3 ms·

> Why not have scientists go off and do their research and rely on a team that is skilled at hoarding grants or publishing works from raw data? Partly because

> Why not have scientists go off and do their research and rely on a team that is skilled at hoarding grants or publishing works from raw data? Partly because science is very specialized. If I work on something new, then it's possible there is very few people in the world that know it or will easily understand it. I couldn't have someone else write a paper or grant proposal based on it - either they wouldn't understand it, or they are a fellow scientist who wouldn't want to do it either. The mistake, in my opinion, is making the number of papers published a metric on which you determine a scientist's worth. Just like most metric-ization of society today, it's a bad idea.

neuromantik8086 10y ago
I disagree on your first point- I think that there could be someone else who could write a grant proposal for me, provided that they a) expressed an interest in grant writing over benchwork / data analysis and b) were fluent enough in the field's literature to be able to translate my ideas into bureaucratic prose more elegantly and efficiently than me. Such a hypothetical individual might have some overlap with the "scientific popularizer" roles that we see cropping up with figures like Neil deGrasse Tyson or Bill Nye, in the sense that they could serve as a conduit between scientists and funding (whereas popularizers serve as links between scientists and the general public). The issue is that this kind of idealized division of labor can't exist in the current climate- there just isn't enough money to distribute a researcher's load across multiple specialists. This results in researchers being expected to juggle advising, grant writing, committee work, publishing, coding, analysis, and teaching, which creates more stress on researchers and poorer quality results. I do agree with you on your second point though- publish or perish is academia's equivalent to rank and yank.
- neuromantik8086 10y ago
  As an aside, I doubt that psychology would be experiencing nearly the number of issues that it is now with replicability if psychologists were able to delegate their analysis to an in-house statistician for verification. Or maybe it would (assuming that the poor application of statistics is more of a socio-political issue, which might very well be the case).
  - nonbel 10y ago
    >"the poor application of statistics" It is actually statisticians who are at the root of the problem. There is no proper way to apply the statistics most people are taught (NHST). However, everyone tries (and fails) to come up with some way to use the stats because they can't believe they were taught to perform such a waste of time ritual. See here for a decent overview: http://library.mpib-berlin.mpg.de/ft/gg/GG_Mindless_2004.pdf http://library.mpib-berlin.mpg.de/ft/gg/GG_Mindless_2004.pdf
hasbroslasher 10y ago
>The mistake, in my opinion, is making the number of papers published a metric on which you determine a scientist's worth. Just like most metric-ization of society today, it's a bad idea. Nodding intensifies It really is just a logical extension of capitalism - defining and measuring productive units so that you can optimize the means. What people fail to see time and agin is that incentivizing metrics gains is incentivizing metrics games (sorry that just came out gross and punny). It's no longer about doing good science/work/music/programming, it's about getting those numbers up so you can look good to the overseers

neuromantik8086 10y ago

I disagree on your first point- I think that there could be someone else who could write a grant proposal for me, provided that they a) expressed an interest in grant writing over benchwork / data analysis and b) were fluent enough in the field's literature to be able to translate my ideas into bureaucratic prose more elegantly and efficiently than me. Such a hypothetical individual might have some overlap with the "scientific popularizer" roles that we see cropping up with figures like Neil deGrasse Tyson or Bill Nye, in the sense that they could serve as a conduit between scientists and funding (whereas popularizers serve as links between scientists and the general public). The issue is that this kind of idealized division of labor can't exist in the current climate- there just isn't enough money to distribute a researcher's load across multiple specialists. This results in researchers being expected to juggle advising, grant writing, committee work, publishing, coding, analysis, and teaching, which creates more stress on researchers and poorer quality results. I do agree with you on your second point though- publish or perish is academia's equivalent to rank and yank.

neuromantik8086 10y ago
As an aside, I doubt that psychology would be experiencing nearly the number of issues that it is now with replicability if psychologists were able to delegate their analysis to an in-house statistician for verification. Or maybe it would (assuming that the poor application of statistics is more of a socio-political issue, which might very well be the case).
- nonbel 10y ago
  >"the poor application of statistics" It is actually statisticians who are at the root of the problem. There is no proper way to apply the statistics most people are taught (NHST). However, everyone tries (and fails) to come up with some way to use the stats because they can't believe they were taught to perform such a waste of time ritual. See here for a decent overview: http://library.mpib-berlin.mpg.de/ft/gg/GG_Mindless_2004.pdf http://library.mpib-berlin.mpg.de/ft/gg/GG_Mindless_2004.pdf

hasbroslasher 10y ago

>The mistake, in my opinion, is making the number of papers published a metric on which you determine a scientist's worth. Just like most metric-ization of society today, it's a bad idea. Nodding intensifies It really is just a logical extension of capitalism - defining and measuring productive units so that you can optimize the means. What people fail to see time and agin is that incentivizing metrics gains is incentivizing metrics games (sorry that just came out gross and punny). It's no longer about doing good science/work/music/programming, it's about getting those numbers up so you can look good to the overseers