https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/dannyholt_hw3_new.html

[giscus[bot]]

https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/dannyholt_hw3_new.html

https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/dannyholt_hw3_new.html

[sp-conway]

Great job Danny! You have a great dataset and interesting questions. I would make sure to make them explicit early in the paper for the final project, but of course this was not stated in the HW3 instructions.

Below are some notes I took as I read. Please let me know if you have any questions!

• Great dataset!
• I like how you describe the dataset. In particular, showing the unique employers, etc.
• Employers in descending order of strikes plot
  • This plot is a bit messy. I think it would be better to plot the following
  • X axis - number of strikes
  • Y axis - count
  • Where you are counting the number of employers that have had EACH number of strikes.
  • Given that most employers have had 1 strikes, and only a handful have had more than 3, it would be a far simpler and more informative plot.
• Could you define the term militancy? It also might even be helpful to the reader to explain HOW a strike occurs. That is, requiring a union vote, etc.
• Number of strikes by State
  • I would remove the alternating colors. It’s a little distracting, and the green text is hard to read.
  • I would also sort the bars from highest to lower number of strikes. You’ll have to count strikes yourself, but it’s fairly straightforward. Something like:
    • strike_st %>% count(States) %>% ggplot(aes(reorder(States,-n),n)) + geom_col()
      • (You might have to debug this slightly, but something like this should work)
    • Also sort bars like this for the “Number of strikes by employer ownership type” plot
• Ownership
  • Very interesting point about the unionization rates in public vs. private sector.
• Unions
  • If you show the union acronym to the reader, I would explain what the acronym is short for! This will help the reader understand things a bit better.
  • I like using the fill based on employer to show the relationship between single unions and multiple employers…I still find it a bit tricky to read at first. I don’t have any suggestions for other ways to improve it, but keep that in mind.
• Start date and end date
  • I’m not sure it makes sense to show a box plot with date data…It makes it seem like there are less strikes in the 2010s or early 90s than in the mid 2000s. Which may be the case, but I think a better plot would be to count the # of strikes per year, and plot that via a bar or line chart.
• Workers
  • Nice job, but I would show a histogram here! Showing the distribution would be extremely informative.
• Days struck
  • Also show histogram
• Change in number of striking workers per year
• Good, but I would use geom_line() instead. It will carry the same information but will be easier to read.
• Change in size of individual strikes by number of workers
  • Re label y axis - “# of workers striking”
  • Use alpha argument in geom_point() to change the transparency. See my post in the “random” Slack channel for details on how to do so.
• Private vs. public sector
  • Nice