search

jueyang / call-me-maybe

Use the issue queue. Dark secrets welcome. (CUNY-J teaching 2015)
3 stars 0 forks source link

Understanding My Data #27

Closed nlizbethlewis closed 9 years ago

nlizbethlewis commented 9 years ago

Hi Jue!

I am working on a data viz chart about american work hours. Basically, I started down this road for my pitch and then realized I was thinking about my data all wrong. I don't have a specific tech question about this project, I really just need to talk the data through and think about other ways I might be able to chart this trend.

Project is due Friday at 10pm. My email is nicole.lewis@journalism.cuny.edu

Thanks!

My pitch: https://docs.google.com/a/journalism.cuny.edu/document/d/1lh7t_Xd7fycGghsVxqzmMms_izyw2iZz9vesBT0CtGE/edit?usp=sharing

jueyang commented 9 years ago

It's awesome you are exploring all the datasets you mentioned in your pitch, and that you are reasoning your choices. Also, kudos to you for inspecting elements!

After diving into BLS a bit myself (with the help of Google and non-help of the BLS database search), I find a few threads that might be of help.

  1. Your logic makes a lot of sense. It is essentially the same as how BLS calculates total working hours in 2014: http://www.bls.gov/respondents/iif/totalhours.htm
  2. BLS decided to use hours at work, as opposed to hours paid, to measure productivity. http://www.bls.gov/mlr/1990/02/art2full.pdf This methodology change came around 1990 (see publication date of the pdf), which makes me think that a 1964-2015 narrative might need to be broken up into smaller pieces (1990-2015 maybe?)
  3. Google led me to a book snippet that brings up Current Population Survey by the BLS, which subsequently led me to the data of persons at work by hours of work. Although this only shows data for 2014, it at least means BLS keeps track of this particular measure.

Your hunch to break up the data into sectors is probably a more interesting angle. Persons at work by hours of work would probably not help much on the sector narrative. I would keep digging for datasets that include persons at work by hours of work and by sector.

Hope this helps. Join the slack channel with your email if you bump into more hurdles. The BLS site (and their methodologies) are not the most user-friendly!

jueyang commented 9 years ago

Hope that was helpful!