Talk:Flesch–Kincaid readability tests/Archive 1

This is an archive of past discussions about Flesch–Kincaid readability tests. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Article written

I hope this is now sufficiently Flesched out. Joyous 05:50, Aug 5, 2004 (UTC)

The article is delightful. The pun, however, may be more than I Kin-stand. - Nunh-huh 05:53, 5 Aug 2004 (UTC)

Ouch. Joyous 13:11, Aug 5, 2004 (UTC)

Nerd humor at it's best, ladies and gentlemen! —Preceding unsigned comment added by 219.89.147.7 (talk) 10:29, 28 October 2008 (UTC)

Question

I guess this test is only applicable to English-language texts. Am I right? If so, shouldn't it be mentioned in the article? – Kpalion (talk) 19:58, 31 Jan 2005 (UTC)

I'm For. --logixoul 17:43, 6 December 2005 (UTC)

I agree 69.14.91.247 (talk) 01:15, 4 February 2009 (UTC)

````My only criticism: " Use of this scale is so ubiquitous that it is bundled with popular..." MW defines ubiquitious as, "

"existing or being everywhere at the same time : constantly encountered" Which doesn't leave room for qualification, it's either ubiquitous or it ain't. —Preceding unsigned comment added by 98.175.96.47 (talk) 14:52, 10 February 2009 (UTC)

But at the same time it gives the example sentence "The company's advertisements are ubiquitous." ^[1] I think the way 'ubiquitous' is used allows for degrees of ubiquity. Zoon van Zaal (talk) 18:26, 8 December 2013 (UTC)

References

^ http://www.merriam-webster.com/dictionary/ubiquitous

Stop changing scores!

Can we stop with the changing of scores? It's completely unnecessary, especially as everyone uses a different calculator anyhow, which is why this article's Flesch-Kincaid keeps changing radically. Just leave at "This article has a Flesch-Kincaid of about ..." and leave it at that. Tejastheory 07:17, 9 November 2005 (UTC)

KWord says that the reading ease of the article is 55. Who has different values and how did they get them? Note that this question is timed and after some time I will probably put in 55 if nobody gives anything else. --logixoul 17:37, 6 December 2005 (UTC)

Done. --logixoul 11:30, 17 December 2005 (UTC)

I imagine that people are getting different values by running a calculator on different versions of the article. Joyous | Talk 13:17, 17 December 2005 (UTC)

The reason everyone gets different scores is because everyone uses a different calculator. KWord will get a score, and an online calculator will get a different score. It's all due to the formula, which includes "syllables" which isn't something that all calculators will count the same. Tejastheory 22:41, 17 December 2005 (UTC)

If what you said is valid, then the article needs to be updated to say that the score isn't an absolute thing and different tools are likely to report different scores. Anybody disagree? --logixoul 13:30, 25 December 2005 (UTC)

The score can obviously not be a static number in this case anyway, since it refers to an article on wikipedia, which by its nature it not a static text. The score should be removed or should contain a clause that notes that it is not necessarily an exact score. --jackohare 04:55, 1 February 2006 (UTC)

I have removed the 'this article's score' altogether, because I don't believe that it's a particularly useful example. General samples are usually far more helpful. Why should this article give a sample of the readability score for any piece of writing. Such specific examples do little to help understanding. It's an inevitable trivium, but hardly an illuminating one. Ingoolemo ^talk 08:55, 1 February 2006 (UTC)

Two, or perhaps even a series of, examples would obviously have greater utility than a single one. Pcb21 Pete 10:12, 1 February 2006 (UTC)

Agreed. --logixoul 12:02, 1 April 2006 (UTC)

For what it's worth, I cut and pasted the paragraph into Word (British English version) and used that programme's readability function, which scored it as: Passive sentences 40% Flesch reading ease 22.0 Flesch-Kincaid grade level 15.1 01.11.07 —Preceding unsigned comment added by 194.176.105.40 (talk) 14:45, 1 November 2007 (UTC)

Seuss's first name

Actually, how is Dr Seuss's first name spelt? Theodor or Theodore? I have seen both in entries. §Peck123 (talk) 14:41, 26 March 2008 (UTC)§

Interesting question, maybe you should look it up on Wikipedia ;P 86.172.63.229 (talk) 18:34, 1 November 2009 (UTC)

US grade levels

Can someone indicate what US grades are? Specifically indicate what student age matches a particular grade (assuming regular schooling starts at 5 years of age). In the UK we have at least three different educational systems. Philip, 22:45, 31 October 2005

Seconded, the section explaining the scoring means nothing to people who are not familiar with the American school system. Ages should be used rather than school 'grades'. Fyorl 19:28, 23 August 2007 (UTC)

With respect to US grade levels: Children generally begin 1st grade between ages 5 and 6. 2nd graders are 6-7, 3rd graders are 7-8, 4th graders at 8-9, 5th graders are 9-10, etc. 12th graders (the final grade) are usually 18-19.

