Data Analysis Tool For All

yellow, green, and blue angled lines

Comparison

Description

Definition: Comparison analysis examines relationships between two or more variables to uncover insights about client conditions or data linkages.
Purpose: Highlights disparities, disproportionalities, differences, relationships, or conditions requiring further investigation.
Examples:

  • Comparing service rates by gender or ethnicity.
  • Analyzing links between poverty levels and case types.

Key Insight: When unexpected differences arise by demographics or legal problem, investigate to understand the data relationships and determine whether advocacy or services targeting multiple conditions are needed.

Example Data Question

How does the racial/ethnic makeup of my organization’s intake compare to the racial/ethnic makeup of the eligible population (less than 200% Poverty Level) in my service area? Are there any racial/ethnic groups who are underrepresented in our intakes?

Recreate This Analysis

Data Sources

U.S. Census Bureau’s American Community Survey (ACS) Public Use Microdata Sample (PUMS) via the Microdata Access Tool Microdata - Census Bureau Datasets (https://data.census.gov/app/mdat/).

Intake data from your case management system:

  • Fields:
    • Demographics about which you are curious, such as Age at Intake, Gender, Race, Ethnicity, With Disabilities, Veteran, Percentage of Poverty, Number in Household, Citizenship Status, Language, Living Arrangement, County, etc.
    • Dates: Date of Earliest, Intake Date, Prescreen Date, Open Date, Close Date, Date of Rejection, and Date of Birth (for use in an age-related formula).
    • Case Information: Legal Problem Code, Problem Code Categories, Close Reason, Rejection Reason, Disposition, Intake Type, etc.
  • Filters/Report Structure:
    • Date Filters: Date of Earliest in last year, or Intake Date or Prescreen Date in last year.
    • Exclude/Filter Out: Test or fake cases and clients and duplicate cases using whichever fields your organizations uses to identify these cases and clients, such as Rejection Reason, Client Name, Case Status, Funding Code, etc.
    • Include: Cases that were closed with service, rejected without service, and remain open.
    • One Row/Record Per Case: Ensure that downloaded data includes just one record (one row) per case. If necessary, text join fields that cause more than one row/record per case.
  • Export Format:
    • If downloading from LegalServer, it is recommended that report results are exported as CSV files (an option that may be set up under Additional Display Format – remember to check Headers in First Row) and saved as Excel files to avoid formatting issues, particularly related to date fields.

Example Analyses Steps

Collecting External Data Via the Microdata Access Tool (MDAT):
  1. Open the MDAT (Microdata - Census Bureau Datasets (https://data.census.gov/app/mdat/).
  2. Select a Dataset & Select a Vintage & click Next. (ACS 1-Year Estimates Public Use Microdata Sample for 2023). 
    1. For information about choosing 5-year or 1-year estimates, click here: Using 1-Year or 5-Year American Community Survey Data (http://census.gov/programs-surveys/acs/guidance/estimates.html)
      null
  3. On the Variables screen, click on Select Geographies and either pick a State or multiple PUMA regions depending on the boundaries of your service area (in this example, a state is selected, but for confidentiality reasons, the state will be referred to as State in the following steps). X out of the screen when you have made your selection(s).
    null
  4. To add variables for analysis, enter search term(s) in the Label box or review the Data Dictionaries for PUMS Files from the PUMS Documentation webpage PUMS Documentation (https://www.census.gov/programs-surveys/acs/microdata/documentation.html).
    1. Race: Select the RAC1P field (or if you need more detailed race data, you may select from among the other race-related fields that appear when you enter “race” in the Label box).
    2. Hispanic: There is no ethnicity field so if you are interested in Hispanic ethnicity, you need to enter “Hispanic” in the Label box. Select the HISP field.
    3. Poverty: Select POVPIP, Income-to-poverty ratio. A message will appear explaining that this field is continuous and that you will have to create groups (i.e., <100% Poverty, <125% Poverty, etc.). Click on X to close that message. See instructions regarding creating groups in the next step.
      null
    4. If any of your variables need to be grouped, click on the Cart link near the top of the screen. If none of your variables need to be grouped, click on View Table in the bottom right of the screen when you have selected all variables. 
  5. On the Cart screen, you can create groups and relabel categories.
    1. RAC1P field:
      1. Suggested Groups: White (White alone), Black (Black or African American alone), AIAN (American Indian alone, Alaska Native alone, & American Indian and Alaska Native tribes specified; or American Indian or Alaska Native, not specified and no other races), Asian (Asian alone), Other (Native Hawaiian and Other Pacific Islander alone (because this is a very small group in this area) and Some other race alone), and Two or More Races (Two or More Races).
      2. Click on Create Custom Group, enter a new label in the Group Label line (type over “Not Elsewhere Grouped”) and select all the race categories that should be included in the new group. Click on Save Group.
        null
      3. Click on the pencil icon next to Not Elsewhere Grouped and repeat the last step until all custom Race groups are created.
        null
      4. Rename the Field by clicking on its current title (Recode for Recoded detailed race code (RAC1P_RC1). Enter Race as the new label. Click Save.
        null
    2. HISP field:
      1. Suggested Groups: Not Hispanic/Latino (Not Spanish/Hispanic/Latino), Hispanic/Latino (all other options).
      2. Click on Create Custom Group, enter a new label in the Group Label line (type over “Not Elsewhere Grouped”) and select all the categories that should be included in the new group. Click on Save Group.
        null
      3. Click on the pencil icon next to Not Elsewhere Grouped and repeat the last step until all custom Race groups are created.
        null
      4. Rename the Field by clicking on its current title (Recode for Recoded detailed Hispanic origin (HISP_RC1). Enter Ethnicity as the new label. Click Save.
        null
    3. POVPIP field:
      1. Suggested Groups: Different poverty levels could be defined, but for this example, because this organization can serve clients up to the Less than 200% Poverty Level, two groups are created: Less than 200% Poverty and 200% Poverty & Above.
      2. Click on Create Custom Group: 
        1. Enter Less than 200% Poverty in the Group Label line (type over “Not Elsewhere Grouped”)
        2. Uncheck the category that begins with N/A and the 501 percent or more category.
        3. Click on the pencil icon next to the Below 501 percent category.
        4. Enter Minimum=0 and Maximum=199 in the pop-up window. Click Save. Click Save Group.
          null
      3. Click on the pencil icon next to Not Elsewhere Grouped:
        1. Enter 200% Poverty & Above in the Group Label line (type over “Not Elsewhere Grouped”)
        2. Uncheck the category that begins with N/A.
        3. Leave Between 200 and 500 and 501 percent or more checked. Click Save. Click Save Group.
          null
      4. Remove Not Elsewhere Grouped: Uncheck Not Elsewhere Grouped because all that remains in this group is the N/A category for whom poverty could not be measured.
        null
      5. Rename the Field by clicking on its current title (Recode for Recoded detailed Income-to poverty ratio recode (POVPIP_RC1). Enter Poverty as the new label. Click Save.
        null

Related Questions You May Ask

  • How do the demographics, referral sources, or intake methods vary for different legal problems?
  • How do the number and characteristics of people requesting assistance compare to the eligible population in my service area?