Lab 1: Census Data Quality for Policy Decisions

Evaluating Data Reliability for Algorithmic Decision-Making

Author

Mackenna Amole

Published

February 24, 2026

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/labs/lab_1/

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

Part 1: Portfolio Integration

Create this assignment in your portfolio repository under an labs/lab_1/ folder structure. Update your navigation menu to include:

- text: Assignments
  menu:
    - href: labs/lab_1/your_file_name.qmd
      text: "Lab 1: Census Data Exploration"

If there is a special character like a colon, 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(tidyverse)
library(tidycensus)
library(scales)
library(RColorBrewer)
library(knitr)


# Set your Census API key
census_api_key(Sys.getenv("CENSUS_API_KEY"))

# 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 because: I gre up here, attended undergrad here, and now am in grad school 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
  medHHin_totalpop <- get_acs(
    geography = "county",
    variables = c(
    total_pop = "B01003_001",       # Total population
    median_income = "B19013_001"   # Median household income
  ),
    state = "PA",
    year = 2022,
    survey = "acs5",
    output = "wide"
  )
# Clean the county names to remove state name and "County" 
# Hint: use mutate() with str_remove()
county_incomepop <- medHHin_totalpop %>%
  mutate(county_name = str_remove(NAME, paste0(" County, Pennsylvania")))

# Display the first few rows
glimpse(county_incomepop)

Rows: 67
Columns: 7
$ GEOID          <chr> "42001", "42003", "42005", "42007", "42009", "42011", "…
$ NAME           <chr> "Adams County, Pennsylvania", "Allegheny County, Pennsy…
$ 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,…
$ median_incomeE <dbl> 78975, 72537, 61011, 67194, 58337, 74617, 59386, 60650,…
$ median_incomeM <dbl> 3334, 869, 2202, 1531, 2606, 1191, 2058, 2167, 1516, 21…
$ 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_incomepop <- county_incomepop %>%
  mutate(MOE_per = median_incomeM/median_incomeE * 100)

county_incomepop <- county_incomepop %>%
  mutate(
    reliability = case_when(
    MOE_per < 5 ~ "High",
    MOE_per > 5 & MOE_per < 10 ~ "Moderate",
    MOE_per > 10 ~ "Low"))

county_incomepop <- county_incomepop %>%
  mutate(Low_Flag = if_else(reliability == "Low", 1L, 0L))

# Create a summary showing count of counties in each reliability category
county_incomepop %>%
  group_by(reliability) %>%
  summarise(count = n(), .groups = "drop")

# A tibble: 2 × 2
  reliability count
  <chr>       <int>
1 High           57
2 Moderate       10

#57 Low
#10 Moderate

# Hint: use count() and mutate() to add percentages

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
high5_uncertainty <- county_incomepop %>%
  arrange(desc(MOE_per)) %>%
  slice(1:5) %>%
    select(
    NAME,
    median_incomeE,
    median_incomeM,
    MOE_per,
    reliability
  )

# Format as table with kable() - include appropriate column names and caption
kable(
  high5_uncertainty,
  col.names = c(
    "County",
    "Median Income",
    "Income Margin of Error",
    "Margin Of Error Percentage",
    "Reliability Category"
  ),
  caption = "Top 5 Counties with the Highest Margin of Error Percentages"
)

Top 5 Counties with the Highest Margin of Error Percentages
County	Median Income	Income Margin of Error	Margin Of Error Percentage	Reliability Category
Forest County, Pennsylvania	46188	4612	9.985278	Moderate
Sullivan County, Pennsylvania	62910	5821	9.252901	Moderate
Union County, Pennsylvania	64914	4753	7.321995	Moderate
Montour County, Pennsylvania	72626	5146	7.085617	Moderate
Elk County, Pennsylvania	61672	4091	6.633480	Moderate

Data Quality Commentary: Counties with lower data quality and/or smaller populations may be more poorly served by these algorithms. Factors such as the accuracy of the median income estimate may contribute to this challenge.

