161sp00memos

SUBSTITUTE FOR FINAL EXAM IN CJ 161.

DATA FILE. You will use the class survey that is on the web site, and that you downloaded for LAB 6.

DEPENDENT VARIABLE. Your dependent variable is confidence in the criminal justice system. This is a four item index that you built in LAB 6 using the syntax box. A higher score indicates more confidence in the criminal justice system. Take a look at the specific items on the questionnaire so you have a specific idea of what is involved with this general index. To learn more about indexes, take 160, or talk to someone who has. The inter-item reliability for this index, by the way, is quite good - .79.

PREDICTORS. You will do a separate-sample t-test for each of the two following predictors:

NONWHITE
Q58FAMGN

For each variable tell me the null hypothesis you are testing. Tell me your alternate or research hypothesis for each predictor. Always state these in full sentences referring to the specific item. Assume throughout that your alternate hypothesis is always two tailed. Describe in NORMAL English what your null and alternate hypotheses are. Set your ALPHA at p < .20. You need to do this to get sufficient statistical power (to keep your Type II error rate relatively low.)
With a two tailed test your t-critical will be 1.303.

10%

Describe the data you are using ( a few sentences). Presume it is a random sample of all 3rd and 4th year undergraduate cj majors on main campus registered for SP01 semester

10%

Generate the histogram for the dependent variable and tell me if it is normally distributed or not; what specific features of the histogram lead you to this conclusion?

10% X 2 = 20%

Describe null and alternate hypothesis for each of of your predictors

10% X 2 = 20%

Describe the t-test results for each predictor - what is t obtained (always use the smaller of the two listed)? Do you accept or reject the null hypothesis?

20% X 2 = 40%

For each predictor, tell me what the results mean in NORMAL, everyday, DAILY NEWS English

extra: 10

EXTRA CREDIT: only attempt this if you are VERY VERY COMFORTABLE with SPSS and stats; otherwise the risks of serious temporary brain damage are extremely high. Selecting JUST the NONWHITE students, (readup on select if in the SPSS book) run a t-test of the effects of Q58FAMGN on CONFICJS. (5 points for being able to do this.) Interpret correctly (for another 5 points). What is going on here?

DUE: LAST DAY OF FINALS, TUESDAY MAY 8, IN KARIMA ZEDAN'S MAILBOX, NO LATER THAN NOON.
PARAMETERS: Typed double spaced throughout. NO MORE THAN 3 pages. ATTACH ALL OUTPUT. Put your SSN at the top of each page of the text, and at the top of each page of output. BY ATTACHING YOUR SSN TO THE TOP OF EACH PAGE OF OUTPUT YOU ARE CERTIFYING THAT YOU GENERATED THOSE OUTPUT ON YOUR OWN.

10%	Describe the data you are using ( a few sentences). Presume it is a random sample of all 3rd and 4th year undergraduate cj majors on main campus registered for SP01 semester
10%	Generate the histogram for the dependent variable and tell me if it is normally distributed or not; what specific features of the histogram lead you to this conclusion?
10% X 2 = 20%	Describe null and alternate hypothesis for each of of your predictors
10% X 2 = 20%	Describe the t-test results for each predictor - what is t obtained (always use the smaller of the two listed)? Do you accept or reject the null hypothesis?
20% X 2 = 40%	For each predictor, tell me what the results mean in NORMAL, everyday, DAILY NEWS English
extra: 10	EXTRA CREDIT: only attempt this if you are VERY VERY COMFORTABLE with SPSS and stats; otherwise the risks of serious temporary brain damage are extremely high. Selecting JUST the NONWHITE students, (readup on select if in the SPSS book) run a t-test of the effects of Q58FAMGN on CONFICJS. (5 points for being able to do this.) Interpret correctly (for another 5 points). What is going on here?