CRJ 520 Midterm Exam This exam has two sections – multiple choice and running/interpreting SPSS procedures

CRJ 520

Midterm Exam                                                            NAME ________________________________

This exam has two sections – multiple choice and running/interpreting SPSS procedures. Be sure you complete each section. The exam is worth a total of 35 points. Good luck! Or, as statisticians say, may your deviation be positive!

Section I. MULTIPLE CHOICE (0.4 points each= 12).

Please circle the best response for each question.

1.     You have a data set on crime and justice issues in the United States that has information on all 50 states (i.e., state is the unit of analysis). You have a variable for the total number of convictions in 2023, and another variable for the number of convictions that resulted in a prison sentence in 2023.  You want to create a new variable that measures the percent of convictions that resulted in a prison sentence in 2023. Which of the following procedures in SPSS would be most appropriate?

1. recode procedure

2. percentifier procedure

3. compute procedure

4. confidence interval procedure

2.     Assume you have created the variable for percent of convictions that resulted in a prison sentence referred to in the previous question. As you are studying its distribution you confirm that this is the type of distribution to which the empirical rule applies. You also have calculated the sample mean to be 33. Based on this information, what else must be true?

1. the distribution is positively skewed

2. the distribution is negatively skewed

3. half of all states have a value of less than 33 on this variable

4. the mean is a poor measure of central tendency for this variable

3.     If the deviation for a particular value in a distribution is 2.9, which of the following is the best interpretation?

1. this value is above the sample mean

2. this value is equal to the sample median

3. this value is below the sample mean

4. this value is statistically significant

4.     In running a crosstab procedure, if the independent variable is assigned the columns (i.e., placed across the top) and the dependent variable is assigned the rows (i.e., down the side), which of the following is true?

1. you should request column percentages in SPSS, and compare them down the table

2. you should request row percentages in SPSS, and compare them across the table

3. you should request column and row percentages in SPSS, and compare them to each other

4. you should request column percentages in SPSS, and compare them across the table

5.     If the mean of a distribution is much greater than the median, this indicates that the distribution is positively skewed and that there are extreme values (outliers) at the high end of the distribution.

1. true

2. false

6.     Which of the following is a measure of central tendency that could be used with a variable that represents which gang someone belongs to (1=X-Tabs Gang, 2=Point O-Fivers, 3=Third Street Chi-Squares, 4=does not belong to a gang)?

1. range

2. mode

3. mean

4. all of the above

7.     Which of the following is a reason given for why it is preferable to use non-directional hypotheses instead of directional hypotheses?

1. non-directional hypotheses are easier to state and the null hypothesis usually makes more sense when using non-directional hypotheses

2. directional hypotheses lead to estimates of statistical significance that are too “conservative” (i.e., it is more difficult to achieve statistical significance), and thus we are too unlikely to reject the null hypothesis when we use directional hypotheses

3. your friends and family will likely make fun of you if you use directional hypotheses

4. none of the above

8.     Using a representative sample of police recruits from across New York state, you are conducting a study to examine the predictors of a recruit’s successful completion of the police academy. You want to check for skewness in one of your primary independent variables – civil service exam scores. How would you do this?

1. calculate the sample mean and sample median for civil service exam score and compare them to each other

2. calculate the sample standard deviation and sample mean for civil service exam score and compare them to each other

3. calculate a 95% confidence interval

4. consult a psychic

9.     What is the unit of analysis for the study in the previous question? 

1. civil service exam scores

2. individual police recruits

3. police academy success

4. New York

10.  You are conducting a study on 150 inmates, and one of your variables measures the highest level of education the inmate completed before coming to prison: 1=less than high school; 2=high school diploma or GED; 3=some college; 4=college degree. What level of measurement is this variable?

1. interval-level

2. ratio-level

3. ordinal-level

4. nominal-level

11.  You are investigating the relationship between whether or not someone had access to SPSS as a child and whether or not they served time in prison as an adult. Which of the following is the independent variable?

1. whether or not they had access to SPSS as a child

2. whether or not they served time in prison as an adult

3. crime

4. none of the above

12.  You want to examine the data from an experiment on the effect of a new drug treatment program for parolees. You have a variable that measures the frequency of drug use for each individual for the six months prior to entering the program, and another variable that measures the frequency of drug use for each individual for the six months after completion of the program. Which of the following procedures would be most appropriate to use in analyzing the data from this experiment?

