Ethics in Statistics and Data Science

STA 199

Bulletin

• this ae is not due for grade.
• Project presentations and repo due Monday lab (by pushing to GitHub)
• Extended deadline: project reports now due Wednesday
• Don’t forget about the statistics experience

Getting started

Clone your ae25-username repo from the GitHub organization.

Today

By the end of today you will…

• critically examine graphics, models and results
• discuss data privacy and redundancy
• analyze a real data example of Simpson’s paradox

Load packages

library(tidyverse)
library(tidymodels)

Guidelines for Discussion

• Respect others. Be charitable towards each other. Speak calmly.

Data Representation

Misleading Data Visualizations¹

Brexit

Brexit

• What is the graph trying to show?

• Why is this graph misleading?

• How can you improve this graph?

Spurious Correlations²

A Spurious Correlation

• What is the graph trying to show?

• Why is this graph misleading?

Statistical modeling

Read, with a critical eye, page 2 and table 1 from Physician–patient racial concordance and disparities in birthing mortality for newborns and chat with your neighbor.

Data privacy

Web scraping³

A data analyst received permission to post a data set that was scraped from a social media site. The full data set included name, screen name, email address, geographic location, IP (Internet protocol) address, demographic profiles, and preferences for relationships. The analyst removes name and email address from the data set in effort to deidentify it.

• Why might it be problematic to post this data set publicly?

• How can you store the full dataset in a safe and ethical way?

• You want to make the data available so your analysis is transparent and reproducible. How can you modify the full data set to make the data available in an ethical way?

Redundancy

• slides

Additional readings

• Why pokemon go’s plan to 3d scan the world is dangerous

• How companies learn your secrets

Discussion questions

• “Simpson’s paradox”, where conclusions drawn from analyzing subgroups differ from conclusions drawn when the groups are combined. Can you demonstrate Simpson’s Paradox with the data below? ⁴

• For further reading see Bickel, Peter J., Eugene A. Hammel, and J. William O’Connell. “Sex Bias in Graduate Admissions: Data from Berkeley: Measuring bias is harder than is usually assumed, and the evidence is sometimes contrary to expectation.” Science 187.4175 (1975): 398-404.

berk = read_csv("https://sta101.github.io/static/appex/data/BerkeleyAdmissionsData.csv")

Rows: 7 Columns: 5
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (1): Dept
dbl (4): MaleYes, MaleNo, FemaleYes, FemaleNo

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

berk

# A tibble: 7 × 5
  Dept  MaleYes MaleNo FemaleYes FemaleNo
  <chr>   <dbl>  <dbl>     <dbl>    <dbl>
1 A         512    313        89       19
2 B         313    207        17        8
3 C         120    205       202      391
4 D         138    279       131      244
5 E          53    138        94      299
6 F          22    351        24      317
7 All      1158   1493       557     1278

Footnotes

1. Source: https://humansofdata.atlan.com/2019/02/dos-donts-data-visualization

2. Source: https://www.tylervigen.com/spurious-correlations Content warning: some examples include death or suicide.

3. Modified from Modern Data Science with R, 2nd Edition

4. Source: https://www.randomservices.org/random/data/Berkeley.html