hi! neat script! the layout seems to be broken, though. I get loads of whitespace between the input box of the script and your last paragraph.

on a sidenote, the comment form looks a bit weird as well -- but it's cool :-) // rck

I find it really pointless.

Nice Script Dave. Been looking for something similar for UK gove websites I work for.

BTW your intro text to the script didn't score very well, whereas the previous posters 'I find it really pointless' probably did very well... which i find strangely humourous... any how respect to you again, and where can i get my hands on this script? Theres some councilspeak out there I gotta deal with.

Paul

very helpful. I am on Linux and their is no feasible readbility checker for me.

It always tells me when something I write is crap (of course I always have a high ego)

oops...that should be "THERE", sorry

Really nice script - I've just linked it in a post, along with another tool (which makes use of some other methods to evaluate texts in Spanish, Swedish, French, ...; still no Portuguese ;) ).

Just one correction: it's the Gunning-*Fog* Index. :)

Handy script! I am working towards a MS degree and my assugnment was just graded using the Flescher Readability score method. I ran the same text through your script which scored rather high, i.e., non-readable for the average reader (9th greade level). I tried rewriting the exercise to see what it took to get a 9th grade level. To my surprise, I could only get it down to 11th grade level and it already read like some moron had written it (lol). Are you sure the script is correct? How do you calculate the number of syllables per word?
Even if not totally accurate, I find the script quite useful. Thanks for providing it.

Oops, assugnment = assignment

ILoveJackDaniels, what is the exact method you use to calculate the number of syllables?

(yes, I do realize this post is over a year old)

<http://www.ilovejackdaniels.com/resources/readability-score/> nav column renders wrong in Explorer 6.0 SP2, ;-(

VG in other browsers.
CONGRATULATIONS for the site

Its a great tool that I can't stop using. Have found another tool at http://resources.aellalei.com/writer/sample.php Only, I keep wondering why the results are that different. Actually, what would be interesting to know is how that model works. Did you check it against your model?

I found the results for the Fog index to be different than my own calculations, but then again I was discounting some words myself that I didn't think should be included - e.g place names.

Still strange though.

As Dezso Papp said, the Obsidian (aellalei) readability analysis gives very different results (e.g., your site gives my text a Flesch-Kincaid Grade Level of 6; theirs, a 3.96). Also, I would enjoy a tool that uses the Powers-Sumner-Kearl formula. Thanks.

Great stuff. We do some statistical investigations using these measures as part of the Year 6 maths course at my school, and this is going to save so much time. Many thanks for making this public.

I am giving a presentation tomorrow on leveled reading. The best part of the presentation will be when I tell the class that at www.ilovejackdaniels.com they can find a handy little readability tool. School should involve more whisky.

Great tool. I'm an editor for a childern's educatonal publisher. I am beginning work on a new series of maths literacy books. I will be using your site regularly to check manuscripts.

Great tool! I use the openoffice.org work suite (website posted as homepage) . I think it's an excellent alternative to MS Office. However, one of the only features the word processor, writer, doesn't have is the is the ability to get stats like grade level and reading ease. This tool solves that little problem. Though if i may make a suggestion, it'd be great if the average sentences per paragraph was included in the notes. Now if only OpenOffice added the ability to drag and drop text, it'd be without a shadow of a doubt the superior office suite.

I absolutely love this tool. I am working on my MA and I use this for every paper I write. I always want to make sure I am writing graduate level work. Thank You!

I really like this script and I see where the problems arrise from the syllable count. If you run the count_syllables() function by itself and then manually count the number of syllables the function usually returns a higher value. There isn't a sufficient rule for words that end in the letter "e". Here are a few 1 syllable words that the function counts as 2:
* have
* house
* hive
* base
* live
* leave
* brave
Here are a few 2 syllable words that the function counts as 2:
* behave
* release
* combine
Additionally, there are a few words in English (At least American English) that look very much like two syllables but are only one like the following:
* lived
* lives
* cares
* rates

Now for the real issue, we have all of those wordes the end with an "e" that should be counted a 1 syllable but then we have others that are supposed to be 2 syllables like the following:
* mile
* tile
* aisle
* isle
* style
One last thought, 'ses and 'ces!
bases and faces are both 2 syllable words but lives and cares are only one syllable.

Any ideas of how we might write a few more rules for these types of words?

Thanks.

Hey dave,

Thanks for the reply. The only reason I bring up this particular issue is doue to the very large number of words in English that end with an 'e'. I don't think that it is important to try to come up with rules for every single situation that a word ends with "e" but it would be nice to get the top 80% which is about what is being missed. The other 20% would work isself out in the averages much like your name:
Dave = 2 syllables
Child = 1 syllable
3 syllables is correct but that was because the script got lucky. ;)

But all things considered, I think that the readability scripts you have put together work very well. There only real weakness from what I can tell is the syllable rocognition system which is difficult to fix since in English, the exception is the rule. lol
I was just wondering if anyone has considered expanding the list of rules to account for a greater number of situations.

vujsa

I need something just like this for school when writing quizzes for Reading Counts. We don't have Internet at our school. Is there a "portable" version of this that I can install on my Library computer?

Is there a "portable" version of this that I can install on my Library computer?

-----

Assuming you have MS Office on your computer, this scale as well as many others can be activated via "Tools > Options > Spelling & Grammer > Readability Statistics".

hey this helped out loads as a student wanting to know if what i was writing was going to be too complicated for examiners it especially helped because i tend to write long and complicated essays that can get out of hand and i dont tend to notice.
so thanks dave and keep up the great work

Excellent tool for nerds. I completely dig it. From a writing perspective, it is difficult to understand how to use the information. E.G. reading level 11.9 readability 42. Low readability, but snobbish M.B.A. profs like it, so who is to say?

