Re: How can I measure reading comprehension? What kinds of music are suitable?

Hi Javier!

Basically, you are asking for a concept on how to test your hypothesis,
that music affects reading speed and comprehension. Planning your
test-series there are several aspects which you will have to consider.

First, as you already recognized: which music-styles you will be using
in your test. I have often read (sorry, no references here, but probably
"Scientific American" would be a good start) that classical music,
especially of the baroque-Era (Bach, Vivaldi), tend to help reading and
concentration capability. Besides this, I think choosing two or three
modern pieces (Rock, Rap, Dancefloor) as a contrast would do. You might
want to take music volume into consideration as another possible aspect.
You definitely should compare your results to comprehension and
reading-speed without music.

Second the actual comprehension test. Here I would rely on an existing
system or at least take a look at some and let them inspire you. For
children in grade-school I can remember that "SRA reading labs" (see:
 http://www.sra4kids.com/teacher/reading/index.html) was a very nice
system, for older students and adults the "test of english as foreign
language - toefl" (see: http://www.toefl.org/rc-pq.html for a few
examples) provides a very good test of reading comprehension as one of
the aspects tested there. If you don't want to rely on an existing test
(or don't have simple access to one), it would be easiest to collect
several short texts (taking 2 to 5 minutes to read; famous speeches or
short stories are good for this) of similar reading level and devise a
set of 5 to 10 comprehension questions. These questions should be
multiple choice and consider comprehension of the basic idea ("The text
deals mainly with...?", "Which of the following adjectives best
describes the text?"), as well as details.

Third the actual testing. I would let every person read a text under
each reading-situation. It would probably be wise to shuffle the texts
and not to use the same four or five for each person.

Meaning:
Person A reads text 1 with classic, text 2 with silence, text three with
Eminem
Person B reads text 3 with classic, text 1 with silence and text five
with Eminem
Person C reads text four with...

This way you prevent the different difficulty of the texts from
influencing your results to strongly. After each text, the testy answers
the questions to that text and then goes on with the next.

Forth, the analysis of your results. For this you should have the scores
for each person and text, as well as the time a person needed for each
of the read texts. Depending on how precise you want the results you can
just compare the mean reading time and mean score for each music style,
or you can compare the mean-values using a statistical test (probably a
T-test according to "Student" would be adequate). I think the last
variant isn't to much more work and will look a lot more professional
:-)

Good luck and have fun!