Navigate the site using the links above. Links to the original articles and videos, along with their comments, can be found in the Comment Sources section. Graphs and charts analyzing the collected data can be found in the Analysis section. Under the Hood goes into detail about the XML and other coding techniques we used to gather, organize, and display our data. Our conclusions and closing remarks can be found in the Conclusions section.

Conclusions

Looking at the data from the Comment Analysis page, we can see some results that are both expected and surprising. Contrary to what we believed would be the case, Youtube comments ended up having more positivity than negativity, and it was the news articles that were overwhelmingly negative. This is possibly due to the fact that when people comment on a news article, they are more likely to give their opinion rather than simply say "good article!". Youtube videos are mostly meant to entertain, and thus promote positive reactions, whereas news articles tend to relay information, which also tends to be bad news or controversial news.

The results of the actual quality of the comments, however, was about what we expected. Youtube comments are just as likely to be written well as they are poorly, with little to no punctuation, typos, and lack of capitalization. Comments from more professional news sites are almost always well-written, if not mostly negative, and also quite lengthy compared to Youtube comments. Again, this goes back to people tending to give their opinions, sometimes very long and winding opinions, on the article, rather than just a quick blurb.

The various other points of measurement also fall in line with expectations, although perhaps not quite as clean cut as the grammar. There were considerably more obscenities and acronyms used in Youtube comments than news article comments (19 instances of obscenities to 7, and 18 instances of acronyms to 10), and there was not a single instance of emotes being used in the news articles, which was an interesting extreme. Excessive punctuation and use of all caps was actually higher in the news article comments, but this is mostly due to the comments themselves being much longer, and therefore having more punctuation and words in general.

At the end of the day, we can see that while your average Youtube comment is of a lower quality than the average NBC news article comment, the actual bile being spewed is no less negative, even if it is written with a better hand.

Some limitations in our study that could be expanded are obviously the amount of comments sampled, first and foremost. because of the time constraints and there only being two of us on the project, we had to limit just how much data we wanted to collect. The original plan called for hundreds of comments from each video or article, with many more videos and articles. This, of course, ended up being unfeasible, and would definitely be something to look into for any future explorations of this project. Another limitation was in the way the videos and articles were sampled. We used a more or less random assortment of videos and articles that were not related to each other in terms of content, but this may have not been the absolute best approach. A different approach might have been to take a particular topic, such as Bill Cosby's assault charges, and find both Youtube and NBC news articles that relate to that same subject, and see how people commenting on the same basic idea present themselves in different outlets. In the end, we still came up with some illuminating results and learned a lot along the way, and hopefully will give others a solid base, both in terms of coding and ideas, to further our research.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License