38 ms·

agree, the sausage-making in science is ugly, but my point is if technology is more accessible, there will be some reliable data points to go by rather than klu

agree, the sausage-making in science is ugly, but my point is if technology is more accessible, there will be some reliable data points to go by rather than kludging together some experimental premise and then making profound pronouncements. Then those findings are then parroted by everyone. Take for example a study done in about 2005 that said "willpower is an exhaustible resource" - They gave some participants some problems to solve - they were cooking cookies at that back so the participants could smell them. Later, half of the participants were given cucumbers and the other half cookies and then observed that people given cookies persisted for longer in solving the problems. This proved "willpower is an exhaustible resources". Is willpower exhaustible? what if the participants while they were giving up early where given an incentive of 500 USD to continue working on the same problem, would they work for much longer than the other group? why? willpower is an exhaustible resource, how can you draw from something that now no longer exists? would you then conclude that willpower can be depleted in humans, but humans have the ability to synthesize it from currency. Ofcorse thats ridiculous. Willpower and expended effort are context based and each person can set their own threshold of when its appropriate to give up - its partly a conscious decision. A few years later the authors of this same study walked back on their claims that "willpower is an exhaustible resource" ( it does not stop people from touting this around as reasoning for whatever point they are trying to make ). My point is bio-monitoring technology with probably VR will atleast give us some verifiable data-points..

tgv 5y ago
Psychology has been using "bio-monitoring" since Helmholtz, although they think of it as involuntary or unconscious (modification of) response, e.g. reaction time, pupil dilation, heart rate. It all has the same problem: the distance between the hypothesis, the experimental manipulation, and the measurement is large and badly understood. Differences between conditions are then ascribed to noise, which enables "creative" interpretations of the results. As an example: in the field in which I worked, it was commonly assumed that the processing mechanism had to interpret a part of the stimulus as belonging to one thing or the other. Then they would manipulate the context to see how they could influence that. There were many results and experimental refinements, leading to an endless stream of contradictory articles, all claiming victory for their theory. But there are models in which this assumption doesn't make sense. Instead, it could belong to both, and the ambiguity would be resolved later. That makes the whole research line I described nonsense. The data is valid (if well documented), but the analyses and conclusions are not. A genetics statistician explained to me that all early genetic research (and that means up until 2000 at least) is highly unreliable, because they used Neyman-Fisher significance testing with high p values, like 0.01. The lack of understanding was (and probably still is) so great, that 0.01 was simply not enough to exclude random results. Nearly all those papers are irreproducible. Nowadays, 5 sigma, which is less than 0.000001, is considered safe, although he himself was convinced that Bayesian is the only way to go. The understanding of the subject in psychology is considerably worse than in biology. No amount of bio-monitoring is going to overcome that.
- dmingod666 5y ago
  I know, it would have been fine if it was just another discipline of science or research that was hard to study.. The amount of interest from regular people and in the subject and the demand for research is so much that people want to hang on to twigs.. most people want to improve themselves and be able to think and do better and take atleast a fleeting interest in psychology... Evidenced by the rise of Jordon Peterson - I recently found Dr.Andrew Huberman to be a voice of reason (check out his Youtube podcast, its very information-dense and pretty good) - He approaches things from a neurobiology perspective. To navigate around getting a deeper understanding yourself, I thing a chemical / hormone / neurotransmitter based discussion at least grounds your feet in real things, from which you can probably explore things for yourself.. It also eliminates a lot of the cruft from the discussion..

tgv 5y ago

Psychology has been using "bio-monitoring" since Helmholtz, although they think of it as involuntary or unconscious (modification of) response, e.g. reaction time, pupil dilation, heart rate. It all has the same problem: the distance between the hypothesis, the experimental manipulation, and the measurement is large and badly understood. Differences between conditions are then ascribed to noise, which enables "creative" interpretations of the results. As an example: in the field in which I worked, it was commonly assumed that the processing mechanism had to interpret a part of the stimulus as belonging to one thing or the other. Then they would manipulate the context to see how they could influence that. There were many results and experimental refinements, leading to an endless stream of contradictory articles, all claiming victory for their theory. But there are models in which this assumption doesn't make sense. Instead, it could belong to both, and the ambiguity would be resolved later. That makes the whole research line I described nonsense. The data is valid (if well documented), but the analyses and conclusions are not. A genetics statistician explained to me that all early genetic research (and that means up until 2000 at least) is highly unreliable, because they used Neyman-Fisher significance testing with high p values, like 0.01. The lack of understanding was (and probably still is) so great, that 0.01 was simply not enough to exclude random results. Nearly all those papers are irreproducible. Nowadays, 5 sigma, which is less than 0.000001, is considered safe, although he himself was convinced that Bayesian is the only way to go. The understanding of the subject in psychology is considerably worse than in biology. No amount of bio-monitoring is going to overcome that.

dmingod666 5y ago
I know, it would have been fine if it was just another discipline of science or research that was hard to study.. The amount of interest from regular people and in the subject and the demand for research is so much that people want to hang on to twigs.. most people want to improve themselves and be able to think and do better and take atleast a fleeting interest in psychology... Evidenced by the rise of Jordon Peterson - I recently found Dr.Andrew Huberman to be a voice of reason (check out his Youtube podcast, its very information-dense and pretty good) - He approaches things from a neurobiology perspective. To navigate around getting a deeper understanding yourself, I thing a chemical / hormone / neurotransmitter based discussion at least grounds your feet in real things, from which you can probably explore things for yourself.. It also eliminates a lot of the cruft from the discussion..