405fa01_hom1112

TO: Students in CJ 405
FROM: R B. Taylor
DATE: 11/6/01
RE: Homework for the rest of the semester

READ THIS PARAGRAPH CAREFULLY A COUPLE OF TIMES AND DO NOT FREAK OUT AND DO NOT SEND ME EMAIL BUT IF YOU DO WANT TO CHAT CALL ME ESPECIALLY IF YOU ARE NOT CLEAR ON HOW TO GET STARTED. FOR THOSE OF YOU WHO DO NOT GET TO CAMPUS IN BETWEEN CLASSES YOU MAY HAVE DIFFICULTY GETTING SOME OF THE READINGS SO CALL ME AND WE CAN CHAT ABOUT ONLINE SEARCHING THROUGH CJ ABSTRACTS AND PROQUEST DIRECT. I outline below a process that will allow you to pull together a multiple regression paper using the GIA data over the next few weeks. It is a lot of work. It means the next few homework assignments are going to be harder than perhaps you had planned on. It also means I may be "weighting up" these assignments and "weighting down" the earlier homeworks. So there may be some discussion about what is suggested here. So I am leaving this somewhat open in the following sense. If you wish to work on these assignments week by week, and not pull together the final paper, you can do so and take the final rather than submit a paper. I will grade the assignment each week based on what I think it would be reasonable to expect to accomplish within a week. Or you can say to me the proof is in the pudding, let me pull together the big paper, and talk about it some, and forego the final. If you do opt to pull all of this together then the weekly grade you receive for each piece is superceded by the quality of that piece in the final product.

I still want you to keep up with the reading about additional related topics (as listed on original syllabus) but I am probably going to present that in the terms of: here are additional analyses you need to think about in your projects.

DEPENDENT VARIABLE

As we work toward the end of the semester, you are going to be working, among other things, on putting together a series of homeworks in which you investigate the determinants of the main outcome that has been of interest to us in the GIA data file, the size of the working firearm collection in gunowning households, and/or a related outcome. Here are the outcomes that you could examine:

NGUNS2 (Q6, transformed)
KIDGUNDU (dummy variable (0,1); 1=got gun as a kid)
PARENDUM (dummy variable (0,1); 1=parents had one or more guns in household while respondent was growing up)
HANDGDUM (dummy variable (0,1) 1= have a handgun)
CARRYDUM (dummy variable (0,1); 1=carry gun for protection)

For all the dummy variables, see how the compute statements work below so that you are sure you understand the outcome.

A couple of points:

if you are working with NGUNS2 as an outcome variable, you will want to do parallel analyses of Q6 so you can get a close "feel" for what is happening with the UNTRANSFORMED variable; that is, you can understand how the slopes affect plain old number of guns.
if you are working with any of these dummy variables as outcomes, it means you will need to be doing logistic regressions, since these are binary outcomes. NO big deal. There are coefficients and t tests just like in plain old regression, but it may mean you have to work a bit harder to understand what is going on. But I know you guys can do this.

So your first step is to pick the outcome that is of interest to you. NGUNS2 is extremely interesting because no one has done much work on this before. CARRYDUM is interesting because you can link it to the literature on guns and protection. HUNTDUM and PARENDUM and KIDGUNDU are interesting because they link to involvement in the gun culture.

THEORETICAL GROUNDING

Second step is to dig into the literature; see at end of assignment for some suggested starting references; follow up with the social science citation literature to build your literature review.

Third step is to begin to develop your list of predictors. That is, given what you have read, you want to begin to describe what predictors link to what outcomes for what reasons. In other words, you want to state null and alternate hypotheses, and rationales, and tie those rationales to specific literature.

PREDICTORS

Those I want you to "force" in as control variables

You want to be sure to include in your models the following predictor variables:

Concept (indicator)

Gender (FEMALE)

Age

Education level (Q106)

Marital status (MARRIED)

Ethnicity (AFRAM or NONWHITE; NOT both)

Political orientation (Q107)

Rural vs. urban (Q101 or Q100; NOT both; check for collinearity with Q107)

Other ones to consider

You may or may not want to include, depending on your theoretical orientation, and your outcome, additional variables such as:

Involvement in the gun culture (GUNCULTR or GUNCULT2)

Number of kids or young kids in the household

Veteran status

Training received

Being a victim of a robbery or attack (VICTMDUM)

Whether or not they hunt (HUNTDUM)

Whether or not they own a gun (OWNGDUM)

Please do NOT go beyond the list of above variables in selecting your predictors without chatting with me first.

