For as long as I've been teaching, I've been changing homework patterns and expectations. That's 30 years of homework flux. Why the variability? Mostly this is true because teaching and learning changes all the time and student groups vary from one group to another.
In light of new research and what I've observed this year with regard to math learning and teaching, I think I'll try homework 1 2 3 next year. One will be a problem that relates to the days learning--a problem I expect students to answer in about 10 minutes. Two will be a short reflection about the day's learning, and 3 will include suggestions for possible further study and practice if a child desires it.
I think this 1 2 3 homework will work well for today's culture for the following reason. First it will inspire students to take a few minutes each day to reflect about their math learning and to practice what we're learning at that time. Next it will be simple work to review and provide some feedback to, and finally the work won't create a lot of home-school challenge since it won't rely solely on computers, family help, or lots and lots of extra time. I'll try to make the daily problem meaningful, interesting, and relevant too.
It's true that in the past I've given homework that's overbearing and unaccessible to some. I don't want to do this anymore particularly in light of the latest research I've read with regard to homework and equity Yet, I want to still promote some daily practice and reflection as it's so important and we simply run out of time in school for that kind of quiet, regular, try it yourself work.