Assignment 2
Measuring Religiosity
 

The objective for this week is to describe in more detail the relationship between different measures of religiosity. After you complete reading and doing Chapters 11, 13-15 in Babbie & Halley’s Adventures in social research, complete the following brief exercises. These questions assume you have already gone through the assigned chapters. Submit your work next week (October 27) at the beginning of the class session. You will need to use the GSS.SAV file supplied with your book and your GSS file in your P-drive. Please do not use the file in the K-drive folder.  

  1. Open your Babbie data file. Many of you wanted to find variables more easily. To reset the variables are listed, toggle the "edit" option in the top left corner of the SPSS page and then select "options." Inside options, notice the choices in the upper right. Toggle the option "display names" have the variables listed by their variable name rather than the default option, long variable labels. Also toggle "alphabetic". You will need to exit the Babbie data file for these changes to initiate when you reopen the file. I recommend that you exit by opening another data file, for example the full year file you have on you P-drive. This is a bit quicker than closing SPSS and restarting it.
  2. Until this process is completed on each machine we use, you will have to repeat this step on each new machine you use that has not be reset.
     

  3. Since a fair number of you reported that you had difficulty building an index (or composite measure), this first task is to repeat parts of Chapter 9 and develop another index from the information available in Babbie’s condensed data set. The new variable will be called RELBELIF. 
  4. First, compute a frequency distribution of PRAY, ATTEND, and POSTLIFE.  

    Next, use the Transform à Compute command and in the upper left window, type in the new variable name "RELBELIF." The variable now exists, but it does not yet have variability (or data). To begin identifying the values of this new variable, enter 0 in the big, upper left numeric expression window and click "OK". We have just assigned everyone a baseline value of zero. (If already confused, refer to Chapter 9, p. 78 for more details on the process of entering data).  

    Select Transform à Compute again, and click the "reset" button at the bottom of this "page." Type RELBELIF into the upper left target field again, and in the upper right type RELBELIF+1 Click the "If" button in the middle of this "page" and notice a new page. Using this page we set the conditions for who will be recoded from a zero to a one. First toggle "Include if case satisfies condition" in the upper right, otherwise the screen will remain gray and you will not be able to select a variable. Select POSTLIFE from the variable list, and after it enters the rectangluar window add the condition = 1 then click "continue" and when the screen changes, click "OK". The program will query "Change the existing variable?" which is exactly what you want to do, so click "OK" again. 

    You have now created RELBELIEF with two values – zero and one. We need to continue. Again select Transform à Compute and click "reset." Again type RELBELIEF into the target field and REBELIEF+1 into the upper right window. Next, again click the "If" button, select "PRAY" from the variable list, and after it enters the window add the condition <3 then click "OK" and again "OK". The condition <3 for PRAY specified that people who had a code of 1 or 2 (prayed several times a day or prayed once a day) be added to the RELBELIEF index. We now have possible scores ranging from zero to 2.  

    One more time, select Transform à Compute and click "reset." Type RELBELIEF into the targer field and RELBELIEF+1 into the other window. Click "If" and select "ATTEND" from the variable list, and set its condition as > 5. [This translates into people who scored 6, 7, 8. The attend nearly every week, or every week, or more often.] 

    You now have a composite index that has a possible range from zero to 3. Compute a frequency distribution and print a copy of the table. 

    Validate your index. Compute a correlation between RELBELIEF, PREMARSX, and HOMOSEX. What are the correlations coeffienents? Are the positive or negative? How do you interpret them (see pp. 229-230 to see the way the two sex variables were coded)?
     

  5. The primary objective for this next task is to describe the relationship between religious denomination and two measures of religiosity. Compute a bivariate table of PRAY2 and RELIG (with denomination the column variable), and a second table for CHATT2 and RELIG. (Pssst: If you did not save your PRAY2 & CHATT2 variables, rebuild them using last week’s guidelines and then proceed. Be sure to correctly label the values; see page And, if you are uncertain about computing bivariate tables, follow the directions in your workbook, pp. 102-104) 
  6. Drawing from the crosstab tables, what ______% of the Catholic respondents rarely or never pray, and ________% of the Protestants pray several times a day. Do Catholic, Protestant, and Jews participate in their religion equally? Who is more involved in their religion, as determined by their extrinsic religiosity? Now, who is more involved as defined by their intrinsic religiosity?
     

  7. Change data sets by simply "opening" your GSS year from your P-drive. (Opening this closes the Babbie data.) To being, compute a simple line chart for age. Print a copy the chart. You print a single chart by selecting (or clicking) the one you want among all the output [Psst: selecting "draws" a box around the chart.] Next select FILE à PRINT and toggle the "print selection" option before you print; otherwise you will print all of the output.
  8. Develop a new variable "AGECAT" by using the Transform à RECODE command. Recode AGE into AGECAT (using the "recode into different variable" toggle) and create five categories – under 25, 25-40, 41-55, 55-70, 71+. Compute a frequency distribution of this variable. Save your files (and thus save your recoded work). 

    Develop another variable "CHATT" using the Transform à Recode command. Recode ATTEND into CHATT with four categories – never (0), rarely (1,2=1), monthly (3,4,5=2), and at least weekly (6,7,8=3). Compute a frequency distribution of this variable. Save your files (and thus save your recoded work). 

    Develop another variable "PRAY2" by recoding pray into three categories – rarely (5,6=1), occasionally (3,4=2), often (1,2=3). Set "all other values" to SYSMIS. Compute a frequency distribution of PRAY and PRAY2. Save your files (and thus save your recoded work).
     

  9. Compute bivariate tables for RELITEN by SEX and RELITEN by AGECAT and RELITEN by CLASS (sex, age category, and class are the column variables). Be sure to include column percents (via the "Cells" button) and request the chi square statistic and the phi--Cramer’s V statistic (via the "Statistics" button). 
  10. What evidence do you have to support the deprivation theory of religiosity? First, is each observed relationship between RELITEN and sex, age, and class statistically significant? (You answer this by reviewing the significance of the chi-square statistic; is it less than .05?) Second, how strong is each observed relationship? (You assess strength of relationship with a measure of association, and I asked you to select the Cramer’s V statistic). 

    Strength of affiliation     c 2(Chi square)     Cramer’s V 
     

      by sex 

      by age 

      by class

      
  11. Compute bivariate tables for PRAYER by SEX and PRAYER by AGECAT, and PRAYER by CLASS. Again include column percents and request the chi square statistic and the phi--Cramer’s V statistic. 
  12. You are assessing what variable? (Look at the variable label and then look at the value labels to see how the variable was coded.) Are the relationships between this variable and sex, age, and class chance relationships or reliable relationships (recall, the chi square statistic determines the significance of the relationship, and for chance relationships the chi square is greater than .05)? According to the Cramer’s V, are the observed relationships weak, moderate, or strong? 

    Strength of affiliation     c 2(Chi square)         Cramer’s V 

      by sex

by age
by class