Featured Post

Homeland Security: The Sworn Duty of Public Officials

Homeland Security: The Sworn Duty of Public Officials     The United States has a unique position amongst the countries of the world;...

Thursday, January 8, 2015

Replication of the Lemiux and Felson Study on Person-Hours/Activity in Relation to Violent Crime Victimization in Austin, Texas

Replication of the Lemiux and Felson Study on Person-Hours/Activity in Relation to Violent Crime Victimization in Austin, Texas

An attempt to replicate the Lemiux and Felson study to determine level of risk for victimization of violent crime localized to Austin, Texas, would not work with the current 2010 Census and (Uniform Crime Reporting) UCR data. The Census data does not include the time/activity data that collected through review of American Time Use Survey (ATUS), and the Part I indexed crimes collated via the UCR do not match the range of crimes labeled as “violent” and gathered through their study of National Crime Victimization Survey (NCVS) data. Lemiux and Felson “Twenty types of violence are included in this analysis, ranging from verbal threats of assault to completed rapes.” (2012, p.p 640-641).
So is there a method to replicate this study otherwise? One category of data for ATUS is under the code METAREA, which reports the metropolitan area in which a household was located. By finding METAREA for the Austin area, and filtering the data, we could use the local ATUS data. 2010 ATUS data can be downloaded from the Bureau of Labor Statistics. (BLS, 2014) Unfortunately, we would not be able to localize the NCVS data, as there is no sorting this data by region; instead, NCVS defines location of residence by urbanicity, or type of urban location (urban, suburban, or rural). (BJS, n.d) We could return to the UCR Table 6 to match the metropolitan area, but then we run into the issue of the indexed crime data that is collected by the UCR versus the range of definitions for violent crime that Lemiux and Felson garnered from the NCVS. We next turn to the Austin Police Department (APD, and find their reported crime statistics. (APD, n.d.) This returns us to the same issue in reporting that using UCR data presents; APD reports the Part I indexed crimes by their own category, and non-indexed crimes as the total of all non-indexed crime, preventing us from replicating the results of the study of Lemiux and Felson.
However, we can approximate their risk assessment for time activity for the indexed crimes reported by either APD or the UCR, keeping in mind that “Any differences between offenses reported in this[APD's] report and the Uniform Crime Report are due to differences in time of report, reporting requirements, and the inclusion of unfounded cases.”(APD, n.d., para. 6). In fact, with not only that consideration but that the UCR Table 6 represents data from the metropolitan area, not not just the City of Austin, we then would use the UCR Table 6 data as the numerator data, and the ATUS data filtered by the METARE A code for the Austin region for the denominator data. Lemiux and Felson discuss their selection of the numerator/denominator on pp. 640-641 for comparison; in our study, we are looking at the UCR Table 6 data, specifying Part I indexed violent crimes for the Austin metropolitan area as the numerator, and the ATUS METAREA data as the denominator.
In conclusion, the violent crime victimization risk assessment that Lemiux and Felson based on time adjusted study cab be replicated for a metropolitan area with the caveat that the violent crime risk will only be assessed for the violent Part I crimes, and not for all violent crimes as defined by Lemiux and Felson.

Appendix 1 – Tables and Charts

Table 1- 2010 Violent Crime by Indexed Crimes
38 Murder and nonnegligent manslaughter
265 Forcible rape
1231 Robbery
2256 Aggravated assault

(Source – Federal Bureau of Investigation, Uniform Crime Reports, 2010

(Source – Federal Bureau of Investigation, Uniform Crime Reports, 2010)

Appendix 2 – Minitab Utilization

To produce a table, and chart from Minitab
1- Input data, either manually, or via cut and paste from the source document (if pasting data into Minitab, be aware of formatting issues and be prepared to delete superfluous rows)
2- Associate data with coding labels; for example, the data value “38” in Table 1, Appendix 1 was entered in the C1 column in Minitab, the label “Murder and nonnegligent manslaughter” was associated in the same row on C2
3- To export the Minitab worksheet into a spreadsheet, select “File->Save Current Worksheet As”, then input filename and the desired format into the option form.
4- To use this data in a Word document, highlight the cells with the desired data, right-click and select “Copy”; in the Word document, move to the “Edit” menu, select “Paste Special”, then select the “calc8” option. Drag the table into the desired layout
5- To create a chart in Minitab, select the “Graph” menu, then choose the type of cart (Area Graph, Bar Chart, or Pie Chart); select the “chart values from a table” option, then assuming your data is in C1, select C1 as the “categorical values” option; and select C2 for the labels option
6-Once the chart has been created, move your mouse to the labels on the chart, edit these labels for clarity
7-To copy the chart into a Word document, right-click the chart and select the “Copy Chart” option; paste into Word and drag the chart into the desired layout
8- To save the worksheet, additional worksheets involved with the data, and all associated charts, select “File → Save Project As”, then select a filename and location.

At this point, the use of Minitab to generate descriptive statistics has not been well illustrated to me; in the case of the 2010 Census data, I felt that the data had to be separated into categories before they provided any useful information/comparison, and only the age category would be better described by the use of descriptive statistics. In the UCR data, I did not feel that generating descriptive statistics for the dataset served any purpose at all, as the data was describing different categories of crime. I don't see any purpose for assigning a mean or standard deviation between the data counting murders and he data counting aggravated assaults.


APD. (n.d.) Crime information. Austin Police Department. Retrieved July 19, 2014 from http://www.austintexas.gov/department/crime-information

APD. (2011). Indexed & Non-indexed offenses by zip code (includes unfounded):01-JAN-10 thru 31-DEC-10. Austin Police Department. Retrieved July 19, 2014 from http://assets.austintexas.gov/police/zipcode/zipcode/indx_nindx_zip_1210.pdf

BJS.gov. (n.d) NCVS victimization anaylsis tool. Bureau of Justice Statistics. Retrieved July 19, 2014 from http://www.bjs.gov/index.cfm?ty=nvat

BLS.gov (2014). American time use survey — 2010 Microdata Files. Bureau of Labor Statistcs.
Retrieved July 19, 2014 from http://www.bls.gov/tus/datafiles_2010.htm

Lemieux, A. M., & Felson, M. (2012). Risk of violent crime victimization during major daily activities. Violence and Victims, 27(5), 635–655. Retrieved July 19, 2014 from http://search.proquest.com.southuniversity.libproxy.edmc.edu/docview/1081338409?pq-origsite=summon

Uniform Crime Reports, Table 6 - Crime in the United States by metropolitan statistical area, 2010. (2010). Federal Bureau of Investigation. Retrieved July 12, 2014 from http://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2010/crime-in-the-u.s.-2010/tables/table-6

No comments:

Post a Comment