Method Overview

Data

What data sets are you using?

We are using the 2023 UW RDS (Respondent-Driven Sampling) and UW PSD (Puget Sound Data) datasets.

What did you do to prepare the data?

We prepared the data by:

Tools (Component Specification)

What software packages, modules, etc., did you use?

We used:

What are the dependencies between these and how did you render them interoperable?

Processes

What does your workflow or pipeline look like?

The workflow was divided into several steps:

  1. Function Creation: Fellows created functions to target specific columns and tasks in the data.
  2. Function Integration: All created functions were consolidated into a single script.
  3. Data Cleaning: The script produced cleaned dataframes, including:
    • A clean UW_RDS dataset
    • A clean UW_PSD dataset
    • A clean combined dataframe called UWRDS-PSD for integrated analysis.

What steps did you follow?

Analyses

What approaches did you try that didn’t work? What analyses did you end up sticking with?

Limitations

What are the shortcomings of your approach? How can your work be improved?

Survey Design and Sampling

Survey Design

The survey was designed to capture comprehensive data on the unsheltered population using the following methodology:

Sampling Methodology

The sampling methodology employed includes:

Survey Phases:

  1. Initial Screener: Approximately 5 minutes, interviewer-administered, with a $20 gift card and two Metro bus tickets.
  2. Individual Interview: About 30 minutes, self-administered via tablet, with an additional $10 gift card as incentive.

Eligibility Criteria

Participants were selected based on the following criteria:

Interviews were conducted in English and Spanish, with real-time translation available for other languages.