Table with alpha values and item statistics for multiple scales

Returns a data.frame with item analyses for the provided scales. Useful for reporting scale reliabilities and item statistics in manuscripts.

Usage

alpha_table(
  data,
  scales,
  labels = NULL,
  round = 2,
  CI = TRUE,
  conf_level = 0.95,
  check_key = TRUE,
  keys = NULL,
  keys_from_weights = TRUE,
  RMSEA = FALSE,
  difficulty = FALSE,
  values = NULL,
  fa = TRUE
)

Arguments

data: A data Frame or tibble containing the item responses.
scales: A list containing vectors with variable names. Each list element defines one scale. Named list elements are used as labels.
labels: Label names for scales (defaults to named list elements in 'scales').
round: Rounds values to given decimal position.
CI: If TRUE confidence intervals are calculated.
conf_level: Confidence level (e.g. 0.95 for 95 percent).
check_key: Check_key for the psych::alpha function.
keys: Optional key argument for the psych::alpha function.
keys_from_weights: If TRUE, tries to define keys from scaledics "weights" parameter.
RMSEA: If TRUE RMSEA is calculated.
difficulty: If TRUE, the difficulty of the item is calculated.
values: Sets maximum and minimum valid values necessary to calculate item difficulty
fa: If TRUE, a one factor exploratory factor analyses is calculated and loadings are reported.

Value

A data frame with concise scale indices.

Details

The function uses the psych::alpha function to calculate Cronbach's alpha and standardized alpha. Additionally, it calculates item means, standard deviations, item discriminations (item-total correlations), and item difficulties (if requested). If the 'fa' argument is set to TRUE, a one-factor exploratory factor analysis is performed, and the minimum and maximum absolute factor loadings are reported. The function can also compute confidence intervals for alpha and standardized alpha. It handles missing data by removing rows with all items missing for each scale. Variables with zero variance are automatically excluded from the analysis, and a message is displayed indicating which variables were dropped. The resulting data frame contains concise scale indices, making it easy to report scale reliabilities and item statistics in manuscripts.

Author

Jürgen Wilbert

Examples

scales <- get_scales(ex_itrf,
  Int = scale == "ITRF" & subscale == "Int",
  Ext = scale == "ITRF" & subscale == "Ext"
)
alpha_table(ex_itrf, scales = scales, difficulty = TRUE, values = list(c(0, 3)), RMSEA = TRUE)
#> ! This function is deprecated. Please use the nice_alpha_table() function from the wmisc package.
#> ! Removed 4 rows because all items were missing.
#> Note. values in brackets depict upper and lower bound of confidence intervals or [min,max] intervals.
#>   Scale    n n items    Alpha CI95% Std.Alph CI95% Homogeneity Discriminations
#> 1   Int 4247      19 .91 [.91, .92] .91 [.91, .92]         .36      [.49, .68]
#> 2   Ext 4616      17 .93 [.93, .93] .93 [.93, .93]         .44      [.39, .74]
#>   Difficulties        Means          SDs |Loadings| RMSEA
#> 1   [.07, .16] [0.21, 0.48] [0.51, 0.77] [.52, .70]  .154
#> 2   [.12, .32] [0.35, 0.95] [0.72, 1.04] [.38, .78]  .196