1. independent samples t-test

2. paired samples t-test

3. crosstabs

4. none of the above are appropriate

13.  You are interested in the average number of inmate disciplinary infractions that corrections officers in New York file per month. Based on a statewide sample of corrections officers, you get a sample mean of 5.2, and you calculate a 95% confidence interval of 3.3 – 7.1. Which of the following is true?

1. the true population value is 5.2, and 95% of samples will yield a value between 3.3 and 7.1.

2. there is a 95% chance the true population value lies between 3.3 and 7.1.

3. 95% of the corrections officers in the population file between 3.3 and 7.1 disciplinary infractions per month.

4. we are 95% confident the true population value is between 3.3 and 7.1

14.  Which of the following is a measure of dispersion?

1. cumulative percent

2. median

3. range

4. none of the above

15.  In which of the following situations do you reject the null hypothesis?

1. when the p-value is less than or equal to .05

2. when the chi-square is less than 0

3. when the standard deviation is greater than +2.5

4. you never reject the null hypothesis – you either accept or fail to accept it

16.  When do you need to be cautious in your interpretation of chi-square results?

1. when more than 20% of the cells in the crosstab have an expected frequency less than five

2. when more than 20% of the cells in the crosstab have an expected frequency greater than five

3. when the chi-square itself is greater than .05

4. when any cells in the crosstab have an expected frequency less than five

17.  You are interested in the relationship between whether or not someone graduated from high school (1=no; 2=yes) and the number of times they have been arrested in adulthood. Which of the following procedures could you use to examine this relationship?

1. independent samples t-test

2. paired-samples t-test

3. confidence intervals

4. none of the above

18.  Which of the following is NOT one of the common errors in statistical analysis?

1. using an inappropriate statistical test

2. confusing statistical significance with substantive significance

3. basing an analysis on a sample that is too small

4. having too many null hypotheses

19.  Based on the empirical rule, which of the following is true?

1. in a normal distribution, 68.26% of cases will be one standard deviation above the mean

2. in a normal distribution, 99.74% of cases will be three standard deviations below the mean

3. in a normal distribution, 31.74% of cases will be further than one standard deviation away from the mean (i.e., above or below)

4. in a normal distribution, 95.46% of cases will be twice as large as the mean

20.  You are conducting an analysis to examine whether or not there is a relationship between a person’s political orientation (liberal or conservative) and their support for capital punishment (yes or no). If you are using non-directional hypotheses, which of the following would be an appropriate alternative hypothesis in this case?

1. a person’s political orientation is related to his or her support for capital punishment

2. liberals will be less likely to support capital punishment

3. conservatives will be less likely to support capital punishment

4. there is no difference between liberals’ and conservatives’ support for capital punishment

21.  Which of the following is NOT one of the properties of the normal distribution?

1. the distribution is symmetric

2. the empirical rule applies

3. 15% of all values are outliers

4. the distribution has only one mode

22.  You are conducting a study of involvement in neighborhood watch groups among a sample of 500 Western New York residents. You measure an individual’s participation in these groups by recording the actual number of meetings he or she attended over the past year. What is the level of measurement of this variable?

1. nominal

2. interval

3. ratio

4. ordinal

23.  You run a contingency table to examine the relationship between gender and support for capital punishment (1=no; 2=yes), using a sample of 300 Niagara University students. If you want to know what percent of all students in your analysis, regardless of gender, support capital punishment, what statistic would you use?

1. marginal percents

2. chi-square

3. the expected count

4. the standard deviation

24.  We know that chi-square compares the EXPECTED and OBSERVED counts in a contingency table. What do the EXPECTED counts refer to?

1. the number of cases you would expect in each cell if the alternative hypothesis is true

2. the number of cases you would expect in each cell if the null hypothesis is true

3. the number of cases you would expect in each cell if chi-square is statistically significant

4. the number of cases you would expect in each cell if you had a normal distribution

25.  You are analyzing data from an experiment to examine whether boot camps lead to lower recidivism than prison. Your two variables are the offender’s sentence (1=boot camp; 2=prison), and whether or not the offender was rearrested in the year after he or she was released (1=no; 2=yes). You run a crosstab and find that 25% of boot camp offenders were rearrested, while 28% of prison offenders were rearrested. The actual chi-square = 2.37, and the p-value=.14.  Based on this information, what is the best conclusion? 

1. reject the null hypothesis, and conclude there is a statistically significant difference between boot camps and prison in terms of recidivism

2. accept the null hypothesis, and conclude there is not a statistically significant difference between boot camps and prison in terms of recidivism

3. fail to reject the null hypothesis, and conclude there is not a statistically significant difference between boot camps and prison in terms of recidivism

4. accept the alternative hypothesis and conclude there is a statistically significant difference between boot camps and prison in terms of recidivism.

