4 ms·

In social sciences cause and effect often intermixed. It is pain in the ass for researchers. From the other hand, it means that it not so important to distingui

by ordu 6y ago

In social sciences cause and effect often intermixed. It is pain in the ass for researchers. From the other hand, it means that it not so important to distinguish cause and effect. If we had a correlation, we would have a predictor, so we could do something about it. Just studying agencies with military gear would change outcomes. Even media discussing this correlation could change things. The role of an observer even more noticeable than with quantum phenomena.

andreilys 6y ago
Cause and effect is important to distinguish. Spurious correlations exist everywhere (ice cream consumption is correlated with drowning, should we ban ice cream before swimming?) From my experience most people don’t understand this difference and assume any narrative around correlation equals causation.
- jmcqk6 6y ago
  >Cause and effect is important to distinguish. Alternatively we should probably abandon the notion that anything at the society level has a simple cause and a simple effect. When dealing with society, we're dealing with a chaotic system. I don't see much in society that can be clearly defined in terms of "if this then that." It seems like we would be much better off by being honest about the complexities of what we're dealing with.
  - andreilys 6y ago
    There is a lot of work being done in the casual space. Sure we might not be able to ever arrive at 100% causation but doesn’t mean we can’t improve our tooling and understanding. Policy decisions based on simple correlations can range from harmless to disastrous.
- ordu 6y ago
  > From my experience most people don’t understand this difference and assume any narrative around correlation equals causation. I know. Our thinking is built around cause and effect and when confronted by a correlational data it start to come to a wrong conclusions, because it have no idea what to do with correlations. It tends to assume that correlation is a causation, because such an assumption allow our thinking to do at least something with the data.

by ordu 6y ago

andreilys 6y ago

Cause and effect is important to distinguish. Spurious correlations exist everywhere (ice cream consumption is correlated with drowning, should we ban ice cream before swimming?) From my experience most people don’t understand this difference and assume any narrative around correlation equals causation.

jmcqk6 6y ago
>Cause and effect is important to distinguish. Alternatively we should probably abandon the notion that anything at the society level has a simple cause and a simple effect. When dealing with society, we're dealing with a chaotic system. I don't see much in society that can be clearly defined in terms of "if this then that." It seems like we would be much better off by being honest about the complexities of what we're dealing with.
- andreilys 6y ago
  There is a lot of work being done in the casual space. Sure we might not be able to ever arrive at 100% causation but doesn’t mean we can’t improve our tooling and understanding. Policy decisions based on simple correlations can range from harmless to disastrous.
ordu 6y ago
> From my experience most people don’t understand this difference and assume any narrative around correlation equals causation. I know. Our thinking is built around cause and effect and when confronted by a correlational data it start to come to a wrong conclusions, because it have no idea what to do with correlations. It tends to assume that correlation is a causation, because such an assumption allow our thinking to do at least something with the data.