7/30/2023 0 Comments New york times party quizSo I had to think about that, what the critique was and how to address it. What would you say is the biggest trend you saw in this data?Īt the start that was the other thing was I was looking for – essentially a correlation between demographic and party ID, all these things have been strong at parts in history, in contemporary data a can you create a ranking of predictors and I ended up emailing a fairly well known social scientist saying what do you think of this idea, and they were like I wouldn’t do this, I don’t think it would make sense to readers. When you answer a question in the quiz the actual size of that graphic grows in height and if you don’t do anything about that you get a skipping effect, so the graphic grows beyond your screen, so there were a lot of little refinements. So then we started thinking out how to make a quiz, and how to move up and down this tree in an interesting way. So then I had a mostly working reel in JavaScript. I spent a couple days figuring out how to do that. I was showing it to my colleagues, who said it would be nice if the width measured something, which ended up being an interesting geometry coding thing as well. There was a bunch of analysis stuff that included recoding, and categories, a lot of reading literature as to what people have used in the past and how people have chosen to do those things. I initially was pretty strong leaning towards the classification approach and got almost all the way there, but I think ultimately the outcome with that is it’s really hard to explain what you’re measuring. I tried a bunch of different things because there’s regression trees and classification trees and I was split on what to use. I wasn’t too worried about reliability because the CCES is widely used and cited in a lot of papers. The AP had this vote cast data that I found in the middle of me working on this, and I played around with that, but the problem was it didn’t have all the questions I was interested in so I ended up not using it. I poked around with those, and I was looking for a large data set, and this one fit the bill and had a couple years of data. I had just been familiar with both of those and I wanted to use them with something but couldn’t come up with the right story. I ended up using the CCES (Cooperative Congressional Election Study) and the ANES (American National Election Studies). I had a bunch of versions of that and went to my editor, and he was like, “It’s kind of a perennial interest, how parties relate to demographic.” So we started chatting about it and he gave me some time for it.Ĭould you tell me about the process for the quiz, in terms of data sourcing, analysis, and visualization? I think I was just working on it with some free time I had, coming up with initial sketches, originally it was more of a flow chart than a tree, and I had a really intricate version of that, I was trying to play with it and work it into a form that seemed more reasonable. In talking about it with my editor and colleagues we were like how do we insert the reader into the story, and that’s where the idea for making the quiz came about. The quiz wasn’t the starting point, the starting point to me was the second graphic, that looks like a tree, that is where I started. I think I was looking vaguely at survey data, some political science and statistics papers, and those came together. I was working on an earlier idea and Amanda Cox brought this idea to my attention and I had looked at some of her work. One is this idea of using a decision tree approach. This idea came about from a couple of different things mixing around in my head. Where did the idea for this story come from? in economics at Harvard University, Chinoy spoke with Storybench about how he found the best way to display the data and draw the reader in. The piece includes a quiz for the reader that predicts if they are a Democrat or a Republican, as well as a decision tree used to show all possible outcomes, and how the voting patterns of different groups have changed over time. Can your race, religion and education predict your political party affiliation? Last month, Sahil Chinoy, a former graphics editor for the opinion section of The New York Times, published an interactive quiz and accompanying article that laid out exactly how demographics influence the way Americans vote.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |