161sp03_lab8.htm

TO: Students in CJ 161 - Section 3 - SP03
FROM: R. B. TAYLOR
DATE: 4/23/03
RE: LAST PAPER ASSIGNMENT

NOTE: THIS PAPER IS DUE AT THE BEGINNING OF CLASS NEXT WEEK, 5/1/03. In class we are going to be having a QUIZ on the t-test, and be doing some lab work that can get you extra credit so you WANT to be at class on time!!!!!!!!!!!!!
PURPOSE: The purpose of this lab, and the associated paper, is for you to engage in hypothesis testing using the chi squared statistical test. I CANNOT use the data from THIS class but there are too few cases. So, instead, we are going to use the data from another CJ stat class from SPRING 01 semester. They look a lot like this class in many respects, although of course I recognize this group is different too in important ways. But we really need cases for this exercise to work, to I have used this earlier data file.

CLICK HERE >> cj161_sp01_ee.sav to get data file

THE DEPENDENT VARIABLE (Q 47 on your questionnaire, Q41 on the 2001 questionnaire):

Q. 41 Would you favor or oppose a law which required private citizens to register handguns they own?
0 . . . NOT in favor
1 . . . IN favor
9 . . . I have NO opinion on this matter [These are coded as MISSING data - there are 4 such cases.
These should be left OUT of your crosstab.]

You are going to look at ONE of several PREDICTORS to see if one of these does or does not influence this outcome. Here are the predictors:

Q. 58FAMGN
Does any member of your immediate family currently own one or more guns of any type (handgun, shotgun, or long rifle)? THIS IS Q59 ON YOUR SURVEY
0 . . . No
1 . . . Yes

Q1 What is your gender?
0 ... Male
1 ... Female

AFRAM African-American?
0 . . . No
1 . . . Yes

You will carry out a chi squared test, and describe specific steps in your hypothesis testing, and interpret your results given the following scenario (see below) .

THE LAB

Download the 2001 class data file
Q41RGSTR is your DEPENDENT variable or your OUTCOME. THIS IS Q47 ON THE CURRENT SURVEY
Run ONE of the following crosstabs: Q58FAMGN as a predictor OR Q1GENDER as a predictor. Be sure to ask for Chi squared statistic, observed frequencies, and column percentages. Be sure Q41RGSTR is your ROW variable and your predictor is your COLUMN variable. ONLY USE ONE PREDICTOR.
Save your output file; save your syntax file if you paste the commands into the syntax box.

THE SCENARIO
You are involved in an on-campus group that is recruiting volunteers to help with a local campaign urging Congress to mandate registration of all handguns. You are interested in getting some of your fellow cj majors in 161 to help with the campaign. You are going to assume that the results obtained in your class apply generally to CJ majors on main campus. But you do not want to just approach your fellow majors randomly. You want to approach those who are more likely to support handgun registration. You are interested in one of two possible predictors: the gender of your classmate (Q1GENDER) or whether your classmate reports a family member having a gun (Q58FAMGN).

[NOTE: IF YOU WISH, you can imagine a REVERSE scenario - you are working for a group trying to DISCOURAGE Congress from passing a law mandating the registration of handguns, and so you are most interested in identifying those who are most strongly OPPOSED to mandatory handgun registration.]

THE PAPER

PARAMETERS
No more than two pages, typed, double spaced, although references may appear on a third, extra page. SSN only on the top of each page. No names anywhere.
PAGE 1
   1. State the level of measurement for both variables
   2. State the null hypothesis; presume for the purposes of this paper that these respondents are a representative random sample of CURRENT CJ majors on main campus. . Remember - your hypothesis always refers back to the population.
   3. State your significance level, and the corresponding chi square critical statistic. FOR THIS EXERCISE set your alpha level at p < .20, and use as your chi squared critical the value: 1.642
   4. IF YOU USE Q1GENDER: describe the portion (or percent) of men supporting mandatory handgun registration; state the portion (or percent) of women supporting mandatory handgun registration. IF YOU USE Q58FAMGN: state the portion (or percent) of those who DO report a gun currently owned by a family member who support mandatory handgun registration; state the portion (or percent) of those who do NOT report a gun currently owned by a family member who support mandatory handgun registration. CAUTION: you need to be sure your contingency table is set up properly, and that you have requested column percentages.
   5. State the chi squared observed
   6. Based on the difference between the chi squared observed, and the chi square critical, state whether you reject or fail to reject the null hypothesis.
PAGE 2
This will be in the form of a one page (or less) memo. Imagine you are writing a memo to the head of the local chapter of the organization. Describe for him/her the practical implications of your contingency table analysis for the organization's recruiting efforts? Whom should they target and why? In backing up your planned strategy, be sure to link your recommendations to specific results you have seen and the results of your hypothesis testing. Stated differently, if you were able to reject your null hypothesis, you are going to be able to make one kind of recommendation. If you were NOT able to reject your null hypothesis, you are going to have to make a different kind of recommendation.

THE GRADING

10%

Correctly state level of measurement for both variables

10%

Correct statement of null hypothesis

5%

Correctly state alpha level and chi squared critical (this is a no-brainer - it is in the assignment)

10%

Correctly describe both of the requested column percentages (from printout) (5% each)

5%%

Use the correct chi squared observed (from printout)

10%

Correctly accept or reject the null hypothesis

40%

Page 2. A = i) come to correct conclusion based on your results; ii) correctly describe if the null hypothesis is accepted or rejected; iii) state in nontechnical terms what the implications of that are and simultaneously correctly translate the conclusion to the appropriate action strategy; A/B = miss one of the above B/C= miss two of the above

10%

Attach YOUR printout from the lab WITH YOUR SSN WRITTEN ON EACH PAGE OF THE PRINTOUT. By putting your SSN on the lab output you are attesting that you yourself generated the lab output, and that it is not anyone else's; any type of misrepresentation here could raise questions of potential academic misconduct. Be sure to save your disk with your output file on it so you have that for verification in case we need it.

HOW TO PREPARE. Read K&F Chapter 17 (pp. 100-105); B&P: 292-310

10%	Correctly state level of measurement for both variables
10%	Correct statement of null hypothesis
5%	Correctly state alpha level and chi squared critical (this is a no-brainer - it is in the assignment)
10%	Correctly describe both of the requested column percentages (from printout) (5% each)
5%%	Use the correct chi squared observed (from printout)
10%	Correctly accept or reject the null hypothesis
40%	Page 2. A = i) come to correct conclusion based on your results; ii) correctly describe if the null hypothesis is accepted or rejected; iii) state in nontechnical terms what the implications of that are and simultaneously correctly translate the conclusion to the appropriate action strategy; A/B = miss one of the above B/C= miss two of the above
10%	Attach YOUR printout from the lab WITH YOUR SSN WRITTEN ON EACH PAGE OF THE PRINTOUT. By putting your SSN on the lab output you are attesting that you yourself generated the lab output, and that it is not anyone else's; any type of misrepresentation here could raise questions of potential academic misconduct. Be sure to save your disk with your output file on it so you have that for verification in case we need it.