Assignment 1: Census Data Quality for Policy Decisions

Evaluating Data Reliability for Algorithmic Decision-Making

Author

Angel Rutherford

Published

September 28, 2025

Assignment Overview

Scenario

You are a data analyst for the Pennsylvania Department of Human Services. The department is considering implementing an algorithmic system to identify communities that should receive priority for social service funding and outreach programs. Your supervisor has asked you to evaluate the quality and reliability of available census data to inform this decision.

Drawing on our Week 2 discussion of algorithmic bias, you need to assess not just what the data shows, but how reliable it is and what communities might be affected by data quality issues.

Learning Objectives

  • Apply dplyr functions to real census data for policy analysis
  • Evaluate data quality using margins of error
  • Connect technical analysis to algorithmic decision-making
  • Identify potential equity implications of data reliability issues
  • Create professional documentation for policy stakeholders

Submission Instructions

Submit by posting your updated portfolio link on Canvas. Your assignment should be accessible at your-portfolio-url/assignments/assignment_1/

Make sure to update your _quarto.yml navigation to include this assignment under an “Assignments” menu.

Part 1: Portfolio Integration

Create this assignment in your portfolio repository under an assignments/assignment_1/ folder structure. Update your navigation menu to include:

- text: Assignments
  menu:
    - href: assignments/assignment_1/your_file_name.qmd
      text: "Assignment 1: Census Data Exploration"

If there is a special character like comma, you need use double quote mark so that the quarto can identify this as text

Setup

# Load required packages (hint: you need tidycensus, tidyverse, and knitr)
library(tidycensus)
library(tidyverse)
library(knitr)
# Set your Census API key
census_api_key("595f8315942e8965e5a91f7cfa62b7210fe62245")
# Choose your state for analysis - assign it to a variable called my_state
my_state <- "PA"

State Selection: I have chosen Pennsylvania for this analysis for its personal relevance since I currently live here.

Part 2: County-Level Resource Assessment

2.1 Data Retrieval

Your Task: Use get_acs() to retrieve county-level data for your chosen state.

Requirements: - Geography: county level - Variables: median household income (B19013_001) and total population (B01003_001)
- Year: 2022 - Survey: acs5 - Output format: wide

Hint: Remember to give your variables descriptive names using the variables = c(name = "code") syntax.

# Write your get_acs() code here
county_data <- get_acs(
  geography ="county",
  variables= c(
     median_income ="B19013_001",
     total_pop="B01003_001"
  ),
  state = my_state,
  year = 2022,
  output = "wide"
)
# Clean the county names to remove state name and "County" 
# Hint: use mutate() with str_remove()
county_data <- county_data%>%
  mutate(county_name = str_remove(NAME, " County, Pennsylvania"))
         
# Display the first few rows
glimpse(county_data)
Rows: 67
Columns: 7
$ GEOID          <chr> "42001", "42003", "42005", "42007", "42009", "42011", "…
$ NAME           <chr> "Adams County, Pennsylvania", "Allegheny County, Pennsy…
$ median_incomeE <dbl> 78975, 72537, 61011, 67194, 58337, 74617, 59386, 60650,…
$ median_incomeM <dbl> 3334, 869, 2202, 1531, 2606, 1191, 2058, 2167, 1516, 21…
$ total_popE     <dbl> 104604, 1245310, 65538, 167629, 47613, 428483, 122640, …
$ total_popM     <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA,…
$ county_name    <chr> "Adams", "Allegheny", "Armstrong", "Beaver", "Bedford",…

2.2 Data Quality Assessment

Your Task: Calculate margin of error percentages and create reliability categories.

Requirements: - Calculate MOE percentage: (margin of error / estimate) * 100 - Create reliability categories: - High Confidence: MOE < 5% - Moderate Confidence: MOE 5-10%
- Low Confidence: MOE > 10% - Create a flag for unreliable estimates (MOE > 10%)

Hint: Use mutate() with case_when() for the categories.

