Some ongoing research projects where I’m looking for collaborators. The list is not exhaustive but covers most of the main lines of research I’m currently involved in. You can contact me at email@example.com.
1) Immigrant outcomes in Western countries
Studies like these:
The studies are quite simple, but bizarrely no one else does them. I already have lots of data for these not analyzed or not published. Here I am primarily in need of writers to help me write up the results and submit them to journals.
2) US Black-White gap over time
There hasn’t been a large-scale review of this for a while (since Roth 2001). We have done a lot of preliminary work, but much data dredging remains to be done. However, there’s not so many surprises here.
Working on this basically just means trying to locate datasets and studies with relevant data, then writing them into a spreadsheet in a standardized format. The analytic work is easy and has already been mostly completed.
3) Genomics of race and IQ
The death blow to the blank slate awaits in the analysis of genomics. We have already done 2 admixture studies (PING study here), both with hereditarian results. There’s a meta-analysis we did of admixture studies of social outcomes, no surprises either.
Unfortunately, the data to do these studies is hard to get because of the usual IRB reasons. Only professors can apply for them etc. etc. We’re working on this already.
4) Old-school studies of race and IQ
John Fuerst wants to do a big review of all the pre-genomics studies of racial admixture and IQ. This is basically the same kind of work as the black-white gap in (2), i.e. skim lots of old studies, report results in a spreadsheet in standardized format. This task is probably harder than (2) because many of these studies are very old, hard to even get a PDF of, and report the wrong statistics we can’t easily meta-analyze (e.g. reports means on a test no one has normals for, and no standard deviations, thus making it not possible to calculate standardized effect sizes).
5) Regional studies of IQ
It’s not too difficult but requires someone to write up the results. In many cases, I have already got the data, but it is unanalyzed or in some foreign language like Spanish/Portuguese which is cumbersome to work with without a good speaker. Primarily need writing help and foreign language help. We are primarily interested in large-n studies, as most of the previous studies have been underpowered to do anything but correlations. Lower level administrative unit data is available for many countries, see Wikipedia’s overview for inspiration.
6) Face analytics
We’re doing a big project on building predictive models of personality etc. traits from face data. We scrape data from various dating sites for the data. Currently, we are in the scraping phase and already have a trial dataset of n = 10k one can begin developing on. This requires Python skills in image analysis. Primarily looking for Python coders who wants to work on the image analysis. We’re looking to do studies like these:
7) Stereotype research
If you’ve done your research, you know that researchers and the media claim all sorts of silly things about stereotypes, whereas the research on them basically shows them to be quite accurate, and when biased, often in the wrong way too, and when not right in size, then under-estimate real differences rather than exaggerate them. We have a program of contributing to this research. See e.g. recent papers:
Of course, these are quite difficult to get out in normie journals, as gatekeepers (=editors and reviewers) realize their potential and up the method requirements accordingly.
Doing more of these requires mainly that someone comes up with an idea and designs the survey. We gather the data online using e.g. prolific.ac/. Help needed here: ideas for studies, survey design, and writing.
8) Cognitive meritocracy
The Bell Curve is the most famous book arguing that society is mostly a cognitive meritocracy. However, lots of more research on this question can be done, especially to contrast the model with competitors like discrimination model. Large public or semi-public datasets with IQ measures and various alternative proposed causes can be used to examine the question. Familial can also be used.
9) Academic and media biases
Since Duerte et al 2015, there has been a strong and growing case for the relevance of studying researcher biases and how they play together with QRPs to distort the scientific process and hence the literature. A similar problem exists in the media, yet it is even less well studied (but see book length review of US data here). A variety of approaches including text-based machine learning methods can be used to study these questions.