Assignment:
Data Analysis Project
For this project you will demonstrate competency in researching economics; that is, creatively designing a research question, locating pertinent and credible data to support an answer, and presenting results in a professional and articulate manner. Furthermore, you will also be applying fundamental statistical and regression concepts to your data sets to more quantitively answer your research questions. Follow these steps to complete the project:
1. Using the data covered in the National Economy, Wealth Income and Poverty, Business Statistics, Labor Statistics, and Government, generate five research questions to study (one from each category). For this project use causal type phrasing (e.g. "Higher taxes cause lower GDP", "Increased worker productivity increases savings", etc.).
2. Excel File:
A. For three of the five research questions create an Excel sheet with your data set, one graph and the statistical metrics listed below. Compile all the statistical metrics below and use a different type of graph for each question. All statistical metrics and graphs are to be calculated/generated in Excel using the functions reviewed in class.
- Mean (weighted, arithmetic, or geometric)
- Median
- Sample Variance
- Standard Deviation
- Coefficient of Variation
- Range
- Percentiles
- Quintiles
- Skewness
B. SINGLE-VARIABLE REGRESSION: For one of the five research questions create an Excel sheet with your data set, one scatterplot graph, and the analysis output. Furthermore:
• Make sure n ≥ 30 (that is you should have at least 30 data points that correlate across time or space)
• Add the R-squared and trendline to the scatterplot; use the functional form with the highest R-squared.
• Use Excel's Data Analysis TookPak regression function to run a single-variable regression and generate the Analysis Output.
C. MULTIPLE-VARIABLE REGRESSION: For one of the five research questions create an Excel sheet with your data set and the analysis output. Furthermore:
• Make sure n ≥ 30 (that is you should have at least 30 data points that correlate across time or space)
• Use at least six independent variables. Remember, all your independent variables should be believed to influence the dependent variable.
• Use Excel's Data Analysis TookPak regression function to run a multiple-variable regression and generate the Analysis Output.
3. PowerPoint Presentation: For each research question (five total), create at least one PowerPoint slide illustrating the pertinent graphs, statistical metrics, regression results, bullet points (up to 3 and optional), and hyperlinks to your data source website (make sure the links work). The PowerPoint should also contain an introduction slide (e.g. name, project #, and class). For the regressions add at least one slide answering each of the following:
• What is the regression R-Squared and what does it mean regarding your data?
• What are the statistically significant coefficients and how did you conclude they were statistically significant?
• What are the statistically insignificant coefficients and how did you conclude they were insignificant?
• Interpret each statistically significant coefficient to determine how your dependent and independent variables are correlated; that is, for a change in each independent variable how does this impact the dependent variable. Make sure you use the proper denominations (e.g. each square foot added to a home increases the home price by $123). If no statistically significant results are found, then you are to do the same thing but indicate that the results cannot be relied upon.