You wish to test the basic assumption that the experience of ever being incarcerated has an impact on one’s current income. In other words, do those who have ever been incarcerated earn more income, less income – or is there any difference at all.
Exercise I
Problem:
For those of you who have taken a course in Corrections, you are probably somewhat aware of the so-called collateral consequences of incarceration – i.e., the costs that go beyond an offender’s direct experience of punishment (e.g., incarceration). A recent executive summary from the Pew Center (Western & Pettit, 2010) documents that one of those presumed collateral consequences of incarceration is long-term economic prospects in the labor market, and social mobility. It is positioned that the incarceration stint is a long-term barrier for those offenders (95%) that are returning to society. As a student of criminal justice and corrections, you decide to test this assumption by using one of the more relevant longitudinal data sets in the field of criminology and the social sciences: The National Longitudinal Study of Youth (NLSY). The NLSY data, collected by researchers at THE Ohio State University since its inception in 1979, represents longitudinal data beginning in 1979 (cohort I) and again in 1997 (Cohort II) specifically examines the long-term labor market consequences of experiences during adolescence and young adulthood. In particular, the NLSY connects data from experiences such as age of first offense/arrest, incarceration, school dropout, to long-term earning potential in the labor market. You have recently acquired a version of the 1997 NLSY data that follows this sample from adolescence (aged 12-17 at time of initial survey) into young adulthood in 2011. Using this data, students will be able to test the assumption that the experience of incarceration has long-term ramifications for an individual’s earning potential and social mobility. The data set – NLSY for Corrections – is ATTACHED!
Instructions:
You wish to test the basic assumption that the experience of ever being incarcerated has an impact on one’s current income. In other words, do those who have ever been incarcerated earn more income, less income – or is there any difference at all. To answer this question, you will be using two variables within the NLSY. The first variable, Incarceration_Prevalence_Dummy is a dummy variable that taps into the lifetime prevalence of incarceration during one’s lifetime (0 = “Never”, 1 = “Yes”. The second variable, Personal Income ranges from zero to infinity.
In particular, you are testing the null hypothesis that the mean level of income for those who have been incarcerated will be THE SAME as those that have not been incarcerated – i.e., the mean difference between these groups will be insignificant. Hint: You are using two variables, so based on Chapter 8 in your SPSS text (23rd edition), you should be familiar with the proper analytical procedure to be employed here.
After performing the relevant analysis, you are to answer the following questions:
1. What is the independent variable here – lifetime incarceration prevalence or current personal income? Dependent variable?
2. What is the level of measurement for the variable Incarceration_Prevalence_Dummy?
a. Is this variable discrete or continuous? Why?
3. What is the level of measurement for the variable personal_income?
a. Is this variable discrete or continuous? Why?
4. After performing the relevant analyses (hint: only one analysis is necessary here), what is your final judgement on the null hypothesis that there is no difference in the mean income of those who have ever been incarcerated versus those who have been incarcerated? How did you make this determination (what test did you perform, and what tests of significance were employed?)?
a. Report the means for both attributes of the incarceration dummy. Is there are a difference? If so, is the difference statistically significant?
5. What does this lead you to speculate with regard to the collateral consequences of incarceration? Are Western and Pettit (2010) on to something here? Why or why not?
6. Can we determine anything causal from these results – i.e., can we assume that the prevalence of incarceration has a causal impact on income, based on the statistical analysis that we performed here?
Exercise II
Introduction:
Several recent studies have found that more than half of all crimes in a city are committed in only a few places. Some criminologists have called these places “hot spots” (Caplan, Kennedy, & Piza, 2013; Carson & Wellman, 2018). Even within crime-ridden neighborhoods, it has been found that crime clusters at a few locations while others remain relatively free of crime. The clustering of violent crime at particular locations suggests there are important features of, or key dynamics at, these locations that give rise to frequent violence. Thus, focused crime-prevention efforts have recently sought to modify these “criminogenic” conditions and reduce violence.
Problem-oriented policing (POP) strategies (similar to community policing) are increasingly used by urban jurisdictions to reduce crime in these “hot spot” crime places. Problem-oriented policing challenges officers to identify and analyze the causes of problems behind a string of criminal incidents. Once the underlying conditions that gave rise to crime problems are known, police officers can then implement appropriate responses to reduce crime. For example, strategies include using community members as information sources to discuss the nature of the problems the community faces, the possible effectiveness of proposed responses, and the assessment of implemented responses. Other strategies target the social disorder problems inherent in these neighborhoods, such as cleaning up the environment of the place and making physical improvements, securing vacant lots, or removing trash from the street.
Instructions:
As the chief of police in Middletown, you are interested in the efficacy of these policing strategies in reducing acts of violence in neighborhoods plagued by high rates of crime. The Middletown PD targets 20 neighborhoods within the city and send out teams of community police officers to implement POP strategies in these neighborhoods. Before the program begins, we obtain the number of arrests for violent offenses (variable pre_pop_arrests) that were made in each neighborhood within the 60 days prior to program implementation. After the program has been in place, we again obtain the number of arrests for violent offenses (variable post_pop_arrests) that were made in each neighborhood for a 60-day period. In this case, having problem-oriented policing in the community is the independent variable, and the number of violent offenses is the dependent variable.
Keep in mind that we are dealing with the same neighborhoods here, we have the number of crimes BEFORE and AFTER the introduction of problem-oriented policing. (Students should know the proper bivariate test to conduct.) We wish to know whether the average number of violent arrests increased or decreased after the police program was implemented. Please keep in mind that we are unsure what the effect of our policing strategy will be – i.e., it might things better, or it might create more crime (this will inform whether we use a directional or non-directional hypothesis)
· Null hypothesis: Mean difference in number of violent arrests pre and post program implementation = 0.
· Alternate hypothesis: Null hypothesis: Mean difference in number of violent arrests pre and post program implementation is significantly different than 0.
The data set – middletown data – is ATTACHED!
After performing the relevant analysis, you are to answer the following questions:
1. What is the independent variable in this analysis?
a. Is this variable discrete or continuous? Why?
2. What is the dependent variable in this analysis?
a. Is this variable discrete or continuous? Why?
3. Did you perform a n independent OR dependent/paired-samples t-test here? Why
a. Did you use a directional (one-tailed) or non-directional (two-tailed) significance test? Why?
4. After performing the relevant analyses (hint: only one analysis is necessary here), what is your final judgement on the null hypothesis that there is no significant difference in the average number of violent arrests before and after the problem-oriented policing program was implemented? How did you make this determination (what test did you perform, and what tests of significance were employed?)?
a. Report the means for arrests for both pre and post problem-oriented policing program implementation. What is the difference?
5. What does this lead you to conclude regarding the relationship between problem-oriented policing programs and violent behavior? Do they seem to work?
6. Can we determine anything causal from these results – i.e., can we assume that the implementation of problem-oriented policing programs has a causal impact on violent crime arrests, based on the statistical analysis that we performed here?
There will be two separate assignment drop boxes: syntax drop box, and Class Project VII drop box. Students should include the syntax files for all procedures used.
Format:
While there is no specific length/page requirement to this assignment, students are asked to answer the questions comprehensively. Assignments MUST be typed, with 1” margins throughout the document. All students must use 12-point, Times New Roman Font. Students should single-space within each answer (including the letters – e.g., 1a, 1b), but double space between answers (e.g., 1, 2, 3).
