- What do you see when you run the program?
- Now that you have ran the program and seen it's basic output, what do you think is going on?
- Were the decisions rendered by the program correct or incorrect? How often were they correct? Why do you think so?
- How does your behavior influence the data of the program?
- Were there any trials that heavily skewed your overall results?
- How can you be profiled/tracked based on your online behavior?
- How is entering random letters on a keyboard similar to surfing the web in the way that your information is tracked?
- If you like everything you see on Instagram, without rhyme or reason, what are some possible implications?
- If your data is used incorrectly, for say ad recommendations, what happen? What are the implications associated with making incorrect biases?
- Can you think of a way to reduce potential biases in the algorithms making inferences about users? Do you feel the government should make companies more liable for incorrectly using and tracking user data?
- While this assignment looks to use the idea of hand bias to portray larger biases in online algorithms, given the technology we use is hand bias a real thing? For instance, Instagrams polls always put the "Yes" option on the left side of the screen. Additionally, most phone users seem to put their commonly used apps on the side of the screen where their dominant hand lies.
To import your generated results.csv file from the program, first go to the class Google Sheet.
Then click File, then Import file, then choose your file.
Choose the options as shown in the picture (Append to current sheet, Detect Automatically).
Your results.csv will then be added to the Google Sheet for future analysis. Does anything stick out to you in the populated sheet?
If you want to use the program to gain insights about the class-wide Google Sheet simply export it as a CSV. Then when you run the program, after you run new trials, you will have the option to read in the "class-wide" CSV file for further analysis.