That should be 17-18 years old for 12th graders, unless you mean to imply that most Americans redo a grade. --jackohare 04:57, 1 February 2006 (UTC)

The rule of thumb is to take the grade level and just add 5 to figure out their approx age. Hence a 12th grader would be 17 years old. 1st grader would be 6 years old. It's not an exact science, but it's a good quick and dirty way to find age equivalency.

Every other reference I've seen has first graders at 6-7 Fanx 01:50, 28 March 2006 (UTC)

Maybe in the southern states. Here in Seattle we start them at 5.

In Milwaukee, Wisconsin, we start first grade at age 6, then usually turn age 7 somewhere in the year. So therefore, in second grade, we start out age 7, then turn age 8, so on so forth. At high school graduation, most students are 18, or will be turning 18 the following summer.

Either way, the article should use ages and not US grade levels. Fyorl 19:30, 23 August 2007 (UTC)

No. Grade levels are equivalent to number of years of post-kindergarten education. For example, a 2nd grader has had two years of education after kindergarten. Not all people learn at the same rate. Some have to repeat grades for various reasons. Some people can skip grades. Grade level is a more accurate indicator of reading ability than age. A person who has an above average IQ and high aptitude for verbal skills will read much better than a person who has both a lower IQ and a lower aptitude for verbal skills. They may be the same age, but their reading skills are far different. (BTW, with regard to snide remarks about "southern states", I'd put any Floridian scores up against those from the Pacific Northwest any day of the week.) —Preceding unsigned comment added by 66.177.186.19 (talk) 04:44, 6 May 2008 (UTC)

The dispute for the ages is that you can start first grade when you are either 5 years old or 6 years old. Therefore, when you finish it, you will be either 6 years old or 7 years old. It all depends on when in the year your birthday is.--190.74.122.104 (talk) 05:52, 11 September 2008 (UTC)

Okay. Nobody in this section has really addressed the original purpose of this section: The article should replace all US grade levels with actual ages. I would do it myself, but I don't really know what ages a "6-7 grade level" would be or mean. Is that 10-11 or 11-12 years of age? --Thorwald (talk) 23:19, 3 December 2010 (UTC)

Like others have pointed out, average age-grade correlation in the USA differs between states. I know from personal experience because I moved to three different states in high school (when my parents got divorced). I skipped a grade in one state and then got held back a grade in another state simply because their education systems were different. Obviously, it also depends on which month you were born. For example, two students born the same year won't graduate from high school at the same age in May if one already turned 18 in April and the other will still be 17 until June. — Preceding unsigned comment added by 70.251.82.150 (talk) 02:14, 10 November 2011 (UTC)

In addition, removing all references to U.S. grade levels would ignore the fact that the F-K Grade Level test is intended to correlate to U.S. grade levels, not a particular age.63.229.1.242 (talk) 09:02, 9 January 2012 (UTC)

Formula Error?

The formula as I found it (with a plus sign before 84.6) was giving answers >150, which is not possible. I went back to the original article referenced in (7): https://web.archive.org/web/20160712094308/http://www.mang.canterbury.ac.nz/writing_guide/writing/flesch.shtml and this is what it said: "Multiply the average sentence length by 1.015. Multiply the average word length by 84.6. Add the two numbers. Subtract this sum from 206.835." Thus, there is a minus sign before 84.6 (because the original formula had parentheses around the two multipliers). Relleh22hctac (talk) 06:52, 27 August 2018 (UTC)

I have tried using the formula for the Flesch-Kincaid Grade Level and suspect an error: although the final additive constant is widely noted on the web as 15.59, one cannot obtain consistency between the ASL, ASW and the scores with this value: given ASL, RE and GL one can invert the equations and determine ASW. Using the given value of 15.59 one obtains two different results (difference several percent) for ASW; however with a value of 15.9 one obtains almost perfect agreement (~<0.1% difference).

