Cronbach's alpha for one or more questionnaires

Computes Cronbach's alpha (a measure of internal consistency) for each questionnaire present in a tallier_export or tallier_study object. Item-level responses are extracted via items_long(), coerced to numeric, and a participant × item matrix is constructed per questionnaire.

Usage

cronbach_alpha(obj, questionnaires = NULL, conf_level = 0.95, min_items = 2L)

Arguments

obj: A tallier_export or tallier_study object, or a data frame as returned by items_long() (must contain columns participant_id, questionnaire_id, item_id, completed_at, response).
questionnaires: Character vector of questionnaire IDs to include. Defaults to all questionnaires present in obj.
conf_level: Numeric. Confidence level for the CI (default 0.95). Uses the exact F-distribution method of Feldt et al. (1987).
min_items: Integer. Minimum number of numeric items required to attempt estimation (default 2).

Value

A data.frame with one row per questionnaire and columns:

questionnaire_id: Questionnaire identifier.
alpha: Cronbach's alpha (standardised scale: −∞ to 1).
ci_lower: Lower bound of the conf_level CI.
ci_upper: Upper bound of the conf_level CI.
n_items: Number of numeric items used.
n_obs: Number of complete observations (participants).
note: NA on success, or a short message explaining why estimation failed.

Details

Non-numeric items (e.g. MCTQ clock times, STOP-BANG yes/no) are silently dropped before estimation. Questionnaires with fewer than 2 numeric items or fewer than 2 complete observations return NA with an explanatory note.

References

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297–334. doi:10.1007/BF02310555

Feldt, L. S., Woodruff, D. J., & Salih, F. A. (1987). Statistical inference for coefficient alpha. Applied Psychological Measurement, 11(1), 93–103. doi:10.1177/014662168701100107

Examples

if (FALSE) { # \dontrun{
study <- read_scoreme_dir("exports/")

# All questionnaires
cronbach_alpha(study)

# Specific subset
cronbach_alpha(study, questionnaires = c("ess", "isi", "phq9"))

# From an items_long() data frame (e.g. already filtered)
items <- items_long(study)
cronbach_alpha(items)
} # }