Proving the effects of gravity by using the scientific method

[zanick]

Its been awhile since I even thought about conducting a scientific experiment, and I do remember that you need a dependent variable.. the observation or effect such as the apple fell from the tree, and then the independent variable, such as gravity. But in order to test the hypothesis, you need to alter the independent variable. (a cause) (and there can only be 2 independent variables). How could you effectively alter gravity? or would you alter its effects by testing its effects through a known resistance or medium and then make changes to that... or would you simulate its force to create similar effects and then change the force .

 

[m4r35n357]

Its been awhile since I even thought about conducting a scientific experiment, and I do remember that you need a dependent variable.. the observation or effect such as the apple fell from the tree, and then the independent variable, such as gravity. But in order to test the hypothesis, you need to alter the independent variable. (a cause) (and there can only be 2 independent variables). How could you effectively alter gravity? or would you alter its effects by testing its effects through a known resistance or medium and then make changes to that... or would you simulate its force to create similar effects and then change the force .

You could measure falling times for an object at various heights, and you could compare the falling times for two different objects from the same height.

 

[jbriggs444]

Its been awhile since I even thought about conducting a scientific experiment, and I do remember that you need a dependent variable.. the observation or effect such as the apple fell from the tree, and then the independent variable, such as gravity. But in order to test the hypothesis, you need to alter the independent variable. (a cause) (and there can only be 2 independent variables). How could you effectively alter gravity? or would you alter its effects by testing its effects through a known resistance or medium and then make changes to that... or would you simulate its force to create similar effects and then change the force .

If you are trying to test the truth of F=mg or working to determine g then one goal is to demonstrate that g is constant. Or, put another way, that no other variable like x, y or z enters in. Another goal is to demonstrate that the dependence on mass is linear.

So you could test gravity at various places, for objects of various masses, for objects at different velocities. For hot objects and cold, for spinning objects, white objects, black objects and red objects. Hollow objects, pyramids, inverted pyramids, etc, etc, etc. Every one of those are independent variables. It is just that in each case, the predicted effect is zero. The experimentally determined effect might be different from zero.

If you are trying to deduce the truth of $F=G\frac{m_1 m_2}{r^2}$ from terrestrial observations then you do indeed have a problem because you cannot easily manipulate $m_2$ or make large modifications to $r$. You might work on a Cavendish experiment, Look at astronomical observations. Or notice the deviations of g from constant in the third digit on the surface of the Earth and try to find a pattern there.

 

[Dale]

But in order to test the hypothesis, you need to alter the independent variable. (a cause) (and there can only be 2 independent variables)

Other people have addressed the actual question re: experimental tests of gravity, however, I wanted to clear up a couple of misconceptions expressed here. Both of the comments in parentheses are incorrect.

Statistics establishes correlation, and correlation is not causation. Because correlation is not causation it is not necessary that the independent variable be the cause. The important thing is that the independent variable be known with high accuracy so that all of the uncertainty can be attributed to uncertainty in the dependent variable.

Also, there can be as many independent variables as you like. Two is a good number, but not a requirement.

 

[sophiecentaur]

How could you effectively alter gravity?

Early work on Gravity involved the study of Pendulums. The timescales are so much longer than for free falling bodies and the same sorts of concussions can be arrived at - e.g. independence of time taken and mass.

PS Altering Gravity is beyond the scope of kitchen table experiments; you would need to go to the Moon or Mars. A less drastic approach could be to go to high latitudes (ideally the N Pole) and compare the pendulum period with what it does at the Equator. This would require some more accurate measurements than doing it on the Moon.

 

[ZapperZ]

Its been awhile since I even thought about conducting a scientific experiment, and I do remember that you need a dependent variable.. the observation or effect such as the apple fell from the tree, and then the independent variable, such as gravity. But in order to test the hypothesis, you need to alter the independent variable. (a cause) (and there can only be 2 independent variables). How could you effectively alter gravity? or would you alter its effects by testing its effects through a known resistance or medium and then make changes to that... or would you simulate its force to create similar effects and then change the force .

I don't quite understand what "hypothesis" that you are trying to test here.