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
selected_counties <- county_incomepop %>%
  filter(
    county_name %in% c("Allegheny", "Forest", "Philadelphia")
  )
# Display the selected counties with their key characteristics
# Show: county name, median income, MOE percentage, reliability category
selected_counties %>%
  select(
    county_name,
    median_incomeE,
    MOE_per,
    reliability
  ) %>%
  kable(
    col.names = c(
      "County",
      "Median Income",
      "MOE Percentage",
      "Reliability Category"
    ),
    caption = "Selected Counties for Neighborhood-Level (Tract) Analysis"
  )

Selected Counties for Neighborhood-Level (Tract) Analysis
County	Median Income	MOE Percentage	Reliability Category
Allegheny	72537	1.198009	High
Forest	46188	9.985278	Moderate
Philadelphia	57537	1.376506	High

Comment on the output: Selected counties are Allegheny (with PA’s lowest MOE %), Forest (with PA’s highest MOE %), and Philadelphia (as an interest because we live and study here).

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
race_vars <- c(
  total_pop = "B03002_001",
  white = "B03002_003",
  black = "B03002_004",
  hispanic = "B03002_012"
)

# Use get_acs() to retrieve tract-level data
# Hint: You may need to specify county codes in the county parameter
# Allegheny = 003, Forest = 053, Philadelphia = 101
tract_demographics <- get_acs(
  geography = "tract",
  state = "PA",
  county = c("003", "053", "101"),
  variables = race_vars,
  year = 2022,      # same ACS year used earlier
  output = "wide"
)

# Calculate percentage of each group using mutate()
# Create percentages for white, Black, and Hispanic populations
tract_demographics <- tract_demographics %>%
  mutate(
    pct_white = whiteE / total_popE * 100,
    pct_black = blackE / total_popE * 100,
    pct_hispanic = hispanicE / total_popE * 100
  )

# Add readable tract and county name columns using str_extract() or similar
tract_demographics <- tract_demographics %>%
  mutate(
    tract_name = str_extract(NAME, "^Census Tract[^;]+"),
    county_name = str_extract(NAME, "(?<=; )[A-Za-z ]+(?= County)")
  )

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
#high5_uncertainty <- county_incomepop %>% arrange(desc(MOE_per)) %>% slice(1:5) %>%select(NAME,median_incomeE,median_incomeM,MOE_per,reliability)
tract_mostlatino <- tract_demographics %>%
  arrange(desc(pct_hispanic)) %>%
  slice(1)

# Calculate average demographics by county using group_by() and summarize()
# Show: number of tracts, average percentage for each racial/ethnic group
county_demographics <- tract_demographics %>%
  group_by(county_name) %>%
  summarise(
    num_tracts = n(),
    avg_pct_white = mean(pct_white, na.rm = TRUE),
    avg_pct_black = mean(pct_black, na.rm = TRUE),
    avg_pct_hispanic = mean(pct_hispanic, na.rm = TRUE),
    .groups = "drop"
  )
# Create a nicely formatted table of your results using kable()
kable(
  county_demographics,
  col.names = c(
    "County",
    "Number of Tracts",
    "Avg % White",
    "Avg % Black",
    "Avg % Hispanic/Latino"
  ),
  caption = "Average Tract-Level Demographics by County"
)

Average Tract-Level Demographics by County
County	Number of Tracts	Avg % White	Avg % Black	Avg % Hispanic/Latino
Allegheny	394	74.45359	15.41641	2.416422
Forest	2	71.18900	13.56075	7.379975
Philadelphia	408	35.38937	39.19500	13.843119

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)
tract_demographics <- tract_demographics %>%
  mutate(MOE_per_white = whiteM/whiteE *100 ) %>%
  mutate(MOE_per_black = blackM/blackE *100) %>%
  mutate(MOE_per_hispanic = hispanicM/hispanicE*100)

