forward.search: Forward search algorithm for outlier detection

Description

The forward search algorithm begins by selecting a homogeneous subset of cases based on a maximum likelihood criteria and continues to add individual cases at each iteration given an acceptance criteria. By default the function will add cases that contribute most to the likelihood function and that have the closest robust Mahalanobis distance, however model implied residuals may be included as well.

Usage

forward.search(
  data,
  model,
  criteria = c("GOF", "mah"),
  n.subsets = 1000,
  p.base = 0.4,
  print.messages = TRUE,
  ...
)
# S3 method for forward.search
print(x, ncases = 10, stat = "GOF", ...)
# S3 method for forward.search
plot(
  x,
  y = NULL,
  stat = "GOF",
  main = "Forward Search",
  type = c("p", "h"),
  ylab = "obs.resid",
  ...
)

Arguments

data: matrix or data.frame
model: if a single numeric number declares number of factors to extract in exploratory factor analysis. If class(model) is a sem (semmod), or lavaan (character), then a confirmatory approach is performed instead
criteria: character strings indicating the forward search method Can contain 'GOF' for goodness of fit distance, 'mah' for Mahalanobis distance, or 'res' for model implied residuals
n.subsets: a scalar indicating how many samples to draw to find a homogeneous starting base group
p.base: proportion of sample size to use as the base group
print.messages: logical; print how many iterations are remaining?
...: additional parameters to be passed
x: an object of class forward.search
ncases: number of final cases to print in the sequence
stat: type of statistic to use. Could be 'GOF', 'RMR', or 'gCD' for the model chi squared value, root mean square residual, or generalized Cook's distance, respectively
y: a null value ignored by plot
main: the main title of the plot
type: type of plot to use, default displays points and lines
ylab: the y label of the plot

Author

Phil Chalmers rphilip.chalmers@gmail.com

Details

Note that forward.search is not limited to confirmatory factor analysis and can apply to nearly any model being studied where detection of influential observations is important.

References

Chalmers, R. P. & Flora, D. B. (2015). faoutlier: An R Package for Detecting Influential Cases in Exploratory and Confirmatory Factor Analysis. Applied Psychological Measurement, 39, 573-574. tools:::Rd_expr_doi("10.1177/0146621615597894")

Flora, D. B., LaBrish, C. & Chalmers, R. P. (2012). Old and new ideas for data screening and assumption testing for exploratory and confirmatory factor analysis. Frontiers in Psychology, 3, 1-21. tools:::Rd_expr_doi("10.3389/fpsyg.2012.00055")