Please see attachments
Your job in this project is rather simple: to operationalize the two central variables in this
research – combat experience and serious violent offending – by creating an index/scale for each variable. To accomplish this task, you will be using data from Wave IV (2008) of the National Longitudinal Study of Adolescent Health (Add-Health), collected by social scientists at The University of North Carolina – Chapel Hill. I have uploaded the relevant codebooks to Canvas (Military Experience and Violent Offending and Victimization), along with the data. Students will be accessing the data directly from the ICPSR website, via the UNA Social and Behavioral Sciences Research Center (we will go through this process together). Once the data has been downloaded, students will be asked to construct and compute two scales in order to
operationalize the independent and dependent variables: (1) a five-item combat experience
scale, which includes a number of relevant combat-related measures (shooting/killing someone
in combat, the experience of being wounded/injured during combat, and witnessing individuals
being wounded or killed); and (2) a scale of serious violent behavior. I will provide students
with considerable assistance in constructing the combat experience scale, but students will be
expected to create their scale of serious violence on their own. Fortunately, The Add Health
has arranged the five combat variables in consecutive order in the Military Codebook (section
10). The five variables comprising the scale are as follows:
R_H4MI14: During your combat deployment have you ever killed, or thought you killed, someone
R_H4MI15: During your combat deployment, were you wounded or injured?
R_H4MI16A: During combat deployment, did you see a friend/ ally wounded, killed, dead?
R_H4MI16B: “” an enemy combatant?
R_H4MI16C : “” a civilian?
Students are simply asked to
create a summated scale of these five items
(TRANSFORM/COMPUTE command within SPSS), entitled COMBAT_EXP. The Variable
Label should be “five-item summated scale for combat experience”. Since the variables are
all measured on the same metric, more or less, there will be no need for performing any recodes on your initial scale.
VERY IMPORTANT: REMEMBER THAT WHEN COMPUTING SCALES AND RECODING VARIABLES STUDENTS SHOULD PASTE ALL OF THEIR PROCEDURES INTO A SYNTAX FILE IN SPSS.
As for your scale on serious violence scale/index, students are to go through the Section 21 Add-
Health Codebook (Violent Offending and Victimization – see uploaded file in Canvas), and
select the items that they think should comprise the serious violent offending scale (Hint: I
think there are six items that “can” be used here, but students are to make their own choices,
and defend them). Please keep in mind that some of these items are on different metrics, (two are dichotomous, the others are on a four-point scale), so recodes may be necessary, as to ensure that we are comparing “apples to apples”. In the event that students need to perform recodes on some of these measures (four of them), I would recommend dummy-coding all scale indicators, as to where the values indicate (0 = have not committed the act at least once in the past year, and 1 = have committed the act at least once in the past year the following). This can be done via the following steps:
recode these items into different variables (
SPSS: TRANSFORM/RECODE INTO DIFFERENT VARIABLES) and change the OLD values (2-3) into NEW values (1). Copy the OLD values 0 and 1. Remember, I recommend renaming your newly recoded variables with the “R” designation.
Therefore, all of the indicators that comprise your soon-to-be created scale will have the values 0 or 1. When creating the summated violence scale in SPSS, I recommend using the following name: SERIOUS_VIOLENCE. I would label the variable “summated scale for serious violent behavior”. Again, students have the freedom to choose the indicators that they prefer, as long as they can defend their choices.
When the scales have been constructed, students are encouraged to run a frequency distribution of their scales, and see if anything looks odd (e.g., are some of the attributes not represented in your scales, and may be in need of being collapsed/merged into another attribute/value?).
Lastly, students are to perform a reliability analysis on their newly-created scales by performing
the following procedures in SPSS: ANALYZE/SCALE/RELIABILITY ANALYSIS. Students use the proper statistics to measure reliability (as indicated in the TWO classroom demonstrations).
After performing these procedures, students are asked to answer the following discussion questions, which will be used as the primary means of assessment for class project V.
1. Why were some of the indicators of your violence scale recoded? Does this choice make sense? Why or why not?
2. Run a frequency distribution of your two newly-minted scales. Do any of your scales need to be recoded? Why or why not?
a. If your scales need to be recoded (I think at least one of them might), how would you recode them and why?
3. Comment on the
reliability of your two scales. In particular, are the two scales reliable?
a. What procedure in SPSS did you use to make this determination?
b. After performing the procedure, how did you come to the conclusion that your measures were reliable, or lacking in reliability? In other words, what threshold did you use to make this determination?
4. Comment on the validity of these two measures. Do you think that you have accurately captured combat-related PTSD, and serious violent behavior? Why or why not (be prepared to defend your choices)?
a. What procedure did you use to make this determination?
b. Are there statistical procedures that can assess validity?
5. Discuss the process by which you moved from the abstract (what does violence really mean, man? What is post-traumatic stress), through conceptualization and operationalization to the final point of real-world measurement. Discuss the “science” behind the procedures you performed, while also giving some attention to the position that this science may still be viewed by some as being “soft”.
· Students should include the syntax files for all procedures used. As a matter of fact, while students will not receive a grade on their syntax submission, the syntax file is of equal importance to your assignment submission, as I will use your syntax file to determine if you have performed all procedures properly. Only the class project VIII assignment drop box will actually be evaluated.
