Week 8 Blog- Furthering My Analysis and Making Conclusions

As I continued to run regressions and calculations this week, I noticed that I was often unable to obtain statistically significant results. After doing some testing to figure out the problem, I realized my degrees of freedom were too low because I was using such a small sample size. Thus, I made the tedious and long decision to expand my data set. Instead of merely starting one year before the beginning of the shutdown, I started my range from January 2022. Although it would have been beneficial to expand my range even further, I believe the outlier of the 2020 pandemic, which prolonged into 2021, would skew the trend of my data. I included this as a limitation in the conclusion of my final product draft.

Additionally, since time has passed since I first began collecting data, the data for the entire first quarter of 2026 has been released, so I was able to incorporate that into my dataset as well, although I had originally only had January of 2026. By making these changes, I was able to produce results that were more meaningful and complex. I found that many variables, based on whether they affect people in the short term or long term, would similarly usually only have a significant p-value for D, a variable showcasing the instance of the shutdown, or D_t, representing the period following the shutdown. While personal income and unemployment claims were instantly changed, variables like consumer confidence and news or sentiment data took longer to gain traction as the effects of the shutdown spread. I was able to talk in detail about this observation in the analysis section of my final product draft.