Contents-ASR God MS

ISC PUBLICATIONS [ Contents | What's New | ISC Home Page]

2AF Secondary Analysis File:

"National Context, Parental Socialization, and Religious Belief: Results From 15 Nations"

Jonathan Kelley and Nan Dirk De Graaf.
American Sociological Review 62 (August, 1997) 639-659.

Abstract

TEXT

For the text of the article, see the August 1997 issue of the ASR. We will put the text on line later, but for now please see the journal.

RAW DATA

Data are from the 1991 'Religion' module of the International Social Survey Programme (ISSP), an international consortium composed primarily of academic survey organizations (Zentralarchiv 1993). Each year, the ISSP creates a module containing exactly the same questions, answer categories, and sequencing for all countries surveyed. This module is then fielded in conjunction with each country's regular annual survey. Each country also collects detailed data on background and demographic variables, using questions appropriate to local circumstances and institutions. The samples are all large, representative national samples of adults. In all there are 19,815 cases: 2,203 in Australia; 984 in Austria; 1,486 in East Germany; 1,257 in Great Britain; 1000 in Hungary; 1,005 in Ireland; 983 in Italy; 1,635 in the Netherlands; 1,070 in New Zealand (Gendall 1992); 838 in Northern Ireland; 1,506 in Norway; 1,063 in Poland; 2,080 in Slovenia; 1,346 in West Germany; and 1,359 in the United States. The ISSP survey also includes data from Israel, the Philippines, and Russia but these are not used in the present analysis, for reasons described in the paper.

Getting the raw data. Available here is a subset of the data, including all the variables used in this paper and some others that might offer interesting new analyses. It includes only the 15 nations used in this paper. Get this subset of the data here [get data]. These are the nations:

V3        Country
Value Label   Value  Frequency
D-W               1      1346
D-E               2      1486
GB                3      1257
NIRL              4       838
USA               5      1359
H                 6      1000
NL                7      1635
I                 8       983
IRL               9      1005
N                10      1506
A                11       984
SLO              12      2080
PL               13      1063
NZ               16      1070
AUS              18      2203
                       -------
              Total     19815

The complete data set is available from the International Social Survey Programme and many data archives. Our SPSS should run on these data without difficulty except that you will need to check that all our variable definitions work OK for Israel, the Philippines, and Russia, which were not in our analysis. (Watch out especially for items not asked in all nations and see notes in our SPSS about religious belief questions in the Philippines.)
If you only want to use the variables in our analysis and the additional ones we selected, and only the 15 nations we analyzed, we recommend that you not use these raw data set but instead use the 'analytic data set' described below -- that will be easier to use.

SPSS: CONSTRUCTION OF VARIABLES

Details on the SPSS file:

(1) This file begins with the raw data from the International Social Survey Program's "Religion" module. (The variables needed, together with some others that might allow useful parallel analyses on related topics are in file GOD3). This SPSS makes all the variables used in our analysis. It would be easy to extend this SPSS to prepare other variables for analysis (either those in the small file GOD3 for in the original full ISSP data set -- the SPSS will work with either).

(2) The variables used in our ASR paper are then saved in file WORK3 (described below). This file would be a convenient starting place for further analyses on the issues raised in the paper, or for analyses on a number of related topics.

(3) The full set of analysis runs (and some additional analyses not used in the published paper) are at the end. They can be run either in the full sequence beginning with the raw data (in file GOD3 or the original ISSP data) or, alternatively, directly from the file WORK3).

(4) A note on style: We have heavily commented the SPSS for clarity. The SPSS is set up for SPSS for Windows and should run 'out of the box' after you change the the two 'GET FILE' statements to reflect where you have put the data files on your computer. All statements end with a period (which is why comments use the exclamation point rather than a period); this is right for PC systems but may need changing on UNIX systems – that is not hard and we have run it on both systems.

(5) Requests:
- (a) We have tried to be very careful to catch programming errors, but there is still a real risk of error in any large program. If you find any errors, please let us know (we and others are building on this SPSS for further work).
- (b) If you develop any good SPSS for this data set that you would be willing to share (careful, well documented, and potentially interesting to others) please send it to us and we will, if we like it, include it in the public version, with appropriate acknowledgment. Useful possibilities include: characteristics of denominations -- for example theology, size, wealth; or characteristics of nations in addition to those we have coded (but be careful: with only 18 nations, there are limits to what can sensibly be done).
(6) E-mail: kelley@international-survey.org and n.degraaf@mailbox.kun.nl

View the SPSS: See the file here.