I have tried to find the original paper or other authoritative source online but failed; maybe one of the original author's or contributors is more connected and can verify the formula? I tried MIL-M-38784 but that doesn't give the formula (only states a maximum Grade Level); the US National Technical Information Service for whom Kincaid did work on "developing readability scores" doesn't go back to his 1975 paper.

I suspect the formula currently provided is just the result of a typo that propagated...

Julian I Do Stuff 15:00, 2 August 2007 (UTC)

There also seams to be a typo in the Flesch reading ease equation. The original "old" formula given by Flesch et al. in the 1975 paper is 206.835-1.015(words/sentences)-.836(syllabels/100 words) therefore 84.6(syllabels/words) should be 83.6(syllabels/words). But maybe even the authors were confused later on. — Preceding unsigned comment added by 89.183.70.219 (talk) 11:05, 23 September 2012 (UTC)

Just Way of Deciding Level of Words by Syllables?

I personally believe that the Flesh-Kincaid Formulas are an easy, programmable function to estimate grade levels, but there are many intricate words in the english language that have one or two syllables and simple words that consist of 4 or more syllables. If the system is to be used as a reliable source of grading a piece of writing, then there should be a list for the "levels of difficulties" of all words. I too realize that this is inpractical, and near impossible, but this system can be completly wrong. You can write a "12th" grade or higher piece of work and give it to a 4th grader and they could understand the paper with ease. —Preceding unsigned comment added by 75.186.61.39 (talk) 23:53, 24 May 2008 (UTC)

It can be beyond wrong. Joyce's Ulysses, according to Amazon's Text Stats, has a Flesch-Kincaid score of 6.8. Show me the seventh grader who can comprehend Ulysses, and I'll pay for his college education. Deltabeignet (talk) 09:23, 25 June 2008 (UTC)

I agree. The following sentence by the Word calculation has readablilty 0 and grade lvl requirement 21.3, though it is not particularly challenging; "The obvious association of Hamlet with Wittenberg and Wittenberg with the University and the University with intelligence is Hamlet’s foremost claim to intelligence, but there are manifest examples nonetheless". 121.218.53.10 (talk) 09:12, 4 June 2009 (UTC)Methalos

lol wut

The Rebel Allaince 'Fleet' was now nearing The enemy Patrol, entering the so caled "Demilitarized Zone" Now no longer Peaceful as The Rebel Fleet of Fishing Ships, waveing Kuronan and Questorian Flags were heading for the enemy Naval Yard. Hiden in the Holds of the Ships were Rebel Troopers.

The Ships soon sighted the enemy Patrol Ships and soon started getting near The enemy Ships. It was almost Time

Somehow this gets a score of 41 and a grade level of 13. Somebody ought to get on the horn and let some people know the equation is borked. Oc t ane ^{[improve me?]} 13.08.08 0851 (UTC)

Yep. And Churchill's We shall fight them on the beaches gets a grade level of 12.6... and still it seemed to do its job of encouraging the British public, not just the academics. At best, F-K a flawed measure. Pgagge (talk) 13:56, 18 June 2010 (UTC)

The reading ease of this article.

Flesch reading ease score: medium

   53.2

Automated readability index: high

   14.8

Flesch-Kincaid grade level: medium

   11.9

Coleman-Liau index: medium

   12.9

Gunning fog index: high

   16.9

SMOG index: high

   14.4

Perhaps we should adjust the content of this article to achieve a more desirable score on these statistics. I suggest taking out introductory sentences and breaking apart arbitrary sentences to increase the readability scores. If words with too many syllables become an issue, we can run it through a thesaurus to replace them with the synonym with the least syllables. By following these guidelines we may increase the pleasurableness of reading this article by a factor between 1.3 and 1.45.

71.123.93.195 (talk) 00:33, 6 November 2008 (UTC)

This article

Microsoft Word gave the article a grade level of 12.0, is that interesting? 173.8.158.41 (talk) 19:20, 27 January 2009 (UTC)

Oh Ye of low faith

Content of this article is similar to things I deal with on daily basis. Lets stop quarelling on how precise those "equations" are.

