Soc2205 SPSS Introductory Exercise-

Read all instructions carefully before proceeding with the assignment! Each step of the assignment and its interpretation will be discussed in class when the relevant chapter is covered. Have your SPSS output ready to bring to class by the relevant dates. It is highly recommended that you complete the SPSS analysis for this assignment in the King's Computer Lab in Wemple during TA's lab hours. PLEASE PRINT THIS ASSIGNMENT FOR USE IN THE COMPUTER LAB!

General Instructions:

You will complete an individual independent statistical analysis that will involve testing three hypotheses using SPSS (statistical software package) after which you will write an interpretation of your findings. The assignment is due in class on Dec. 7th.  The completed assignment will consist of the SPSS printouts of your analysis accompanied by a brief written discussion of your results that includes (i) your hypotheses for steps 2, 3, and 4, (ii) the alpha level used and the decisions that you made, and (iii) a brief interpretation of your decisions and results. The assignment should be handed in personally in class, or if late, deposited, in an envelope addressed to me to the green box in the Annex. E-mailed copies will not be accepted. The written part of the assignment should be typed, double-spaced, on 8½ X 11 paper, and, together with the printouts, should be stapled with a cover page indicating your name and student number (no plastic covers or binders).  Late assignments are penalized at 5%/day.

Readings and Data Set to be Used for This Assignment:

For this assignment, please read Healey and Prus 1st, 2nd or 3rd Canadian Editions, Appendix F and G, as well as the SPSS Demonstration section at the end of each relevant chapter in Healey. To do the assignment, you will need to use the shortened version of 2009 General Social Survey (GSS) data set supplied for Healey and Prus 2rd Can. Ed.  The data set can be retrieved from the Nelson website by clicking on Data Sets and Codebooks in the sidebar at:

http://www.cengage.com/cgi-wadsworth/course_products_wp.pl?fid=M20b&product_isbn_issn=0176651365&template=NELSON

You will need the SPSS software to open the data set, so simply save it and then open it when you have accessed SPSS.

The SPSS analysis for this assignment can be completed using SPSS in the King's computer lab in W143 or remotely from home using the King's Scotty server. The TA will provide instructions for working from home if necessary but it is highly recommended that you complete the assignment in the lab. Please see the TA in the lab during her hours posted on the main webpage.

Variables and Hypotheses to be Tested:

For this assignment, you will be examining your chosen variables and statistically testing hypotheses that you create according to the guidelines presented in class.  You will be using variables from the shortened version of the 2009 GSS data set. This data set has 1512 cases and 48 variables. You can find a description of all of the variables in the dataset by clicking on the link to the GSS User's Guide (the Codebook) further down on the Nelson website; however for your assignment, you are limited to the variables listed below:

Dependent Variable:

Average number of evening activities respondent goes out in a month (numevact_c)

Independent Variables:

Sex of respondent (sex)

Urban/Rural indicator (luc_rst)

Self Reported Health (In general, would you say your health is...) (srh_110)

Marital status of the respondent (marstat)

When you have completed the SPSS portion of the assignment, it is recommended that you check with me or the TA before continuing with the assignment, in order to avoid having to make corrections later on.

The purpose of this assignment is to try to explain what factors cause variability how often respondents go out in the evening. The dependent variable (DV) for all steps is "Average number of evening activities respondent goes out in a month" (numevact_c ). Step 1 of the assignment involves describing your variables using frequency tables and graphs. Step 2 is a hypothesis test to test the mean number of evening activities for the sample against a known population parameter. Step 3 involves testing for significant differences in the DV by the independent variables, "Sex of respondent" (sex) and "Urban/Rural indicator" (luc_rst). Step 4 tests for significant differences by "Self Reported Health (In general, would you say your health is...)" (srh_110) and "Marital status of the respondent" (marstat).

Printing Your Output:

Make sure that you save all of your output as well as any changes you make to your data set on a USB flash drive in case you need to edit or redo parts of it later on. If you are working at the computer lab, you can print your output directly from SPSS for a small charge. You can also copy and paste your output into Word at the lab and then edit it before printing it or emailing it to yourself and printing it at home. If completing the analysis remotely from home, the easiest way to print your results is to use the "export" command in SPSS to first save your results as an .rtf (Word) file. You can then save the .rtf file on your home computer or email the file to yourself . You will be able to use Word to open the file at home and print it.

