Week #3 Commentary
a.
* It is interesting that it is still hard to tell who is right even after the fact.
* It is also interesting that we have no reliable way to measure counterfactuals.
* It is interesting that the author states that Aristotle's theory is false and Galileo's theory is true.
b.
* Why do we have to test theories ideally through controlled experiments before they are accepted as reliable?
* Are theories reliable if they are intuitive?
* How do we generalize the results from experiments to a larger scale? It is true as the physicist says?
c.
No one can predict thing 100% correctly, because all the things or events are too complicated, there are so many elements involved in predicting the results. Thus, we start to use controlled experiments to get closer to the right track. But things are way more complex than we think, even if you do numerous experiments. That only can apply to some situation, then the question comes up which is how do we generalize those experimental results to a larger scale or how do we randomize these consequences ? Still, we cannot get the perfect answer. For example, "multisite RFTs" experiments worked pretty well before, but after it was tested in 12 of the programs, it pretty much failed. We also fail to use the experiments to interpret the reason of crime. Later on we use internet to do tons of experiments to get better results. The businesses find out that they can only deal with problems of casual density by scaling up the testing process. All the experiments above are social-science experiments, we know three things from them. First of all, few programs can be shown to work in properly randomized and replicated trials.Secondly, within this universe of programs that are far more likely to fail than succeed, programs that try to change people are even more likely to fail than those that try to change incentives. Finally, there is no magic, it is true that we can get to the certain level of the result, but we do not get the perfect truth. Experiments are absolutely changing our way of making better decision. And we are at the beginning of a social-science experimental revolution, as the time goes on, we will eventually get the unimaginable discoveries.
* It is interesting that it is still hard to tell who is right even after the fact.
* It is also interesting that we have no reliable way to measure counterfactuals.
* It is interesting that the author states that Aristotle's theory is false and Galileo's theory is true.
b.
* Why do we have to test theories ideally through controlled experiments before they are accepted as reliable?
* Are theories reliable if they are intuitive?
* How do we generalize the results from experiments to a larger scale? It is true as the physicist says?
c.
No one can predict thing 100% correctly, because all the things or events are too complicated, there are so many elements involved in predicting the results. Thus, we start to use controlled experiments to get closer to the right track. But things are way more complex than we think, even if you do numerous experiments. That only can apply to some situation, then the question comes up which is how do we generalize those experimental results to a larger scale or how do we randomize these consequences ? Still, we cannot get the perfect answer. For example, "multisite RFTs" experiments worked pretty well before, but after it was tested in 12 of the programs, it pretty much failed. We also fail to use the experiments to interpret the reason of crime. Later on we use internet to do tons of experiments to get better results. The businesses find out that they can only deal with problems of casual density by scaling up the testing process. All the experiments above are social-science experiments, we know three things from them. First of all, few programs can be shown to work in properly randomized and replicated trials.Secondly, within this universe of programs that are far more likely to fail than succeed, programs that try to change people are even more likely to fail than those that try to change incentives. Finally, there is no magic, it is true that we can get to the certain level of the result, but we do not get the perfect truth. Experiments are absolutely changing our way of making better decision. And we are at the beginning of a social-science experimental revolution, as the time goes on, we will eventually get the unimaginable discoveries.
Comments
Post a Comment