For this project use GSS 2016 data posted on the Blackboard in the data file. The codebook for the data is also available there. This is a full data set (not the student version you have used for the homework). Be attentive when selecting your variables. Run the frequency tables first to see the distribution of the variables you are going to use for the T and F tests.

Project EXAMPLE and book assignment clarification:

This is an example. It is not perfect :-) Select one dependent and one independent variable for each of the two tests (nottwo dependent and two independent variables, as the book suggests). Word limit is 650 words, excluding the SPSS tables.

Perform the analysis for ONE pair of variables for each of the two tests: T and F. Do NOT use the variables analyzed in the example. Here is the list of variables that you should not use in your project: polviews, age, sex, prestige80, and class. Read the example (including the attachment part) carefully before working on the project.

Part A: T-Test

As stated in the assignment on p. 455 of your book the independent variable should logically be the cause of your dependent variable. Come up with a simple theory, for example: gender, the independent variable, influences ones viewpoint towards the joy of watching children grow up, the dependent variable. Then state the null and the research hypothesis, and then use t test to test your hypothesis.

Use only IR variable as your dependent variable and a nominal variable as your independent variable. Do NOT use ordinal variable as dependent variable as the book suggests.

For a t-test, independent variables can have ONLY TWO categories (two groups). For variables with more than two categories (relig, for example), you can meet this requirement by using Define Groups button in the Grouping Variable box to select specific categories of a variable. You can, for example compare Protestants and Catholics on relig by choosing scores of 1(protestants) and 2 (Catholics). You can also recode a variable with several categories into a new variable with two categories or recode an interval-ratio variable. For example number of children in the family can be recoded into the nominal variable with two categories: "Do you have children?" - Yes/No. Do one-tail hypothesis test when possible.

Part B: Analysis of Variance Test (F test).

Use interval ratio variable as your dependent variable and ordinal variable as your independent variable. If you find that at least one of the means of your sample groups is significantly different, run the Post Hoc Analysis. If you find that there are no significant differences between the means of the sample groups, then you dont run the Post Hoc test, as it does not make sense to look for a group(s) that is different, when the ANOVA test shows an overall insignificant difference between the group means. The Post Hoc Analysis is not shown in the example.

RESEARCH MEMO

Introduction

The purpose of the subject project was to apply statistical concepts and SPSS program software for several hypothesis testing methods including the Two-Sample t test and Analysis of Variance (ANOVA) test.

Part A: Independent Samples T Test

I have used the Independent-Samples T test for the variables polviews (independent variable) and age (dependent variable) to determine if people who think of themselves as conservative are significantly older than those who think of themselves as liberal.

The five step model to test for significance

Step 1: Assumptions and Test Requirements

Step 2: State Null Hypothesis

Step 3: Select Sampling Distribution and Establish Critical Region

Step 4: Compute Test Statistic Using SPSS

Step 5: Make Decision and Interpret Result

Part B: One-Way ANOVA Test

Step 1: Assumptions and Test Requirements

Step 2: State Null Hypothesis

Step 3: Select Sampling Distribution

Step 4: Compute Test Statistic Using SPSS

Step 5: Make Decision and Interpret Result

Since the ANOVA test finds the difference in the occupational prestige score by social class, we have to run the Post HOC analysis to determine which group(s)mean is different.