This lab focuses on misleading and biased graphics. In order to recognize when you are seeing them in the broader world, it can be helpful to make them yourself.

Pre-Lab Readings

https://www.horace.org/blog/wp-content/uploads/2012/05/How-to-Lie-With-Statistics-1954-Huff.pdf

Note: This book is a product of the 1950s, some of the language used, especially when referring to minorities in inappropriate. I tried to skip those pages while selecting the pages that drive home the point of how to lie with stats.

  1. Read the intro: Pages 7-9
  2. Chapter 1: Pages 11-13
  3. Chapter 2: Pages 27-36 (all of it)
  4. Chapter 3: Pages 37-40
  5. Chapter 5: Pages 60-65, (all of it)
  6. Chapter 6: Pages 68-70
  7. Chapter 7: Pages 74, 80-83 (Trigger warning 81)
  8. Chapter 8: Page 87 (correlation is not causation)
  9. Chapter 9: Pages 100, 102-105, 110-111, 118-119
  10. Total pages: 3+3+10+4+6+3+1+4+1+1+4+2+2=44

\(~\)

Directions (Please read before starting)

  1. This lab is not as coding-intensive as labs we’ve had over the past few weeks. With that in mind, your written answers should be thoughtful and well articulated. Lack of effort in answers or unconvincing arguments (see below) will earn a non-satisfactory for this lab.
  2. Please record your answers and any related code for all embedded lab questions. I encourage you to try out the embedded examples, but you shouldn’t turn them in.
  3. Please ask for help, clarification, or even just a check-in if anything seems unclear.

\(~\)

Preamble

Packages and Datasets

library(tidyverse)
library(ggplot2)

This lab focuses on misleading and biased graphics. In order to recognize when you are seeing them in the broader world, it can be helpful to make them yourself. If you did the readings, you should have learned about ways to manipulate data including statisticulating (103), 1 dimensional figures (68-69), and shifting axes. Today you are going to work with data to tell contradicting stories.

\(~\)

Lab: Friday

At this point you should begin working on the lab. You are free to work in groups of 1-3 (your choice of partner). If you do not choose to have a partner you must explain your answers to questions 1 and 2 with Prof. Friedrichsen or the course mentor before moving on.

\(~\)

Question 1 (10min) Working with your partner: what are other techniques that you learned about for misleading/lying with statistics from the reading?

\(~\)

Question 2 (10min) Working with your partner or the nearest other group or someone around you: What are examples of misleading graphics/figures/etc that you have seen recently in the news? Try to find these examples again.

\(~\)

Question 3 Using the Diamonds dataset, and any outside sources that you want to use, come up with statistically based arguments for the following: (that is, use statistical techniques, graphics, etc)

  1. Diamonds are horribly expensive to use in engagement rings (>$5000)
  2. Diamonds are reasonably priced to use in engagement rings (<$1500)

Extra info: https://www.rdocumentation.org/packages/ggplot2/versions/3.3.3/topics/diamonds

Note: Your argument should be constructed in a way that distorts the story or uses claims to mislead. Straight lying or changing data is not the intent.

\(~\)

Question 4 (As time permits) Using any data sources that you can find, create two figures with associated 1 paragraph persuasive arguments:

  1. Inflation in the US is a serious problem right now and should be a high priority.
  2. Inflation in the US is not a problem and we should be worried about other things.

\(~\)

Question 5 NOT GRADED: Start thinking about figures in the news that you could recreate to be less misleading for Next Monday.

The following has been added for Monday

Question 6 (The entire lab). Find 3-5 examples of misleading graphics in the news. For each graphic do the following

  1. Provide a link to it (citations are important)
  2. Find a way to insert the image into RMarkdown
  3. Explain what is misleading about it (2 sentences - 1 paragraph)
  4. Give your opinion on why the information was presented in that way (who benefits from the slant) (1 paragraph)
  5. Explain how you would make the image more neutral (1-2 sentences)
  6. Explain how you could use the same data to tell the opposite story (1-2 sentences)
  7. If possible, create the figures for 5 and 6