This post has the following scores:
Flesch Reading Ease: 60.74
Flesch-Kincaid Grade Level: 6.00
Automated Readability Index: 6.00
“What in the world are these?” you might ask? Read on.
This post is all about readability tests, some of which I recently discovered and re-discovered.
There are numerous readability tests available, some of the more familiar (at least to me and what I’ve read so far) are:
- SMOG (Simple Measure Of Gobbledygook)
- Flesch-Kincaid Readability Test
- Fry Readability Formula
- Automated Readability Index (ARI)
- Gunning-Fog Index
And the ones which I’ll be focusing more here are Flesch-Kincaid Grade Level (FKGL), Flesch Reading ease (FRE) and ARI, since they’re the ones used by Internet search giant Google in their Google Documents. Readability tests are really mathematical in nature in that they use predetermined constants to give a readability ‘grade’ to a particular work. Note that there are differences among the 3 (FKGL, FRE, ARI) and having a very readable grade with one test doesn’t necessarily mean it will also have a good readable score on the other two. The first two tests were developed by Rudolf Flesch, an author and a readability expert. The U.S. Department of Education has the following education levels with their corresponding age brackets:
|9th Grade (Freshman)||14-15|
|10th Grade (Sophomore)||15-16|
|11th Grade (Junior)||16-17|
|12th Grade (Senior)||17–18|
In acquiring the number of words, sentences, etc. in your work, here are the rules to be followed: periods, exclamation points, question marks, colons, and semi-colons are considered end-of-sentence marks. Each group of continuous non-whitespace characters counts as a word. Each vowel in a word counts as one syllable subject to the following sub-rules: Ignore final -ES, -ED, -E (except for -LE)
Words of three letters or less count as one syllable. Consecutive vowels count as one syllable. Although there are many exceptions to these rules, it works in a remarkable number of cases.
This test tells you what the minimum educational attainment is necessary for a reader to comprehend your work, and is based on the educational grade level of the U.S. For example, a grade of 9 means that the reader has to be at least a 9th grader (or has a 9th grader’s education/supposed intellect) in order to comprehend your work. Or if you get a grade of 9.8, the minimum educational attainment is the floor of that value, which is 9 or the 9th grade still. To calculate your work’s FKGL grade level, the formula is as follows:
FKGL = [ (.39 x ASL) + (11.8 x ASW) – 15.59 ]
ASL = average sentence length (total number of words divided by total number of sentences)
ASW = average number of syllables per word (total number of syllables divided by total number of words)
For this test, the higher a score you get, the easier it is to read your work, and is based on a 100-point scale. The formula for FRE is as follows:
FRE grade = [ 206.835 – (1.015 x ASL) – (84.6 x ASW) ]
where ASL and ASW are the same as the ones above.
ARI unlike the former two, is based on the number of characters per word. ARI is also based on the U.S. grade level like the other two above. This test is usually less palatable to readability test critics, but it is the easiest one to implement on computer programs since all the program has to do is count the characters instead of syllables. The formula is:
Grade Level = (4.71 x TCW) + (0.5 x TWS) – 21.43
TCW = Total number of characters divided by the total number of words
TWS = Total number of words divided by total number of sentences
Implementation/Testing the Tests
For the sample sentence “Hello World!”, Google Document (when viewing your document, click File > Word count…) gives the following grades:
Characters (no spaces): 11
Characters (with spaces): 12
Pages (approximate): 1
Average sentences per paragraph: 1.00
Average words per sentence: 2.00
Average characters per word: 5.50
Average words per page: 2.00
Flesch Reading Ease: 77.91
Flesch-Kincaid Grade Level: 3.00
Automated Readability Index: 5.00
Which is what you’d get if you count the words, syllables, sentences etc. and manually solve the grade levels from the formulas given above.
Lewis Carroll’s famous poem Jabberwocky in the famous children’s story Through the Looking-Glass and what Alice found there get the following score:
Flesch Reading Ease: 91.89
Flesch-Kincaid Grade Level: 1.00
Automated Readability Index: 2.00
Very readable and highly suitable for children it would seem. What’s a bit surprising is that the poem even has a lower score than the sentence “Hello World!”, making the poem easier to read?
As for the New York Times book review of The God Delusion, the review gets these grades/scores:
Flesch Reading Ease: 57.98
Flesch-Kincaid Grade Level: 8.00
Automated Readability Index: 8.00
Pros of Readability Tests
The original reason for implementing the above mentioned tests was to give teachers, librarians, etc. an overview of the apparent reading difficulty level of a given work. The acquired test grades or scores will then allow them (teachers, librarians, etc.) to choose appropriate materials for a specific group of individual’s age/educational background/attainment. At a glance this reason would seem quite reasonable to many of us, especially if we’re not really familiar with the implications of the rules and the formulas, as well as the entire written work that is being graded or scored.
Cons of Readability Tests
Also, at a glance, especially among people familiar with the tests and their limitations, readability tests are of little or no use at all some times. This uselessness is still subjective, after all, what may be intelligible to a group of people, may be absolutely understandable to others. The obvious (and most wide-spread) criticism of readability tests is that they only test the “surface” or the physical layout of the work. Readability tests gives little (at best) to none at all (at worst) information on the meanings being conveyed by the work, the grammatical structure, as well as the different biases, nuances, and between-the-line impressions or expressions of the author. One could’ve easily written “World Hello!” and the 3 scores for that sentence are identical to the scores of the sentence “Hello World!”.
In conclusion, I think people should be wary when placing readability test results on their works, or when seeing some else’s readability test scores. These scores can easily help clarify or more often times mystify the reader about the true readability of a work. People should not immediately believe what they see, and always keep a skeptical eye on notions as a readability scores of written works. Are readability tests useless in my opinion? I don’t think so. They still have some uses such as when, assuming the writer is well versed in writing or communicating his/her thoughts on paper/documents, then these tests provide a very helpful glimpse as to what audience might be able to appreciate the work.
* A computer algorithm to acquire Flesch-tests related information, including the adapted rules mentioned above, http://portal.acm.org/citation.cfm?id=10583&dl=ACM&coll=GUIDE
* free FKGL, FRE, and ARI test scores for your documents, http://docs.google.com/
* More on readability tests, http://csep.psyc.memphis.edu/cohmetrix/readabilityresearch.htm
* Flesch Reading Ease table, http://www.utexas.edu/research/accessibility/resource/readability/manual/forcast-versus-flesch-English.html
* Lewis Carroll’s poem, Jabberworcky, http://www.jabberwocky.com/carroll/jabber/jabberwocky.html
* New York Times book review of The God Delusion, http://www.nytimes.com/2006/10/22/books/review/Holt.t.html
* A concise but less precise (in terms of citation) source for readability tests, http://en.wikipedia.org/wiki/Readability_test
* More info on mr. Flesch, http://en.wikipedia.org/wiki/Rudolf_Flesch
* A concise but less precise (in terms of citation) source for mr. Flesch’s tests, http://en.wikipedia.org/wiki/Flesch-Kincaid_Readability_Test
* U.S. education levelling, http://en.wikipedia.org/wiki/U.S._grade_level#School_grades