1 of 29

​

INTRODUCTION

​

Welcome to SICSS South Florida

​

2 of 29

What to expect this week

  • This week we will meet here
    • Guest lectures, technical talks, group work

​

  • Lightning talks on Friday

​

  • Next week you will be working in your project groups

​

  • Schedule and materials available on the main page of the website

​

​

​

3 of 29

What is computational social science?

4 of 29

Anything that’s cool.

5 of 29

  • Definitions have been wrong

​

  • We aim to dissolve disciplinary silos

​

  • We want to keep the field creative

​

6 of 29

Sure but what is it, Kat?

  • Computational and social scientific

​

  • Involves ethical and privacy questions that are often considered complex

​

  • Toolkit of methods that allows researchers to use large, unstructured, or previously hard to access data

​

​

  • Combines readymades and custommades

​

7 of 29

Readymades vs. Custommades

  • Taking something that was created for one purpose and turning it into something different (creative re-purposing)
    • Digital Trace data
    • Data Scientists

​

  • Data created intentionally
    • Social science

Cleo the Clownfish from the Shedd Aquarium

8 of 29

Is computational social science a fad?

9 of 29

​

No.

​

10 of 29

Bit by Bit figure 1.1 based on Hilbert and Lopez (2011)

11 of 29

Bit by Bit figure 1.1 based on Hilbert and Lopez (2011)

12 of 29

Evidence

13 of 29

Important things to know

14 of 29

Important things to know

  1. You can use any coding language you want
  2. Failure is often more informative than success
  3. Chat GPT is NOT cheating

15 of 29

Ways to use Chat GPT

  • Ask it questions
  • Ask it help generate code
  • Ask it to proof your code for errors
  • Use it to debug your code
  • Use it to optimize your code

​

16 of 29

Ex: Clayton county court records

​

Goal: turn a group of json files into a bar graph that shows which judges hear the most cases in a magistrate court

​

17 of 29

18 of 29

19 of 29

20 of 29

combined_df['judge'] = combined_df['parties'].apply(lambda x: [party['judge'] for party in x])

​

21 of 29

22 of 29

23 of 29

24 of 29

25 of 29

26 of 29

27 of 29

Suggestions

​

  1. Be specific and thorough in your requests

Ex: please write code to make a bar graph with matplot using variable ‘judge’ from combined_df, use times new roman font, green bars, and save it to a .png file with 300 dpi called top5judges.png

​

  • ChatGPT is not infallible

​

  • Work one step at a time

​

28 of 29

29 of 29

Should I just skip straight to ChatGPT?