Schmid-Leiman transformation

SL() has been superseded by efa_schmid_leiman(), which is the recommended interface going forward. It remains available so existing code keeps working.

Usage

SL(
  x,
  Phi = NULL,
  type = c("EFAtools", "psych", "SPSS", "none"),
  method = c("PAF", "ML", "ULS", "MINRES"),
  g_name = "g",
  ...
)

Arguments

x

An oblique efa_fit() or psych::fa() solution, a pattern matrix (give its factor intercorrelations in Phi), or a second-order CFA solution from lavaan.

Phi

matrix. The factor intercorrelations of the pattern matrix in x. With an efa_fit() or psych::fa() solution in x, Phi replaces the fitted factor intercorrelations. Not used for a lavaan solution.

type

character. One of "EFAtools" (default), "psych", "SPSS", or "none". This is used to control the procedure of the second-order factor analysis. In efa_schmid_leiman() it is set through the type of the estimate_control() object.

method

character. The estimator for the second-order factor analysis; passed to efa_schmid_leiman() as its estimator argument. One of "PAF", "ML", "ULS", or "MINRES".

g_name

character. The name of the general factor of a lavaan second-order solution. Default is "g".

...

The estimation tuning knobs of the second-order efa_fit() (init_comm, criterion, criterion_type, max_iter, abs_eigen, start_method). They are repacked, together with type, into an estimate_control() object, so that they tune that fit exactly as they always did. Other efa_fit() arguments have no effect, and se, b_boot, ci, and seed give an error. The estimator is selected with method.

Value

A list of class c("efa_schmid_leiman", "SL"), identical to the value of efa_schmid_leiman(); see there for the components.

See also

efa_schmid_leiman()