# Calculate MOE percentage and reliability categories using mutate()
county_data <-county_data%>%
  mutate(MOE_PCT=(median_incomeM/median_incomeE)*100,
  reliability = case_when(
    MOE_PCT < 5 ~ "High Confidence (<5%)",
    MOE_PCT <= 10 ~ "Moderate Confidence (5%-10%)",
    TRUE ~ "Low Confidence (>10%)"
    )
  )
county_data
# A tibble: 67 × 9
   GEOID NAME    median_incomeE median_incomeM total_popE total_popM county_name
   <chr> <chr>            <dbl>          <dbl>      <dbl>      <dbl> <chr>      
 1 42001 Adams …          78975           3334     104604         NA Adams      
 2 42003 Allegh…          72537            869    1245310         NA Allegheny  
 3 42005 Armstr…          61011           2202      65538         NA Armstrong  
 4 42007 Beaver…          67194           1531     167629         NA Beaver     
 5 42009 Bedfor…          58337           2606      47613         NA Bedford    
 6 42011 Berks …          74617           1191     428483         NA Berks      
 7 42013 Blair …          59386           2058     122640         NA Blair      
 8 42015 Bradfo…          60650           2167      60159         NA Bradford   
 9 42017 Bucks …         107826           1516     645163         NA Bucks      
10 42019 Butler…          82932           2164     194562         NA Butler     
# ℹ 57 more rows
# ℹ 2 more variables: MOE_PCT <dbl>, reliability <chr>
# Create a summary showing count of counties in each reliability category
# Hint: use count() and mutate() to add percentages
reliability_summary <- county_data%>%
  count(reliability) %>%
  mutate(
    percentage = round(n/sum(n) * 100, 1),
    total = sum(n)
  )
reliability_summary
# A tibble: 2 × 4
  reliability                      n percentage total
  <chr>                        <int>      <dbl> <int>
1 High Confidence (<5%)           57       85.1    67
2 Moderate Confidence (5%-10%)    10       14.9    67

2.3 High Uncertainty Counties

Your Task: Identify the 5 counties with the highest MOE percentages.

Requirements: - Sort by MOE percentage (highest first) - Select the top 5 counties - Display: county name, median income, margin of error, MOE percentage, reliability category - Format as a professional table using kable()

Hint: Use arrange(), slice(), and select() functions.

# Create table of top 5 counties by MOE percentage
top_5_counties <- county_data %>%
  arrange(desc(MOE_PCT))%>%
  slice_head(n=5)

# Format as table with kable() - include appropriate column names and caption 
top_5_counties %>%
  select(county_name, median_incomeE, median_incomeM, MOE_PCT, reliability)%>%
    kable(
      col.names = c("County Name", "Income ($)", "MOE", "MOE Percentage (%)", "MOE Reliability"),
      caption = "Top 5 Counties Exhibiting High Uncertainty",
     digits=c(0,0,1,1,0)
)
Top 5 Counties Exhibiting High Uncertainty
County Name Income ($) MOE MOE Percentage (%) MOE Reliability
Forest 46188 4612 10.0 Moderate Confidence (5%-10%)
Sullivan 62910 5821 9.3 Moderate Confidence (5%-10%)
Union 64914 4753 7.3 Moderate Confidence (5%-10%)
Montour 72626 5146 7.1 Moderate Confidence (5%-10%)
Elk 61672 4091 6.6 Moderate Confidence (5%-10%)

Data Quality Commentary:

These results highlight that although all five counties are all considered to indicate moderate confidence in their income estimations, their margins of error vary significantly. For instance, the county with the largest margin of error, Forest County, lies on the cusp of exhibiting low confidence while the smallest margin of error, Elk County, is much closer to being high confidence than low. These discrepancies, in addition to general uncertainty, indicate a need for additional oversight when using income data from these counties for decision-making.

