These notes are for the ACRL2019 Thursday 2:40pm-3:10pm paper:

https://tinyurl.com/ACRLweedpaper
https://tinyurl.com/ACRLweedslides
#ACRLweed
paper in ACRL 2019 Conference Proceedings

Introduction #

  • Auraria Library serves 3 different schools through a single library, almost 50,000 students between them all
  • Paper resulted from usual lesson planning, to discuss keyword development as a result of consultation with English faculty
  • Recreational marijuana use’s legalization was often discussed by students, so it was easy to talk about synonyms, then they had the general sense that “cannabis” would be used more with health studies while “marijuana” would be used more in criminal justice disciplines
  • Questions about “query expansion” with their databases
  • Political context around marijuana has changed dramatically since about 1996, when California legalized (medical?) use, then Colorado later did recreational legalization, now 33 states + DC have some form

  • Almost all of the states where it’s illegal (except one) voted for Trump in the last presidential election (I think Kevin said the exception is Virginia?)

  • Academic Search Premier, which allows you to turn off the “apply related words” feature (the query expansion), also lets you save citations to CSV format (which lets them download & use the results at scale)
  • 10,479 articles, 5,643 articles on “cannabis”, 4,836 articles on “marijuana,” after they vetted the results according to a few parameters
  • Hand coded representative sample of 540 articles, which is about 5.2% total dataset; ended up being 90 work hours
  • Coded whole set separately, then combined
    categories:
    1. geography
    2. academic disciplines
    3. application of use (i.e. agriculture, etc.)
    4. sentiment expressed (negative, positive, mixed)

  • Explosion of research on this topic across many disciplines

  • Longitudinally, the word used predominately in academic studies has changed across last two decades
  • Substance Use (discipline to do with addiction, etc.) writes the most about this topic but often uses different words
  • important because discipline most likely to discuss it hasn’t settled on the term; even examples of two papers in the same issue of a journal using different terms
  • calls into question any time you tell a student “well, the right term to use for that is…”
  • There hasn’t been significant medical research on this (me: I think that’s because of structural / legal constraints?)

Discussion & Implications #

  • Words Matter; words that we think mean the same thing don’t necessarily mean the same thing
  • They’d like vendors to inform users that a related term has been applied
    • tell users what related terms are
    • allow users to turn off query expansion
  • Proquest doesn’t let you turn query expansion off

  • Google does in advanced, if you use “verbatim”

  • Shoutout to Dr. Safiya Noble’s Algorithms of Oppression

  • a vendor they contacted to see if “marijuana” and “cannabis” are synonyms said yes, but wouldn’t let them see the synonym thesaurus because that’s proprietary
  • We need to talk with students about how algorithms work, and this is an easily comprehensible way to
  • emphasize that language is politicized, as the keynote today said
  • refugees & immigrants
  • They ask audience to think about identities that collide with politics, think about other examples / scenarios around that

Questions & Answer #

Q: In Oregon, a term often used is “CBD”. Where did that fall in here?

A: They’d have to search again and code responses again

Q: Did many articles use the terms interchangeably?

A: For granularity reasons, they mostly coded just the abstract, so they can’t speak to that in too much detail

Q: Any plans to expand this to more subject-specific databases? I’m a medical librarian and this is really relevant, also curious about how this works in primary databases.

A: They were limited by Ebsco being the only database that lets you turn off query expansion

Q: Curious about selection process, was it random?

A: They alphabetized by title by year, then chose every 20th one from that year

Q: Excluding articles not written in english, is it possible that the discrepancy in international article is due to British English, because of colonialism?

A: That could be a reason. Also her journalist training tells Rachel that marijuana is heavily preferred in US journalism

original etherpad