Multiple scaling -- the overlapping way

overlap() returns a overlapped version (either extended, or reversed, or both) of the specified msdf.

overlap(
  msdf,
  mrit_min = NULL,
  negative_too = FALSE,
  overlap_with = "fragment",
  sclvals = NULL,
  use = "pairwise.complete.obs"
)

Arguments

msdf	a multiple scaled data frame (built with disjoint()).
mrit_min	a numeric constant of length 1 to specify the marginal corrected item-total correlation. Its value is in the range of 0-1. The default is set to .3.
negative_too	a logical constant indicating whether reversed items are included in the analysis. The default is set to FALSE.
overlap_with	a string telling overlap() the set of items for the extension. To build up on all variables of a fragment use fragment, for the core-only option type core. The default is set to "fragment".
sclvals	a numeric vector of length 2 indicating the start- and endpoint of a scale. Use something like c(min,max).
use	an optional string to specify how missing values enter the analysis. See use in cor for details. The default is set to pairwise.complete.obs.

Value

A multiple scaled data frame (msdf).

Details

use clarifies how to set up a correlation matrix in the presence of missing values. In a typical scaling process this happens at least twice. First, when determining the core items (the two items in the correlation matrix with the highest linear relationship). Second, when an item is proposed for an emerging scale.

Note that overlap() uses cor's default method pearson.

References

Müller-Schneider, T. (2001). Multiple Skalierung nach dem Kristallisationsprinzip / Multiple Scaling According to the Principle of Crystallization. Zeitschrift für Soziologie, 30(4), 305-315. https://doi.org/10.1515/zfsoz-2001-0404

Examples

# Build a msdf
msdf <- disjoint(mtcars, mrit_min = .4)

# Use positive correlations (extend on fragments)
overlap(msdf, mrit_min = .6)
#> $scl_1
#>           mrit rbar alpha
#> disp, cyl 0.90 0.90  0.95
#> wt        0.89 0.86  0.95
#> hp        0.79 0.81  0.94
#> carb      0.55 0.70  0.92
#> 
#> $scl_2
#>          mrit rbar alpha
#> gear, am 0.79 0.79  0.89
#> drat     0.74 0.74  0.89
#> mpg      0.63 0.66  0.89
#> vs       0.63 0.54  0.86
#> 

# Use positive correlations (extend on cores)
overlap(msdf, mrit_min = .6, overlap_with = "core")
#> $scl_1
#>           mrit rbar alpha
#> disp, cyl 0.90 0.90  0.95
#> wt        0.89 0.86  0.95
#> hp        0.79 0.81  0.94
#> 
#> $scl_2
#>          mrit rbar alpha
#> gear, am 0.79 0.79  0.89
#> drat     0.74 0.74  0.89
#> mpg      0.63 0.66  0.89
#> vs       0.63 0.54  0.86
#> 

# Include negative correlations
overlap(msdf, mrit_min = .7, negative_too = TRUE, sclvals = c(-3,3))
#> 
#> overlap() reversed the following item(s):
#> - mpg
#> - vs
#> - wt
#> - disp
#> - cyl
#> - hp
#> $scl_1
#>           mrit rbar alpha
#> disp, cyl 0.90 0.90  0.95
#> wt        0.89 0.86  0.95
#> hp        0.79 0.81  0.94
#> carb      0.55 0.70  0.92
#> mpg       0.87 0.72  0.94
#> vs        0.75 0.71  0.94
#> 
#> $scl_2
#>          mrit rbar alpha
#> gear, am 0.79 0.79  0.89
#> drat     0.74 0.74  0.89
#> mpg      0.63 0.66  0.89
#> vs       0.63 0.54  0.86
#> wt       0.89 0.59  0.90
#> disp     0.89 0.63  0.92
#> cyl      0.91 0.65  0.94
#> hp       0.79 0.63  0.94
#> 

# Change the treatment of missing values
overlap(msdf, mrit_min = .6, use = "all.obs")
#> $scl_1
#>           mrit rbar alpha
#> disp, cyl 0.90 0.90  0.95
#> wt        0.89 0.86  0.95
#> hp        0.79 0.81  0.94
#> carb      0.55 0.70  0.92
#> 
#> $scl_2
#>          mrit rbar alpha
#> gear, am 0.79 0.79  0.89
#> drat     0.74 0.74  0.89
#> mpg      0.63 0.66  0.89
#> vs       0.63 0.54  0.86
#>