Part 3: Neighborhood-Level Analysis

3.1 Focus Area Selection

Your Task: Select 2-3 counties from your reliability analysis for detailed tract-level study.

Strategy: Choose counties that represent different reliability levels (e.g., 1 high confidence, 1 moderate, 1 low confidence) to compare how data quality varies.

# Use filter() to select 2-3 counties from your county_reliability data
# Store the selected counties in a variable called selected_counties
# Display the selected counties with their key characteristics
# Show: county name, median income, MOE percentage, reliability category
selected_counties <- county_data %>% 
  filter(county_name %in% c("Allegheny", "Carbon", "Forest"))   %>%
  select(county_name, median_incomeE, MOE_PCT, reliability)

selected_counties
# A tibble: 3 × 4
  county_name median_incomeE MOE_PCT reliability                 
  <chr>                <dbl>   <dbl> <chr>                       
1 Allegheny            72537    1.20 High Confidence (<5%)       
2 Carbon               64538    5.31 Moderate Confidence (5%-10%)
3 Forest               46188    9.99 Moderate Confidence (5%-10%)

Comment on the output: Because my data possessed no findings with low confidence, I chose the top county from my “Top 5 Counties Exhibiting High Uncertainty” as my low confidence observation. A gradient in reliability, however, is still achieved as, as previously observed, the margins of error between both counties with estimations of moderate confidence range drastically.

3.2 Tract-Level Demographics

Your Task: Get demographic data for census tracts in your selected counties.

Requirements: - Geography: tract level - Variables: white alone (B03002_003), Black/African American (B03002_004), Hispanic/Latino (B03002_012), total population (B03002_001) - Use the same state and year as before - Output format: wide - Challenge: You’ll need county codes, not names. Look at the GEOID patterns in your county data for hints.

# Define your race/ethnicity variables with descriptive names

# Use get_acs() to retrieve tract-level data
# Hint: You may need to specify county codes in the county parameter
find_codes <- county_data %>% 
  select(county_name, GEOID)%>%
  filter(county_name %in% c("Allegheny", "Carbon", "Forest"))   %>%
  mutate(county_code = str_remove(GEOID, "42"))
find_codes
# A tibble: 3 × 3
  county_name GEOID county_code
  <chr>       <chr> <chr>      
1 Allegheny   42003 003        
2 Carbon      42025 025        
3 Forest      42053 053        
tract_data <- get_acs(
  geography ="tract",
  variables= c(
     white ="B03002_003",
     black="B03002_004",
     hispanic="B03002_012",
     total_pop="B03002_001"
  ),
  state = my_state,
  county =c(003, 025, 053),
  year = 2022,
  output = "wide"
)

tract_data
# A tibble: 413 × 10
   GEOID       NAME   whiteE whiteM blackE blackM hispanicE hispanicM total_popE
   <chr>       <chr>   <dbl>  <dbl>  <dbl>  <dbl>     <dbl>     <dbl>      <dbl>
 1 42003010301 Censu…    755    108   1086    125        60        47       2028
 2 42003010302 Censu…   3283    254    678    174       269       147       4631
 3 42003020100 Censu…   2915    319    541    142       192       102       4310
 4 42003020300 Censu…   1170    201     15     23        44        32       1471
 5 42003030500 Censu…    724    179   1821    451       135       101       3044
 6 42003040200 Censu…    947    234    493    140        53        30       1843
 7 42003040400 Censu…   1192    216     61     49        17        21       1629
 8 42003040500 Censu…   2342    404     79     63        88        56       2840
 9 42003040600 Censu…   1928    417     69     63        67        60       2302
10 42003040900 Censu…   2496    409    609    248        92        63       3722
# ℹ 403 more rows
# ℹ 1 more variable: total_popM <dbl>
# Calculate percentage of each group using mutate()
# Create percentages for white, Black, and Hispanic populations
# Add readable tract and county name columns using str_extract() or similar

unreadable= str_split(tract_data$NAME, ";", simplify =  TRUE)

tract_data <-tract_data%>% 
  mutate(white_pct=(whiteE/total_popE)*100,
         black_pct=(blackE/total_popE)*100,
         hispanic_pct=(hispanicE/total_popE)*100,
         tract_name=unreadable[,1],
         county_name=str_remove(unreadable[,2], "County")
        
