GradReports Sentiment Analysis

Author

Brody Spears

Introduction

Something that I have been interested in all semester has been colleges and more specifically how students review them. I think there is a lot of power and information that can be learned from student reviews of colleges they have attended or even just toured, as students are the primary people who leverage college to help them with their future careers. GradReports is a website that compares colleges on a variety of factors including annual tuition, salary coming out of college, and also has individualized webpages for schools like Xavier University with reviews. What drove the following analysis was my interest in comparing both Xavier University and the University of Cincinnati to see what kinds of things people were saying about each university, and does one university have more positive reviews than the other?

Data Dictionary

I scraped the webpages on GradReports for reviews on both Xavier University and the University of Cincinnati and compiled them into a csv file to analyze. Here’s a short data dictionary that gives some context as to what each variable means and is used for.

  • Name: The name of the person who left a review.

  • Date: The date that the person left their review.

  • University: The specific school that the reviewer was leaving a review on.

  • Review: The content of text included in their written review on the university.

If you are interested in looking at this file yourself, you can find it here.

Analysis

Question 1

How many positive versus negative words are there for reviews of Xavier University and the University of Cincinnati?

Looking at this graph, it appears that there are more positive words around UC reviews than Xavier reviews, and Xavier reviews appear to have more negative words.

Question 2

How do the positivity scores compare over different days of the week?

For this it seems like just about every day UC has Xavier beat on positivity score besides Thursdays. It would be interesting to do some additional analysis on why so many people wrote positive reviews on Monday and Tuesday for UC, and why Xavier’s best review days were Wednesday through Friday.

Question 3

How do Xavier and UC compare in terms of emotional sentiment?

From this graph it appears that UC has more words in general than Xavier, which means that people are writing more and leaving more reviews on UC as opposed to Xavier. It is interesting to see the significance in the difference in anticipation words with UC having over double the amount.

Conclusion

It was quite interesting to dive into some analysis on what kinds of words were being said about both Xavier University and the University of Cincinnati. Something that definitely threw off the analysis a bit was just fact that UC had more reviews than Xavier, so it is hard to say necessarily whether the overall sentiment of UC really is higher than Xavier. Nonetheless, this was a very interesting topic to dive into and analyze.