Maybe someone else finds the readability scale pointless, but I find it useful. I use it to get a rough idea of whether or not I'm keeping what I write simple.

Your layouts borked. Validation might help: http://validator.w3.org/check?uri=http%3A%2F%2Fwww.ilovejackdaniels.com%2Fresources%2Freadability-score%2F

Nice idea though.

Thanks for a great tool and a fun website.

I'd like to add a request: could the results include Total # of Words? And while I'm asking, how about a recommended minimum text to submit?

I think some of my students are trying to use samples that are too small. Thanks again.

Excllent tool,now bookmarked as a study aid. Nice to know I'm writing at Post-grad level!
Would never 'ave thunk it....

Just need a clarification - If I have s script which contains Tables,Tabular tags as well as Text,how the scores will be computed.whether it ignores or considers those Tables and tags.

How does your calcualtion differ from the Readibility and Flesch-Kincaid Grade Level tests in Tools in Microsoft word. The Grade level, words per sentence and readibility levels differ.

its nice but i wish you include readability of a paragraph using gunning fog index.thanks..

I think I found why vujsa was having problems with the syllable count.

In the "count_syllables" function there is two lists of syllables. One for adding to the count, one for subtracting.

If you look at the values in those lists, they appear to be Regular Expressions.
However the function those values are being passed too "strpos" takes a literal string argument, not a regular expression.

I'm getting expected results for syllables using the following loops instead of the "strpos" loops in that function.

foreach ($subsyl as &$syl) {
$syllables -= preg_match('~'.$syl.'~', $word);
}
foreach ($addsyl as &$syl) {
$syllables += preg_match('~'.$syl.'~', $word);
}

Note that in PHP < 5 "as &$syl" needs to be "as $syl" because PHP < 5 doesn't support byref in foreach loops.

Hi! Using the code on the two pages linked from here (copy paste, even), I get different results for the Gunning-Fog Index. Did you change something in your live code but not the example? Regardless -- great code! Thanks for sharing.

I have question about the constant in readability formula. Where they come from?

Like Flesch readability formula:

ReadingEase = 206.835− (1.015 * ASL) − (0.846 *AWL)

Also, Can we apply it to other language like French ?

I am using a couple of other sites as well to decide the readability level (mainly FK) of several sites am using for my research. The problem is ilovejackdaniels, juicystudio, and readability.info, all three show three different grades for the same text. Is there any way I can justify this in my research? And which one shall I consider?

Dave,

I have a question. I am trying to find the readability of questions ona reading test. I am typing each writing prompt into readability checkers such as yours. I get a different level for the Flesch Kincaid and Gunning-Fog depending on the checker I use. There are trends that show a common relationship among the increasing and decreasing difficulties from one question to another, but each question checks in at a different level. Can you explain why this may be?

Love the new design. I noticed after your site scores the text the hints next to the scores are missing, the ones that hinted where you would want the score to be around. Is there a reason you took them off? Is it possible to get those back, it helps me know how close I was.

Great to see this tool still available online.

Only minor glitch I've encountered - when assessing copy, I get "Flesch-Kincaid Reading Ease" twice - should the lower number be the Grade?

Thanks again and keep up the great work!

Dave or anyone with an answer; I just have 2 questions:
1. Where can I find the explanation of the different scores. I know what the FK Readability scores mean but what is the Gunnig Fog and the others listed on your readability scores?
2. Also, could anyone explain how an FK Reading Ease score of 31 means college graduate but the Grade Level score is still 12?
Thanks for this gadget, I am beginning to learn how to use the APA writing format and tools like this have really helped me adjust my writing skills.

Very cool! I teach technical writing to high school students and one lesson is on calculating readability. They just love to pack a paragraph full of jargon and say, "Hey, it sounds technical!" This will save me from bashing my head against that bit of concrete that block my students' understanding. We'll use this in class this fall. Thanks!

I think it is great that you displayed so many different readability formulas. I am a long time teacher and I use leveled text with students. I decided to compare the readabilities here with text that had been leveled using Reading Recovery, IRI, and F&P materials that have been leveled by having students read the text. I knew the levels of the text I put in. What this showed me was that none of the readability formulas is accurate all of the time because a large part of reading depends on the readers interaction with the text. Flesh-Kincaid was accurate 50% of the time and when it is off it is usually but not always high; SMOG was accurate about 40% of the time for all but two of the 40 samples I used it was high when it was off; The Gunning and Coleman indexes were accurate 0% of the time generally they were 3-4 years high but sometimes as much as 7 years high and then on a couple of occasions low; The automated readibility was accurate 50% of the time on text over 4th grade when it was not accurate it was equally incorrect on the high side half the time and the low side half the time. On text below 4th grade reading level it was only accurate 5% of the time and generally high.

My suggestion to anyone leveling text is that you may be able to use an index of this nature to get an apporximate level - especially on text at a 4th grade level or above, but then you really need to have real people whose reading level is known read the text you are checking to get an accurate level. it is certainly more work but the only way to really make sure that what you are writing will fit your audience. If you are a teacher trying to level a text the same is true, this can give you an approximate level, but then you have to check the text against your students actual reaing levels.

By the way on when the site reports the results it says Flesh-Kincaid Reading ease for the reading ease % and then it says Flesh -Kincaid Reading Ease again when it should just say Flesh-Kincaid level or Flesh-Kincaid Score.

Something is wrong in the output:
Flesch-Kincaid Reading Ease: 58.9 (?)
Flesch-Kincaid Reading Ease: 10.9 (?)

I think the correct output is:
Flesch Reading Ease: 58.9
Flesch-Kincaid Grade: 10.9