)

glimpse(tract_data)
Rows: 413
Columns: 15
$ GEOID        <chr> "42003010301", "42003010302", "42003020100", "42003020300…
$ NAME         <chr> "Census Tract 103.01; Allegheny County; Pennsylvania", "C…
$ whiteE       <dbl> 755, 3283, 2915, 1170, 724, 947, 1192, 2342, 1928, 2496, …
$ whiteM       <dbl> 108, 254, 319, 201, 179, 234, 216, 404, 417, 409, 57, 150…
$ blackE       <dbl> 1086, 678, 541, 15, 1821, 493, 61, 79, 69, 609, 1450, 132…
$ blackM       <dbl> 125, 174, 142, 23, 451, 140, 49, 63, 63, 248, 404, 422, 3…
$ hispanicE    <dbl> 60, 269, 192, 44, 135, 53, 17, 88, 67, 92, 3, 0, 84, 98, …
$ hispanicM    <dbl> 47, 147, 102, 32, 101, 30, 21, 56, 60, 63, 8, 11, 73, 54,…
$ total_popE   <dbl> 2028, 4631, 4310, 1471, 3044, 1843, 1629, 2840, 2302, 372…
$ total_popM   <dbl> 61, 198, 368, 206, 535, 235, 220, 423, 446, 452, 401, 377…
$ white_pct    <dbl> 37.2287968, 70.8918160, 67.6334107, 79.5377294, 23.784494…
$ black_pct    <dbl> 53.5502959, 14.6404664, 12.5522042, 1.0197145, 59.8226018…
$ hispanic_pct <dbl> 2.9585799, 5.8086806, 4.4547564, 2.9911625, 4.4349540, 2.…
$ tract_name   <chr> "Census Tract 103.01", "Census Tract 103.02", "Census Tra…
$ county_name  <chr> " Allegheny ", " Allegheny ", " Allegheny ", " Allegheny …

3.3 Demographic Analysis

Your Task: Analyze the demographic patterns in your selected areas.

# Find the tract with the highest percentage of Hispanic/Latino residents
# Hint: use arrange() and slice() to get the top tract
highest_hispanic_pct <- tract_data %>%
  arrange(desc(hispanic_pct))%>%
  slice(1)
glimpse(highest_hispanic_pct)
Rows: 1
Columns: 15
$ GEOID        <chr> "42003120300"
$ NAME         <chr> "Census Tract 1203; Allegheny County; Pennsylvania"
$ whiteE       <dbl> 33
$ whiteM       <dbl> 38
$ blackE       <dbl> 1678
$ blackM       <dbl> 494
$ hispanicE    <dbl> 343
$ hispanicM    <dbl> 499
$ total_popE   <dbl> 2093
$ total_popM   <dbl> 681
$ white_pct    <dbl> 1.576684
$ black_pct    <dbl> 80.172
$ hispanic_pct <dbl> 16.38796
$ tract_name   <chr> "Census Tract 1203"
$ county_name  <chr> " Allegheny "
# Calculate average demographics by county using group_by() and summarize()
# Show: number of tracts, average percentage for each racial/ethnic group
summary_by_county <- tract_data %>%
  group_by(county_name) %>%
  summarize(
    number_of_tracts=n(),
    mean_white_pct=round(mean(white_pct, na.rm = TRUE),1),
    mean_black_pct=round(mean(black_pct, na.rm = TRUE),1),
    mean_hispanic_pct=round(mean(hispanic_pct, na.rm = TRUE),1)
  )
summary_by_county
# A tibble: 3 × 5
  county_name   number_of_tracts mean_white_pct mean_black_pct mean_hispanic_pct
  <chr>                    <int>          <dbl>          <dbl>             <dbl>
1 " Allegheny "              394           74.5           15.4               2.4
2 " Carbon "                  17           89              1.9               6.4
3 " Forest "                   2           71.2           13.6               7.4
# Create a nicely formatted table of your results using kable()
summary_by_county%>%
 kable(
      col.names = c("County Name", "Total Tracts Per County", "Average of White Residents(%)", "Average of Black Residents(%)", "Average of Hispanic Residents (%)"),
      caption = "Average Demographics Per County",
     digits=c(0,0,1,1,1)
)
Average Demographics Per County
County Name Total Tracts Per County Average of White Residents(%) Average of Black Residents(%) Average of Hispanic Residents (%)
Allegheny 394 74.5 15.4 2.4
Carbon 17 89.0 1.9 6.4
Forest 2 71.2 13.6 7.4

