Week 6- Continuing Analysis and Beginning to Make Interpretations

This week, I continued to perform linear regressions with my data. Similar to my hardships last week, I continued to face formatting issues that would not allow my calculations to run smoothly, but after making multiple further changes to my data, I got the hang of it. Though the software I am using can be complex, from the tutorials I’ve watched, I’ve been able to implement a regression simple enough for beginners. I am analyzing my dataset as a time series because I am observing my variables, such as consumer spending and unemployment patterns, on a monthly basis and treating October and November of 2025 as a shock point, since that is the time immediately following the initiation of the 2025 government shutdown. Thus, I am not just using the single event of the shutdown as a regressor, but rather specifically using the period of time following the shutdown as a regressor against my multitude of dependent variables.

So far, I have found extremely low p-values, meaning that there is a statistically significant discontinuity in the trend of my variables following the beginning of the shutdown. This is not enough information to establish a causal relationship between my dependent variables, as there are hundreds of confounding factors I can think of, such as seasonal jobs or expenses, that could influence my regression. However, I do think that my calculations still illustrate a strong correlation and can be used to predict the outcomes of future shutdowns. Gretl also has a feature that has allowed to create visual plots of my data with a line showcasing when the shutdown started. I hope to incorporate these graphs into my final paper and presentation so that readers and audience members can easily visualize the conclusions that I make as I explain them.