More important is what do they tell us. What is important about them is not 2X+13Y/3.

sentences with small amount of words are easier to read
sentences with complicated (more syllables) words are harder to read
complex-sentences(?) are harder to read

These tests are made to by done by machines or by calculator. They cannot be precise due to statistical error (people they tested), "simplicism" of concise math, names of Comics villains ...

It does not mean they are not desirable. Lets look at other players in this game.

readers have common vocabulary (of words they know) Example? 850 words of Basic english are easy to learn. This makes talking and writing easier.
readers have little time to read it
simple tools force You to be "smart"
people act differently when seen by others or are tested :). See Quantum theory ;)

And so on... 86.61.232.26 (talk) 22:29, 27 April 2009 (UTC)

I think they used common text. I think they used common people to test. Is there how much is common texts complicated? It'd be nice to make graph from it. 86.61.232.26 (talk) 22:44, 27 April 2009 (UTC)

Move to "Flesch-Kincaid readability test"

Flesch-Kincaid Readability Test → Flesch-Kincaid readability test — This article should be moved to Flesch-Kincaid readability test according to WP:CAPS. However, there is already a redirect there to here. That must be removed before this can be moved there. —Preceding unsigned comment added by J. Finkelstein (talk • contribs) 17:20, 24 June 2009 (UTC)

The Flesch Reading Ease and the Flesch Kincaid are two different formulas. The Flesch-Kincaid is a later version done in contract with the U.S. Military to produce a grade-level score. I think there should be a separate page for each of the formulas

This page should refer to the page on Readability, which I am working on and which gives the research background for both formulas.Bdubay (talk) 04:03, 13 November 2010 (UTC)

Score range?

What happens if it scores 80-90? That range isn't in the table. —Preceding unsigned comment added by SME (talk • contribs) 10:06, 1 November 2009 (UTC)

Same with 40 - 60. What gives? (zzyss (talk) 10:32, 27 November 2011 (UTC))

Presumably, the passage is easily understood by a student between "11" and "13 to 15," if the score is between 40 and 60. If the score is between 80 and 90, it is best understood by a person with an experience level between that of a "13- to 15-year-old student" and a "university graduate." Seeing as they're rough guidelines to begin with, I don't think it's terribly nonsensical that the intermediate ranges aren't laid out in detail. 63.229.1.242 (talk) 09:10, 9 January 2012 (UTC)

Low Range

There is cetainly a low range, in response to the article questioning this. It is zero, as the subsequent sentence illustrates. A miniscule argument (action taken against the thesis at hand) pertaining abundances of verisimilitude regarding obfuscation dependency (preponderances of antidisestablishmentarianism), most certainly necessitates propositions of truculence. This also hjas a grade level of 28.1, whatever grade that implies. ExardromaExardroma (talk) 17:36, 19 January 2011 (UTC)

grade level revisited

why does it feel that whoever has written this article has zero understanding of U.S. grade levels? Also, why is the reading ease score example of "year 7" in UK terms? Flesch did his research in the U.S. —Preceding unsigned comment added by Ingridjames (talk • contribs) 14:57, 25 June 2010 (UTC)

Globalise

Eg "Many government agencies require documents or forms to meet specific readability levels." Which governments? All the agencies listed are American. -- PBS (talk) 02:20, 10 July 2010 (UTC)

I live in Surrey, British Columbia, Canada. I have a municipal government, a provincial government, and a federal government. That would be multiple governments. 96.49.151.216 (talk) 16:31, 8 January 2013 (UTC)

Agreed - and the use of the term "college" to mean "post-secondary education" or "university" is very American as well. 198.103.104.11 (talk) —Preceding undated comment added 13:19, 18 May 2017 (UTC)

Removed Google Docs

Google Docs no longer supports the Kincaid test. I had to remove it from the Wiki entry. —Preceding unsigned comment added by 76.102.9.187 (talk) 21:22, 10 August 2010 (UTC)

Table error?