26     A criminologist compares average stress levels of correctional officers in two different prisons. The officers in each prison are different individuals.

Which statistical test is most appropriate?

1. Paired-sample t-test

2. Independent-sample t-test

3. Chi-Square test

4. One-sample t-test

27.  A study comparing average fear-of-crime scores between men and women reports:

t(98) = 2.45, p = .016

What is the correct interpretation?

1. There is no significant difference between men and women.

2. Women and men differ significantly in fear-of-crime scores.

3. The difference is large but not statistically significant.

4. The sample size is too small to interpret.

28. A study comparing average fear-of-crime scores between men and women reports:

t(99) = 2.65, p = .061

What is the correct interpretation?

1. There is no significant difference between men and women.

2. Women and men differ significantly in fear-of-crime scores.

3. The difference is large but not statistically significant.

4. The sample size is too small to interpret.

29.  A study examines whether type of neighborhood (urban, suburban, rural) is related to type of crime reported (property vs violent crime).

Chi-Square results:
χ²(2) = 9.85, p = .007

What is the correct interpretation?

1. Neighborhood type and crime type are independent.

2. There is a significant relationship between neighborhood type and crime type.

3. Crime type causes neighborhood differences.

4. The expected frequencies are too small.

30.  In a normally distributed dataset, approximately 95% of the observations fall within which range?

1. Mean ± 1 standard deviation

2. Mean ± 2 standard deviations

3. Mean ± 3 standard deviations

4. Mean ± 4 standard deviations

Section II. Running and Interpreting SPSS Procedures. There are FOUR questions in this section. Be sure you answer ALL of them. (The number of points each item is worth is listed next to the item).

IMPORTANT: To make sure you use the correct version of each data set, use the ones that are in the “Midterm Exam Data Sets” folder on Canvas.

Question #s 1 and 2 rely on the “felons on probation” data set.

1.     Run frequencies on the following variables and answer the associated questions:

1. v122 – # prior felony arrests

   1. What cumulative percentage of offenders had fewer than three prior felony arrests? (1 points)

       

2. v39 – community service ordered? (yes/no)

   1. Only considering those for whom we know this information (i.e., non-missing cases), was community service ordered for most probationers? (1 points)

       

   2. What is the raw number of offenders for whom community service was ordered? (1 points)

       

3. v150 – any arrests while on probation? (no/yes)

   1. What percentage of offenders, for whom we know this information (i.e., non-missing cases), were arrested while on probation? (1 points)

2.     Run a contingency table (crosstab) to find out the following:

1. Of those offenders for whom community service was ordered, what percent were arrested while on probation? (2 points)

    

2. Of those offenders for whom community service was not ordered, what percent were arrested while on probation? (2 points)  

    

3. Run a chi-square analysis to determine whether or not these results are statistically significant. Report the chi-square value and p-value, and state whether or not the results are statistically significant. (3 points)

Question #3 relies on the “weapons related murders 2011” data set.

3.     Create a variable to measure the percent of murders that were committed with a rifle and answer the following questions.

1. Run a frequency on this new variable. What are the minimum and maximum values? (1 points)

   Minimum                     ____________

   Maximum                    ____________

2. What are the mean, median, and standard deviation for this variable? (1 points)

   Mean                           _____________

   Median                        _____________

   Standard Deviation      _____________

3. Run a 95% confidence interval on this variable and answer the two questions below. (2 points)

   1. Confidence Interval = _________  to  ___________

   2. What does this tell us about the true population mean?

Question #s 4 relies on the recidivism on parole data set.

4.     Using the appropriate statistical procedure, examine whether or not there was a statistically significant change in the offenders’ behavior from before to after incarceration. Use the variables for total number of arrests in the first year prior to incarceration (tot1pre) and total number of arrests in 12 months after release (toff12pos) as your two variables.

1. If you are using non-directional hypotheses, what are the null and alternative hypotheses? (1 points)

2. What is the mean for each variable? (1 points)

3. What is the t-score and p-value? Are your results statistically significant? Did offenders’ behavior get better or worse after incarceration, or was there no change? (2 points)  

Question #s 5 relies on the recidivism on Criminology dataset. Conduct independent sample t-test and answer the followings:

1. Use SPSS to calculate independent sample t-test between Gender and GPA. (2 point) 

2. Us APA format to report the results as you learned in class. (2 point)