Friday, May 25, 2012

Likert Scales and Rating Books

I don't like Likert scales.

What are Likert scales, you ask?  Likert scales are forms of measurement often used on questionnaires or surveys to elicit an individual's perspective on whatever is being measured, without any measurable difference between the answer choices.  For example:

With 1 meaning "strongly agree" and 8 meaning "strongly disagree," rate your level of agreement with the following statement:

"Aliens are likely to colonize my planet and eat my brains in the foreseeable future."

Another example:

Yes, Goodreads employs a Likert scale.  If you roll over the stars on their website, they stand for the following:

1 star: "didn't like it"
2 stars: "it was ok"
3 stars: "liked it"
4 stars: "really liked it"
5 stars: "it was amazing"

One could also argue that there's an invisible 6th choice of not rating the book at all, which could mean "OH DEAR LORD SO BAD" or "I don't know, it's too hard to say."  Now here's my beef with this.  Ignoring the fact that I think the neutral response (it was okay) should be in the middle position and therefore anybody who has ever answered this kind of scale before is likely to mark books they feel were just okay as "liked it" because that's the reasonable thing to do - wait what was the point of this sentence?  Oh yeah.  I just don't think my feelings or anybody else's reasonably complex feelings about books (or anything else for this matter) can be broken down into such neat categories.  For example, what if I thought a book was really well written and had a great plot (5 stars, right?) but I couldn't engage with any of the characters and so, did not like it very much (2 stars, maybe?).  Do I split the difference and go with 3 stars?  Or 4?  Yes, I know that this is why there's a little box underneath where you can write out your feelings in more detail but let's face it, who reads those?  Okay, I do, but only because I don't trust the stars.

And that's the thing.  No matter how much I tell myself that those little rating scales don't mean anything, I can't help seeing them.  I can't help but have my decision on whether to read something influenced by them.  And, in doing my own ratings, I constantly get hung up on this.  I didn't particularly like Peace Like a River and I really despised Twilight, but I can't have them sharing the same category, because one is reasonably well done and has some substance to it while the other is drivel that offends me as a woman and a human.  Likewise, I really like Harry Potter but how can that begin to occupy the same space as Oryx and Crake in one of these scales?  It just can't.  One is entertaining and emotional and has me coming back for more and the other is out of this world but actually very realistic and gives you nightmares but in a good way.  How can one scale hope to cover both?

I think my issues with these kind of scales started the day I came home from school and said "Mommy, I got a 96 on my Lord of the Rings paper" (yes I wrote a paper on The Lord of the Rings in high school) and she asked me what that meant.  It meant nothing.  There's no objective scale that will suffice for these things.  What's the difference between a 96 and a 95 or a 97?  A misplaced comma?  Well, conveniently enough for my teacher's grading scale, there were exactly 100 commas in that paper to grade me on.  If only.

The point of this long ramble is that I often run into this issue on my blog.  In the end, I think that I personally have 3 categories: like, dislike, and meh.  But I would never try to tell you that that is a reasonable look at the quality of the book because there are so many factors at play here: quality of the writing (as seen by me), quality of the characterization (as seen by me), personal emotional reaction, and the context in which I am reading the book.

In the end, any sort of rating I do is for me alone and subject to change.  I have on occasion gone to my "read" shelf on Goodreads and actually changed ratings that I gave books in the past.  And that's without having reread the book or really revisited it in any substantial way - in that moment, I just remember it in a different way than I experienced it at the time of rating and it seemed reasonable to change my original rating.

In the end, I'd like to do away with these scales all together.  Let my words be what is used to assess my opinion, not my arbitrary assignment of stars (or bullets or numbers of whatever a specific Likert scale is using).  Of course, scales are easier than actual thoughtful responses because responses can be added or averaged and give new readers a simple (though horribly oversimplified) view of the "quality" of the book.  Or whatever it is we're rating.  And because it's easier, it's here to stay.  And because it's there, I can't help but give that scale on Goodreads, even though I know how ridiculous it is.  On the blog though, I'll just stick with sentences and paragraphs, because that's where the real meaning is found.


  1. I agree with you. Mostly. It's for that reason that I don't have a rating scale on my own blog since the "personal emotional reaction" always figures strongly into my reviews. Just because I didn't like a book doesn't mean it's a bad book, and vice versa.

    I wish Goodreads had more stars in their rating system, because the 3stars = "I liked it" doesn't mean so much to most people. Authors get all upset when you rate their books with 3 stars when you've just said you liked it by giving it 3 stars. Then again, maybe that says more about the author's ego than the Goodreads system. But I have had authors ask me why I've rated their books so low when giving them 3 stars.

  2. I agree that a Likert scale (thanks for the definition) doesn't allow for nuances that really distinguishes a good book from a great book from an awful book. It's one of the reasons I don't have a rating system on my blog. However, I do find them helpful. If I want a quick look at what do people think about a book, it's a good way to get a pulse. If I want to learn more, I'll look up reviews. Yes, they're oversimplified, but sometimes, that's what we need. They shouldn't be the be-all, end-all but they're a good jumping off point.