Grammarly Versus Readability

Westminster Abbey, London / Photograph by Jim Richardson / National Geographic's Photo of the Day

Before anything else, this is not a website review. I am only expressing my observations. I do not intend to tear down any website’s reputation. This is what I do for fun, and I know what you’re thinking. Yes, I’m such a geek.

It looks like this week is all about rants about good writing. Mostly, they are coming from the fact that I’m training another tutor. One thing that I usually consider when training is if the trainee’s comments are easy to understand, which made me recall a website that I played with a few months back. It’s a text readability scoring system on AddedBytes.com, “Text Readability Score.” As I was thinking about it, my boss approached me and told me to go to a website. He said, “Type grammarly.com,” and I did it. He instructed me to use the service, and I was initially impressed by the service since it points out the major issues of a piece of writing. I did see the issues that it detected in one of my trainee’s compositions, so I decided to test my English prowess on it.

I put this text in:

I want to go to Japan. I picked that country because of its food and folklore. I always loved Japanese food. My favorite dish from the Japanese is sushi. In terms of folklore, I am fascinated by their mythical creatures. I am intrigued with dragons, nine-tailed foxes, and kappas.

Grammarly gave my paragraph a score of 78 out of 100, indicating that I had 2 critical writing issues, both involving the use of passive voice. At this point, my score is 74.90 in reading ease and that it can be understood easily if a person has studied for 6.26 years. According to the website, the higher the score for reading ease, the easier it is to understand the text, and of course, a lower number of formal education years required.

Righteous indignation overtook me when I saw that score from Grammarly. I felt appalled that I had a mistake. I refuse to get such a score, so I revised it to this:

I want to go to Japan. I picked that country because of its food and folklore. I always loved Japanese food. My favorite dish from the Japanese is sushi. In terms of folklore, their mythical creatures fascinate me. dragons, nine-tailed foxes, and kappas intrigue me.

The score changed, definitely. At that point, Grammarly gave my paragraph a score of 88 out of 100, indicating that I had 1 critical writing issue, which involves capitalization. When I checked on the readability measurement, my score went down to 74.00 in reading ease, needing 6.54 years of formal education.

In my mind, this is a no-brainer. I can easily spot the mistake now, so the finally draft looks like this:

I want to go to Japan. I picked that country because of its food and folklore. I always loved Japanese food. My favorite dish from the Japanese is sushi. In terms of folklore, their mythical creatures fascinate me. Dragons, nine-tailed foxes, and kappas intrigue me.

Hooray for me! Grammarly gave 100 out of 100, and the readability scores didn’t change. However, I did notice some ideas worth pointing out:

The revision away from what Grammarly calls “passive voice” (usually a “to be” verb and the past participle of another verb) also took away the focus off of me, which is supplied by the “I” in my original draft.

I tried to get more details on where I made a mistake, and when I asked for more information, the website offered me a 7-day free trial. However, I needed to supply a credit card. I know that I won’t be charged anything before the 7 days end, but I also know that if I forget to cancel, I will get charged, so I didn’t take the risk. I’m not sure what this means to other people, but I usually follow the “better safe than sorry” rule.

It took me approximately 20 minutes to figure out what I did wrong without the help of Grammarly. Considering that I’m “learned” in grammar and style, this might be a wake up call to be more conscientious, but I wonder: how long would a person who do not know much about grammar take to fix the sentence?

Most importantly, I noticed that my final draft is relatively “harder” to read and required “more” years of study. I only have one set of data, so I cannot put this under a statistical analysis to check if the variation in scores is negligible or not, but I wonder what will happen if we compiled data on this.

Again, I don’t intend to say anything bad about Grammarly. I think that it’s a cool service. This is just a post on what I do for fun, but I’d like to get your thoughts on this subject, so feel free to hit that comment button.

P.S.

This post contained 60 sentences, with 839 words (13.98 per sentence) made up of 1192 syllables (1.42 per word).

This entire post scored 71 out of 100 in Grammarly. It has 26 critical writing issues:

Contextual spelling check: 1 issue, 1 count in commonly confused words
Grammar: 10 issues, 1 count in comparing two or more things, 1 count in modal verb usage, and 3 counts in faulty sentence structure (press ctrl+F and type “do” to find what it missed)

At this point, I think that Grammarly and I share one thing in common. We don’t do math well.