Larkin’s Blog

Well, I had come accross the Flesch-Kincaid readability tests before - both the 0 to 100 and US grade school levels - but I never put them into practice or thought too long about them. But, can it be a useful tool for a writer?

I searched the web for some articles where people had applied the tests to their writing to see if they were deducing anything from the results.

Flesch-Kincaid: Threat or Menace
Ruminations on writing and the Flesch-Kincaid reading scale
How Easy is it to Read The Wages of Wins?
I like Mitch Albom a lot…but those books of his…yeesh
Does Mitch Albom Think He’s Jesus?

I found two distinct bands of people. The first band revel in the fact that their work scores highly on Flesch-Kincaid. It’s as if they think their writing must be of good quality if they can only be read by someone at least in the 10th or 12th grade. I’m afraid they have missed the point. The other band are a pragmatic bunch of people who want to be read and would like to make the best-sellers list. These people realise that you need to be aiming your work at the 4th to 6th grades. I think the first band fail to realise that a grade score of 5 means that a 5th grader will just about be able to read the text comfortably, but that an adult will find the text very readable. That doesn’t mean the text is dumbed down, it just means that your work is more accessible tot he average person on the street - people who will buy your book at the airport or at their newsagent. Not academics who will be looking at the “literary” value of your work. I believe they make up less than 1% of your potential market. If you want to be a commercial success, the lower the grade the better (without going too low of course).

I applied the metrics to some of my own work and to the entire Moby Dick book, which produces interesting results.

1) An IT article on MySQL Stored Procedures, Triggers and Views published in Linux Magazine, Aug 2006 (see here)

Word Count: approx 4,500
Style: Very Technical with computer code, plenty of industry-specific terminology

Characters per word: 4.7
Passive Sentences: 20%
Flesch Reading Ease: 49.9
Flesch-Kincaid Grade Level: 11.6

The sentences are necessarily verbose and technical, so it is to be expected that it would not be that easily read, particularly by children.

2) An unpublished short story I wrote - the first really serious short story I wrote.

Word Count: 4,264
Style: Emotional rollercoaster. A fair amount of dialogue.

Characters per word: 4.2
Passive Sentences: 6%
Flesch Reading Ease: 78.0
Flesch-Kincaid Grade Level: 5.6

3) About half of the first chapter of what may turn out to be the first novel I write to a finish.

Word Count: 1,600
Style: Horror / Thriller. Very little dialogue to this point. Shorter punchier sentences that 2) above.

Characters per word: 4.5
Passive Sentences: 7%
Flesch Reading Ease: 74.5
Flesch-Kincaid Grade Level: 5.4

4) A real rough piece of work entitled “Stranger at the pier”, which could be complete in itself, or it could lead to a longer story.

Word Count: 728
Style: Pulpy horror. A fair bit of really colloquial dialogue. Short punchy sentences.

Characters per word: 4.4
Passive Sentences: 0%
Flesch Reading Ease: 79.9
Flesch-Kincaid Grade Level: 4.5

Amazingly 0% passive voice. Now, it’s only 728 words long, but still!

5) Moby Dick, by Herman Melville - downloaded legitimately from the Gutenburg catalog

Words: 215,000

Characters per word: 4.57
Passive Sentences: unknown (had to use Google docs, whcih does not calculaulate passive use)
Flesch Reading Ease: 82.17
Flesch-Kincaid Grade Level: 3.45

—

So, when it comes to straight fiction, it’s pretty obvious that my readability is consistent, regardless of the style I employ. The more I descend into pulp (e.g. Stranger at the pier), the lower the grade score, and the higher the readability. But, not by too much it seems.

If we look at the ideal scores according to James V. Smith, Jr. (courtesy of Burnstein):

No more than 4.25 characters per word.
No more than 5% passive voice
No less than an 80% readability on the Flesch Reading Ease scale.
A Flesch-Kincaid Grade Level of 5 (although on the chart, he lists that as 4-6).

Smith deduced these scores from the averages he saw from best-selling authors like Stephen King, Danielle Steele, and John Grisham. Those names won’t feature on the average literary professors top 10 lists, but they certainly will on Average Joe’s. And Average Joe is 99% of the market.

Judging by other articles, that 80% readability is a little tough to get, so I would say that 75% might be more realistic. Passive voice reduces the readability of your work. Just read Strunk and White to find out more on that. 5% passive voice is difficult to make, but 6% or 7% will do for me. Maybe the 0% in my shorter piece is an anomoly.

I can’t say that the Flesch-Kincaid metrics are the be-all and end-all, but I will certainly measure the readability of my work from here on. One needs to know one’s audience (or readership).

Just in case you didn’t know, the readability stats can be accessed in Word. Click on Help and search for readability. Then follow the instructions. Google docs has the stats also.

Technorati Tags: readability, flesch-kincaid, writing process, writing metrics