Sciencemadness Discussion Board

Graph info to find correlation

Yttrium2 - 10-10-2019 at 18:33

Lets say I want to make a graph

x axis is amount of water I drink


y axis is the hours I study



this will show a "correlation" between water drank & hours studied





how can I throw in a 3rd element, such as miles walked, to see how it affects hours studied?--


Is there a way to collect data on several "elements" to see how they affect hours studied?

(seems they'd have to be done one at a time...)


--


--

--


I'm trying to narrow down what affects how many hours I study.
Seems like I'd have to keep things constant
And manipulate 1 "element" (variable?) - at a time.
I've heard "experiments" can be done when 2 variables get manipulated.
But then how is there "control", or a determination of what causes what?


I'm thinking more exercise, less food, losing weight, drinking plenty of water will all contribute to hours studied.





Any info on the topic appreciated


p.s. maybe I've gotten ahead of myself by trying to throw in a 3rd variable, perhaps I should be more concerned with how to design the graph, and how much of a distance there is between increments for water drank. (I'm bad at excel, one day I'll take a course but It'll probably come more handy when I'm finished with math)





[Edited on 10/11/2019 by Yttrium2]

j_sum1 - 10-10-2019 at 20:08

The word you are after is multivariable analysis.
Unless you have the mathematical understanding to support your investigation, it is a lot simpler to consider just two variables at a time.

For the three variables you have chosen, I suspect any correlations you find will be pretty meaningless anyway.
The dependent variable, hours studied, is likely to be dependent on many many variables including motivational factors that are difficult to quantify.
You are also going to have a rather small data set -- there is a rule of thumb that suggests a minimum n of 30 is needed to make any meaningful conclusions. If you consider a data set larger than 30 you also need to include time dependence as one of yoru variables.
What is the point of all this? Quit procrastinating and just do your work. (Even if it is your 31st birthday. <cough>)

[Edited on 11-10-2019 by j_sum1]

SWIM - 11-10-2019 at 10:50

Try keeping your schoolbooks in the bathroom.

This will encourage a direct relationship between your frequency of study and your fluid intake.

You can probably get in a lot more bathroom study time if you give up all roughage in your diet, but this practice may prove to be a bit of a strain:mad::mad::mad::

a more realistic graph

sodium_stearate - 12-10-2019 at 07:47

Why don't you try making a graph that shows
the correlation between how many inane posts
you make here, vs how pissed off everyone else
here is getting from reading them?


unionised - 12-10-2019 at 09:21

Quote: Originally posted by Yttrium2  
Lets say I want to make a graph

x axis is amount of water I drink


y axis is the hours I study



this will show a "correlation" between water drank & hours studied





how can I throw in a 3rd element, such as miles walked, to see how it affects hours studied?--


Is there a way to collect data on several "elements" to see how they affect hours studied?

(seems they'd have to be done one at a time...)


--


--

--


I'm trying to narrow down what affects how many hours I study.
Seems like I'd have to keep things constant
And manipulate 1 "element" (variable?) - at a time.
I've heard "experiments" can be done when 2 variables get manipulated.
But then how is there "control", or a determination of what causes what?


I'm thinking more exercise, less food, losing weight, drinking plenty of water will all contribute to hours studied.





Any info on the topic appreciated


p.s. maybe I've gotten ahead of myself by trying to throw in a 3rd variable, perhaps I should be more concerned with how to design the graph, and how much of a distance there is between increments for water drank. (I'm bad at excel, one day I'll take a course but It'll probably come more handy when I'm finished with math)





[Edited on 10/11/2019 by Yttrium2]


This may help
https://en.wikipedia.org/wiki/Analysis_of_variance

elementcollector1 - 12-10-2019 at 14:00

In practice, this is generally done by setting your third variable to a constant (e.g. '5 miles walked', '10 miles walked', etc.) and graphing your first two variables given that constant. Multiple lines for multiple values of the third variable are a very good indicator of its relationship to the other two.