Outside Event #2

.docx

School

Virginia Tech *

*We aren’t endorsed by this school

Course

5114

Subject

Statistics

Date

Feb 20, 2024

Type

docx

Pages

2

Uploaded by dinoman639

Report
Outside Event #2 The name of the event that I attended was “What are the most important statistical ideas of the past 50 years?” and the speaker for this event was Dr. Andrew Gelman. Dr. Gelman is the Higgins Professor of Statistics and Professor of Political Science at Columbia University. This event took place on Friday, March 5 th , 2021 from 10:30 to 11:30 am. I did not attend the event but watched the livestream on the YouTube channel on March 19 th , 2021. Dr. Gelman started the event by sharing a story about how he worked with someone from Wikipedia regarding the experiments they have done to try and improve the rate at which people give money. He then quotes his old teacher, “the most important aspect of a statistical method, is not what it does with the data but what data it uses” and then states that what follows is “what makes a statistical method effective is that it allows you to use more data effectively.” The quote from his teacher is intriguing. When I look at statistics, I mainly look at the result and this can sometimes lead me and others down the wrong path when the data that it uses is flawed in some way. To get accurate results and statistics, the data that is used must have the qualities of a good dataset. My question for Dr. Gelman is regarding the quote he made. I don’t understand what he means by the effectiveness of a statistical method is reliant on whether it can use more data effectively. I understood the first part of that statement, but the second half left me clueless. Dr. Gelman goes on to share a screenshot of the results of the A/B Testing Statistics that the person from Wikipedia sent him. They compared the differences between a control and square corners for the donation box but there were blatant errors in the trails. The control had around 500,000 fewer trails than the square corners did. Dr. Gelman told them that their hypothesis was wrong, and their experiment was not a completely randomized. Later on, he talks about how it took decades for the establishment of statistics to be comfortable with regularization. At first, I did not understand what Dr. Gelman meant by this, so I took the time to research what regularization was. After understanding that regularization is the technique of generalizing a model, I was surprised that it took so long for it to be accepted. When I see a plot on points on a graph, I can quickly generalize what the function would look like to create that plot but back then, it seems that it was not the case. Nearing the end of the event, Dr. Gelman started talking about Exploratory Data Analysis and Modeling. As a data
science student, anything that can give me further insight on what I will be doing is something I greatly appreciate. It was interesting to hear that that he believes better EDA can help one make better models and vice versa. He says that “EDA is all about discovering the unexpected” which I find very interesting. Another thing I wish I could have asked him is how relevant is Data Science in what he does. Learning more about my major is very important to me as I still have a lot of questions about it.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help