In the section on the Flesch Reading Ease test, the table appears to be incorrect. The text says "higher scores indicate material that is easier to read," but the table shows the low score as "easily understood by an average 11-year-old student" and the high score as "best understood by university graduates." Does anyone know enough about the topic to correct it? I would like to share this article to writers I work with, but in its present form the table will undermine its credibility. —Preceding unsigned comment added by Dbaurac (talk • contribs) 16:29, 15 October 2010 (UTC)

Link ref 1

A better link for the paper by Luo Si and Jamie Callan is http://portal.acm.org/citation.cfm?id=502695

This paper cites http://www.itl.nist.gov/iaui/ovrt/people/sressler/Persp/Views.html as the source for the claim that Flesch-Kinkaid is a US DOD standard ... this paper (Perspectives on Electronic Publishing) in turn cites United States Government Department of Defense standard DOD MIL-M-38784B. This standard was "Canceled Without Replacement" according to http://www.dsp.dla.mil/APP_UIL/content/policy/docs/MilStds_Requiring_DSC_Approval.pdf

~~jiHymas@himivest.com —Preceding unsigned comment added by 65.95.94.247 (talk) 01:52, 21 October 2010 (UTC)

What recent study?

From the article: "A recent study shows that the accuracy of the test is about 12%."

Doesn't the statement above need a citation? —Preceding unsigned comment added by 74.90.120.194 (talk) 17:55, 10 December 2010 (UTC)

Globalized tag

I don't see any discussion here of the "Globalize" tag. The tag doesn't really make sense--what is it that is missing about other countries? I mean, I could add something that says "The test is used in many countries to assess the difficulty of English reading texts, including for non-native English learners," but is that necessary? I'm going to remove the tag, as that type of tag is usually supposed to have a description on talk explaining why it is there. Qwyrxian (talk) 11:31, 21 January 2011 (UTC)

Another Question

If:

Score                   Notes
90.0–100.0    easily understandable by an average 11-year-old student
60.0–70.0     easily understandable by 13- to 15-year-old students
0.0–30.0     best understood by university graduates

Then how about scores from 71-89 and 31-59?

120.61.129.94 (talk) 09:51, 25 April 2011 (UTC)

Some criticism of these tests

In general, the results of any these tests do not change if you alter the order of the words in the phrases. So, a short text like this,

"Jimmy was a little stunned by her courteous manners.  But he had passed all his childhood in a boarding school for orphans and had not been properly trained in the fine manners of society."

Gives us a grade in Flesch-Kincaid of 5.8

Then, you can change the order of the words, "Jimmy was by her courteous manners a little stunned. He had not been in the fine manners of society properly trained for he had passed all his childhood in a boarding school."

So, it has the same results of 5.8. But if the text is apparently grammatical, the awkward order of the words make the text more difficult to understand.

And if the text has not a proper grammar gives also the same result.

Of course, you can use the test to measure how easy is to read a text in another language.

John Galaor (talk) 21:44, 28 June 2011 (UTC)John Galaor

Doesn't work for poetry

At least, it doesn't work in MS Word with poetry. It does if you take the line breaks out. . . — Preceding unsigned comment added by 72.67.14.133 (talk) 14:26, 28 August 2011 (UTC)

Key for levels of Flesch Reading Ease

30-60 and 70-90 are not specified in the table — Preceding unsigned comment added by 141.157.170.202 (talk) 01:30, 16 March 2012 (UTC)

Academic journal articles typically have a Flesch Reading Ease scale of less than 30. My guess is that a score 30 and over means you must be a graduate, 25 - 29 you need a Masters degree 20 - 24 you need a PhD and below 20 well there I think the words Noble Prize Winner is the closest I can come.

Any thoughts? Eric — Preceding unsigned comment added by 86.146.166.76 (talk) 22:28, 17 August 2012 (UTC)

Wikipedia articls can only include information that can be verified in reliable sources, not things that we deduce on our own--we call that original research, which is not allowed. Qwyrxian (talk) 02:05, 18 August 2012 (UTC)

Automatic Flesch scores

http://www.readabilityofwikipedia.com/ calculates scores for texts.

Bad grammar in example sentence

We write: "The sentence, "The Australian platypus is seemingly a hybrid of a mammal and reptilian creature" is a 24.4 as it has 26 syllables."

Say it's a hybrid of mammalian and reptilian creatures, or of a mammal and a reptilian creature instead. I don't know how that would affect the score or I'd do the edit myself.