Part 4: Comprehensive Data Quality Evaluation

4.1 MOE Analysis for Demographic Variables

Your Task: Examine margins of error for demographic variables to see if some communities have less reliable data.

Requirements: - Calculate MOE percentages for each demographic variable - Flag tracts where any demographic variable has MOE > 15% - Create summary statistics

# Calculate MOE percentages for white, Black, and Hispanic variables
# Hint: use the same formula as before (margin/estimate * 100)

# Create a flag for tracts with high MOE on any demographic variable
# Use logical operators (| for OR) in an ifelse() statement
tract_data <-tract_data%>%
  mutate(
    white_moe_pct=(whiteM/whiteE)*100,
    black_moe_pct=(blackM/blackE)*100,
    hispanic_moe_pct=(hispanicM/hispanicE)*100,
    moe_flag=ifelse(
    white_moe_pct > 15 |
    black_moe_pct > 15 |
    hispanic_moe_pct > 15,
    TRUE, FALSE)
)

# Create summary statistics showing how many tracts have data quality issues
 quality_summary <- tract_data%>%
  count(moe_flag) %>%
  mutate(
    percentage = round(n/sum(n) * 100, 1),
    total = sum(n)
  )
quality_summary
# A tibble: 1 × 4
  moe_flag     n percentage total
  <lgl>    <int>      <dbl> <int>
1 TRUE       413        100   413

4.2 Pattern Analysis

Your Task: Investigate whether data quality problems are randomly distributed or concentrated in certain types of communities.

# Group tracts by whether they have high MOE issues
# Calculate average characteristics for each group:
# - population size, demographic percentages

# Use group_by() and summarize() to create this comparison
summary_by_moe <- tract_data %>%
  group_by(moe_flag) %>%
  summarize(
    number_of_tracts= n(),
    mean_population=round(mean(total_popE, na.rm = TRUE),0),
    mean_white_pct=round(mean(white_pct, na.rm = TRUE),1),
    mean_black_pct=round(mean(black_pct, na.rm = TRUE),1),
    mean_hispanic_pct=round(mean(hispanic_pct, na.rm = TRUE),1)
  )
summary_by_moe
# A tibble: 1 × 6
  moe_flag number_of_tracts mean_population mean_white_pct mean_black_pct
  <lgl>               <int>           <dbl>          <dbl>          <dbl>
1 TRUE                  413            3190             75           14.8
# ℹ 1 more variable: mean_hispanic_pct <dbl>
# Create a professional table showing the patterns
summary_by_moe%>%
 kable(
      col.names = c("Quality Issues (MOE > %15)", "Total Tracts Detected", "Average Total Population", "Average of White Residents(%)", "Average of Black Residents(%)", "Average of Hispanic Residents (%)"),
      caption = "Average Demographics Per County",
     digits=c(0,0,0,1,1,1)
)
Average Demographics Per County
Quality Issues (MOE > %15) Total Tracts Detected Average Total Population Average of White Residents(%) Average of Black Residents(%) Average of Hispanic Residents (%)
TRUE 413 3190 75 14.8 2.6

Pattern Analysis: All of the tracts investigated possessed high margins of error and predominantly white communities. This may suggest that demographic data from less diverse communities is more likely to be unreliable.

