6
Methods
The sample used for this study consists of approximately 600 NCVS sexual victimization
incidents collected from years 1997 through early 2000
6
. These incidents are based on “raw
data” from the NCVS incident report questionnaire pages stored at the Bureau of Justice
Statistics Offices. Such pages contain narrative summaries and approximately 25 additional
closed-ended responses (approximately 55 variables for those collected by Computer Assisted
Telephone Interviewing). Each set of questionnaire pages also contain a “header” that contains
manually added variables from the initial screener questionnaire which include victim’s age, sex
and most importantly, a transcribed short-form narrative. The screener narrative is less detailed
than the incident report summary, but often provides additional and sometimes contradictory
detail for comparison.
The findings are based on both a quantitative and qualitative analysis of narrative data.
Analyzing narrative data quantitatively -- converting qualitative detail into numerical codes for
statistical analysis -- helped to establish the overall frequencies of sexual victimization incidents
and key contextual variables. A qualitative exploration of the narratives was used to uncover any
additional situational details and unanticipated patterns found within victims’ descriptions of
incidents. Qualitative findings are integrated into the following sections to illustrate and
underscore key quantitative findings.
6
As of this writing, I am currently in the process of completing the coding and analysis of all of these cases. The
actual findings presented in this paper are from a preliminary coding of 126 cases. Findings will be updated prior to
the summer conference based on results from the complete data set of 600 cases from 1997 - 2000.
Convention