No dataset is perfect, but good data helps improve the good work already being done to help people experiencing homelessness across Washington. Data is important, and there is a report for that. It is the Data Quality Report (DQR).

What is the DQR?

The DQR reviews data quality for many HMIS data elements and helps us understand how our programs and projects are doing with data quality. For an overview of the DQR, check out this article.

Why is the DQR important?

The DQR is important because it helps to identify errors in your HMIS projects. Remember to run the report for all projects and then rerun it after corrections are made. Some projects are at a higher priority than others.  For example, prioritize Permanent Housing (PH), Emergency Shelter (ES) and Transitional Housing (TH) projects first. Then focus on Homelessness Prevention (HP), followed by any additional projects. This order can be different depending on your funding.

Once identified, errors from the report can be used to correct the HMIS records that contain errors. After corrections are made, rerun the report and see the progress made to improve your agency’s data quality scores.

A bonus of handling data issues proactively is that fixing the data quality issues you notice in the DQR will help us prepare for the Longitudinal System Analysis (LSA). The LSA is the third component of the Annual Homeless Assessment Report (AHAR). Commerce submits this report for the entire Balance of State each year.

Pro Tip: Run the DQR monthly or quarterly to understand data quality issues better and make fixes more manageable over time.

A few important items from the DQR to check and resolve:

Q1. Report Validation Table

• Does the Total number of persons served seem reasonable?
• Are there unexpected household types for this project/component?
• The correct number of persons with unknown age?

Q2 Personally Identifiable Information (PII)

• For non-Victim Service providers: The average consent refused rate is 15-20%. If the score for Name, SSN and DOB is greater than that (or much lower), check on data entry and consent procedures.
• The error rate for 4 Race, 3.5 Ethnicity and 3.6 Gender shouldn’t exceed 5%.

Q3. Universal Data Elements

• Correct 10 Project Entry Date, 3.15 Relationship to HoH or 3.16 Client Location errors. Check into Veteran Status or Disabling Condition errors and correct if possible.

We hope this article helps you love your data a little more!

* * * * *

Are you interested in learning more Excel tips and running reports? Check out other articles on the HMIS blog.

Want to sign up for our newsletter? Go to the HMIS website, scroll down until you see the ‘HMIS E-mail Updates’ heading in the right column, add your email address, then click ‘Submit.’