Part 5: Policy Recommendations

5.1 Analysis Integration and Professional Summary

Your Task: Write an executive summary that integrates findings from all four analyses.

Executive Summary Requirements: 1. Overall Pattern Identification: What are the systematic patterns across all your analyses? 2. Equity Assessment: Which communities face the greatest risk of algorithmic bias based on your findings? 3. Root Cause Analysis: What underlying factors drive both data quality issues and bias risk? 4. Strategic Recommendations: What should the Department implement to address these systematic issues?

Executive Summary:

Overall Pattern Identification Across all analyses, the smaller the community being surveyed, the less accurate the data would become. Although the counties that were chosen to be further analyzed at the tract level represented a gradient in income estimation reliability, they all shared similar unreliability in demographic data. Allegheny, the county with the smallest margin of error in income estimation, possessed the largest population of the chosen counties while Forest county, the county with the largest margin of error, possessed the smallest population. These counties similarly, however, possessed a disproportionately large number of white residents compare to other racial and ethnic groups and returned high margins of error in their demographic data.

Equity Assessment The county data here emphasizes that margins of error vary depending on the type of data collected and are influenced by both the absolute size of a population and its proportion within the broader population. More specifically, smaller groups such as counties with small populations and racial or ethnic minorities are more prone to data inaccuracies and thus more vulnerable to bias in algorithmic decision-making. For instance, the tract with the largest hispanic population, the smallest average population across all tracts, had an even larger margin of error. Decisions made based on the this demographic data may vastly misrepresent the true needs of the hispanic population.

Root Cause Analysis The misrepresentation of vulnerable populations such as small or spatially uneven populations in data can result in algorithms using inaccurate inputs and thus, inherently reproducing non-representational outputs. Furthermore, the subjective choices made when analyzing data can also skew outputs.

Strategic Recommendations To mitigate these risks, the Department should identify margins of error and create thorough thresholds for interpretations of margins of errors. Certain thresholds should call for human oversight and require additional social, historical, and spatial explanations.

6.3 Specific Recommendations

Your Task: Create a decision framework for algorithm implementation.

# Create a summary table using your county reliability data
# Include: county name, median income, MOE percentage, reliability category

# Add a new column with algorithm recommendations using case_when():
# - High Confidence: "Safe for algorithmic decisions"
# - Moderate Confidence: "Use with caution - monitor outcomes"  
# - Low Confidence: "Requires manual review or additional data"