Get the SPSS (variable definitions and full specification of all analyses): Download the SPSS command file by pointing to the following link and telling your browser to download file Relig92P.sp2 -- it is about 40,000 bites in size. Be careful: browsers sometimes want to rename it "Relig92P.EXE" -- if yours does, just change the name back.

Conditions of use: Permission to use this SPSS, and the data file it creates, for non-commercial purposes is freely granted, PROVIDED that use is acknowledged by a bibliographic citation to: Kelley, Jonathan and Nan Dirk De Graaf. 1997. "National Context, Parental Socialization, and Religious Belief: Results from 15 Nations" American Sociological Review 62 (August):639-659.

ANALYTIC DATA FILE

What is the 'analytic' data file? The SPSS first defines all the required variables (as described above). Those variables, together with the original raw (unrecorded) variables, make up what we have called the 'analytic file'.

It contains all the variables needed to replicate our analysis (our SPSS analysis commands are all in the SPSS file, so you can see exactly what we have done) and a number of other variables besides.

What is the 'analytic' data file good for?

Replication. The analytic file allows easy replication or revision of our analysis. All the variables needed are there, in a form ready for analysis, and you have only to specify the new analyses you want.
Extension. The analytic file also allows convenient extension of the analysis we have done -- for example to extend the analysis by running it separately for men and women, or for old and young, or for well-educated and poorly educated.

New analyses. The analytic file is also useful for quite different analyses -- for example, church attendance, or attitudes toward abortion or pre-marital sex.
- For example, most of the variables used in "Moral Reasoning and Political Conflict: The Abortion Controversy". Jonathan Kelley, M.D.R. Evans, and Bruce Headey. British Journal of Sociology 44 (1993): 589-612, are available for a dozen nations, which should provide a useful extension of their analysis of Australian data.

Get the 'analytic' file: Download the file [here]. It is about 1.1 mb. (If your browser renames it, just change the name back.) This file is compressed in "ZIP" format -- there are lots of free "un-zip" programs on the internet if you don't already have one. After you decompress it you will have file "Work3.POR" (about 6 mb). This is an SPSS "portable" (or "export") file. It should run on any standard SPSS system, including both PCs and mainframes. As a check to be sure all is OK, your data should match this:

COUNTRY   Nation                        
Value Label             Value  Frequency
Germany-West =======     1.00      1346 
Germany-East =======     2.00      1486 
Great Britain ======     3.00      1257 
N Ireland ==========     4.00       838 
USA ================     5.00      1359 
Hungary ============     6.00      1000 
Netherlands ========     7.00      1635 
Italy ==============     8.00       983 
Ireland=============     9.00      1005 
Norway =============    10.00      1506 
Austria ============    11.00       984 
Slovenia ===========    12.00      2080 
Poland =============    13.00      1063 
New Zealand ========    16.00      1070 
Australia ==========    18.00      2203 
                                 -------
                        Total     19815

Conditions of use: Permission to use these data for non-commercial purposes is freely granted, PROVIDED that use is acknowledged by a bibliographic citation to: Kelley, Jonathan and Nan Dirk De Graaf. 1997. "National Context, Parental Socialization, and Religious Belief: Results from 15 Nations" American Sociological Review 62 (August):639-659.

ANALYSES

Complete details on the analyses in our article are available:

All the variables used are defined in the SPSS file 'Relig92p.sp2' described above.

The SPSS syntax specifying all the analyses is in Relig92p.sp2 too -- toward the end and reasonably well documented.

The results of these analyses are in the following files:

For tables 1 to 3 -- about 90 kb

For tables 4 & 6, the main non-linear regression runs -- about 100 kb.

For table 5 -- about 6 kb

For appendix table A, lots of alternative non-linear regression models -- about 80 kb.

For appendix table B with OLS for each nation separately -- about 45 kb.

Hits since 8 August Hit Counter

ISC PUBLICATIONS [ Contents | What's New | ISC Home Page]