Respondents’ marital status was marked as either single or other.
Divorced/separated, married, and engaged individuals were consolidated in the other
category in order to create a group large enough for analysis. Self-perceived school
ability and intelligence were found to be highly correlated (r = 0.738); school ability was
selected over intelligence due to the fact that its distribution was less skewed. For school
ability, respondents classified themselves as below average, average, or above average on
a seven-point scale. This was converted to a five-category scale—below average,average, slightly above average, and far above average. The finer above average
categories were maintained because the distribution was somewhat skewed in this
direction. Dummy variables were created based upon whether individuals indicated that
they were likely to attend only a two-year college, only a four-year college, both, or
neither.