Analysis of Variance

This assignment gives you practice with analysis of variance using the GSS data set. It will also require you to use several SPSS commands and to answer some questions as you proceed. The basic problem is to analyze the determinants of pro-choice attitudes (a continuous variable) based on whether the respondent has ever had a born again experience (a discrete variable with two categories) and region of the country (a discrete variable with multiple categories). After some preliminary computations, there will be three main parts to the computer analysis: (1) constructing a difference of means test (using SPSS command T-TEST) and a one-way of analysis of variance (using SPSS command ANOVA) for pro-choice attitudes based on whether one has had a born-again experience or not, and comparing the results of the two; (2) constructing a one-way analysis of variance (using SPSS command ANOVA) for pro-choice attitudes based on region of the country in which one lives; and (3) constructing a two-way analysis of variance (using SPSS command ANOVA) for pro-choice attitudes based on both whether one has had a born-again experience or not and region of the country in which one lives. You will then be asked to answer several questions about these analyses.

NOTE: For this assignment, make sure you hand in a) your syntax file; b) your output file; and c) all written responses and hand calculations asked for in the questions at the end .

The key variables from the GSS data set you will need for this analysis are:

ABANY – Respondent’s belief that it is wrong for a woman to have an abortion for any reason

ABDEFECT – Respondent’s belief that it is wrong for a woman to have an abortion if there is a

possible birth defect in the child

ABHLTH – Respondent’s belief that it is wrong for a woman to have an abortion if the woman’s

health is seriously endangered

ABRAPE – Respondent’s belief that it is wrong for a woman to have an abortion if the woman has

been raped

ABSINGLE – Respondent’s belief that it is wrong for a woman to have an abortion if the woman

is single

REBORN – whether Respondent has ever had a "born-again" experience

REGION – region of the country where Respondent lives

1. Use SPSS command FREQUENCIES to get a frequencies distribution and descriptive statistics for each of the seven variables.
   We need to create proxy variables for the first five variables with the COMPUTE command.
2. Use SPSS command COMPUTE to create proxy variables ABANY X , ABDEFECX, ABHLTHX, ABRAPEX, ABSINGLX, and REBORNX.
   The first six variables are dichotomous. Each of them presents two problems for our analysis: (a) they have various missing values codes that are treated as valid data; (b) they are coded as 1, 2 and we would prefer them to be recoded (and re-keyed) as 0,1. We can address both of these issues with the RECODE subcommand.
3. Now, use SPSS command RECODE to recode these proxy dichotomous variables so that codes of 2=0, 1=1, and else= sysmis (the system missing value code ). This will re-key the variables so that each of the abortion variables now express Respondent’s favorability toward abortion (=1) and the born again variable expresses whether the Respondent has had a born-again experience (=1).
4. Use SPSS command FREQUENCIES to get a frequencies distribution and descriptive statistics for each of the re-keyed proxy variables: ABANY X , ABDEFECX, ABHLTHX, ABRAPEX, ABSINGLX, and REBORNX. Confirm that the recoding was successful and that the variables are re-keyed as desired. (Compare to results from Step 1 above.)
   Now you want to create a scale called PROCHOIC to represent pro-choice attitudes toward abortion. This combines the five pro-abortion variables into a single measure which we expect will provide a more robust measure of people’s pro-choice attitudes.
5. Use SPSS command COMPUTE to create a variable called PROCHOIC that consists of adding together the five pro-suicide proxy variables: ABANY X , ABDEFECX, ABHLTHX, ABRAPEX, and ABSINGLX .
   Now, check to see what the distribution of the new variable looks like.
6. Use SPSS command FREQUENCIES to get a frequencies distribution and descriptive statistics for PROCHOIC. Confirm that the creation of the pro-choice scale was done correctly.
   Now, let’s see how the sub-group means for PROCHOIC compare across the two categories of REBORNX and the nine categories of REGION.
7. Use SPSS command MEANS to get the sub-group means of PROCHOIC (a) by REBORNX and (b) by REGION. From a visual inspection, does it appear that there is a difference in pro-choice attitudes based on whether or not one has had a "born-again" experience? How so? Does it appear that there is a difference in pro-choice attitudes based on the region of the country in which one lives? How so?
   Now, after these preliminary steps, we are ready to do the analysis portion of the problem. First, we’ll do analysis number one which compares the difference of means format to the analysis of variance format. Test the difference of means in pro-choice attitudes based on whether or not one has had a born-again experience or not.
8. Use the SPSS command T-TEST to conduct a difference of means test for PROCHOIC based on REBORNX . You can use similar syntax to what you used in the previous lab assignment.
9. Use the SPSS command ANOVA to do a one-way analysis of variance for PROCHOIC based on REBORNX.
   Now, we’ll do analysis number two, which runs a one-way analysis of variance of pro-choice attitudes based on region of the country in which the Respondent lives.
10. Use the SPSS command ANOVA to do a one-way analysis of variance for PROCHOIC based on REGION.
    Now, we’ll do analysis number three, a two-way analysis of variance of pro-choice attitudes based on (a) whether Respondent has had a born-again experience and (b) region of the country in which Respondent lives.
11. Use the SPSS command ANOVA to do a two-way analysis of variance of PROCHOIC by REBORNX and REGION. You can emulate the syntax that was used in the handout on analysis of variance.

Now, based on your computer results, answer the following questions:

1. Write a brief research report discussing the results of analysis one. Discuss the null hypothesis, the research hypothesis, the alpha level, and critical values that you will use. Discuss the results of your findings, including the substantive implications of your results. Demonstrate why the results of the t-test and the analysis of variance produce exactly the same results.
2. Write a brief research report discussing the results of analysis two. Discuss the null hypothesis, the research hypothesis, the alpha level, and critical values that you will use. Discuss the results of your findings, including the substantive implications of your results.
3. Write a brief research report discussing the results of analysis three. Discuss the null hypothesis, the research hypothesis, the alpha level, and critical values that you will use. Discuss the results of your findings, including the substantive implications of your results. Make sure to provide adequate discussion of the main effects, the interaction effects, and the total effects of the model.

Price: $34.02

Solution: The downloadable solution consists of 17 pages, 1702 words and 14 charts.
Deliverable: Word Document