Move to plural

Wikipedia:Naming conventions (plurals) would seem to indicate that this article should be at Flesch–Kincaid readability tests because it covers a group (of two). ENeville (talk) 15:29, 13 June 2013 (UTC)

That would seem quite sensible. I'm not even going to suggest another couple of capital letters. What a shame it's not an unreadable acronym, even in Czech. unlike the other dozen or so listed at readability test, or is it like? Martinevans123 (talk) 17:54, 13 June 2013 (UTC)

So moved. ENeville (talk) 16:10, 18 June 2013 (UTC)

Example of a low score

Would Sandy's mouthful in Chapter XXII of A Connecticut Yankee in King Arthur’s Court be a better example of a low score? I ran the score manually and got a score of -209.83: 1 sentence with 299 words and 400 syllables.

The language used reduces the reading ease further. Judging from the context, I believe this is Mark Twain having another go at German sentence construction.

198.45.3.26 (talk) 14:46, 1 August 2013 (UTC)Paul Rowe

Original Research for low score sentence

The article states "One particularly long sentence about sharks in chapter 64 of Moby-Dick has a readability score of -146.77" and links to the full Project Gutenberg book as a reference. I'll change this to a link to [chapter 64 http://etc.usf.edu/lit2go/42/moby-dick/745/chapter-64-stubbs-supper/]

The sentence is "Though amid all the smoking horror and diabolism of a sea-fight, sharks will be seen longingly gazing up to the ship's decks, like hungry dogs round a table where red meat is being carved, ready to bolt down every killed man that is tossed to them; and though, while the valiant butchers over the deck-table are thus cannibally carving each other's live meat with carving-knives all gilded and tasselled, the sharks, also, with their jewel-hilted mouths, are quarrelsomely carving away under the table at the dead meat; and though, were you to turn the whole affair upside down, it would still be pretty much the same thing, that is to say, a shocking sharkish business enough for all parties; and though sharks also are the invariable outriders of all slave ships crossing the Atlantic, systematically trotting alongside, to be handy in case a parcel is to be carried anywhere, or a dead slave to be decently buried; and though one or two other like instances might be set down, touching the set terms, places, and occasions, when sharks do most socially congregate, and most hilariously feast; yet is there no conceivable time or occasion when you will find them in such countless numbers, and in gayer or more jovial spirits, than around a dead sperm whale, moored by night to a whaleship at sea."

Running that sentence though an online calculator, I get

 Gunning Fog index :	 94.88
 Coleman Liau index :	 11.01
 Flesch Kincaid Grade level :	 90.56	
 ARI (Automated Readability Index) :	 114.12	
 SMOG :	 29.27	
 Flesch Reading Ease :	 -148.15

Copying the entire chapter into the calculator, I get a more accurate assessment

 Gunning Fog index :	 9.68 
 Coleman Liau index :	 7.65
 Flesch Kincaid Grade level :	 8.18	
 ARI (Automated Readability Index) :	 8.12	
 SMOG :	 9.30	
 Flesch Reading Ease :	 70.44

It seems that the Coleman Liau index is less distorted by a long sentence than the other indices. It would be good to find a reliable third party article talking about the problems with F/K and how some tests are better than others. -- Aronzak (talk) 12:00, 16 August 2013 (UTC)

How to automatically calculate it?

The Gnu link is opaque to me. The blue link on that page leads to a directory. What am I supposed to do with that? The directory items, when clicked on do not lead to a Flesh readability page. Please put in a link to a page that opens simply to a place to enter a text to be analyzed.211.225.33.104 (talk) 02:17, 8 January 2014 (UTC)

you mean the link under external links? Same for me. I am going to delete it. EMsmile (talk) 07:59, 14 August 2017 (UTC)

Swann's Way?

What is that primary-sourced "web" evidence WP:OR doing there, snack dab in the middle of this article? Ok, so Marcel likes a long sentence or two. But Marquez's The Autumn of the Patriarch has sentences that last almost for the entire chapter? And exactly how would most of Finnegans Wake fare in this test? How do we do decide what "extreme examples" should be offered? At least the Moby-Dick example score has a calculated score. Martinevans123 (talk) 22:32, 16 September 2017 (UTC)

[1] ttp://www.merriam-webster.com/dictionary/ubiquitous

[1]