You can't just say that you want to test the "effect of gravity". This is vague! The sign of a good experiment is that it knows what quantities should be measured, and what theoretical description is being tested by measuring those quantities.

In many General Physics labs, students often make a measurement of the Earth's "g" value using a simple pendulum, or even via dropping a mass at various heights. Aren't these experiments to test the "effect of gravity"? Have you never done such experiments before?

If not, then you need to be clearer on what exactly you are trying to test.

Zz.

 

[zanick]

so, this would prove a constant acceleration, or that gravity is the cause (the independent variable)because the times for different objects will be "near" the same... however, because they are not exactly the same, is that a problem... (unless we put them in a vacuum). a feather and a bowling ball, wouldn't really provide any useful data unless we got into measuring drag values for each . however, a bowling ball and a small rock might.

You could measure falling times for an object at various heights, and you could compare the falling times for two different objects from the same height.

 

[zanick]

If we are trying to show that the independent variable, (the cause) is gravity, we have to be able to vary it in a set of experiments. the dependent variable, i that the apple falls from the tree (effect)... if we change the apple to a rock change or the height from which it falls, are we not changing the "dependent" variable, and not the independent variable? this is why i suggested ways to measure drag and change that factor, and then deduce that gravity is the force causing the apple to fall.

I think your last paragraph is the the real issue. If we are deducing the real cause being gravity, it is hard to change gravity as an independent variable. (as you say, such as changing "r" or "m") that's why i thought if we can vary the medium and the drag forces , all apart of why an object falls at the rate it does, that might be a possibility.

thanks for the comment

If you are trying to test the truth of F=mg or working to determine g then one goal is to demonstrate that g is constant. Or, put another way, that no other variable like x, y or z enters in. Another goal is to demonstrate that the dependence on mass is linear.

So you could test gravity at various places, for objects of various masses, for objects at different velocities. For hot objects and cold, for spinning objects, white objects, black objects and red objects. Hollow objects, pyramids, inverted pyramids, etc, etc, etc. Every one of those are independent variables. It is just that in each case, the predicted effect is zero. The experimentally determined effect might be different from zero.

If you are trying to deduce the truth of $F=G\frac{m_1 m_2}{r^2}$ from terrestrial observations then you do indeed have a problem because you cannot easily manipulate $m_2$ or make large modifications to $r$. You might work on a Cavendish experiment, Look at astronomical observations. Or notice the deviations of g from constant in the third digit on the surface of the Earth and try to find a pattern there.

 

[zanick]

you said both comments in parenthesis, is incorrect, but as i remember , the independent variable is the cause, the observation is the effect... is this not true... I've seen , heard and remember the following:

"Statements about association are usually stated in terms of a relationship between an independent and a dependent variable. The idea is that one variable is the effect of another variable or, to say it another way, that one variable precedes and/or causes another.

The dependent variable is the variable to be explained (the ‘effect”).

The independent variable is the variable expected to account for (the “cause” of)
>>>>>>>
i agree and subscribe to "correlation doesn't substantiate causation"

so, if we are seeing an observation of that an apple falls.. we then point to gravity as the cause (the independent variable).. is this not correct? but if it is, how do we vary it to prove that it is? or is there another way as i suggested?

Other people have addressed the actual question re: experimental tests of gravity, however, I wanted to clear up a couple of misconceptions expressed here. Both of the comments in parentheses are incorrect.

Statistics establishes correlation, and correlation is not causation. Because correlation is not causation it is not necessary that the independent variable be the cause. The important thing is that the independent variable be known with high accuracy so that all of the uncertainty can be attributed to uncertainty in the dependent variable.

Also, there can be as many independent variables as you like. Two is a good number, but not a requirement.

 

[zanick]

this is what i was thinking as well. would you need to go to the equator and prove things weigh (force due to gravity ) less there due to centripetal force of the Earth's spin on the mass, OR, go to a high altitude and see the predictive force to be less? But , my thought of changing the mass or the time of the falls is altering the dependent variable , not the independent variable. (the cause being gravity)

Early work on Gravity involved the study of Pendulums. The timescales are so much longer than for free falling bodies and the same sorts of concussions can be arrived at - e.g. independence of time taken and mass.

