diff --git a/08 EFA/EFA.R b/08 EFA/EFA.R index e27f32d..8426918 100644 --- a/08 EFA/EFA.R +++ b/08 EFA/EFA.R @@ -17,16 +17,11 @@ round(cor(Harman74), 2) # hard to visually "see" structure in the data... check_factorstructure(Harman74) -## Raw or scaled? -------- -# Now we first need to decide: -# Are we conducting a FA on the raw scales? Or the standardized scales? - - ## Scree plot -------- screeplot( - prcomp(Harman74, scale. = FALSE), # set scale. = TRUE for standardized scales + prcomp(Harman74, scale. = TRUE), npcs = 10, type = "lines" ) # where is the elbow? @@ -35,7 +30,8 @@ screeplot( ## Other methods -------- ns <- n_factors( - Harman74, # for standardized scales, standardized the data first! + Harman74, + type = "FA", algorithm = "pa", rotation = "oblimin" ) # This function calls many methods, e.g., nFactors::nScree... Read the doc!