# Create a flag for tracts with high MOE on any demographic variable
# Use logical operators (| for OR) in an ifelse() statement
tract_demographics <- tract_demographics %>%
  mutate(high_moe_flag = ifelse(
    MOE_per_white > 15 | MOE_per_black > 15 | MOE_per_hispanic > 16,
    "High MOE",
    "Acceptable"
  ))
# Create summary statistics showing how many tracts have data quality issues
tract_demographics %>%
  summarise(
    total_tracts = n(),
    tracts_with_issues = sum(high_moe_flag == "High MOE", na.rm = TRUE),
    percent_with_issues = 100 * mean(high_moe_flag == "High MOE", na.rm = TRUE)
  )

# A tibble: 1 × 3
  total_tracts tracts_with_issues percent_with_issues
         <int>              <int>               <dbl>
1          804                804                 100

#100% ...

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
tract_demographics <- tract_demographics %>%
  mutate(
    high_moe_white = ifelse(MOE_per_white > 15, "High MOE", "Acceptable"),
    high_moe_black = ifelse(MOE_per_black > 15, "High MOE", "Acceptable"),
    high_moe_hispanic = ifelse(MOE_per_hispanic > 15, "High MOE", "Acceptable")
  )

tracts_by_race_moe <- tract_demographics %>%
  summarise(
    white_high_moe = sum(high_moe_white == "High MOE", na.rm = TRUE),
    black_high_moe = sum(high_moe_black == "High MOE", na.rm = TRUE),
    hispanic_high_moe = sum(high_moe_hispanic == "High MOE", na.rm = TRUE),
    total_tracts = n()
  ) %>%
  mutate(
    pct_white_high_moe = 100 * white_high_moe / total_tracts,
    pct_black_high_moe = 100 * black_high_moe / total_tracts,
    pct_hispanic_high_moe = 100 * hispanic_high_moe / total_tracts
  )

# Calculate average characteristics for each group:
# - population size, demographic percentages

#white
white_summary <- tract_demographics %>%
  group_by(high_moe_white) %>%
  summarise(
    n_tracts = n(),
    avg_population = mean(total_popE, na.rm = TRUE),
    avg_pct_white = mean(pct_white, na.rm = TRUE),
    avg_pct_black = mean(pct_black, na.rm = TRUE),
    avg_pct_hispanic = mean(pct_hispanic, na.rm = TRUE)
  ) %>%
  mutate(subgroup = "White")

#black
black_summary <- tract_demographics %>%
  group_by(high_moe_black) %>%
  summarise(
    n_tracts = n(),
    avg_population = mean(total_popE, na.rm = TRUE),
    avg_pct_white = mean(pct_white, na.rm = TRUE),
    avg_pct_black = mean(pct_black, na.rm = TRUE),
    avg_pct_hispanic = mean(pct_hispanic, na.rm = TRUE)
  ) %>%
  mutate(subgroup = "Black")

#hispanic
hispanic_summary <- tract_demographics %>%
  group_by(high_moe_hispanic) %>%
  summarise(
    n_tracts = n(),
    avg_population = mean(total_popE, na.rm = TRUE),
    avg_pct_white = mean(pct_white, na.rm = TRUE),
    avg_pct_black = mean(pct_black, na.rm = TRUE),
    avg_pct_hispanic = mean(pct_hispanic, na.rm = TRUE)
  ) %>%
  mutate(subgroup = "Hispanic")


# Use group_by() and summarize() to create this comparison
pattern_summary <- bind_rows(white_summary, black_summary, hispanic_summary) %>%
  select(subgroup, everything()) %>%
  arrange(subgroup, desc(n_tracts))

# Create a professional table showing the patterns
kable(pattern_summary, digits = 1,
      caption = "Average Tract Characteristics by Racial/Ethnic Subgroup and MOE Flag")

