4 ms·

When taking science classes in school, I never understood why a hypothesis had to take a side. It could not be a question, like "hypothesis: does hot water free

by zck 8y ago

When taking science classes in school, I never understood why a hypothesis had to take a side. It could not be a question, like "hypothesis: does hot water freeze faster than cold water?", but had to take a position on the issue: "hypothesis: hot water freezes faster than cold water". I never got why this was the case; it sets the experiment up for failure: "oh no, I proved my hypothesis wrong". Who cares? As long as you show something (even not reflecting the null hypothesis is _something_) with a well-designed experiment, why does it matter if your guess is right or not? I'd much rather prove a hypothesis like "a chalk circle drawn around sugar does nothing to inhibit ants" wrong, than to prove it right! That would be much more interesting!

dragonwriter 8y ago
> When taking science classes in school, I never understood why a hypothesis had to take a side. It could not be a question, like "hypothesis: does hot water freeze faster than cold water?", Because a hypothesis is a falsifiable answer, not a question. > I never got why this was the case; it sets the experiment up for failure: Setting up an explanation for falsification is exactly why. But falsification isn't failure of the experiment, only the hypothesis.
automatoney 8y ago
I'm not sure what level of schooling you're talking about, but I think there's two reasons to take a side. First, I think in early school it prepares you to think in a way that makes sense with null/alternative hypotheses and all of those z and t stats hypothesis tests. The second reason is to have you think about what you expect to happen. If you're doing an experiment, you likely have some information to come up with what you think will happen, and what the explanation for that is. If you get a result that goes against what your reasoning was beforehand, then you have a misconception and have to adjust what you think you know. If you don't take a side, then you could potentially say that you believed that would happen all along, even if it doesn't make sense with how you understand things. Taking a side makes you confront your confusion when it doesn't work out as you thought.
- analog31 8y ago
  >>> If you don't take a side, then you could potentially say that you believed that would happen all along How does that harm anything? The result is the result. Pretending to have a belief so it can potentially be disproven seems contrived.
  - automatoney 8y ago
    That part was probably more about using the scientific method in personal life, I got a bit off track. But to answer you, the idea is that you should have a belief beforehand, and the experiment is how you test beliefs. The experiment should be designed to falsify a specific idea. Like say for example your code doesn't work. You have to have an idea of where it could be broken in order to test that belief. So you know there's a method with complex decisions, and it should call foo, but you believe that it isn't. To test this you throw a print at the top of foo, and see what happens. The belief is necessary to create the experiment - you won't make efficient progress if you put print statements after every line. The experiment is designed to differentiate between "this idea wrong" and "this idea is probably right".
    - analog31 8y ago
      I suppose in the case of programming, there should be a belief in what a function does, because somebody wrote the function in the first place. I'm thinking about my own work, in areas such as physics and electronics. On one fine day, I measured the characteristics of about 100 Zener diodes, to inform a design decision with about 3 or 4 possible outcomes. I don't see where pinning myself down to a particular opinion of the expected outcome before making the measurement would have changed the outcome.
hydrox24 8y ago
You need a statement because a question is not falsifiable. Asking students to come up with a statement rather than a question is one step closer to coming up with a falsifiable proposition. Your high school self stumbled onto a serious debate in the philosophy of science of course. What I have said is fairly accepted, but if you're interested in the philosophy & arguments behind it you should read some Karl Popper[0]. [0]: https://plato.stanford.edu/entries/popper/ https://plato.stanford.edu/entries/popper/
- graeme 8y ago
  Isn’t answering a question functionally the same as falsifying one of the possible hypotheses held in he question? Yes falsifies one half, no the other.
maxlybbert 8y ago
I’m no scientist, so this is really conjecture. I remember being confused as well; and in elementary and even middle school, many students were convinced the hypothesis had to be the right answer, so they wouldn’t fill it out until after the experiment. But, at least at my schools, we never did any real experiment. We did demonstrations and tried to shoehorn them into the scientific method. We never started with an observation (say, mice live longer when given reduced calorie diets), worked out a hypothetical explanation for that observation, and then designed an experiment to prove or disprove our hypothesis. I’m sure that’s done at some level, but not in any science class I took.
dnautics 8y ago
This is just a matter of perspective. Usually the hypothesis you set up is indeed the null hypothesis, and an observation to disprove the null is what you "seek".
citrablue 8y ago
Scientific theory is based around the idea that you can never, 100% prove anything absolutely true.* The best we can do is to be "less wrong" by learning how we're wrong. It's a perspective on acquiring knowledge that prioritizes false negatives over false positives. You don't prove yourself right, you just fail to prove yourself wrong. Trying to prove yourself wrong is a much more reliable way for a human, with all our biases, to end up closer to the truth. * mathematicians disagree

