In this thread I will tweet some of the errors observed while reading Catherine Bliss's new book about social science genomics, Social by Nature. (I'm interviewee #9 in the book, btw.)
Ben Domingue is in Stanford's School of Education. @bendomingue 
It's Rehkopf. @drehkopf
Freese and Fletcher did not overlap as faculty at Wisconsin and have never co-authored with one another.
I don't think Fletcher has had a School of Medicine appointment, and his CV and webpage don't say anything about one.
Christopher Ellison was a prominent sociologist of religion at Texas when he collaborated with grad student Matt Bradshaw (subsequently hired at Duke, now I think at Baylor).
RAND mini-med started in 1998.
The American Journal of Sociology had a special issue in 2008.
Jesus wept. Quoting an interviewee here, who was maybe talking about "ESTIMATING SYSTEMS of equations."
Quoting an interviewee here, talking about "NULL hypothesis testing"
I'm not exactly sure what is meant by "medical partnership" here, but this is not the first.
It's deCODE.
More men than women, but more like 60/40 than >75/<25. (Google "IGSS 2016" or "IGSS 2017" and you can see the conference participant lists from the last couple years.)
"This agency" refers here to NSF, but the workshops are actually funded by NIMH.
"Genetic methodological validity" is not a thing.
This is me being quoted. I would be very surprised if I did not say "IN THE BRAIN," which is important for the point I'm making.
This doesn't make sense as worded, and I think might be mashing together a couple of separate complaints that this economist informant had.
For better or worse, this has it backwards in terms of current practice. Analyses are more likely to separate major race/ethnic groups in analysis than to separate men vs. women (although the latter is also common).
A few minutes of Googling news stories turns up dissent fairly easily.
Just one example of several in which interviewee is quoted at length and then some summary proclamation made about it that has no obvious relationship with what the interviewee said.
E.O. Wilson was 17 in 1946. Wikipedia credits John Paul Scott with introducing the term in 1948.
(Not stopping because that's the end, but need to do other things, so maybe more later.)
Not commuted (still on death row at least as of spring 2017); the genetic angle was mentioned only once and incidentally in decision; name is actually Kenneth Eugene Barrett. (This is the only one of the three case that I looked up.)
What Albert actually says in the Fox News story: "It would be premature to use this finding to screen children to determine who should receive intervention."
Meanwhile, there are various examples of interviewees quoted as saying weird or dumb things for which it's hard to know where between interviewee brain and published quote the muddle is.
Not a thing. If what's meant is "econometrician," not correct. (Economist.)
I taught a "full course" on this multiple times at Northwestern as part of their Society, Biology, & Health cluster, and already have at Stanford with @bendomingue. [ht: @jasonmfletcher]
Obviously incorrect. Many events each year, not to mention exhibitor booths, at professional meetings for the studies in question, at which the existence of genetic data is incidental.
One of several examples in which she suggests folks doing social genomics are megalomaniacs, but then the quote she provides as her best evidence doesn't imply that at all.
As anyone who read the article would see from its first page, Caspi and Moffitt were at Wisconsin and the Institute of Psychiatry in London. (They did move to Duke ~5 years later.)
While @danbelsky is an expert on many things, "health policy" is neither his job title nor research focus.
This is referring to me and the Wisconsin Longitudinal Study. I don't lead WLS, @pamela_herd does. Elsewhere Bliss gives me sole credit for a grant on which Pam is co-PI.
As for the "early 2000s," WLS did not have any genetic data available for analysis until 2009.
This makes no sense. The interviewee presumably said "number of SEXUAL partners."
Ludicrous assertion that is easily refuted by a simple literature search.
C'mon! Does it really seem plausible that a group of economists and sociologists sit around and see no difference between educational attainment and Alzheimer's disease? (I talk about the different explanatory structure of social outcomes at length in my 2008 AJS.)
Pleiotropy (single gene affecting different phenotypes) is roughly the opposite of gene-gene analysis (different genes moderating each other on single phenotype).
Of course this one is minor, but still: It's a "stressful LIFE event." A stressful live event would be like watching the Super Bowl go into overtime or something.
Utterly standard social science methods of measuring educational attainment presented as "new metrics" that are somehow transforming "social measures into genetic ones."
There may be some poli sci folks who talk like this sometimes, but the idea that this is a core tenet or point of pride for the subfield is a bizarro assertion to me. (And I'm an interviewee identified as central in the subfield. Wouldn't I know?)
One instance of how the book will make some claim and then offer a long quote that's supposedly an example, but then, if you actually read the quote, it contradicts the claim.
I suspect demographers and those who work with secondary data will understand why this is irksome without explanation. (In any case, once a cohort is genotyped, you could make an analogous claim about the genome side of subsequent research as well.)
Another one that seems to me like, "Hey, wait, doesn't this quote say the opposite of what you are saying it says?"
The last part of this makes me smile too much to really quibble about how I was no longer at Wisconsin by this time.
Of course, Jeremy must be exaggerating when he says the book accuses researchers of being unwitting pawns in a dystopian project.
Trivial, but telling? As @MichelleNMeyer mentioned, the very first word of the book is a mistake.
Adjacent sentences that directly contradict one another. (Admittedly, you can tell from context which one is correct.)
The example is not an example of differential susceptibility theory (it's just a phenotype-environment interaction). Moreover, the example is taken from a newspaper article that never refers to differential susceptibility theory.
So famous in the field that I had forgotten that it existed. (Less than 100 citations in Google Scholar, not even in the top 20 cited works for the editor in question.)
First half of highlighted sentence is incorrect (as systematic inspection of literature would show), whereas second half is inscrutable.
Example of sprawling text-quote introduced as evidence that, when read correctly, actually contradicts what is asserted about it both before and afterwards.
I have never heard anyone who does social genomic research describe their work as "directing us to a more utopian society." Have heard people offer ways work may contribute downstream to social good; not sure if this is just a tactic for imbuing that with a creepy cast.
Some of the things that the book asserts are just so weird, and then there isn't any cite or anything.
Anyhoo, still other errors of varying scope that I noted here and there, but that might be it for tweets based on my markup of my initial read of book.
Made note to look this up for my sport class. Alas, entity no longer exists and hasn't for some time; looks like Stanford study/group rather than "company"; focus on sports injury rather than "athletic prowess"; doesn't appear there was really a "collaboration" with 23andme.
