Large differences between groups when it comes to spelling, grammar, average reading age and even swearing
With just 12 months to go until Britain exits the European Union, the fierce debate amongst 'Leave' and 'Remain' camps rages on. And it rages nowhere more passionately than in Facebook Groups.
To see what we could learn about each camp, we scraped and analysed nearly 30,000 comments made by members of two private Facebook Groups: the pro-leave Brexit HQ, and the pro-remain We Are #StopBrexit.
Facebook Groups are forums for members to talk and share links about common interests like sports teams, church, or in this case Brexit
Both groups are ‘closed groups’ – you have to apply to join, answering a handful of screening questions from the moderator. Once approved you can then post new topics, comment on existing ones and (most importantly here) view other members’ comments.
Each respective group doesn’t allow the ‘other’ camp to join, and if you do sneak through the screening questions and later reveal yourself, you’re usually quickly removed from the group. While that’s probably not ideal for healthy debate, these echo chambers provide a fascinating insight into two political extremes of a divided Britain.
Let’s get to the good stuff.
Jump to charts:
Based on the first name of each commenter in the sample, we can make a reasonable assumption about their gender. Both groups have an almost identical gender split, with a skew towards men.
As this is based purely on comments posted, it doesn't necessarily mean the members of the group skew men, it could be that they're slightly more vocal in the group compared to women.
When it comes to spelling, those in the Leave group made an average of 39.4% more mistakes than those in the Remain camp. In other words, for every 10 spelling mistakes Remainers made, Leavers made 14.
As with all these charts, for such a large sample size this is statistically significant, but what the data unfortunately doesn't tell us is why this is the case.
The Leave Facebook group made 59.8% more grammatical errors than those in Remain.
We used the Flesch–Kincaid Grade Level Formula to find the typical minimum age a person needs to be to understand each set of comments.
This formula is used in education to score the difficultly of English sentences, looking at word length, sentence length and syllables per word.
Leaver comments are typically able to be read by a 10 year old, whereas the minimum age needed to read Remain comments is 14 years old.
We checked each comment for any mention of around 40 common swear words as defined in Ofcom research.
The sweariest group are by far the Leavers, triggering our profanity filter 4.8 times more often than Remainers.
Word Cloud - Leave
A word cloud shows the frequency of single words in a body of text, with words mentioned more often shown larger. Very frequent everyday words like 'a', 'is', 'and' etc. are removed.
Interesting standouts here are "country", "time" and "Blair".
Word Cloud - Remain
There are few obvious difference between the Remain and Leave word clouds. This is likely because the same major topics are discussed in both groups.
Interesting standouts here are "idiots", "Labour" and "sovereignty".
What do you make of the differences? Let's all argue in the comments below, but please watch your spelling 👀
How this data was collected and analysed:
- Using a specially created tool, we scraped approximately 15,000 of the most recent comments from each closed Facebook Group
- For a like-for-like comparison, we truncated each body of text to approximately 200,000 word sets. Actual counts as follows: Leave group, 199,982 words, Remain group, 199,963 words
- Data collection took place between 6th March 2018 and 8th March 2018
- Spelling and grammar statistics using Microsoft Word’s built-in tool, with United Kingdom dictionary
- Reading age using Flesch–Kincaid Grade Level Formula
- Gender classification using first name of commenter, via Genderize.io
- Swearing analysis completed using swear words defined in this Ofcom research.