WGU D606 Advanced Data Analytics - Task 2 Capstone
Report
Predictive Analysis of State Correctional Spending: A Multi-Variable
Regression Study of Recidivism, Reentry Investment, and Operational
Cost Factors
Student Name: Shanikwa Haynes
Course: D606 – Data Science Capstone
Date: July 24, 2025
Capstone Project Name: Predictive Analysis of State Correctional Spending: A Multi-
Variable Regression Study of Recidivism, Reentry Investment, and Operational Cost
Factors
Project Topic: Statistical Analysis of Correctional Spending Patterns and Recidivism
Outcomes Across U.S. States
IRB Status: ☒ This project does not involve human subjects research and is exempt from
WGU IRB review.
A. Research Question
Summary of Research Question
The research question identified for this capstone project is: “What factors significantly
predict total correctional spending across U.S. states, and how do recidivism rates,
reentry investments, and operational cost components contribute to overall
correctional expenditures?”
Justification for Research Question
This research question is justified by the critical need for evidence-based policy decisions
in correctional management. State correctional systems face mounting pressure to reduce
recidivism rates while managing increasing inmate populations and limited budget
resources. With annual correctional spending exceeding $80 billion nationally,
understanding the relationship between spending patterns and outcomes is critical for
policy development and resource optimization.
This analysis addresses a significant gap in correctional policy research by providing
quantitative evidence of how different spending categories relate to both total
expenditures and recidivism outcomes. The research will benefit state correctional
,administrators, policymakers, and budget analysts by identifying which spending
strategies provide the best return on investment for reducing recidivism and improving
public safety. Current policy debates about prison reform and rehabilitation funding lack
comprehensive statistical analysis of spending effectiveness across multiple states. This
study will provide evidence-based recommendations for optimal resource allocation in
correctional systems, potentially influencing policy decisions that affect thousands of
inmates and millions of taxpayer dollars annually.
Description of Context
The context for this research exists within the broader landscape of criminal justice reform
and correctional policy in the United States. State correctional systems are confronted
with multiple competing priorities: maintaining public safety, reducing recidivism rates,
managing increasing inmate populations, and operating within constrained budget
environments. The financial stakes are substantial, with correctional spending
representing a significant portion of state budgets and directly impacting taxpayer
resources.
The research context is particularly relevant given ongoing policy debates about the
effectiveness of traditional incarceration versus rehabilitation-focused approaches. States
vary significantly in their spending patterns and approaches to correctional management,
providing a natural laboratory for examining which strategies produce the most cost-
effective outcomes. This analysis capitalizes on this variation to identify evidence-based
best practices for correctional resource allocation.
Discussion of Hypothesis
Primary Hypothesis: This study will test the hypothesis that spending patterns in specific
correctional categories (reentry investment, general operations, medical costs, and
educational programs) significantly predict total state correctional spending, with reentry
investments showing an inverse relationship to recidivism rates.
Statistical Hypotheses:
Null Hypothesis (H₀): There is no significant relationship between recidivism rates, reentry
investment spending, operational costs (general operations, medical, educational), and
total correctional spending across U.S. states. - Formally expressed as: β₁ = β₂ = β₃ = β₄ = β₅
=0
Alternative Hypothesis (H₁): At least one of the predictor variables (recidivism rates,
reentry investment spending, general operations, medical costs, or educational program
spending) is a significant predictor of total correctional spending across U.S. states. -
Formally expressed as: At least one β ≠ 0
This hypothesis is plausible given existing criminological research suggesting that
comprehensive rehabilitation programs, including reentry support, are associated with
reduced recidivism rates and potentially more cost-effective correctional systems. The
, hypothesis aligns with evidence-based practices in correctional management and is
testable using multiple linear regression analysis with the available state-level data.
B. Data Collection
Description of Data Collected
The dataset utilized for this analysis contains correctional spending and recidivism data for
all 50 U.S. states. The comprehensive dataset includes seven key variables providing a
complete picture of state correctional operations and outcomes:
1. State: Categorical identifier for each of the 50 U.S. states
2. Recidivism_Rate: Percentage of released inmates returning to incarceration within
three years (continuous variable)
3. Reentry_Investment: State funding for reentry programs and services (millions
USD)
4. General_Operations: Basic facility operations and staffing costs (millions USD)
5. Medical_Costs: Healthcare expenditures for inmates (millions USD)
6. Educational_Programs: Vocational and academic training investments (millions
USD)
7. Total_Spending: Combined correctional spending across all categories (millions
USD)
The data represents fiscal year spending amounts and standardized three-year recidivism
measurement criteria, providing a comprehensive view of state correctional spending
patterns and outcomes. This data is compiled from publicly available state correctional
department reports and federal Bureau of Justice Statistics publications, consisting of
aggregate, non-personal information routinely published by state governments for public
transparency and accountability purposes.
Advantage and Disadvantage of Data-Gathering Methodology
Methodology Used: Secondary data analysis of officially reported state correctional
expenditures and recidivism statistics compiled from state departments of corrections
annual reports, federal Bureau of Justice Statistics publications, and standardized state
financial reporting systems.
