Absolutely, methods matter. But there appears to be a very real limit at which changing the method doesn't notably change the results. I don't have a quick cit

Absolutely, methods matter. But there appears to be a very real limit at which changing the method doesn't notably change the results. I don't have a quick cite list. But it's been consistent across literally every single study I've seen that controls for parental involvement/cultural value of education. It's more predictive than parents' education level, neighborhood, curriculum, socioeconomic status, etc. The only off-hand references I have to get you started are what the Michigan Department of Education has been beating my sister-in-law over the head with since she's enrolled my niece in public school. http://docs.google.com/viewer?a=v&q=cache:n3iDh53x4ywJ:www.michigan.gov/documents/Final_Parent_Involvement_Fact_Sheet_14732_7.pdf+educational+study+parents+more+importan+than+schools&hl=en&gl=us&pid=bl&srcid=ADGEESi1Zf8LcSmR43Wnxipj42z7ixpTyG8-rvBRrI-TrZKhKvWYI0OSbgB7PRg0VPNR38poxOKygnqH6Qsty4QS__X-Y6oWU11Mcmqe95elMRX4Hm6volQBcfD892V6Z5OZ-nJh6J6L&sig=AHIEtbQ9gadEneSnbwTjpWyQ1a0MqwZGgg http://docs.google.com/viewer?a=v&q=cache:n3iDh53x4ywJ:w... http://pdonline.ascd.org/pd_online/whatworks/marzano2003_ch13.html http://pdonline.ascd.org/pd_online/whatworks/marzano2003_ch1...