PS Altering Gravity is beyond the scope of kitchen table experiments; you would need to go to the Moon or Mars. A less drastic approach could be to go to high latitudes (ideally the N Pole) and compare the pendulum period with what it does at the Equator. This would require some more accurate measurements than doing it on the Moon.

 

[zanick]

the hypothesis is that gravity cause objects to fall toward the surface of the earth.

If someone asks , how do you know". then it is good to use the scientific method to explain/test it.
If something is falling, it could have a force from somewhere else acting on it, it could be self propelled, it could magnetized, etc... If we are trying to prove that gravity is the force that makes something fall , what is the best way to use the scientific method to do this?

Yes, i have done the experiments many times for gravity. we get acceleration rates and then deduce a force required to follow A=f/m.

So, if we are testing the "effects" of gravity, that seems to be easier. Gravity creates a force that causes acceleration of objects to fall at 9.8 m per second^2. this is where i get a little confused. If we start changing the mass of the objects and do our observations, are we not changing the dependent variable? ( we see the object fall( dependat variable)... why does it fall... if it is gravity (cause) , isn't that the independent variable? if so, how to we change /alter it to prove that it is? )

.

I don't quite understand what "hypothesis" that you are trying to test here.

You can't just say that you want to test the "effect of gravity". This is vague! The sign of a good experiment is that it knows what quantities should be measured, and what theoretical description is being tested by measuring those quantities.

In many General Physics labs, students often make a measurement of the Earth's "g" value using a simple pendulum, or even via dropping a mass at various heights. Aren't these experiments to test the "effect of gravity"? Have you never done such experiments before?

If not, then you need to be clearer on what exactly you are trying to test.

Zz.

 

[jbriggs444]

the hypothesis is that gravity cause objects to fall toward the surface of the earth

You need a better hypothesis. Preferably something quantitative. As it stands, hot air balloons defy the hypothesis.

 

[Dale]

the independent variable is the cause, the observation is the effect... is this not true

It is not true. There is no causal relationship between dependent and independent variables. In fact, when you take a statistics course they usually spend a lot of effort to make that clear, usually by showing relationships with a high degree of correlation but no causation. My favorite one is ice-cream sales and shark attacks. They are highly correlated, but increased ice-cream sales do not cause increased shark attacks, instead they are both a result of increased temperatures.

 

[zanick]

You need a better hypothesis. Preferably something quantitative. As it stands, hot air balloons defy the hypothesis.

good point... is refining it better, or something completely different? objects fall due to gravity, if their density is greater than density of the medium they are falling through?

 

[Nugatory]

good point... is refining it better, or something completely different? objects fall due to gravity, if their density is greater than density of the medium they are falling through?

Dunno... it's your thread and your experiment. What are you trying to accomplish here?

 

[zanick]

It is not true. There is no causal relationship between dependent and independent variables. In fact, when you take a statistics course they usually spend a lot of effort to make that clear, usually by showing relationships with a high degree of correlation but no causation. My favorite one is ice-cream sales and shark attacks. They are highly correlated, but increased ice-cream sales do not cause increased shark attacks, instead they are both a result of increased temperatures.

Dale, i thought that was the entire purpose of the scientific methodology. to find cause and effect through the dependent and independent variable relationship. your example: shark attacks (dependent variable) ice cream sales (independent variable) . strong correlation, but not the cause... opening your car door might cause a moth to react , but its the light coming on, not the door opening.. how do you change the independent variable to show causation ? is the answer in the "control variables"?

 

[zanick]

Dunno... it's your thread and your experiment. What are you trying to accomplish here?

trying to use the scientific method to prove that gravity is the reason the apple falls from the tree. any ideas? this was born from my son and friend's question regarding gravity. the friend didn't believe in gravity and was hold me to prove it using the scientific method. They had been studying the scientific method in school, and i was going off memory!

 

[jbriggs444]

prove that gravity is the reason

Science is not about proving that hypotheses are true. It is about trying and failing to prove them false.

If your son and friend doubt that gravity is responsible, ask them what is responsible instead. Then with the two competing hypotheses in hand see if either of them makes a testable prediction that differs from the other.

If they do make a testable prediction that differs then test it!

