Analyzing Presidential Voting and Racial Injustice Views with SPSS Manual

Task 1:

• Open up your ANES2016-2020 datafile

• As before, WEIGHT by the variable POST2020

�

Task 2:

Let�s turn V47 (2020 Presidential Vote) into a Binary (Biden/Trump) by eliminating (SYSMIS) those who voted for other candidates. We�ll create a new variable, TWOPARTY:

RECODE V47 ( 1=1)(2=2)(ELSE=SYSMIS) into TWOPARTY.

Run a Frequency on that new variable to make sure it aligns with this output. Category �1� indicates a vote for Biden, �2� for Trump. Biden did much better in this survey than in the actual election, even considering sampling error:

Task 3:

New Procedure � CROSSTABS (revised SPSS MANUAL, CROSSTABS)

• Many of you, in your initial hypothesis, treated the DV as if it was a direct assessment of �how� someone voted, not their feelings towards the candidates individually. We are now giving you the chance to directly test for that IV-Vote relationship.

• Produce a crosstabulation/contingency table (CROSSTABS) with the new variable TWOPARTY as your dependent variable (defining the rows�it can be reversed, but this way you are all using the same format) and your previously chosen binary IV as your independent variable.

• Ask for column percentages and chi-square.
• Your original binary IV (RACE or V41 or AGE65 or AGEMEDIAN or EDUCATION or INCOME) will be your independent variable

• As this is a 2x2 table (df=1) use the (Yates) Continuity Correction Chi-square value.

�

Task 4 :

• Go back to your original (A1) or revised (A2) hypothesis. Answer the following questions :

• Does your analysis descriptively confirm or disconfirm your original (or revised) hypothesis(es)? Why or why not? Be complete in your answer.

• Does your analysis inferentially confirm or disconfirm your original (or revised) hypothesis(es)? Why or why not? Be complete in your answer.

New Procedure � controlled CROSSTABS (revised SPSS MANUAL, CROSSTABS)

You will now use a third variable as a control. This control can be used for either an explanatory link (why are your IV and DV connected as they are?) or interactive/specification effect (what are the effects of two variables on the outcome). Repeat the steps in Section A for each of your controlled crosstabs.

• If RACE is your binary IV:

Task 1: Are differences in presidential voting between Black and Non-Black citizens explained by their different views towards racial injustice? Use V99 (Do police treat blacks or whites better?) by recoding into a new binary variable POLICE (too few listed category 3, so combine it with 2):

RECODE V99 (1=1)(2,3=2) INTO POLICE.

Your new variable, POLICE, should list as (run a FREQUENCY):

Task 2:

Run a controlled CROSSTABS with RACE as your IV, TWOPARTY as your DV, and POLICE as your control. Again, ask for COL percentages and CHISQUARE. What do these tables and chi-square statistics tell you? Discuss in detail. Are your original racial differences altered? Are they less or more pronounced for each category? Fully explain, with both percentage differences and chi-square what you have found.

Task 3:

Run a controlled CROSSTABS with RACE as your IV, TWOPARTY as your DV, and INCOME as your control. Again, ask for COL percentages and CHISQUARE.