Average Tract Characteristics by Racial/Ethnic Subgroup and MOE Flag
subgroup	high_moe_white	n_tracts	avg_population	avg_pct_white	avg_pct_black	avg_pct_hispanic	high_moe_black	high_moe_hispanic
Black	NA	795	3548.6	55.2	26.9	8.2	High MOE	NA
Black	NA	9	2705.2	19.8	68.6	5.1	Acceptable	NA
Hispanic	NA	803	3534.1	54.8	27.4	8.2	NA	High MOE
Hispanic	NA	1	7591.0	63.0	6.1	5.3	NA	Acceptable
White	High MOE	564	3406.2	42.1	36.8	10.5	NA	NA
White	Acceptable	240	3851.7	83.4	6.1	2.9	NA	NA

Pattern Analysis: Since ACS provides an estimate (as opposed to a more exact count from the Census) and census tracts are a pretty granular level to analyze, it makes sense that there are some substantial margins of error. However, I was surprised to see that all 804 census tracts in this analysis were flagged for having at least one MOE of any race greater than 15%. In further pattern analysis, Black & Hispanic tracts have significantly more flagged as having High Margins of Error (there is only 1! Latino tract with an Acceptable MOE!). This means Black and Latino residents of these 3 counties are significantly worse represented in the data.

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: A common pattern across the four analyses is that data quality issues are not random, but are concentrated around certain communities. Black and Latino communities and tracts with smaller populations have significantly higher margins of error. By initially flagging for a tract having potentially unreliable data when any of the three analyzed racial groups’ (white, Black, Latino) margin of error was greater than 15%, this encompassed all 804 included tracts, suggesting a high-level of data unreliability. However, further steps looking into the data more granularly very quickly highlighted that the margin of error is extremely high for Latino (>99%) and Black (>98%) tracts. This means, amongst other concerns, that data quality issues specifically related to these racial groups may be overlooked or unaddressed. Following this, Black and Latino communities face the greatest risk of algorithmic bias. The data attributed to these groups in the ACS surveys are estimates with extremely high margins of error and rates of margins of error. The level of uncertainty and how widespread this quality issue spreads both signify that these data are unlikely to reflect the true population. And, with a less accurate picture of these communities, plans and community action are more likely to overlook and/or plan without adequate consideration for these groups. Underlying factors that may drive data quality issues and/or bias risk include the data collection methods. As previously stated, the ACS is a survey of citizens conducted more frequently than the decennial census. Thus, the ACS data are based on a sample of the complete population. When estimating the demographics of, for example, smaller communities (the tract level is relatively small and hard to encapsulate) can have larger margins of error just based on the sample size and how the sample is translated to different geographies. Additionally, groups that have been historically marginalized are less likely to be well-covered in this data and therefore may be marginalized further. Despite its challenges, ACS data is a significant opportunity for understanding change over time and dates more frequent than the Census. It is, however, absolutely necessary for the Department to address these systematic issues, in order to utilize this data as a tool without perpetuating certain groups’ marginalization. The department can make use of the flagging mechanisms used here to set a threshold for when a Margin Of Error is significant; when data are flagged for having high margins of error, this can be noted and taken into account in planning. Specifically, when high MOEs are flagged, either further data collection/research can be produced or the data can be not the primary driver of any planning decisions.

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

county_incomepop

# A tibble: 67 × 10
   GEOID NAME    total_popE total_popM median_incomeE median_incomeM county_name
   <chr> <chr>        <dbl>      <dbl>          <dbl>          <dbl> <chr>      
 1 42001 Adams …     104604         NA          78975           3334 Adams      
 2 42003 Allegh…    1245310         NA          72537            869 Allegheny  
 3 42005 Armstr…      65538         NA          61011           2202 Armstrong  
 4 42007 Beaver…     167629         NA          67194           1531 Beaver     
 5 42009 Bedfor…      47613         NA          58337           2606 Bedford    
 6 42011 Berks …     428483         NA          74617           1191 Berks      
 7 42013 Blair …     122640         NA          59386           2058 Blair      
 8 42015 Bradfo…      60159         NA          60650           2167 Bradford   
 9 42017 Bucks …     645163         NA         107826           1516 Bucks      
10 42019 Butler…     194562         NA          82932           2164 Butler     
# ℹ 57 more rows
# ℹ 3 more variables: MOE_per <dbl>, reliability <chr>, Low_Flag <int>