WHAT I AM EXPECTING WHEN

11/14	WHAT YOU ARE WORKING TOWARD: Statement of your theoretical model that will describe the outcome you have chosen, and a list of the predictors. For each predictor state the null and alternate hypotheses, being careful to state these correctly; be sure to include descriptions of rationales that refer to processes. Be sure to have literature citations. WHAT I WANT BY THIS DATE: statement of your outcome; description of it; completing this process for as many predictors as you can in the time allowed.
11/21	WHAT YOU ARE WORKING TOWARD: Completing the regression results, and correctly interpreting the impacts of each predictor. WHAT I WANT BY THIS DATE: Printout with all the relevant diagnostics showing your results; discussion of as many predictors as you can muster by this date.
11/28	WHAT YOU ARE WORKING TOWARD: Interpreting your regression results in the context of the relevant literature. WHAT I WANT BY THIS DATE: A brief statement about how your findings on each predictor fit or do not fit with the literature to date.
12/3	Those doing a "big" paper do short presentation on what they know to date
12/10	Those doing a "big" paper turn it in; those doing the final take it

* BACKGROUND ON HOW TO COMPUTE THE DUMMY VARIABLES ON YOUR OWN
* YOU SHOULD BE ABLE TO JUST CUT AND PASTE THIS INTO A SYNTAX BOX
* AND CREATE THE NEW VARIABLES. REMEMBER TO SAVE THE NEW FILE WITH A
* DIFFERENT FILE NAME
* .

COMPUTE HANDGDUM=0 .
COMPUTE RIFLEDUM=0 .
COMPUTE HUNTDUM=0 .
COMPUTE CARRYDUM=0 .
COMPUTE VICTMDUM=0 .
COMPUTE PARENDUM=0 .
COMPUTE OWNGDUM=0 .
COMPUTE KIDGUNDU=0 .
COMPUTE TRAINDUM=0 .
COMPUTE ARMYDUM=0 .
EXECUTE .
IF ((Q23 GE 1) AND (Q23 LE 3)) HANDGDUM=1 .
IF ((Q23 GE 4) AND (Q23 LE 7)) RIFLEDUM=1 .
IF (Q31=1) HUNTDUM=1 .
IF (Q37=1) CARRYDUM=1 .
IF (Q51=1) VICTMDUM=1 .
IF (Q92=1) PARENDUM=1 .
IF (Q93=1) OWNGDUM=1 .
IF ((Q94 GE 5) AND (Q94 LE 21)) KIDGUNDU=1 .
IF (Q95=1) TRAINDUM=1 .
IF ((Q108=1) OR (Q108=2)) ARMYDUM=1 .
EXECUTE .
COMPUTE GUNCULTR=MEAN.2(HUNTDUM,CARRYDUM,PARENDUM,OWNGDUM,KIDGUNDU,TRAINDUM,ARMYDUM) .
EXECUTE .
*******************
* ALSO DOING A SECOND GUNCULTURE VARIABLE THAT LEAVES OUT OWNING AND CARRYING
******************
* .
COMPUTE GUNCULT2=MEAN.2(HUNTDUM,PARENDUM,KIDGUNDU,TRAINDUM,ARMYDUM) .
EXECUTE .

REFERENCE SECTION

You already have read the piece (click here to get the adobe file) by Ludwig on gender bias in reporting of gun collections. Re-read it, see if there are some relevant references. Below are some further references on guns and gun ownership. The 1980 piece by Lizotte and Bordua is very important and useful and highly recommended.

Azrael, D. and D. Hemenway (2000). "In the safety of your own home: Results from a national survey on gun use at home." Social Science and Medicine 50: 285-291.

Cook, P. and J. Ludwig (1995). Guns in America . Washington , DC , Police Foundation.

Decker, S. and S. Pennell (1995). Arrestees and guns: Monitoring the illegal firearms market. NIJ Research Preview. Washington, DC, National Institute of Justice.

DeJong, C. (1997). "Differential indicators of defensive gun ownership: A Comparison by gender." Journal of Criminal Justice 6: 517-526.

Kleck, G. (1998). "What are the risks and benefits of keeping a gun in the home?" Journal of the American Medical Association 280: 473-475.

Lizotte, A. and D. Bordua (1980). "Firearms ownership for sport and protection." American Sociological Review 45: 229-244.

Lizotte, A., G. Howar, et al. (1997). "Patterns of Illegal Gun Carrying Among Young urban Males." Valparaiso University Law Review 31: 375-393.

Marciniak, L. and C. Loftin (1991). "Measuring Protective Handgun Ownership." Criminology 29: 531-540.

O'Brien, S. P. and D. P. Haider-Markel (1998). "Fueling the fire: Social and political correlates of citizen militia activity." Social Science Quarterly 79: 456-465.

Roth, J. (1994). Firearms and Violence. Washington D.C. , National Institute of Justice.

Smith, T. W. and R. J. Smith (1995). "Changes in firearms ownership among women: 1980-1994." Journal of Criminal Law and Criminology 86: 133-149.

Wright, J. D. (1995). "Ten Essential observations on Guns in America ." Society march/april 1995: 63-68.

Young, J. T., D. Hemenway, et al. (1996). "The polls - trends: Guns." Public Opinion Q. 60: 634-649.