Quote:
Originally Posted by Sunlite
I'm with desertgrandma in this. Even if the average boy is poorer at reading than the average girl; why use the gender loosely connectewd with the wanted characteristic instead of the characteristic? If you want to help poor readers, take poor reader into your test group. It may turn out that the group then contains mostly or even only boys, but the decision point should not be the gender.
|
If you're conducting scientific research, then you always try to minimise the number of "variables" in the experiment. If you want to see if an eBook Reader increases reading ability, then to start with, say, one hundred 12 year old boys and measure their "before and after" reading abilities will give you more useful information than to start with a random mix of 12 year olds of both genders, because on average you know that the girls will do a lot better than the boys.
Of course you could equally well perform the experiment with 100 girls - the point is to minimize the variations by removing gender as a variable in the experiment.
Pretty much everyone who does educational experiments of this nature does so with same-aged, single-sex study groups.