Regression Interpretation
This assignment is intended to provide you practice conducting regressions in R and interpreting the regression output. Along with this Word file in the Assignment in Canvas, you will find an Excel file with the data needed to complete this assignment, and another Word file with the regression equations to use. Do not copy and paste from R and RStudio. Construct your answers as if you were preparing a report for a supervisor. When submitting your assignment, include a Word file with your answers to the questions and an R Script file with the code you used to complete the assignment. This assignment is due by 11:59PM, March 6, 2022, and is to be submitted through Canvas.
1. Load the data into R. (No documentation needed for this part)
2. With the full data frame, run the regression equation for absolute value abnormal accruals. In a table report the coefficient estimates, the t values, and p values for the intercept and the independent variables. Under this table report the adjusted R-squared. How much of the variation in absolute value abnormal accruals do the independent variables explain? Which independent variables are statistically significant? Which independent variables are associated with higher values of absolute value abnormal accruals? Which independent variables are associated with lower values of absolute value abnormal accruals?
3. With the full data frame, run the regression equation for the natural log of audit fees. In a table report the coefficient estimates, the t values, and p values for the intercept and the independent variables. Under this table report the adjusted R-squared. How much of the variation in the natural log of audit fees do the independent variables explain? Which independent variables are statistically significant? Which independent variables are associated with higher audit fees? Which independent variables are associated with lower audit fees?
4. Absolute value abnormal accruals is a measure of audit quality. The higher the number the poorer the audit quality. The closer that number is to zero the better the audit quality. The natural log of audit fees in a measure of the audit fees paid. Higher values indicate higher audit fees were paid. The independent variable of interest for this model is PCAOB. PCAOB is an indicator variable valued 1 if the audit firm for the company is registered with the Public Company Accounting Oversight Board (PCAOB) in the U.S. and valued at 0 if the audit firm for the company is not registered with the PCAOB.
(a) What does the first regression say about the audit quality of companies audited by firms registered with the PCAOB compared to companies audited by firms not registered with the PCAOB?
(b) What does the second regression say about the audit fees paid by companies audited by firms registered with the PCAOB compared to companies audited by firms not registered with the PCAOB?
5. Abnormal accruals can be either positive or negative. Abnormal accruals are calculated as a comparison of what the expected accruals are for a company to the company’s actual accruals. Positive abnormal accruals are income increasing and negative abnormal accruals are income decreasing. The variable AbnAcc_W is the signed (positive or negative) abnormal accruals for each observation.
(a) Create a data frame where signed abnormal accruals are only positive. With this new data frame run the regression equation for both positive and negative abnormal accruals. In a table report the coefficient estimates, the t values, and p values for the intercept and the independent variables. Under this table report the adjusted R-squared. Which independent variables are statistically significant? Which independent variables are associated with more positive abnormal accruals? Which independent variables are associated with less positive abnormal accruals?
(b) Create a data frame where signed abnormal accruals are only negative. With this new data frame run the regression equation for both positive and negative abnormal accruals. In a table report the coefficient estimates, the t values, and the p values for the intercept and the independent variables. Under this table report the adjusted R-squared. Which independent variables are associated with more negative abnormal accruals? Which independent variables are associated with less negative abnormal accruals?
6. How do the regression results from Part 5 differ from the regression results from Part 2?
6. Based on the results from the regressions in Part 5, how does contracting with an audit firm that is registered with the PCAOB impact audit quality? Please be specific, do not just answer that it improves audit quality, it does not improve audit quality, or it makes audit quality worse.
7. Based on the analysis you performed in this assignment, if you were in charge of recommending an audit firm to contract with for the company you work for, would you recommend an audit firm registered with the PCAOB? What are the benefits of contracting with a PCAOB registered audit firm? What are the detriments of contracting with a PCAOB registered audit firm? Why did you make the recommendation you did?
Variable Definitions
AbnAcc_W – The signed value of abnormal accruals for the company
AbsAbnAcc – The absolute value of abnormal accruals for the company
Big4 – An indicator variable valued 1 if the company is audited by a Big 4 affiliated firm
CFO_W – Operating cash flows divided by total assets
CrossOTH – An indicator variable valued 1 if the company is listed on a stock exchange outside
China aside from the United States
CrossUS – An indicator variable valued 1 if the company is listed on a stock exchange in the
United States
DTA_W – A measure of leverage calculated as total long-term debt divided by total assets
FOREN – An indicator variable valued 1 if the company reports foreign operations
GC – An indicator variable valued 1 if the company received a going concern report
id – The company identifier
INVREC_W – Inventory and receivables divided by total assets
LIQ_W – The company’s current ratio
LnTOTALfee_W – Total audit fees paid
MAO – An indicator variable valued 1 if the company received a modified audit opinion
PCAOB – An indicator variable valued 1 if the company is audited by a firm registered with the
PCAOB
ROA_W – Return on assets calculated as net income divided by total assets
Size_W – The natural log of total assets
Switch – An indicator variable valued 1 if the company has a new auditor
SWITCHYEAR – An interaction variable calculated as Switch * PCAOB
TOP10 – An indicator variable valued 1 if the company is audited by one of the top 10 audit
firms in China
——–
Interpretation of Regression
This project is designed to give you practice running regressions in R and understanding the results. You will find an Excel file with the data needed to complete this assignment, as well as another Word file with the regression equations to use, in the Assignment in Canvas, along with this Word file. Copy and paste from R and RStudio is not permitted. Prepare your responses as though you were writing a report for a boss. Include a Word file with your answers to the questions and a R Script file with the code you used to complete the assignment when submitting your assignment. This assignment is due at 11:59 p.m. on March 6, 2022, and must be submitted via Canvas.
1. Import the data into R.
