Summarize all of your analysis and results in the form of a memo.Discuss the strengths, and limitations, of this kind of analysis for demonstrating a cause-effect relationship.

Regression Analysis (memo format)

In this project, you will do some hands-on analysis of real data to examine a possible cause-effect relationship with policy implications. You will use data from the 2019 US Cities Sustainable Development Report, which is provided to you already formatted and ready to use.

You’ll use Excel to analyze the data. Here’s what you need to get started:

Data

The data for this project comes from the 2019 US Cities Sustainable Development Report, a UN-affiliated initiative that compiled 57 indicators of city performance across a wide-variety of topics, from incarceration to clean water, and from early education to transit, for the 105 largest US cities. Two files are important for understanding and analyzing these data:

• The report, which provides background on the project and (importantly) a codebook that defines each of the indicators (see especially the Annex: Sources and Definitions of Indicators, pp. 38-49). This information is important for understanding the variables (indicators) you will be working with.

• The data, which is a CSV file with 57 indicators (the columns) for 105 cities (the rows).

Instructions

For this project, you will use real data and software to identify a possible cause-effect relationship (X-Y) between two variables of interest to you.

Specifically, you will use the US cities data (see above) and Excel to do the following:

1. Choose two variables in the data that represent a possible causal relationship (X-Y) of interest to you. Begin by visualizing the relationship using a scatter plot.

2. Next, run a simple (bivariate) regression analysis of the relationship. Interpret the slope, statistical significance (t-test, p-value), and R-squared.

3. Finally, identify a third variable that may be a potential common cause of both your X and Y variables and include this variable in a multiple regression analysis (so that you “control for” its influence). Interpret the difference between these new results and your simple (bivariate) results.

Summarize all of your analysis and results in the form of a memo. Be sure to discuss the strengths, and limitations, of this kind of analysis for demonstrating a cause-effect relationship.

Last Completed Projects