Gary' s inspection of reading test answers is on the mark.  Some tests -
like the Nelson-Denny do NOT have corrections for guessing so if you just
answer more questions, you're bound to get a higher comprehension score.  

Other tests have corrections for quessing -  don't remember how the
Stanford test scoring system works. They used to figure you'd get a
percentage right just by guessing -  one out of five or whatever. 

But CAVEAT EMPTOR -  alsways read the test manual or ask your local
psychometrist  to help you figure out how your tests work. 
In the old days there were tests like the COOP Reading  where you got both
a total comprehension score ( number right out of number of items
completed) and a power score  -  number of items correct out of the first
20 or so items so you could identify the  slow readers with good

Martha M. 

<Gary Probst wrote:

<In grading a reading test, I always check the percent of questions the
<students answered correctly.  I have always found a small number of
<students who correctly answer 90 percent or more of the questions -- but
<because they read and work slowly only answer a few questions.  When the
<timed test is graded, these students are placed into reading because of
<a low reading score. These students do not need a course in improving
<their vocabulary and comprehension skills.  What these students need to
<do is improve their reading speed. The total correct reading score which
<is usually used for placement can be very misleading.

At our college I have noticed that students answer on the Standford
Reading test 30 to 40 of the 60 comprehension questions.  They usually
get 75 percent of the questions they answer correct.  After a 15 week
reading course, they answer 45 to 60 comprehension question and still
get 75 percent correct.  This leads me to believe that reading speed is
the cause of the improvement in reading score and not an improvement in
comprehension skills.