Study Guide: Research on Homelessness and Youth Programs

Within-Subjects Design

Same participants experience all conditions.

Statistical Power

Reduces variability from individual differences.

Fewer Participants

Efficiency in recruitment and analysis.

Maturation

Participant changes over time (e.g., longitudinal adolescent studies).

Regression

Extreme scores normalize.

Selection

Pre-existing differences affect outcomes (e.g., quasi-experimental studies).

Attrition

Dropout affects results (e.g., longitudinal designs).

Instrumentation

Changes in measurement tools.

Testing

Repeated testing changes outcomes.

History

External events impact results.

Paired t-Test

Related groups (siblings, twins, within-subjects designs).

Independent-Sample t-Test

Separate groups (between-subjects designs).

Factorial Design

Simultaneous testing of 2+ independent variables.

2x2 Design

Indicates 2 IVs, each with 2 levels.

Social Problem Solving (SPS)

Study by Toro et al. (1989) with matched 86 children with LD and 86 without, ages 7-11.

Behavioral Approaches to Weight Loss

Study by Twentyman et al. (1981) with 70 married men and women, 15% overweight.

Peer Advocacy vs. Psychiatric Rehabilitation

Study by Toro et al. with a 2x2 factorial design.

Homelessness and Substance Use

Study by Toro et al. (1995) with groups of currently homeless, previously homeless, and never-homeless poor adults.

HALO Project

Study by Toro et al. (1997-2008) comparing 250 homeless adolescents and 148 housed peers.

Community Placement of Mentally Ill Patients

Study by Christenfeld et al. (1985) with 113 patients releasable within 3 months.

Factorial Design Table Example

2x2 Design with IV1: Spouse Involvement (yes/no) and IV2: Buddy Program (yes/no).

Threats to Validity Mnemonic

MRS. HIT (Maturation, Regression, Selection, History, Instrumentation, Testing).

DPATH1

1992-94 survey with 297 participants across 41 sites.

DPATH2

2000-02 survey with 220 participants across 29 sites.

DPATH3

2019 study interrupted by the pandemic in 2020.

Interviews Length

3-4 hours.

Age Distribution 2000-02

42% aged 40-49.

Age Distribution 1992-94

27% aged 18-29.

Duration of Homelessness

Majority experiencing homelessness for less than a year.

Health Symptoms

Participants with 30+ symptoms rose from 6% to 13%.

Mental Illness

Increased prevalence from 27% (1992-94) to 35% (2000-02).

Substance Abuse

Remained high at 74% (1992-94) and 71% (2000-02).

Dual Diagnosis

Approximately 26% of participants had dual diagnoses.

Social Network Index (SNI)

Smaller family and friend sizes in 2000-02.

Client Data vs. Key Informant Data

Comparison of needs between clients and informants.

Top Client Needs

Permanent housing, medical services, job placement prioritized.

Top Informant Priorities

Transportation and affordable housing ranked higher.

Youth Homelessness Demonstration Project (YHDP)

Aims to coordinate care for youth homelessness.

HUD Funding

First round in 2017 for 10 communities.

YHDP Objectives

Prevent youth homelessness and build partnerships.

Evaluation Design

Longitudinal, multi-site comparative case study approach.

Evaluation Questions

Focus on community implementation and youth perceptions.

Homeless Adult Samples

Older participants showed increased mental health concerns.

Youth Inclusion

Effective strategies involve youth in decision-making.

Program Development

Systematic process for improving human services.

Needs Assessment

Identifies service needs at early development stage.

Qualitative Methods

Focus groups and interviews for data collection.

Program Monitoring

Tracks ongoing processes through management systems.

US Context

14-40% of homeless adults previously in foster care.

Michigan Aging Out Study

816 cases studied; 50% faced homelessness.

BRIDGE Program

Prevents homelessness for youth aging out of care.

Poland Pilot Evaluation

BRIDGE youth had zero homelessness cases.

Psychological Distress

High levels of distress reported among youth.

Outcome Evaluation

Examines program impact and effectiveness.