This remarkably thin personal blog post seems to be getting some traction, so I will remark (subject to editing in a moment) that the classic way to figure out "why is everyone like their parents?" was to do a study with a "genetically sensitive design." The ways to do that include studying identical (monozygotic, or "MZ") twins brought up in differing households, usually the result of adoptions splitting up twin pairs, sometimes from each parent of a twin pair bringing up one child after a divorce. More modern studies include designs such as studying the children of co-twins, that is studying first cousins who each have an uncle or aunt who is an MZ twin with one of their parents. There are many variations of studies like these.

Professor Eric Turkheimer, who I think is the current president of the Behavior Genetic Association (certainly he is very active in the association) does the scholarly community a favor by putting direct links to most of his papers on behavior genetics on his faculty webpage.

http://people.virginia.edu/~ent3c/vita1_turkheimer.htm

You can learn a lot about what makes children similar to--and different from--their parents by reading those papers. Another author who puts many of his papers on behavior genetics up on the Web is the younger researcher Lars Penke,

http://www.larspenke.eu/en/publications.html

who has published as a co-author with some of the leading researchers on human behavior genetics.