Exercise 1. Party identification and presidential vote
Party identification usually is strongly related to voting behavior. This relationship is discussed in the section on voting behavior. To see how strongly party identification was related to the 2004 presidential vote, you should generate a table that shows the relationship between the individual's party identification (V007) and his or her presidential vote (V002).
The table that you just generated may be a little difficult to interpret because:
Party identification has seven categories. This makes the table fairly complicated. If party identification were simplified so that it had fewer categories, this would make the table less complicated.
The presidential vote includes those who voted for minor party candidates. Since so few people voted for minor candidates, it would simplify the table to eliminate them.
A cleaner and simpler table can be generated by:
Recoding party identification to just three categories (Democrats, independents, and Republicans).
Recoding presidential vote so that we have only the two major-party candidates.
General information about why it can be desirable to recode variables is in the section on principles of data analysis.
Helpful hints on using DAS/SDA to recode variables
Generate a new, simpler table (Table 1B) by using recoded versions of party identification and presidential vote. Compare Table 1B to Table 1A.