Re: readability

From: MarkH38514@aol.com
Date: Fri Jul 04 1997 - 07:08:09 EDT


Return-Path: <nifl-health@literacy.nifl.gov>
Received: from literacy (localhost.nifl.gov [127.0.0.1]) by literacy.nifl.gov (8.8.5/8.8.5) with SMTP id HAA24108; Fri, 4 Jul 1997 07:08:09 -0400 (EDT)
Date: Fri, 4 Jul 1997 07:08:09 -0400 (EDT)
Message-Id: <970704070720_-1126161529@emout19.mail.aol.com>
Errors-To: lmann@literacy.nifl.gov
Reply-To: nifl-health@literacy.nifl.gov
Originator: nifl-health@literacy.nifl.gov
Sender: nifl-health@literacy.nifl.gov
Precedence: bulk
From: MarkH38514@aol.com
To: Multiple recipients of list <nifl-health@literacy.nifl.gov>
Subject: Re: readability
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Status: RO

The article that Audrey is referring to is: 
Mailloux, S.L., Johnson, M.E., Fisher, D.G., et al (1995) How reliable is
computerized assessment of readability? Computers in Nursing, 13(5), 221-225.


They found in their comparison of four software programs (Corporate Voice,
Grammatix (sic) IV, MS Word and RightWriter) that each gave a somewhat
different estimate--the range of scores for the Flesch-Kincaid ranged from
5.6 to 7.2. The authors noted that readability formulas used in the software
were provided for three of the four programs, and although the formulas were
identical, the grade results were different. They concluded that "This
finding is difficult to explain because if the formula (sic) were truly
identical, no discrepancy should be found" (p. 224) What they don't explain
is that even if the formulas were identical, the programmers for each of the
four software programs still had to give the program a way of identifying and
counting words and syllables. 

There are different ways of counting syllables, including the number of
vowels per word, the number of consonants per word, or the number of letters
per word. Based on the information in the article, I'm not convinced that the
authors really had an in-depth understanding of the software that they were
working with. 

If a sentence is separated by a colon or semi-colon, some programs will count
it as one sentence, some as two sentences. Most readability programs
calculate the number of sentences by counting periods. Consider the sentence
"Research shows that 1.2% of patients are likely to suffer from angina (i.e.,
chest pain). There are four periods, which will calculate as four sentences
(1 sentence of 4 words, one of 10 words, one of 1 word and one of 2 words).
Instead of one sentence of 15 words you might get 4 sentences averaging 4.25
words/sentence. 

Readability users must be careful to remove periods that do not come at the
end of a sentence.  From a readability standpoint, the sentence should read
that "Research shows that 12% of patients are likely to suffer from angina
(ie, chest pain)." if it's to be counted as one sentence. 

This means that the user must be careful to set the options for each software
program. As I've reviewed the readability literature, I can't recall a single
article in which the authors described the "settings" that they used, or the
strategies that they used to clean files with respect to periods,
semi-colons, colons, dashes, etc. 

Mark Hochhauser
3344 Scott AVenue North
Golden Valley, MN  55422-2748
Phone/Fax: (612) 521-4672
MarkH38514@aol.com



This archive was generated by hypermail 2b29 : Tue Jan 11 2000 - 13:23:46 EST