Nicholas Felton

Introduction to Data Visualization: From Data to Design

Hi Sara! Super to see your process. Coincidentally, for my project, I'm doing a gender representation analysis of the "Society, Culture and Media" (which is similar to your section but a bit more general since it's a small paper) -section of Die Tageszeitung here in Berlin, Germany. What's interesting is that I could go on more than names since the German language always declares gender in the grammar. So simplified, if it says "The murderer said …" I know it's a man because otherwise it would say "The murderess". I'm not entirely sure yet (still counting) how to go about it though. For example, case 1, if it's a book review, the author's name is mentioned all the time: "J. Doe wrote a book…", "What's common in Doe's book is…" or "Books are swell, says J.". If it's a report on an event however, case 2, I find the names more often than not substituted for synonyms such as "The perpetrator", "The murderer" or "The asailant". Would you count both those cases as three mentions of a person, or three in the first case and none in the second, or three in the first case and once in the second? I see that you counted only "proper names included in the article" but what would you do in my case when all words are gendered? Also, did you count each instance of the same name as a new mention? Thanks for the input! Great job! Olof.

Thanks for the kind words, Olof! Your project sounds really exciting! And that’s so interesting the German language has this aspect we don’t have as much in English.


In my dataset, I only counted a name once. For example, one article was about David Bowie’s new album, and he was mentioned by name, and as ‘he’ or ‘an elusive rock star,’ but I only counted him one time. 

This made me wonder if I should ‘weight’ words. In my analysis, one name can be used 10 times and another word used once and they essentially count as the same. I decided against it because I’m awful at math and didn’t feel like it would change the conclusions in a meaningful way. But I’m still curious about that layer of information when you count individual instances. 

Early on, I also considered counting use of she/her/his/hers/he/him but the dataset would have become too big and the process was going to be tedious even using some online tools. Here again I thought my conclusions would be similar even with the extra data, so I decided to keep it simple with just proper names counted once.

Hm, if all words are gendered, that’s tricky, and not just because I don’t really understand German. One thing I tried to keep in mind was to make my chosen dataset as clear and easy to explain as possible because the more you add, the more complexities appear. This is good in some ways, but when it comes to visualizing it can be a little overwhelming.  I would probably start on the simpler end and see what insights you get from that data. Maybe you can do a small sample set where you add in all of the gendered words and see if there’s a difference. (Ideally there’s an online tool to help with this because it otherwise sounds like a ton of work).  

Not sure if this helps but all the best with your project! Look forward to seeing it!
Sara

Hi Sara - This is a terrific exploration, particularly for someone who is not a graphic designer! Thanks for sharing all the steps of your investigation. Your scatter plot is particularly revealing. The gender seems especially biased in this presentation, and demonstrates that visualizing mentions alone does not show the full picture. Thanks!

Thank you so much, Nicholas! I really appreciate the feedback. Excited to keep advancing this project and I'll definitely be returning to your course for more ideas.

Interesting to see that the only time women are in the majority is when pictured. Almost as if they're considered better as decoration than as contributors. Maybe a bit far fetched because of the small data set, but it makes you wonder because of the other clearly male-centric data.
Another interesting statistic to visualise would be how many of the mentions of men are by men, and vice versa.

Yes! I wonder do the (male/female writers, regardless of whether the subject of their article is m/f) tend to use more of one kind of proper noun or reference? Also is there a correlation between men covering 'male' subjects and women cover 'female' subjects? (Not sure how the editorial process at NYT works and if articles are assigned). Would love to look at these topics and more over a longer time period. That's one of the interesting aspects of this exercise is that it raises even more questions!

The desire to see how a visualization changes week over week has been a common thread for this project. When I offered the workshop this fall, one group found that 0 words in the sports section were written by women, which left us all wondering if this was a fluke or the status quo for the newspaper.

This was a really fun project!&nbsp;
I decided to look at the Arts section on 1/5/16 of my local paper, The New York Times. What I wanted to learn was how&nbsp;often men and women are represented both as authors and within the article content itself.&nbsp;
Here's an overview of my process:
<img src="https://static.skillshare.com/cdn-cgi/image/quality=80,width=1000,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/97cc959c" class="zoomable rounded-image-wrapper" alt="Gender Representation in the NYT Arts Section - image 1 - student project" srcset="https://static.skillshare.com/cdn-cgi/image/quality=80,width=600,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/97cc959c 600w, https://static.skillshare.com/cdn-cgi/image/quality=80,width=1000,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/97cc959c 1000w" sizes="(max-width: 600px) 600px, 100vw" loading="lazy">
I looked at some of the more sophisticated online data tools but they weren't really supporting the question I had. So I kept the analysis pretty manual. Hope to use them more in the future though.
Here's my first pass. Note: I'm not a graphic designer, but may work on giving it some more design love in another iteration.
<img src="https://static.skillshare.com/cdn-cgi/image/quality=80,width=1000,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/a4d896f9" class="zoomable rounded-image-wrapper" alt="Gender Representation in the NYT Arts Section - image 2 - student project" srcset="https://static.skillshare.com/cdn-cgi/image/quality=80,width=600,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/a4d896f9 600w, https://static.skillshare.com/cdn-cgi/image/quality=80,width=1000,format=auto/uploads/project/de9ac30d2e87d6f5c105203ab7199b35/a4d896f9 1000w" sizes="(max-width: 600px) 600px, 100vw" loading="lazy">

Gender Representation in the NYT Arts Section

Company

Community

Teaching

Mobile

English

English