Acct 426 Project 2

The purpose of this analysis is to determine the relation between spending and math proficiency scores in West Virginia. Used data includes standardized test scores from the 2022 school years and spending data from the 2022 school year. I believed that spending will be highly positively correlated to test scores. However, after running a regression, it was evident that there is a little correlation between the two variables, posting an R^2 adjusted value of only .30. Based on this, a recommendation can not be made at this time.

Data

The model used the following variables to predict proficiency scores. Spending: Found by dividing total expenditures by enrollment. Enrollment: The number of students enrolled in the school. Total State Revenue: The total amount of revenue generated by the state. Total Local Revenue: The total amount of revenue generated by the local area. Total Expenditures: The total amount spent on schools. Per Pupil Expenditures: The total amount spent on each student. In regards to the test score data, it was truncated to only include overall math proficiency scores. While for the spending data, specific vendors were focused on to identify spending.

Methods

This regression model sought to determine the relationship between spending and math proficiency scores in WV. Using proficiency (test scores) as our dependent variable, we used spending, enrollment, state revenue, local revenue, and total expenditures as our independent variables. For a second method, k-means was implemented to determine if there were any clusters of counties that had similar spending and proficiency scores. Two loose clusters were found, with one cluster having a higher proficiency score and spending than the other.

## 
## Regression Results
## =========================================
##                       Dependent variable:
##                       -------------------
##                           Proficiency    
## -----------------------------------------
## Spending               20.929*** (5.213) 
## Enrollment               0.462 (0.316)   
## State Revenue           -0.002 (0.003)   
## Local Revenue           0.001* (0.0003)  
## Total Expenditure      0.001*** (0.0002) 
## Per Pupil Expenditure -0.0003** (0.0001) 
## -----------------------------------------
## Observations                  54         
## R2                           0.368       
## Adjusted R2                  0.302       
## =========================================
## Standard errors in parentheses

Limitations

The main limitation is that no socioeconomic factors were incorporated into the analysis. Further tests should be run with said data before any conclusion is made.

References

Copilot was used to generate correlation plot and stargazer code for this analysis.