99

Units of Analysis in Research

  • Important to identify different units of analysis in research questions.

  • Seven distinct units of analysis stated or implied:

    1. Gangs

      • Related to item 1.

    2. Gang members

      • Associated with items 2 and 3.

    3. Jurisdiction (city or county area)

      • Reflected in items 1, 2, and 3.

    4. Street sales of drugs

      • Connected to item 3.

    5. Drug types

      • Referenced in item 3.

    6. Agency

      • Mentioned in item 4.

    7. Special unit

      • Highlighted in item 4.

Quotes from 1997 Survey Report

  • Reflect on quotes from the National Youth Gang Center (1999) report regarding units of analysis:

    • Quote 1: "Fifty-one percent of survey respondents indicated that they had active youth gangs in their jurisdictions in 1997."

      • Analysis: Accurately reflects jurisdiction-level data.

    • Quote 2: "Thirty-eight percent of jurisdictions in the Northeast, and 26 percent in Middle Atlantic regions reported active youth gangs in 1997."

      • Analysis: Correctly presents regional jurisdiction data.

    • Quote 3: "Estimated 30,533 youth gangs and 815,986 gang members active in the U.S. in 1997."

      • Analysis: Provides a quantitative summary of gang activity.

    • Quote 4: "The percentage of street sales of crack cocaine, heroin, and methamphetamine conducted by youth gang members varied by region...."

      • Analysis: Illustrates the relationship between drug sales and gang activity in specific regions.

    • Quote 5: "The majority (66 percent) of respondents indicated that they had some type of specialized unit to address gang problems."

      • Analysis: Offers insight into the existence of specific agencies addressing gang issues.

Evaluating Statement Accuracy

  • Some statements accurately reflect the survey's units of analysis, while others may not.

  • Recognizing misleading statements improves understanding of research context and reliability.

Types of Surveys and Studies

  • Different types of studies provide varied insights:

    • Census Study: Aimed at describing the U.S. population at a specific point in time.

    • National Crime Victimization Survey (NCVS): A cross-sectional study estimating crime victimization over a set period.

    • Cross-Sectional Exploratory Study: Conducted by police departments to understand residents' perceptions of crime sources.

Limitations of Cross-Sectional Studies

  • Inherent issues in drawing causal relationships due to the nature of cross-sectional studies:

    • Measurements are taken at one point in time, limiting the ability to infer causality.

    • Example: Asking about burglary experiences and lock installations samples simultaneity, preventing clarity on causative factors (e.g., whether locks were installed post-burglary or failed to prevent the crime).