POLI 300                                                 PROBLEM SET #5B                                      Due 04/06/09

SPSS FREQUENCY DISTRIBUTIONS

Note:   These problems require work in the Student PC Labs in the Engineering Building, Library, or elsewhere. Remember to take the SPSS Handouts and the SETUPS/ANES Codebook with you. You will probably want to do all the SPSS work in one sitting, take printed output with you, and then answer the substantive questions later. If possible, attach a copy of your SPSS output to your answers. If this is not possible, copy the frequency numbers off the monitor screen and put them in hand-drawn frequency tables.

1.         Construct a frequency distribution for Question 33 (REPUBLICAN PARTY IDEOLOGY) in the Student Survey data. Then look in the SETUPS Codebook to find the corresponding ANES variable. Go to a PC Lab and run the SPSS FREQUENCIES procedure on this SETUPS variable. (Also ask for an SPSS bar chart, if you wish.) Does comparison between the student survey and NES frequency distributions turn up any interesting similarities or differences? Is this comparison in some ways improper, i.e., a comparison of “apples and oranges”?

2.         Construct a frequency distribution for V25 (BUSH IDEOLOGY) in the Student Survey data. Then figure out how to get a SETUPS/NES frequency distribution that matches (as well as possible) this student survey data. (Note 1: you will have to use the Select Cases feature. See the handout on Using Setups 1972-2004 NES Data and SPSS for Windows, Section VII on “How to Select a Subset of Cases.” Note 2: for some reason, a straightforward Select Cases expression such as “1988 < V01 < 2000” does not work; you must use an expression such as “v01 > 1988 & v01 < 2000” instead.) Run this frequency and compare the NES and student survey distributions. (Merged bar charts may facilitate this comparison.)

3.         Sentence #17 in Problems Set 1 says: "Party identification in the electorate has weakened over the last generation." Use SPSS and the SETUPS/NES 1972-2004 data to assess the empirical truth of this proposition. (See the handout on Using Setups 1972-2004 NES Data and SPSS for Windows, Section X on “How to Quickly Get Frequencies or Crosstabulations For Each Presidential Election Year Separately.”)

4.         Go to some reference source to find the official national popular vote percentages (percent of vote won by each candidate) for all Presidential elections from 1972 through 2004. (You might use SPSS and the MASTER.sav data file to calculate these percentages, but you can get this information much more readily from a textbook, website , Information Please Almanac, etc.) How representative do the ANES 1972-2004 samples appear to be with respect to reported vote for President? (To answer this questions, you will have to use SPSS and the SETUPS/ANES data. See the handout on Using Setups 1972-2004 NES Data and SPSS for Windows, Section X on “How to Quickly Get Frequencies or Crosstabulations For Each Presidential Election Year Separately.”)