There are many analysis commands, some of
which are described in brief here.
DISTS
DISTS gives detailed information about your model's
district-level results. It enables you to estimate the partisan makeup of
individual districts and estimate the probability that each district will elect
a Democrat or a Republican. (For example, you might wish to evaluate the
partisan makup of districts that have incumbents, or ``minority districts,''
drawn to adhere to the Voting Rights Act. Or, you could make vote and seat
predictions under hypothetical situations, such as if minority incumbents
retire.) You can also evalute the fit of your model to the data, or its
success in predicting future election results, by comparing lists or plots of
observed and expected (or predicted) district-level results.
SVCURVE
This creates a seats-votes curve with standard errors (or
optionally lists the numerical values that would be used to graph such a
curve). The seats-votes curve is a plot of the proportion of legislative seats
Democrats would be expected to win for a range of given values of the average
Democratic district vote.
SEATS
This command gives results similar to those of the SVCURVE
command, except that with SEATS you can request seat predictions for particular
values of the average district vote. (SVCURVE gives you this in graphical
form, or listed form, but is evaluated at values equally spaced between
pre-chosen bounds.) For example, to evaluate alternative redistricting plans
you might wish to present a table with expected seats (and its standard error)
under each plan for several plausible average district votes. You might issue
this command for a prediction analysis under each redistricting plan: SEATS 0.45 0.50 0.55;. This will give the expected seats for the Democrats
given these three average district Democratic votes.
SUM
On the basis of the data already read in, the SUM command estimates
four summaries of the seats-votes curves and their standard errors: (1) two
measures of responsiveness, the extent to which changes in a party's
average district vote influences its proportion of seats, and (2) two measures
of partisan bias, the proportion of seats each party receives due only to
asymmetry in the allocation of seats to parties. It is also possible to
calculate the contribution of various regions (or other groups of districts) to
each of these quantities. Substantive details appear in reference section
8.23.
FREQ
This command enables you to estimate the number of seats any group
will win (such as the number of blacks or incumbents to win), or the proportion
of individual district results expected to fall within a range of vote
proportions (such as the expected proportion of districts within a range of
expected vote totals).
PROB
With this command you can estimate the actual number of districts
with specified probabilities of a candidate winning, such as the number of
``marginal'' seats.
