The Flathead Valley’s Leading Independent Journal of Observation, Analysis, & Opinion

 

25 January 2012

Issues pages of Steve Daines & Rob Stutz have best readability scores

The worst readability scores belong to Dave Strohmaier and Diane Smith, but that could change: I have yet to analyze Sam Rankin’s page. If he files, I’ll test his prose. And when Jason Ward, who has filed, opens his website, I’ll check its readability.

Before discussing the numbers, a caveat or two. Readability formulas — Flesch-Kincaid is probably the best known — assume that long sentences, especially long sentences with long words, are harder to read than shorter sentences with shorter words. But readability scores are blind to logic, grammar, word choice, typography, graphic design, and many other things that professional editors consider in forming an opinion on readability.

So why use readability scores? In my experience, they’re a quick way to flag prose complexity that’s incommensurate with the author’s target audience. If you’re writing for eighth-graders, your prose should not generate college level readability scores. If it does, it’s time for some blue-penciling.

Keep that in mind while you examine the following graph, which includes both House candidates’ scores and scores for selected Presidential speeches and several local writers: InterLake editor Frank Miele, Beacon editor Kellyn Brown, and Whitefish Superintendent of Schools Kate Orozco.

I generated my Flesch-Kincaid scores with a cross-platform, open-source application, Flesh. Microsoft Word also generates Flesch-Kincaid scores, but many versions of Word truncate the grade-level at 12 for no good reason. If you’re interested in the Gunning Fog Index and other readability tests, the donationware application Word Counter generates more than 50 statistics. Depending on how these applications are configured, the readability scores will vary slightly.

Table 1
Flesch-Kincaid Readability Scores

 
Candidate
Grade
level
Reading
ease
Sent-
ences
 
Words
Syllables
per word
Words per
sentence
Rob Stutz 9.2 51.9 77 990 1.68 12.9
Steve Daines 10.1 53.2 26 441 1.61 17.0
Melinda Gopher 11.6 39.0 29 455 1.80 15.7
Kim Gillan 12.4 43.6 52 1,098 1.68 21.1
Franke Wilmer 12.6 37.9 74 1,392 1.77 18.8
Dave Strohmaier 13.0 41.3 84 1,855 1.69 22.1
Diane Smith 14.9 33.5 30 764 1.74 25.5
Obama Inaugural 8.8 65.7 126 2,375 1.44 18.9
WJC 1st Inaugural 8.4 62.9 103 1,610 1.51 15.6
JFK Inaugural 11.4 59.1 52 1,346 1.44 25.9
FDR 1st Inaugural 10.3 54.7 101 1,884 1.57 18.7
FDR 4th Inaugural 7.9 71.3 30 550 1.38 18.3
Lincoln 2nd 7.8 68.5 40 654 1.44 16.4
Frank Miele 10.3 57.4 61 1,238 1.52 20.3
Kellyn Brown 11.8 50.0 28 625 1.59 22.3
Kate Orozco 15.2 36.4 25 709 1.67 28.4

What accounts for the difference between the scores for lawyers Smith and Stutz? Sentence length, primarily. Stutz uses bullet lists — short sentences — to present his positions on issues. Smith does not. She writes discursively. Her sentences are nearly twice as long as Stutz’s.

Another factor leading to dangerously high readability scores for most of Montana’s Democratic candidates for the U.S. House is under-funded campaigns. Big budget campaigns employ top flight advertising agencies with highly skilled writers to produce their websites and write speeches, position papers, press releases, and so forth.

Low budget campaigns, such as those of Stutz, Gillan, et al, seldom have access to writers and copy editors the candidate respects and trusts. Consequently, a lot of what appears on low budget websites is written by the candidates, who are highly educated, high powered people accustomed to writing for their peers, and therefore are susceptible to unwittingly writing over the head of the average voter.

That doesn’t happen with Barack Obama. His inaugural address’ Flesch-Kincaid grade level was 8.8 with a reading ease score of 65.7. According to Slate, most of his speeches are in that range. That’s consistent with the old newsroom rule of writing for the milkman with the eighth grade education.

The milkman I write for graduated from high school. Below, the statistics for the ten paragraphs above, generated by Word Counter.

Table 2
Flathead Memo readability scores

Name Value
Words 486
Unique Words 292
Characters 3122
Characters In Words 2541
Sentences 29
Paragraphs 10
Average Word Length 5.2
Average Sentence Length 87.6
Average Words Per Sentence 16.8
Long Words (7 or more characters) 133
Short Words (3 or fewer characters) 154
Syllables 810
Syllables per word 1.67
Monosyllabic Words 297
Polysyllabic Words (3 or more syllables) 88
Automated Readability Index 11.6
Coleman Index 13.2
Coleman-Liau Index 15.0
Dale Chall Readability Index 9.7
Dale Chall Readability Grade Level 13 - 15
Degrees Of Reading Power (DRP) 72
Degrees Of Reading Power Grade Level Greater than 12
Fang Easy Listening Formula 11
Fang Easy Listening Interpretation Easy
Farr-Jenkins-Patterson Score 49
Farr-Jenkins-Patterson Reading Ease Difficult
Flesch Reading Ease Score 49
Flesch Reading Ease Difficult
Flesch Reading Ease Grade Level 13 to 16 (College)
Flesch-Kincaid Grade Level 10.6
Flesch-Kincaid Reading Age 15.6
Fry Readability Grade Level 12
Fry Readability Reading Age 17
FORCAST Grade Level 10.8
FORCAST Reading Age 15.8
Gunning Fog Index (FOG) 13.9
Gunning Fog Reading Age 18.9
Henshall formula 869.4
Johnson Readability 38.9
Johnson Grade Level 8 or higher
Lexical Density 60.1
Laesbarhedsindex (LIX) Index 44.0
Laesbarhedsindex (LIX) Readability Standard
Laesbarhedsindex (LIX) Grade Level 8
Linsear Write Readability 11.4
McAlpine EFLAW© Test 22.1
McAlpine EFLAW© Readability Quite Easy
Miyazaki EFL Readability Index 34.6
Power-Sumner-Kearl Grade Level 6.7
Power-Sumner-Kearl Reading Age 11.7
Rate Index (RIX) 4.6
Rate Index (RIX) Grade Level 10
Raygor Readability Grade Level College
SMOG Score 13.1
SMOG Index 13.0
SMOG Reading Age 18.0
Spache Readability Index (Original) 6.1
Spache Readability Index (Revised) 5.5
Wheeler Smith Index 47.2
Wheeler Smith Grade Level Greater than 4