This analysis examines the association between financial reporting quality and women on boards. You should rely on two data sets to analyze this issue. One data contains information about firm financials and the other data contains information about board structures. Both data are shared in Google Drive. Your write up should aim at providing high quality analyses to help a reader to understand how and why financial reporting quality is associated with women on boards. In your write up, please include the following main sections.

Section 1: What is your prediction about the impact of women on boards on firms’ financial reporting quality? Do you believe gender diversity is helpful to mitigate misreporting or improve financial reporting quality? Why? Please note that, to answer the above questions, you are proposing a causal relationship. Please discuss the limitations of a regression approach to test your predictions.

Section 2: Know your data. 2.1 What information is provided in the two data sets? What unique IDs should you use to merge the two datasets? 2.2 After merging the two datasets, provide some sample descriptions. For example, what variables can you use to measure financial reporting quality (possible candidates include Mscore, Fraud indicator, FSD_SCORE, and extreme accruals, etc.)? What is your sample period? How many unique firms are in your sample with non-missing information? How many total observations do you have with all available information? 2.3 To reduce the impact of extreme values, winsorized all continuous variables. Then, provide descriptive statistics for main variables you use such as financial reporting quality variables, women on boards variables, other board characteristics, firm size, leverage, profitability, growth potential, etc. Please include mean, median, and standard deviation for each variable. Are the distributions of these main variables consistent with your expectations? 2.4 After deciding which variables to use to measure financial reporting quality, construct your test variable. Test variable is the independent variable you are mainly interested in. For example, how do you measure women on board? Do you plan to use an indicator variable (1 if at least one woman on board) to capture the existence of female directors? Or a continuous variable such as percentage of female directors? Please note that your choice of the test variable should be consistent with your main argument in Section 1. Please clearly explain your rationale behind your choice of the test variable.

Section 3: Data Analyses 3.1 What are the correlations among your financial reporting quality measures, women on boards measures, and other board characteristics? Are these correlations consistent your expectations? 3.2 Run a simple linear model of Financial Reporting quality on your Women on Boards test variable (only specify two variables in your model). If you have multiple test variables or multiple financial reporting variables, you should run the linear model separately. How do you interpret the results? What omitted correlated variables might bias the estimated coefficient on the Women on Boards variable (i.e. what factors affect financial reporting quality, are NOT in your model, but are potentially correlated with the Women on Boards variable)? 3.3 We often include “control” variables in a linear model? What is the purpose of adding “control” variables? What control variables should be included in your above model to make it more complete? Why? 3.4 What do the results look like after you add the control variables? How to you interpret these results? Are the coefficient estimates of these control variables consistent with your expectations?

Section 4: Conclusions. 3.1 What do you learn from the above data analysis? What do the overall results suggest? 3.2 Recognizing association tests do not answer causal questions, it is still useful to first understand the statistical relationship of your data. Can you provide more discussions about why your reported results CANNOT answer the question “Do women on boards improve financial reporting quality”? What additional information is helpful? Can you propose alternative tests or analyses so that you can actually answer the above question?