FLAS: Raymond's Foreign Language Attitude Scale (FLAS) Data
In mmaechler/miP: Multiple Imputation Plots

Description Usage Format Details References Examples

Raymond (1987) describes data that were collected to investigate the usefulness of a newly developed instrument, the Foreign Language Attitude Scale (FLAS), for predicting success in the study of foreign languages. The dataset contains missing values.

1	data(FLAS)

A data frame with 279 observations on the following 12 variables.

lang: foreign language studied, a factor with 4 levels ("french", "spanish", "german", "russian").
age: age group, a two-level factor with levels "-19" = less than 20, and "20+" = 20 and more.
priC: the number of prior foreign language courses; as a 3-level factor w/ levels c("none", "1--2", "3+").
sex: M=male, F=female.
FLAS: score on foreign language attitude scale.
MLAT: Modern Language Aptitude Test, fourth subtest score.
vSAT: verbal score of Scholastic Attitude Test.
mSAT: math score of Scholastic Attitude Test.
eng: score on Penn State English placement exam.
HGPA: high scool grad point average.
CGPA: current college grad point average.
gradeB: final grade in foreign language course as binary (hence the “B” in the variable name) factor w/ levels "B-" (B or lower), and "A".

This dataset is an adjusted version of the original dataset by Mark Raymond. The adjustments can be looked up in Schafer (1997, p.200). Where Schafer called the variables Y_1, Y_2, ..., Y_{12}, then in the original miP package they had all-capital colnames, see FLAS.nms in the example, these have been slightly modified by Martin Maechler; also, the 5 factor variables got meaningful levels (instead of just integer codes 1, 2, ...).

Schafer, J.L. (1997) Analysis of Incomplete Multivariate Data Chapman and Hall, London.

data(FLAS)
str(FLAS)
summary(FLAS) # summary *includes* number of NA's

## contingency table of three interesting factors -- including NA's
with(FLAS, table(sex, age, gradeB, useNA = "ifany"))


### The original (2011 `miP` CRAN package) data "labeling" and coding:

##  1) labeling all factors with integers; this is exactly
which(iF <- sapply(FLAS, is.factor)) # 1  2  3  4  12
FLAS.n <- FLAS
FLAS.n[iF] <- lapply(FLAS[iF], function(.) as.factor(as.integer(.)))
##  2) and using all capital variable names:
FLAS.nms <- c("LAN",  "AGE",  "PRI", "SEX",  "FLAS", "MLAT",
              "SATV", "SATM", "ENG", "HGPA", "CGPA", "GRD")
names(FLAS.n) <- FLAS.nms
## -> FLAS.n  is what 'FLAS' used to be in 2011's miP version