# Format as a professional table with kable()
county_summary <-county_data%>%
  mutate(
    rec_for_algorithm= case_when(
    reliability == "High Confidence (<5%)" ~ "Safe for algorithmic decisions",
   reliability == "Moderate Confidence (5%-10%)" ~ "Use with caution - monitor outcomes",
    reliability == "Low Confidence (>10%)" ~ "Requires manual review or additional data"
    )
  )%>%
  select(county_name, median_incomeE, MOE_PCT, reliability, rec_for_algorithm)%>%
  arrange(desc(MOE_PCT))%>%
  kable(
    col.names = c("County Name", "Median Income ($)", "MOE Percentage(%)", "Reliability", "Algorithm Recommendation)"),
      caption = "County-level Algorithmic Recommendations",
     digits=c(0,0,1,0,0)
  )
county_summary
County-level Algorithmic Recommendations
County Name Median Income ($) MOE Percentage(%) Reliability Algorithm Recommendation)
Forest 46188 10.0 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Sullivan 62910 9.3 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Union 64914 7.3 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Montour 72626 7.1 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Elk 61672 6.6 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Greene 66283 6.4 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Cameron 46186 5.6 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Snyder 65914 5.6 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Carbon 64538 5.3 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Warren 57925 5.2 Moderate Confidence (5%-10%) Use with caution - monitor outcomes
Pike 76416 4.9 High Confidence (<5%) Safe for algorithmic decisions
Wayne 59240 4.8 High Confidence (<5%) Safe for algorithmic decisions
Juniata 61915 4.8 High Confidence (<5%) Safe for algorithmic decisions
McKean 57861 4.7 High Confidence (<5%) Safe for algorithmic decisions
Huntingdon 61300 4.7 High Confidence (<5%) Safe for algorithmic decisions
Indiana 57170 4.6 High Confidence (<5%) Safe for algorithmic decisions
Bedford 58337 4.5 High Confidence (<5%) Safe for algorithmic decisions
Potter 56491 4.4 High Confidence (<5%) Safe for algorithmic decisions
Lycoming 63437 4.4 High Confidence (<5%) Safe for algorithmic decisions
Clarion 58690 4.4 High Confidence (<5%) Safe for algorithmic decisions
Adams 78975 4.2 High Confidence (<5%) Safe for algorithmic decisions
Fayette 55579 4.2 High Confidence (<5%) Safe for algorithmic decisions
Crawford 58734 3.9 High Confidence (<5%) Safe for algorithmic decisions
Clinton 59011 3.9 High Confidence (<5%) Safe for algorithmic decisions
Wyoming 67968 3.9 High Confidence (<5%) Safe for algorithmic decisions
Columbia 59457 3.8 High Confidence (<5%) Safe for algorithmic decisions
Fulton 63153 3.6 High Confidence (<5%) Safe for algorithmic decisions
Mercer 57353 3.6 High Confidence (<5%) Safe for algorithmic decisions
Armstrong 61011 3.6 High Confidence (<5%) Safe for algorithmic decisions
Bradford 60650 3.6 High Confidence (<5%) Safe for algorithmic decisions
Blair 59386 3.5 High Confidence (<5%) Safe for algorithmic decisions
Venango 59278 3.4 High Confidence (<5%) Safe for algorithmic decisions
Mifflin 58012 3.4 High Confidence (<5%) Safe for algorithmic decisions
Jefferson 56607 3.4 High Confidence (<5%) Safe for algorithmic decisions
Cambria 54221 3.3 High Confidence (<5%) Safe for algorithmic decisions
Tioga 59707 3.2 High Confidence (<5%) Safe for algorithmic decisions
Monroe 80656 3.2 High Confidence (<5%) Safe for algorithmic decisions
Perry 76103 3.2 High Confidence (<5%) Safe for algorithmic decisions
Susquehanna 63968 3.1 High Confidence (<5%) Safe for algorithmic decisions
Lawrence 57585 3.1 High Confidence (<5%) Safe for algorithmic decisions
Franklin 71808 3.0 High Confidence (<5%) Safe for algorithmic decisions
Clearfield 56982 2.8 High Confidence (<5%) Safe for algorithmic decisions
Somerset 57357 2.8 High Confidence (<5%) Safe for algorithmic decisions
Centre 70087 2.8 High Confidence (<5%) Safe for algorithmic decisions
Lebanon 72532 2.7 High Confidence (<5%) Safe for algorithmic decisions
Northumberland 55952 2.7 High Confidence (<5%) Safe for algorithmic decisions
Butler 82932 2.6 High Confidence (<5%) Safe for algorithmic decisions
Lackawanna 63739 2.6 High Confidence (<5%) Safe for algorithmic decisions
Erie 59396 2.6 High Confidence (<5%) Safe for algorithmic decisions
Schuylkill 63574 2.4 High Confidence (<5%) Safe for algorithmic decisions
Washington 74403 2.4 High Confidence (<5%) Safe for algorithmic decisions
Luzerne 60836 2.4 High Confidence (<5%) Safe for algorithmic decisions
Beaver 67194 2.3 High Confidence (<5%) Safe for algorithmic decisions
Dauphin 71046 2.3 High Confidence (<5%) Safe for algorithmic decisions
Cumberland 82849 2.2 High Confidence (<5%) Safe for algorithmic decisions
Lehigh 74973 2.0 High Confidence (<5%) Safe for algorithmic decisions
Westmoreland 69454 2.0 High Confidence (<5%) Safe for algorithmic decisions
Northampton 82201 1.9 High Confidence (<5%) Safe for algorithmic decisions
Lancaster 81458 1.8 High Confidence (<5%) Safe for algorithmic decisions
York 79183 1.8 High Confidence (<5%) Safe for algorithmic decisions
Chester 118574 1.7 High Confidence (<5%) Safe for algorithmic decisions
Berks 74617 1.6 High Confidence (<5%) Safe for algorithmic decisions
Delaware 86390 1.5 High Confidence (<5%) Safe for algorithmic decisions
Bucks 107826 1.4 High Confidence (<5%) Safe for algorithmic decisions
Philadelphia 57537 1.4 High Confidence (<5%) Safe for algorithmic decisions
Montgomery 107441 1.3 High Confidence (<5%) Safe for algorithmic decisions
Allegheny 72537 1.2 High Confidence (<5%) Safe for algorithmic decisions

