130-132
PUTTING IT ALL TOGETHER
Understanding the necessity of measuring race for studying racial profiling in traffic enforcement.
Challenges associated with measuring race arise from conceptualization.
CONCEPTUALIZING RACE
General Concept: Race often perceived as genetic or biological.
Research indicates significant genetic variation within race groups (Winker, 2004).
Self-Designation: Individuals define their race for personal and cultural relevance.
Approach used by the Census Bureau; problems arise when categories do not match heritage (Swarns, 2004).
Hispanic or Latino individuals often select "some other race" due to limited category options.
Observational Appearance: Conceptualizing race based on physical appearance can lead to biases.
Importance of understanding how police might measure race in practice for profiling.
OPERATIONALIZING RACE AND ETHNICITY
Two approaches to operationalize: self-report and observed race.
Most studies rely on observation leading to challenges in identifying certain racial categories, particularly Hispanic drivers.
Different methodologies demonstrated:
Pennsylvania Study: Observers located on roadways (Engel et al., 2004).
North Carolina Study: Teams of researchers as mobile observers.
New Jersey Study (2005): Used digital cameras to photograph drivers.
Photos analyzed by teams to classify race: classified as white, black, Asian, American Indian, or other.
Self-reporting method used on the New Jersey Turnpike:
Drivers exiting had to report race/ethnicity among six choices.
Police recorded race of stopped drivers.
Data summary indicates variations across three measures:
Measure
White
Black
Hispanic
Other
Photo Coding
66%
15%
11%
9%
Tollbooth Survey
52%
29%
10%
10%
Police Stop Data
58%
26%
7%
9%
RELIABILITY AND VALIDITY
Interrater Reliability: Important for observational race coding.
In Pennsylvania, agreement required among observers; only 2.7% had classification disputes.
Digital photographs in New Jersey added reliability but faced challenges.
High-speed image captures resulted in 24% unusable images.
Remaining images required consensus of at least 2 of 3 coders (valid for 89% of usable images).
Discusses Validity in relation to measures of driver profiling.
Appearance of race considered more valid for assessing police stops than self-identified race.
TYPOS
Composite measures can summarize data across multiple indicators into a single numerical score.
Typology: Classifying people by combining measures to create categories is essential.
Example of court experience typology derived from survey questions regarding jury duty and witness testimony.
Complex Typologies: Multiple measures may be combined to track patterns over time.
Rolf Loeber's research: Defined categories of delinquency within longitudinal studies:
Categories encompassed varying levels of delinquency (None, Minor, Moderate, Serious).
Operationalized data into classifications reflecting changes in offense seriousness over time.
INDEX OF DISORDER
Questions the meaning of disorder from Wesley Skogan's reflections.
Explores links between crime, fear, and social issues (public drinking, drug use, litter, etc.).