4 ms·

Cambridge University Fires a Young Scholar for 'Problematic Research'

7 points by vixen99 8y ago

nabla9 8y ago
Not problematic research. For racist pseudoscience published in OpenPsych pseudojournals. Noah Car is not an edge case.
- klohto 8y ago
  How exactly is research backed by data racist? https://openpsych.net/files/papers/Carl_2016b.pdf https://openpsych.net/files/papers/Carl_2016b.pdf
  - nabla9 8y ago
    Before even looking at the article I must ask: is simple counterexample of using data correctly important argument? Or should we consider the whole body of work. Generally speaking must look the relevance and validity of data and methods used in each case. Even if it's correct it can be collected and used to argue intentionally misleading way. "Using data" is not enough for science.
    - educationdata 8y ago
      You answered wrong question. klohto never asked you whether using data is enough for science. he was asking how is this racist. Do you mean if it is not enough for science then it must be racist?
- educationdata 8y ago
  The university never made such claim. What evidence do you have? You should not throw "racist", "pseudoscience" around like this.
  - nabla9 8y ago
    >The panel found that Dr Carl had put a body of work into the public domain that did not comply with established criteria for research ethics and integrity. In any event, it considered that the poor scholarship of this problematic body of Dr Carl’s work, among other things, meant that it fell outside any protection that might otherwise be claimed for academic freedom of speech.
    - educationdata 8y ago
      Which part of your quote has anything to do with "racist", "pseudoscience" ?
IshKebab 8y ago
I don't think you need to be an expert in "psychology, intelligence research (a sub-field of psychology), and/or economics" to understand this paper. I read it and it was easy to follow. It also has nothing to do with intelligence research or economics. My biggest issue with the paper is that he uses a log-log plot ("to reduce skewness"?) and then does a linear regression on that (and we all know everything looks correlated on a log-log plot). Also the resulting correlation is claimed to be very strong, but I think that psychologists so rarely see any correlation that "very strong" to them is "there's definitely some correlation!". He shouldn't have been fired though.

7 points by vixen99 8y ago

nabla9 8y ago

Not problematic research. For racist pseudoscience published in OpenPsych pseudojournals. Noah Car is not an edge case.

klohto 8y ago
How exactly is research backed by data racist? https://openpsych.net/files/papers/Carl_2016b.pdf https://openpsych.net/files/papers/Carl_2016b.pdf
- nabla9 8y ago
  Before even looking at the article I must ask: is simple counterexample of using data correctly important argument? Or should we consider the whole body of work. Generally speaking must look the relevance and validity of data and methods used in each case. Even if it's correct it can be collected and used to argue intentionally misleading way. "Using data" is not enough for science.
  - educationdata 8y ago
    You answered wrong question. klohto never asked you whether using data is enough for science. he was asking how is this racist. Do you mean if it is not enough for science then it must be racist?
educationdata 8y ago
The university never made such claim. What evidence do you have? You should not throw "racist", "pseudoscience" around like this.
- nabla9 8y ago
  >The panel found that Dr Carl had put a body of work into the public domain that did not comply with established criteria for research ethics and integrity. In any event, it considered that the poor scholarship of this problematic body of Dr Carl’s work, among other things, meant that it fell outside any protection that might otherwise be claimed for academic freedom of speech.
  - educationdata 8y ago
    Which part of your quote has anything to do with "racist", "pseudoscience" ?

IshKebab 8y ago

I don't think you need to be an expert in "psychology, intelligence research (a sub-field of psychology), and/or economics" to understand this paper. I read it and it was easy to follow. It also has nothing to do with intelligence research or economics. My biggest issue with the paper is that he uses a log-log plot ("to reduce skewness"?) and then does a linear regression on that (and we all know everything looks correlated on a log-log plot). Also the resulting correlation is claimed to be very strong, but I think that psychologists so rarely see any correlation that "very strong" to them is "there's definitely some correlation!". He shouldn't have been fired though.