county_incomepop <- county_incomepop %>%
  mutate(
    algorithm_recommendation = case_when(
      reliability == "High" ~
        "Safe for algorithmic decisions",

      reliability == "Moderate" ~
        "Use with caution – monitor outcomes",

      reliability == "Low" ~
        "Requires manual review or additional data",

      TRUE ~ NA_character_
    )
  )
county_algorithm <- county_incomepop %>%
  select(county_name,median_incomeE,MOE_per,reliability,algorithm_recommendation)

# 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()
kable(
  county_algorithm,
  digits = 1,
  caption = "County-Level Data Reliability and Algorithm Implementation Recommendations"
)

County-Level Data Reliability and Algorithm Implementation Recommendations
county_name	median_incomeE	MOE_per	reliability	algorithm_recommendation
Adams	78975	4.2	High	Safe for algorithmic decisions
Allegheny	72537	1.2	High	Safe for algorithmic decisions
Armstrong	61011	3.6	High	Safe for algorithmic decisions
Beaver	67194	2.3	High	Safe for algorithmic decisions
Bedford	58337	4.5	High	Safe for algorithmic decisions
Berks	74617	1.6	High	Safe for algorithmic decisions
Blair	59386	3.5	High	Safe for algorithmic decisions
Bradford	60650	3.6	High	Safe for algorithmic decisions
Bucks	107826	1.4	High	Safe for algorithmic decisions
Butler	82932	2.6	High	Safe for algorithmic decisions
Cambria	54221	3.3	High	Safe for algorithmic decisions
Cameron	46186	5.6	Moderate	Use with caution – monitor outcomes
Carbon	64538	5.3	Moderate	Use with caution – monitor outcomes
Centre	70087	2.8	High	Safe for algorithmic decisions
Chester	118574	1.7	High	Safe for algorithmic decisions
Clarion	58690	4.4	High	Safe for algorithmic decisions
Clearfield	56982	2.8	High	Safe for algorithmic decisions
Clinton	59011	3.9	High	Safe for algorithmic decisions
Columbia	59457	3.8	High	Safe for algorithmic decisions
Crawford	58734	3.9	High	Safe for algorithmic decisions
Cumberland	82849	2.2	High	Safe for algorithmic decisions
Dauphin	71046	2.3	High	Safe for algorithmic decisions
Delaware	86390	1.5	High	Safe for algorithmic decisions
Elk	61672	6.6	Moderate	Use with caution – monitor outcomes
Erie	59396	2.6	High	Safe for algorithmic decisions
Fayette	55579	4.2	High	Safe for algorithmic decisions
Forest	46188	10.0	Moderate	Use with caution – monitor outcomes
Franklin	71808	3.0	High	Safe for algorithmic decisions
Fulton	63153	3.6	High	Safe for algorithmic decisions
Greene	66283	6.4	Moderate	Use with caution – monitor outcomes
Huntingdon	61300	4.7	High	Safe for algorithmic decisions
Indiana	57170	4.6	High	Safe for algorithmic decisions
Jefferson	56607	3.4	High	Safe for algorithmic decisions
Juniata	61915	4.8	High	Safe for algorithmic decisions
Lackawanna	63739	2.6	High	Safe for algorithmic decisions
Lancaster	81458	1.8	High	Safe for algorithmic decisions
Lawrence	57585	3.1	High	Safe for algorithmic decisions
Lebanon	72532	2.7	High	Safe for algorithmic decisions
Lehigh	74973	2.0	High	Safe for algorithmic decisions
Luzerne	60836	2.4	High	Safe for algorithmic decisions
Lycoming	63437	4.4	High	Safe for algorithmic decisions
McKean	57861	4.7	High	Safe for algorithmic decisions
Mercer	57353	3.6	High	Safe for algorithmic decisions
Mifflin	58012	3.4	High	Safe for algorithmic decisions
Monroe	80656	3.2	High	Safe for algorithmic decisions
Montgomery	107441	1.3	High	Safe for algorithmic decisions
Montour	72626	7.1	Moderate	Use with caution – monitor outcomes
Northampton	82201	1.9	High	Safe for algorithmic decisions
Northumberland	55952	2.7	High	Safe for algorithmic decisions
Perry	76103	3.2	High	Safe for algorithmic decisions
Philadelphia	57537	1.4	High	Safe for algorithmic decisions
Pike	76416	4.9	High	Safe for algorithmic decisions
Potter	56491	4.4	High	Safe for algorithmic decisions
Schuylkill	63574	2.4	High	Safe for algorithmic decisions
Snyder	65914	5.6	Moderate	Use with caution – monitor outcomes
Somerset	57357	2.8	High	Safe for algorithmic decisions
Sullivan	62910	9.3	Moderate	Use with caution – monitor outcomes
Susquehanna	63968	3.1	High	Safe for algorithmic decisions
Tioga	59707	3.2	High	Safe for algorithmic decisions
Union	64914	7.3	Moderate	Use with caution – monitor outcomes
Venango	59278	3.4	High	Safe for algorithmic decisions
Warren	57925	5.2	Moderate	Use with caution – monitor outcomes
Washington	74403	2.4	High	Safe for algorithmic decisions
Wayne	59240	4.8	High	Safe for algorithmic decisions
Westmoreland	69454	2.0	High	Safe for algorithmic decisions
Wyoming	67968	3.9	High	Safe for algorithmic decisions
York	79183	1.8	High	Safe for algorithmic decisions

