Readability question? (take II)

[Milan Davidovic]

--- Geoff Hart <ghart at videotron.ca> wrote:
> For these parameters to be meaningful, you must  
> prove that two separate assertions are reasonable:
> that longer words are always more difficult to
> understand than shorter words, and that longer
> sentences are always harder to understand than
> shorter sentences.

what about:

Given two texts, one of which is judged (by means
other than readability formulae)  to be more difficult
than the other, the following statements would have to
be true:

- The proportion of longer words is greater in the
more difficult text than in the easier one.
- The proportion of longer sentences is greater in the
more difficult text than in the easier one.

(Hmmm... I think I could have written that a bit more
economically, but my brain's a bit tired at this
hour.)

I'm not sure how we would handle the relative length
of these texts; should they be of equal length --
something not too short for the advanced audience but
not too long for the less advanced audience? They'd
have to be long enough to provide a reasonable sample.

Ultimately, it's texts that readability formulae work
on, not individual words or sentences. I think
therefore that our test of the formulae should use texts.

Milan
http://altmilan.blogspot.com
http://www.terminus1525.ca/studio/view/2758

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com