Page 7 of 25
Re: Experimental Design B/C
Posted: February 4th, 2014, 6:03 pm
by Phys1cs
jwalkotten wrote:A couple questions if anyone could help
1) it says to give a measure of central tendency (for division C). What does it want here? Because mean, median, and mode, are all measures of central tendency, but we have already given these...
2) what does "other appropriate statistic used" mean? some people have said to get a regression analysis, but how do you get that for all three levels of IV? because your regression analysis is going to be for the line of best fit for your data, but your statistics are supposed to be comparing each of you trials in each level of IV (at least that's what I thought...)
Thanks if anyone could help!
For the other measure of central tendency, Standard Deviation and the Variance (because you need the variance for standard deviation) are common. Those are what I use, because Standard deviation is used to find the general range the data should be falling in.
Regression Analyses are more interpreting your regression, or line of best fit.
Re: Experimental Design B/C
Posted: February 9th, 2014, 6:48 pm
by tsinzitari
So Me and 2 other people just got thrown in to do this event in Comp (DIV B) any advice
Re: Experimental Design B/C
Posted: February 10th, 2014, 12:30 am
by fozendog
tsinzitari wrote:So Me and 2 other people just got thrown in to do this event in Comp (DIV B) any advice
Go over this rubric with your partners and divide the sections that each of you will be doing during the experiment.
http://www.soinc.org/sites/default/file ... 121213.pdf
Our team has a writer, doer, and a helper:
The writer covers sections a-d and n.
The doer covers sections h-k.
The helper covers sections e-g and l-m.
After you separate the sections have your coach give you a set of materials and do an experiment.
Have your coach grade the experiment, then go over it with you and see where you all can improve in your sections.
Then do another experiment to build a relationship with your team and cover any details you need to add.
Good luck at your competition!
Re: Experimental Design B/C
Posted: February 11th, 2014, 6:38 pm
by jwalkotten
Anyone else have any ideas about what we're supposed to do for "other appropriate statistic used" for Div. C.
Thanks!
Re: Experimental Design B/C
Posted: February 11th, 2014, 7:44 pm
by fozendog
jwalkotten wrote:Anyone else have any ideas about what we're supposed to do for "other appropriate statistic used" for Div. C.
Thanks!
The statistics we use are:
-Mean
-Median
-Mode
-Range
-Standard Deviation
-Linear Regression (just the slope of the line)
That will get you all the points if done correctly for each IV.
Re: Experimental Design B/C
Posted: February 12th, 2014, 2:48 am
by Flavorflav
Do you do standard deviation for each level, or for all of your levels together?
Re: Experimental Design B/C
Posted: February 12th, 2014, 3:52 am
by Phys1cs
Flavorflav wrote:Do you do standard deviation for each level, or for all of your levels together?
each level
Re: Experimental Design B/C
Posted: February 24th, 2014, 2:12 pm
by jwalkotten
fozendog wrote:jwalkotten wrote:Anyone else have any ideas about what we're supposed to do for "other appropriate statistic used" for Div. C.
Thanks!
The statistics we use are:
-Mean
-Median
-Mode
-Range
-Standard Deviation
-Linear Regression (just the slope of the line)
That will get you all the points if done correctly for each IV.
How do you do linear regression for each IV? b/c your linear regression is going to compare all three levels of IV correct? so you can't really do linear regression for each level of IV
Re: Experimental Design B/C
Posted: February 24th, 2014, 2:27 pm
by Phys1cs
jwalkotten wrote:fozendog wrote:jwalkotten wrote:Anyone else have any ideas about what we're supposed to do for "other appropriate statistic used" for Div. C.
Thanks!
The statistics we use are:
-Mean
-Median
-Mode
-Range
-Standard Deviation
-Linear Regression (just the slope of the line)
That will get you all the points if done correctly for each IV.
How do you do linear regression for each IV? b/c your linear regression is going to compare all three levels of IV correct? so you can't really do linear regression for each level of IV
A LinReg for each IV can show you how different the values are.. if you have a slope of one and two, one obviously doubles. But does that mean that there is more weight...higher temerature.. what is the significance of the doubling with regards to your lab? That's what I've always used it for. You then can use that data point in your conclusion and analysis to get those points.
Re: Experimental Design B/C
Posted: February 24th, 2014, 3:41 pm
by fozendog
We just do linear regression for the three averages that are plotted on the graph, not for each IV.