If they do not make a testable prediction that differs then science is mute about which hypothesis to accept. [However, Occam's razor may provide a preference]

Edit: If the hypothesis from son and friend fails to make any testable predictions at all then one can reject it out of hand as being both useless and unscientific. For instance, "because magic fairy dust" does not predict anything unless it comes with a description of the effects of magic fairy dust.

 

[russ_watters]

trying to use the scientific method to prove that gravity is the reason the apple falls from the tree. any ideas? this was born from my son and friend's question regarding gravity. the friend didn't believe in gravity and was hold me to prove it using the scientific method. They had been studying the scientific method in school, and i was going off memory!

Is this for real? I don't see what the problem is, unless the guy is a crackpot who thinks science has it wrong on how gravity works. "Gravity" is just a name for an observed phenomena. There are several theories on how it works, which take various observations and model behaviors. If the models work, the theory is considered valid.

But in the apocryphal story, Newton saw an apple fall from a tree and he called the effect that caused it "gravity". It is what is is - there's nothing to prove.

 

[Dale]

how do you change the independent variable to show causation

You cannot. Causation comes through the theory. If your theory says A causes B and you have lots of good evidence to support the theory then you assert the causation. If you merely have a single experiment or a data set then you do not assert causation, merely correlation.

trying to use the scientific method to prove that gravity is the reason the apple falls from the tree

As @russ_watters wisely pointed out, it is just a definition. Whatever caused the apple to fall from the tree is called gravity. That is just a matter of definition and is completely uninteresting. What we want to do with the scientific method is make a theory about how gravity works and then test that. For example, we propose that it is a force that follows the equation $F=G m_1 m_2/r^2$. That is no longer just a definition and is something interesting. Then we can apply that to the apple and determine how long it takes to fall and so forth.

the friend didn't believe in gravity

Just tell the kid he is being silly and will fail the test. No need to take a silly statement like that seriously.

 

[zanick]

Science is not about proving that hypotheses are true. It is about trying and failing to prove them false.

If your son and friend doubt that gravity is responsible, ask them what is responsible instead. Then with the two competing hypotheses in hand see if either of them makes a testable prediction that differs from the other.

If they do make a testable prediction that differs then test it!

If they do not make a testable prediction that differs then science is mute about which hypothesis to accept. [However, Occam's razor may provide a preference]

thats the answer i was looking for and a great couple of points. " see if one offers a testable prediction that differs..."

 

[zanick]

You cannot. Causation comes through the theory. If your theory says A causes B and you have lots of good evidence to support the theory then you assert the causation. If you merely have a single experiment or a data set then you do not assert causation, merely correlation.

As @russ_watters wisely pointed out, it is just a definition. Whatever caused the apple to fall from the tree is called gravity. That is just a matter of definition and is completely uninteresting. What we want to do with the scientific method is make a theory about how gravity works and then test that. For example, we propose that it is a force that follows the equation $F=G m_1 m_2/r^2$. That is no longer just a definition and is something interesting. Then we can apply that to the apple and determine how long it takes to fall and so forth.

Just tell the kid he is being silly and will fail the test. No need to take a silly statement like that seriously.

makes sense. ... yes, you want to just respond to the kids in that way, but because they are asking honestly and thinking about the scientific methodology, i wanted them to figure this out for themselves with what they learned so far.

 

[zanick]

If you are trying to test the truth of F=mg or working to determine g then one goal is to demonstrate that g is constant. Or, put another way, that no other variable like x, y or z enters in. Another goal is to demonstrate that the dependence on mass is linear.

So you could test gravity at various places, for objects of various masses, for objects at different velocities. For hot objects and cold, for spinning objects, white objects, black objects and red objects. Hollow objects, pyramids, inverted pyramids, etc, etc, etc. Every one of those are independent variables. It is just that in each case, the predicted effect is zero. The experimentally determined effect might be different from zero.

If you are trying to deduce the truth of $F=G\frac{m_1 m_2}{r^2}$ from terrestrial observations then you do indeed have a problem because you cannot easily manipulate $m_2$ or make large modifications to $r$. You might work on a Cavendish experiment, Look at astronomical observations. Or notice the deviations of g from constant in the third digit on the surface of the Earth and try to find a pattern there.

excellent... thanks!