Detailed Step-by-step Assignment Instructions

Step 1- Frequency Tables, Statistics, and Charts (Resource: SPSS explanations at end of chapters in Healey 1st Cdn. Ch. 2-4 or Healey 2nd or 3rd Cdn. Ch. 2-3):          

(a) Use SPSS to generate frequency tables, statistics and charts for all of your variables (go to ANALYZE>DESCRIPTIVE STATISTICS> FREQUENCIES). Get appropriate descriptive statistics for each variable, as well get an appropriate bar, pie chart or histogram (click normal curve too) for each (click STATISTICS and CHARTS.)  Print your output.

(b) Write a brief description of the results (including the graphs) for each variable. ' (suggested completion Sept. 28)

Step 2 - One Sample T-test (Resource: SPSS explanations at end of chapter in Healey 1st Cdn. Ch. 8 or 2nd & 3rd Cdn. Ch. 7)

Your first hypothesis involves testing the mean value of your DV "Average number of evening activities respondent goes out in a month" against a population mean value of 24.5 evening activities per month. Use the 5-step method: your SPSS analysis is Step 4.

(a) State your null and alternate hypothesis and the alpha level you will use. (Hint: the sample data set has 1500+ cases)

(b) Test your hypothesis using the "ONE SAMPLE T-TEST" command (go to ANALYZE>COMPARE MEANS>ONE SAMPLE T-TEST) to test whether there is a significant difference between the sample mean and the population mean of 24.5 (this is your TEST VALUE). Click on OPTIONS to change your confidence interval to the appropriate ALPHA level.)

(c) Write a brief description of the results of the above analysis. To interpret the One Sample T-test, follow the guidelines for hypothesis testing discussed in class, using the five step method. Bring your SPSS printout to class on Nov. 2, when we will discuss how to interpret it.

Step 3 - Independent Two Sample T-Test (Resource: SPSS explanations at end of chapter in Healey 1st Cdn. Ch. 9 or 2nd or 3rd Cdn. Ch. 8)

This step involves conducting 2 independent sample T-Tests to investigate whether there are significant differences in "Average number of evening activities respondent goes out in a month" by the independent variables, Sex and Urban/Rural. Your SPSS analysis is Step 4 of the 5-step method.

(a) State your null and alternate hypothesis and alpha level for each test.

(b) Test both of your hypotheses using the "INDEPENDENT SAMPLES T-TEST" command (go to ANALYZE>COMPARE MEANS>INDEPENDENT SAMPLES T-TEST) to test whether there are significant differences between the group means on your dependent variable (your DV is at the interval-ratio level) using nominal level IV's with two categories (Sex, Urban/Rural.) Print.

(c) Write a full description of the results of the above analysis. The T-Test on your printout out is displayed in two rows and you will first need to examine the Levene's test (first two columns of the output chart) to decide which row of the T-test to use. If the p-value for the Levene's test is > than your alpha level, use the top row of the T-test and if it is lower than your alpha, use the bottom row. To interpret the T-test, follow the guidelines for hypothesis testing discussed in class and the five step method. Compare the results of the two t-tests. (Bring SPSS printout to class on Nov. 9)

Step 4 - One-Way ANOVA Test (Resource: SPSS explanations at end of chapter in Healey 1st Cdn. Ch. 10 or 2nd or 3rd Cdn. Ch. 9)

This is an ANOVA test to investigate whether there are significant differences in the group means of "Average number of evening activities respondent goes out in a month" by the independent variables, Marital Status and Self Reported Health, both of which have 5 categories. Use the 5 step method (your SPSS analysis is step 4.)

(a) For each test, state your null and alternate hypothesis and alpha level.

(b) Test your hypotheses using the "ONEWAY ANOVA" command (go to ANALYZE>COMPARE MEANS>ONEWAY ANOVA) to test whether there are significant differences between group means of your dependent variable (your interval-ratio level variable, Number of Evening Activities) using a nominal/ordinal level variable with at least three categories as your independent variable (in this case, Marital Status and Health.) Use the Tukey's b test (click POST HOC>TUKEY'S B) to identify exactly which group(s) are different from one another. Use the appropriate alpha level (the significance level is found at the bottom of POST HOC dialogue box.) Print.

(b) Write a brief description of the results of the above analyses. Compare the results. (Bring SPSS printout to class on Nov. 16)

Step 5 - Summary

Briefly summarize (just a paragraph or two) the results of steps 2 - 4 above. Which independent variables affect your dependent variable (i.e. which relationships are significant?) Can you identify one independent variable which seems to affect the dependent variable most strongly? Try to make some sense of your findings!