abortion {lcca} | R Documentation |
This dataset, which was extracted from the 2006 General Social Survey, reports the responses of adults in the United States to six questions about legalized abortion. The questions began, “Please tell me whether or not you think it should be possible for a pregnant woman to obtain a legal abortion if...”
abortion
a data frame with 4510 rows and 8 variables:
SEX
respondent's sex (factor with two levels)
ABANY
“...The woman wants it for any reason?” (factor with five levels)
ABDEFECT
“...If there is a strong chance of serious defect in the baby?” (factor with five levels)
ABHLTH
“...If the woman's own health is seriously endangered by the pregnancy?” (factor with five levels)
ABNOMORE
“...If she is married and does not want any more children?” (factor with five levels)
ABPOOR
“...If the family has a very low income and cannot afford any more children?” (factor with five levels)
ABRAPE
“...If she became pregnant as a result of rape?” (factor with five levels)
WTSSNR
analytic weight adjusted for subsampling of initial nonrespondents (numeric)
Because of the split half-sample design used in the 2006
GSS, only about half of the sampled adults were asked the abortion
questions. The response code "NAP"
(not applicable) indicates
that the question was not asked.
The other response codes are
"YES"
, "NO"
and "DK"
(Don't know). A missing
value (NA
) indicates that the
question was asked but no answer was given.
Analysts should recode "NAP"
to
a missing value. Whether "DK"
should be
converted to a missing value is debatable.
Although the GSS has a complex multistage area sampling plan, it was
designed to be self-weighting in the sense that every adult in the
sample frame had an approximately equal chance of being selected.
However, many sampled persons did not respond to the initial interview
request. To help reduce nonresponse bias, a random sample of these
nonrespondents were selected for aggressive followup attempts. Those
who were successfully interviewed in this followup procedure ought to
be assigned greater weight, because they need to represent
those who were not selected for followup. The variable
WTSSNR
is a weight that adjusts for this nonresponse followup
procedure, and the GSS documentation recommends that this weight be used
in analyses.
Latent-class analyses of abortion questions from earlier GSS surveys were reported by McCutcheon (1987) and by McCutcheon and Nawojczyk (1987).
Davis, J.A. and Smith, T. W. (2007) General Social Surveys, 1972-2006 (machine-readable data file). Chicago: National Opinion Research Center (producer). Storrs, CT: The Roper Center for Public Opinion Research, University of Connecticut (distributor).
McCutcheon, A.L. (1987) Sexual morality, pro-life values, and
attitudes toward abortion: a simultaneous latent structure analysis
for 1978-1983. Sociological Methods and Research, 16, 256-275.
McCutcheon, A.L. and Nawojczyk, M. (1995) Making the break: popular
sentiment toward legalized abortion among American and Polish Catholic
laities. International Journal of Public Opinion Research, 7,
232-252.
For example analyses of this dataset using functions in
the LCCA package, see the manual LCCA Package for R, Version 1
in the subdirectory doc
.