by zck 8y ago

dragonwriter 8y ago

> When taking science classes in school, I never understood why a hypothesis had to take a side. It could not be a question, like "hypothesis: does hot water freeze faster than cold water?", Because a hypothesis is a falsifiable answer, not a question. > I never got why this was the case; it sets the experiment up for failure: Setting up an explanation for falsification is exactly why. But falsification isn't failure of the experiment, only the hypothesis.

automatoney 8y ago

I'm not sure what level of schooling you're talking about, but I think there's two reasons to take a side. First, I think in early school it prepares you to think in a way that makes sense with null/alternative hypotheses and all of those z and t stats hypothesis tests. The second reason is to have you think about what you expect to happen. If you're doing an experiment, you likely have some information to come up with what you think will happen, and what the explanation for that is. If you get a result that goes against what your reasoning was beforehand, then you have a misconception and have to adjust what you think you know. If you don't take a side, then you could potentially say that you believed that would happen all along, even if it doesn't make sense with how you understand things. Taking a side makes you confront your confusion when it doesn't work out as you thought.

analog31 8y ago
>>> If you don't take a side, then you could potentially say that you believed that would happen all along How does that harm anything? The result is the result. Pretending to have a belief so it can potentially be disproven seems contrived.
- automatoney 8y ago
  That part was probably more about using the scientific method in personal life, I got a bit off track. But to answer you, the idea is that you should have a belief beforehand, and the experiment is how you test beliefs. The experiment should be designed to falsify a specific idea. Like say for example your code doesn't work. You have to have an idea of where it could be broken in order to test that belief. So you know there's a method with complex decisions, and it should call foo, but you believe that it isn't. To test this you throw a print at the top of foo, and see what happens. The belief is necessary to create the experiment - you won't make efficient progress if you put print statements after every line. The experiment is designed to differentiate between "this idea wrong" and "this idea is probably right".
  - analog31 8y ago
    I suppose in the case of programming, there should be a belief in what a function does, because somebody wrote the function in the first place. I'm thinking about my own work, in areas such as physics and electronics. On one fine day, I measured the characteristics of about 100 Zener diodes, to inform a design decision with about 3 or 4 possible outcomes. I don't see where pinning myself down to a particular opinion of the expected outcome before making the measurement would have changed the outcome.

hydrox24 8y ago

You need a statement because a question is not falsifiable. Asking students to come up with a statement rather than a question is one step closer to coming up with a falsifiable proposition. Your high school self stumbled onto a serious debate in the philosophy of science of course. What I have said is fairly accepted, but if you're interested in the philosophy & arguments behind it you should read some Karl Popper[0]. [0]: https://plato.stanford.edu/entries/popper/ https://plato.stanford.edu/entries/popper/

graeme 8y ago
Isn’t answering a question functionally the same as falsifying one of the possible hypotheses held in he question? Yes falsifies one half, no the other.

maxlybbert 8y ago

I’m no scientist, so this is really conjecture. I remember being confused as well; and in elementary and even middle school, many students were convinced the hypothesis had to be the right answer, so they wouldn’t fill it out until after the experiment. But, at least at my schools, we never did any real experiment. We did demonstrations and tried to shoehorn them into the scientific method. We never started with an observation (say, mice live longer when given reduced calorie diets), worked out a hypothetical explanation for that observation, and then designed an experiment to prove or disprove our hypothesis. I’m sure that’s done at some level, but not in any science class I took.

dnautics 8y ago

This is just a matter of perspective. Usually the hypothesis you set up is indeed the null hypothesis, and an observation to disprove the null is what you "seek".

citrablue 8y ago

Scientific theory is based around the idea that you can never, 100% prove anything absolutely true.* The best we can do is to be "less wrong" by learning how we're wrong. It's a perspective on acquiring knowledge that prioritizes false negatives over false positives. You don't prove yourself right, you just fail to prove yourself wrong. Trying to prove yourself wrong is a much more reliable way for a human, with all our biases, to end up closer to the truth. * mathematicians disagree