger-Swineford (1939) subscales have equal reliability?
# The same students take all three, so the fits are paired (G-way, df = 2).
subs <- list(visual  = c("x1", "x2", "x3"),
             textual = c("x4", "x5", "x6"),
             speed   = c("x7", "x8", "x9"))
fits <- lapply(subs, function(v)
  alpha(dat.holzinger1939[, v], estimator = "nt_fiml"))
do.call(alpha_test, c(fits, list(paired = TRUE)))
#> 
#> 	Paired (dependent) test of equal alpha across 3 fits
#> 
#> data:  visual, textual, speed
#> X-squared = 75.994, df = 2, p-value < 2.2e-16
#> sample estimates:
#>    visual   textual     speed 
#> 0.6261171 0.8827069 0.6884550 
#> 

# Real missing data with groups: psych::bfi. The same 2800 respondents take
# all five Big Five scales; reverse-key the negatively-worded items first.
data(bfi, package = "psych")
neg <- c("A1", "C4", "C5", "E1", "E2", "O2", "O5")
bfi[neg] <- 7 - bfi[neg]

# Dependent: are Neuroticism and Extraversion equally reliable? The same
# respondents answer both, so the estimates are paired.
N <- paste0("N", 1:5)
E <- paste0("E", 1:5)
alpha_test(N = alpha(bfi[, N], estimator = "nt_fiml"),
           E = alpha(bfi[, E], estimator = "nt_fiml"),
           paired = TRUE)
#> 
#> 	Paired (dependent) test of equal alpha across 2 fits
#> 
#> data:  N, E
#> X-squared = 27.493, df = 1, p-value = 1.576e-07
#> sample estimates:
#>         N         E 
#> 0.8138131 0.7615000 
#> 

# Independent: is Conscientiousness equally reliable across the two genders?
g <- split(seq_len(nrow(bfi)), bfi$gender)
C <- paste0("C", 1:5)
alpha_test(
  men   = alpha(bfi[g[["1"]], C], estimator = "nt_fiml"),
  women = alpha(bfi[g[["2"]], C], estimator = "nt_fiml"),
  paired = FALSE)
#> 
#> 	Independent-sample test of equal alpha across 2 fits
#> 
#> data:  men, women
#> X-squared = 0.0082484, df = 1, p-value = 0.9276
#> sample estimates:
#>       men     women 
#> 0.7249358 0.7231688 
#>