Data Visualizations of a Deeply-flawed Dataset

Having spent some time adding to The Big List, I wanted to make some use,or at least some sense, of it. So I set out to make some data visualizations from it.

First, this is a hugely unscientific dataset. I have been the major contributor, which means that the list has evolved with my whims. I have missed things. I have privileged certain categories of things over others. For instance, I’m writing a piece about Luke’s adaptive reuse of the diner, so I paid extra attention to any scenes where he talked about the process of turning his dad’s hardware store into a diner. Emily Gilmore’s involvement in the DAR is mentioned frequently but with almost no additional content, so I have rarely noted it. Early on I listed literary references, since they seemed public history adjacent. Then I didn’t. So there was quite a bit of data clean-up to do.

Data clean-up

Here are steps I took to prepare my spreadsheet:

  • Limited the entries to the first three seasons of Gilmore Girls as I’ve been the most thorough with those
  • Removed several extraneous entries (literary references, panning shots of buildings without character involvement)
  • Changed “Lorelai ‘Rory’ Gilmore’” to “Rory Gilmore” and “Lorelai ‘Named Her Daughter on Demerol’ Gilmore” to “Lorelai Gilmore.” Those names were funny when I created the form, but it made for a visual mess. This change has now been made to the Google Form too..
  • Went through the incidents I’d added to the spreadsheet and added a few characters that I hadn’t previously included on the Google Form (e.g. Andrew). They’re now on the Form for future additions.
  • I came up with six categories of public history incident that each data point fit into (e.g. Antiques, Reenactment, etc.). These categories are the least objective part of this project.

Here’s the spreadsheet I was working with:

What to Chart ?

Now I had to figure out how to display my data. I have a bunch of incidents that involve multiple characters over three seasons (65 episodes), categorized into six categories. I could create a bar graph for each character’s mentions (Figure A). That was a start.


I thought about how to use a timeline to show the distribution of these incidents but couldn’t quite make it work in my head and none of the timeline tools I saw provided any solutions. I could do line graphs showing character’s references per season, however (Figure B).


I also wanted to show how involved certain characters were in certain categories. For example, Taylor is involved in a lot of incidents of various kinds, but he’s also involved in nearly all of the reenactment activities. Mrs. Kim is usually involved in discussion of antiques, but she’s not involved in all of them (Figure C).


And the categories I identified? How do they compare in frequency? (Figure D).


The Tool.

I wanted to make several simple charts and I found that Beam ( as you can see in the watermarks below) suited my needs. A simple spreadsheet interface worked well with my very small sample size and my er…hands-on approach (meaning I manually counted stuff). There are many better visualizing tools, but this one was simple.


So what, if anything, can we learn from these visualizations?

Figure A

This one’s kind of interesting. I ordered the x-axis according to the number of episode credits on, which gives dimension to a couple outliers. Taylor Doose, Mrs. Kim, and Andrew (who is pretty far down the IMDB list) all have high public-history-per-appearance rates (of course). Lane Kim, on the other hand, despite living in an antiques store, rarely actively engages with the past. However, this last conclusion is a bit misleading because the first six characters are all in every episode of the show. If Lane gets shifted over next to Sookie (who appears in only one fewer episode), she doesn’t seem like such an aberration. Factor in that the four Gilmores and Luke enjoy far more screen-time in each episode, and Lane and Sookie seem closer to the Stars Hollow norm.

eac58b2c-dc61-4eab-b406-a7efaf9aa405 (1)
Figure B

This chart probably has a lot more to do with the flaws in the data sample. Going in to my latest run-through, I was really psyched to document every minor public history occurrence. By Season 2 I’d found a good balance of watching and making notes, and by Season 3, well, I was mostly multitasking with other stuff I had to do. Alas. I imagine, however, that a more thorough sample size (help me out people!) over seven seasons might reveal interesting ebbs and flows in the characters’ (and writers!) historical consciousness.

Figure C

This is the simplest chart here, but it still manages to be a bit misleading. If I’d logged every instance of Mrs. Kim selling an antique or every scene set in the story, the red portion would be much larger. Instead I limited myself to scenes where Mrs. Kim engaged with her customers about issues of authenticity or provenance. On the other hand, I was less picky about other antiques references. I also elided a scene (in this study) in which guests at the Independence Inn chatted about antiquing simply on the grounds that no major character was involved. Future analyses could take a more broad definition of antiques. Or perhaps leave them out altogether. I included them because I see them as an engagement with the past a la the respondents to Thelen and Rosenzweig’s 1990s survey. Thoughts?

Figure D

These categories are so broad as to be almost useless. Within Reenactment, I’ve included pretty much anything with costumes and a notion of the past. Tourism and Museum are interpreted very loosely. Antiques, I’ve discussed a bit above. Miscellaneous is, well…

And I mentioned up at the beginning that I’m working on a piece about Luke’s adaptive reuse of the diner building, which goes pretty far in explaining the number of data points there.

Not sure that this break down is worth much, but a larger sample size with narrower categories might reveal some patterns or tendencies.


I think the first two charts have the most potential. Cross-referencing either of them with characters’ share of the screen (unfortunately, no one on the internet has dedicated their life to this task) could yield some interesting conclusions. I suspect (without any evidence to back this up) that the latter seasons, in which Rory spends quite a bit of time away from Stars Hollow, would contain fewer public history-ish moments. If we quantify by duration, however, I suspect that “A Year in the Life” would rank high, given the length of “Stars Hollow: The Musical,” which could be fodder for a dissertation or at least a witty blog post.

The most important takeaway, however, is that these sorts of references are hard to quantify. They don’t really make good data. So while I will encourage you all to contribute to The Big List (and feel free to recommend changes to the Google Form), I think that will ultimately be more helpful as bookmarks for scholarship along more qualitative lines of inquiry.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s