Number of factors to retain

Parallel analysis (Philox normals), Velicer's MAP (squared or fourth powers), Kaiser, scree optimal coordinates and acceleration factor (nScree rules), variance explained, and AIC/BIC of maximum likelihood factor models.

Usage

EfaNfactors(
  data,
  method = c("parallel", "map", "map4", "kaiser", "scree", "af", "variance", "aic",
    "bic"),
  nsim = 100,
  seed = 42,
  quantile = 0.95,
  threshold = 0.7,
  max_factors = NULL
)

Arguments

data

Numeric data (rows with missing values are dropped).

method

One of "parallel", "map", "map4", "kaiser", "scree", "af", "variance", "aic", "bic".

nsim

Parallel-analysis simulations.

seed

Philox seed.

quantile

Parallel-analysis quantile.

threshold

Proportion of variance for method "variance".

max_factors

Largest model for AIC/BIC.

Value

list(n_factors, method, eigenvalues, ...criterion values).

References

Horn, J. L. (1965). Psychometrika 30, 179-185. Velicer, W. F. (1976). Psychometrika 41, 321-327. Raiche, G. et al. (2013). Methodology 9, 23-29.

Examples

X <- cbind(1:20, 2 * (1:20) + (1:20) %% 3, 3 * (1:20) - (1:20) %% 2, (7 * (1:20)) %% 5)
EfaNfactors(X, method = "kaiser")$n_factors
#> [1] 1