Key Recommendations:

Your Task: Use your analysis results to provide specific guidance to the department.

  1. Counties suitable for immediate algorithmic implementation: Allegheny, Montgomery, Bucks, Chester, Delaware, Lancaster, York, Northampton, Lehigh, Cumberland, Dauphin, Beaver, Luzerne, Washington, Schuylkill, Erie, Lackawanna, Butler, Lebanon, Centre, Somerset, Clearfield, Franklin, Lawrence, Susquehanna, Perry, Monroe, Tioga, Cambria, Jefferson, Mifflin, Venango, Blair, Bradford, Armstrong, Mercer, Fulton, Columbia, Wyoming, Clinton, Crawford, Fayette, Adams, Clarion, Lycoming, Potter, Bedford, Indiana, Huntingdon, McKean, Juniata, Wayne, Pike

Justification: Low margins of error indicate reduced risk in algorithmic bias

  1. Counties requiring additional oversight: Sullivan, Union, Montour, Elk, Greene, Cameron, Snyder, Carbon, Warren

Justification: Although these counties fall in the moderate confidence range, outcomes should be monitored to ensure appropriateness.

  1. Counties needing alternative approaches: Forest

Justifications: Although Forest county falls in the moderate confidence range, it borders the low confidence range and likely should avoid automated decisions until the quality of data improves.

Questions for Further Investigation

  1. How small or strict should data reliability ranges be?
  2. Would it be statistically meaningful to compare margin of error percentages of different variables?

Technical Notes

Data Sources: - U.S. Census Bureau, American Community Survey 2018-2022 5-Year Estimates - Retrieved via tidycensus R package on 09/26/2025

Reproducibility: - All analysis conducted in R version R 4.5.1 - Census API key required for replication - Complete code and documentation available at: https://musa-5080-fall-2025.github.io/portfolio-setup-aruth3/

Methodology Notes: County selection was determined base on numerical values of reliability rather than categorical indications of reliability due to a lack of categorical variation.

Limitations: Certain population percentage estimates returned values of infinity due to division by denominators with a value of zero. These results should be interpreted with caution as they do not represent meaningful statistical outcomes.

Submission Checklist

Before submitting your portfolio link on Canvas:

Remember: Submit your portfolio URL on Canvas, not the file itself. Your assignment should be accessible at your-portfolio-url/assignments/assignment_1/your_file_name.html