If you follow the links further, Bryan's case seems a little more solid.
Claim 1: ...The claim is implausible (if such parenting can make the kids neurotic, other personality changes must certainly be possible), and the link is to an article where the blog author is not convinced by a book arguing the contrary of this claim. The argument is roughly analogous to saying the heliocentric theory has zero credibility because "De revolutionibus orbium coelestium" didn't contain enough measurements.
Bryan Caplan cites his own blog post ("Implausible Wimps"), which explains that it's not a problem of too few measurements. From his post:
"...a massive twin and adoption literature on personality finds that family environment has little or no effect on personality...my complaint isn't just that Marano's case isn't airtight; my complaint is that she doesn't even try to rebut the hereditarian presumption long-established in personality psychology."
The book "Unequal Chances" is cited, which allegedly reviews the field and shows that personality is primarily hereditary.
Regarding Claim 2: Caplan cites his post "Don't miss the invisible gorilla", which in turn cites hist post "Magic Potencies", which cites the book "How People Learn":
This particular paper attempts to measure (among other things) "how the knowledge learned from a particular teacher influences student learning of new material."
I think Bryan Caplan is mainly guilty of too many self links, which obscures the original sources he is basing his opinions on.
Maybe claim 1 is true, but I still think it's badly supported here. Without having read "Unequal Chances", its subtitle "Family Background and Economic Success" seems to indicate that it's not mainly concerned with personality and character issues. The least that Caplan could do is source the relevant part of this book directly, instead of linking to his own blog post, which only mentions the book in one sentence with hardly any detail.
Again, about claim 2, if there is better support available, why not quote it directly? It's hardly good writing to make an important claim that requires the reader to go through reference to reference to reference and read a whole book in order to find some passage that may or may not be what Caplan's claim is based on.
Regarding claim 3, I still fail to see how that has any bearing on teacher's influence on character. Even that weak claim is not supported, much less the central issue whether parental influence (presumably much stronger, at least in this context) does, or does not, affect the formation of character and personality in their children, particularly in the case when parents choose to be very intrusive.
Amazon's preview feature suggests Chapter 7 ("Personality and the Intergenerational Transmission of Economic Status") is the relevant one. The introduction also discusses "genetically based behavioral characteristics" and how they relate to labor market success. Skimming chapter 1, it also appears that statistics on transmission of personality traits are given.
Again, about claim 2, if there is better support available, why not quote it directly?
Agreed, Caplan should cite his sources better.
As for Claim 3, I guess it depends on what one means by "character". Caplan's source suggests that whatever character effects do exist, they do not significantly effect schooling.
(This would imply that character traits like conscientiousness and a desire to appease authorities are not influenced much by environment, but perhaps not a trait like being kind-natured would be.)
Claim 1: ...The claim is implausible (if such parenting can make the kids neurotic, other personality changes must certainly be possible), and the link is to an article where the blog author is not convinced by a book arguing the contrary of this claim. The argument is roughly analogous to saying the heliocentric theory has zero credibility because "De revolutionibus orbium coelestium" didn't contain enough measurements.
Bryan Caplan cites his own blog post ("Implausible Wimps"), which explains that it's not a problem of too few measurements. From his post:
"...a massive twin and adoption literature on personality finds that family environment has little or no effect on personality...my complaint isn't just that Marano's case isn't airtight; my complaint is that she doesn't even try to rebut the hereditarian presumption long-established in personality psychology."
The book "Unequal Chances" is cited, which allegedly reviews the field and shows that personality is primarily hereditary.
Regarding Claim 2: Caplan cites his post "Don't miss the invisible gorilla", which in turn cites hist post "Magic Potencies", which cites the book "How People Learn":
http://www.nap.edu/openbook.php?record_id=6160
This book appears to talk about more than just learning Latin, though it also uses Latin as an example.
Regarding claim 3, the underlying paper being cited is here: http://www.google.com/url?sa=t&source=web&cd=4&v...
This particular paper attempts to measure (among other things) "how the knowledge learned from a particular teacher influences student learning of new material."
I think Bryan Caplan is mainly guilty of too many self links, which obscures the original sources he is basing his opinions on.