reliabilityAll: Compute Cronbach's Alpha and McDonald's Omega for a set of...
In GiulioCostantini/markerIndex: markerIndex: computes the marker index proposed by Perugini and Gallucci (2007); includes also a collection of utility functions

Description Usage Arguments Details Value Author(s) References See Also Examples

View source: R/reliabilityAll.R

One often needs to estimate reliability for a large set of scales. This function does that in a convenient way, exploiting regularities in the naming of the items belonging to each scale.

1 2	reliabilityAll(data, pattern, simplify = TRUE, recode = FALSE, recString = "_r", omega = TRUE)

`data`	A dataframe.
`pattern`	A list of named items. Each element of the list identifies a set of items composing a scale, through a vector of strings. Each scale for which reliability is computed is identified by a vector of strings. All items whose names include all the strings specified in the vector are included in the scale.
`simplify`	Logical. If true, format the results in a table, otherwise return the raw results from functions `alpha` and `ci.reliability`.
`recode`	Logical. Whether reverse-coded items should be recoded (i.e., multiplied by -1) before estimating reliability. See argument `rec`.
`recString`	A string that identifies reverse-coded items, by default "_r". Be careful: if `recode = TRUE`, all items including this string in their names will be recoded before estimatting reliability.
`omega`	Logical. Whether coefficient omega should be estimated.

Cronbach's alpha is estimated through alpha using default settings. McDonald's Omega is estimated through ci.reliability using argument type == NULL, which results in estimating hierarchical omega for continuous items and categorical omega for categorical items.

if simplify = TRUE, a data.frame including the following items for each scale

`scale`	the name of the scale
`nitem`	the number of items in the scale
`raw_alpha`	the raw alpha value from `alpha`
`std_alpha`	the std alpha value from `alpha`
`omega`	the omega estimate from `ci.reliability`
`errorsAlpha`	error messages returned by function `alpha`
`warningsAlpha`	warnings returned by function `alpha`
`errorsOmega`	error messages returned by function `ci.reliability`
`warningsOmega`	warnings returned by function `ci.reliability`

Otherwise, a list including the output of alpha and ci.reliability for each scale.

Giulio Costantini

Cronbach, L.J. (1951) Coefficient alpha and the internal strucuture of tests. Psychometrika, 16, 297-334.

Kelley, K. & Pornprasertmanit, P. (2016). Confidence intervals for population reliability coefficients: Evaluation of methods, recommendations, and software for homogeneous composite measures. Psychological Methods 21(1), 69-92. https://doi.org/10.1037/a0040086

McDonald, R. P. (1999). Test theory: A unified approach. Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers.

Revelle, W. (in preparation) An introduction to psychometric theory with applications in R. Springer. (Available online at https://personality-project.org/r/book).

alpha, ci.reliability

library(psychTools)
library(lavaan)
data(bfi)

# compute alpha and omega for dataset bfi in psychTools. Warnings and negative values
# of alpha reflect the fact that items were not reverse-scored
reliabilityAll(bfi,
               pattern = list(
                 "Openness" = c("O"),
                 "Conscientiousness" = c("C"),
                 "Extraversion" = c("E"),
                 "Agreeableness" = c("A"),
                 "Neuroticism" = c("N")))



# bfi.keys include information on reverse-scoring. In particular, items with the string
# "-" are reverse-scored. We can exploit this information to reverse-score all items
bfi2 <- bfi
names(bfi2)[1:25] <- c("-A1", "A2",  "A3",  "A4",  "A5",
                       "C1",  "C2",  "C3",  "-C4", "-C5",
                       "-E1", "-E2", "E3",  "E4",  "E5" ,
                       "N1", "N2", "N3", "N4", "N5",
                       "O1",  "-O2", "O3",  "O4", "-O5")

reliabilityAll(bfi2, pattern = list(
  "Openness" = c("O"),
  "Conscientiousness" = c("C"),
  "Extraversion" = c("E"),
  "Agreeableness" = c("A"),
  "Neuroticism" = c("N")),
  recode = TRUE,
  recString = "-")