Advantage: The primary advantage of this data-gathering methodology is the high
reliability and standardization it provides through official government verification. This
approach ensures data credibility through established governmental reporting
mechanisms while enabling comprehensive cross-state comparisons using consistent
measurement criteria and reporting periods. The methodology eliminates potential bias
Report
Predictive Analysis of State Correctional Spending: A Multi-Variable
Regression Study of Recidivism, Reentry Investment, and Operational
Cost Factors
Student Name: Shanikwa Haynes
Course: D606 – Data Science Capstone
Date: July 24, 2025
Capstone Project Name: Predictive Analysis of State Correctional Spending: A Multi-
Variable Regression Study of Recidivism, Reentry Investment, and Operational Cost
Factors
Project Topic: Statistical Analysis of Correctional Spending Patterns and Recidivism
Outcomes Across U.S. States
IRB Status: ☒ This project does not involve human subjects research and is exempt from
WGU IRB review.
A. Research Question
Summary of Research Question
The research question identified for this capstone project is: “What factors significantly
predict total correctional spending across U.S. states, and how do recidivism rates,
reentry investments, and operational cost components contribute to overall
correctional expenditures?”
Justification for Research Question
This research question is justified by the critical need for evidence-based policy decisions
in correctional management. State correctional systems face mounting pressure to reduce
recidivism rates while managing increasing inmate populations and limited budget
resources. With annual correctional spending exceeding $80 billion nationally,
understanding the relationship between spending patterns and outcomes is critical for
policy development and resource optimization.
This analysis addresses a significant gap in correctional policy research by providing
quantitative evidence of how different spending categories relate to both total
expenditures and recidivism outcomes. The research will benefit state correctional
,administrators, policymakers, and budget analysts by identifying which spending
strategies provide the best return on investment for reducing recidivism and improving
public safety. Current policy debates about prison reform and rehabilitation funding lack
comprehensive statistical analysis of spending effectiveness across multiple states. This
study will provide evidence-based recommendations for optimal resource allocation in
correctional systems, potentially influencing policy decisions that affect thousands of
inmates and millions of taxpayer dollars annually.
Description of Context
The context for this research exists within the broader landscape of criminal justice reform
and correctional policy in the United States. State correctional systems are confronted
with multiple competing priorities: maintaining public safety, reducing recidivism rates,
managing increasing inmate populations, and operating within constrained budget
environments. The financial stakes are substantial, with correctional spending
representing a significant portion of state budgets and directly impacting taxpayer
resources.
The research context is particularly relevant given ongoing policy debates about the
effectiveness of traditional incarceration versus rehabilitation-focused approaches. States
vary significantly in their spending patterns and approaches to correctional management,
providing a natural laboratory for examining which strategies produce the most cost-
effective outcomes. This analysis capitalizes on this variation to identify evidence-based
best practices for correctional resource allocation.
Discussion of Hypothesis
Primary Hypothesis: This study will test the hypothesis that spending patterns in specific
correctional categories (reentry investment, general operations, medical costs, and
educational programs) significantly predict total state correctional spending, with reentry
investments showing an inverse relationship to recidivism rates.
Statistical Hypotheses:
Null Hypothesis (H₀): There is no significant relationship between recidivism rates, reentry
investment spending, operational costs (general operations, medical, educational), and
total correctional spending across U.S. states. - Formally expressed as: β₁ = β₂ = β₃ = β₄ = β₅
=0
Alternative Hypothesis (H₁): At least one of the predictor variables (recidivism rates,
reentry investment spending, general operations, medical costs, or educational program
spending) is a significant predictor of total correctional spending across U.S. states. -
Formally expressed as: At least one β ≠ 0
This hypothesis is plausible given existing criminological research suggesting that
comprehensive rehabilitation programs, including reentry support, are associated with
reduced recidivism rates and potentially more cost-effective correctional systems. The
, hypothesis aligns with evidence-based practices in correctional management and is
testable using multiple linear regression analysis with the available state-level data.
B. Data Collection
Description of Data Collected
The dataset utilized for this analysis contains correctional spending and recidivism data for
all 50 U.S. states. The comprehensive dataset includes seven key variables providing a
complete picture of state correctional operations and outcomes:
1. State: Categorical identifier for each of the 50 U.S. states
2. Recidivism_Rate: Percentage of released inmates returning to incarceration within
three years (continuous variable)
3. Reentry_Investment: State funding for reentry programs and services (millions
USD)
4. General_Operations: Basic facility operations and staffing costs (millions USD)
5. Medical_Costs: Healthcare expenditures for inmates (millions USD)
6. Educational_Programs: Vocational and academic training investments (millions
USD)
7. Total_Spending: Combined correctional spending across all categories (millions
USD)
The data represents fiscal year spending amounts and standardized three-year recidivism
measurement criteria, providing a comprehensive view of state correctional spending
patterns and outcomes. This data is compiled from publicly available state correctional
department reports and federal Bureau of Justice Statistics publications, consisting of
aggregate, non-personal information routinely published by state governments for public
transparency and accountability purposes.
Advantage and Disadvantage of Data-Gathering Methodology
Methodology Used: Secondary data analysis of officially reported state correctional
expenditures and recidivism statistics compiled from state departments of corrections
annual reports, federal Bureau of Justice Statistics publications, and standardized state
financial reporting systems.
Advantage: The primary advantage of this data-gathering methodology is the high
reliability and standardization it provides through official government verification. This
approach ensures data credibility through established governmental reporting
mechanisms while enabling comprehensive cross-state comparisons using consistent
measurement criteria and reporting periods. The methodology eliminates potential bias