Key Recommendations:

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

Counties suitable for immediate algorithmic implementation: PA has primarily counties with high confidence based on these parameters. Allegheny and Philadelphia counties both fall into this category. These counties’ data are solid at the county level, but as seen in this analysis, planners must be aware of the groups who are hidden and whose data is not flagged at the county level despite its low confidence at smaller geographies.
Counties requiring additional oversight: Cameron, Carbon, Elk, Forest (ANALYSIS), Greene, Montour, Snyder, SUllivan, Union, and Warren counties have moderate reliability data. This means that the data may be okay, but planners relying on this data should be aware of this flag and be prepared to adapt or complete further data collection.
Counties needing alternative approaches: PA has no counties with low confidence based on these parameters. However, since all 804 PA tracts analyzed across three counties were found to be unreliable, planners should be aware of the subgroups within counties that may be overlooked or misrepresented by seemingly high-confidence data.

Questions for Further Investigation

To what extent is it possible to most-completely overcome algorithmic bias? How much does the confidence TRULY increase at a scale such as county as compared to tracts without continuing to miss the same overlooked, marginalized groups?

Technical Notes

Data Sources: - U.S. Census Bureau, American Community Survey 2018-2022 5-Year Estimates - Retrieved via tidycensus R package on in February 2026

Reproducibility: - All analysis conducted in R version 2025.05.01+513 - Census API key required for replication - Complete code and documentation available at: https://mamole27.github.io/PPA_Portfolio/

Methodology Notes: Thresholds and parameters followed explicitly given guidelines in this assignment.

Limitations: Analysis based on a small selection of counties for no particular reason.

Submission Checklist

Before submitting your portfolio link on Canvas:

All code chunks run without errors
All “[Fill this in]” prompts have been completed
Tables are properly formatted and readable
Executive summary addresses all four required components
Portfolio navigation includes this assignment
Census API key is properly set
Document renders correctly to HTML

Remember: Submit your portfolio URL on Canvas, not the file itself. Your assignment should be accessible at your-portfolio-url/labs/lab_1/your_file_name.html