Experimental Design B/C

BoldlyGoingNowhere · Post by **BoldlyGoingNowhere** » Sat Feb 11, 2012 8:42 am

How do you give example calculations for the quantitative data?

Post by **foreverphysics** » Sat Feb 11, 2012 10:15 am

Just, y'know, put down all the math you did to solve.

EpicFailure · Post by **EpicFailure** » Sun Feb 12, 2012 9:03 am

foreverphysics wrote:Just, y'know, put down all the math you did to solve.

You could also write the formulas and put down the math for one set of data. For example, an example calculation for average would look something like:

Average: mean of all numbers = (Trial 1 + Trial 2 + Trial 3)/3 = (21.1+20.5+22.5)/3 = 21.4

Post by **foreverphysics** » Mon Feb 13, 2012 6:13 am

How long would you suggest practicing with your partner(s)? Seeing as it's 12 days before Regionals and I haven't yet practiced with either of them, this is a pretty big problem.

piisamazing · Post by **piisamazing** » Mon Feb 13, 2012 12:12 pm

foreverphysics wrote:How long would you suggest practicing with your partner(s)? Seeing as it's 12 days before Regionals and I haven't yet practiced with either of them, this is a pretty big problem.

not really. i didnt practice with my partners, and we made 1st place at reigionals. All you need to really do is make sure you can write a hypothesis correctly. the only practice ever do with my partners is to have sample bags and just come up with ideas of possible experiments

Flavorflav · Post by **Flavorflav** » Thu Feb 16, 2012 3:01 am

piisamazing wrote:
kayken13 wrote:For standard deviation, are we supposed to use find the population standard deviation or sample standard deviation?

Also, are we supposed to find standard deviation for each trial set up, for the final numbers, or both?
As far as I know you find standard deviation for the final numbers so you can avoid having to deal with outliers and mistrials. Also, you find population standard deviation.

Why do you say that? I would think sample s.d. would be more appropriate.

piisamazing · Post by **piisamazing** » Thu Feb 16, 2012 1:05 pm

Flavorflav wrote:
piisamazing wrote:
kayken13 wrote:For standard deviation, are we supposed to use find the population standard deviation or sample standard deviation?

Also, are we supposed to find standard deviation for each trial set up, for the final numbers, or both?
As far as I know you find standard deviation for the final numbers so you can avoid having to deal with outliers and mistrials. Also, you find population standard deviation.
Why do you say that? I would think sample s.d. would be more appropriate.

I say population s.d. would be apropriate ONLY if you are finding your s.d using the final numbers. it also depends on what you are testing. The population s.d. would be used in a trial something like this: you are testing how far a ball will launch from the slingshot and the variables are the rubber band lenghth and the angle of launch. you do each trial 5 times each. when you find s.d. for a certain angle, you find the mean of each of the 5 trials for the rubber band length. this fits the purpose of population standard deviation because you have more data, but you use only the means of data groups.

Clarent_sword of ice · Thu Feb 16, 2012 2:01 pm

ok am i reading this right????
jello???? seriously?

deezee · Post by **deezee** » Thu Feb 16, 2012 2:22 pm

piisamazing wrote:
foreverphysics wrote:How long would you suggest practicing with your partner(s)? Seeing as it's 12 days before Regionals and I haven't yet practiced with either of them, this is a pretty big problem.
not really. i didnt practice with my partners, and we made 1st place at reigionals. All you need to really do is make sure you can write a hypothesis correctly. the only practice ever do with my partners is to have sample bags and just come up with ideas of possible experiments

For us, we really didn't practice, but we sat down with the rubric and split up roles (i.e, I do Data and Graphs, Bob does problem, hypothesis, conclusion etc, and Larry does exeriment with me and procedure, materials, stuff like that)

Flavorflav · Post by **Flavorflav** » Fri Feb 17, 2012 2:11 am

piisamazing wrote: I say population s.d. would be apropriate ONLY if you are finding your s.d using the final numbers. it also depends on what you are testing. The population s.d. would be used in a trial something like this: you are testing how far a ball will launch from the slingshot and the variables are the rubber band lenghth and the angle of launch. you do each trial 5 times each. when you find s.d. for a certain angle, you find the mean of each of the 5 trials for the rubber band length. this fits the purpose of population standard deviation because you have more data, but you use only the means of data groups.

This makes no sense to me. Are you suggesting you find the standard deviation of the means of different launch angles? That would be a really, really bad idea.

One uses population s.d. when one has data on the whole population. In your example, that would mean that you have shot the ball all of the times that it is possible for it to be shot. Since that is obviously never going to be the case, you are always going to be dealing with a subset of the possible data and you should use sample s.d. Note that as the number of data points rises the difference between pop s.d. and sample s.d. disappears, which is really the point - if you have only done it five times, you really don't know what the true variance is and you allow a little more room.

Scioly.org

Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Re: Experimental Design B/C

Connect

Learn

Get Involved

About

Disclaimer