Corellation Coefficient - Calculate??? :?

ChemHickey

Hey guys, I'm a Leaving cert student (LC 2012) and, with this new project maths course, we have to do a whole new pile of stats and probablility... Now, I have a Sharp EL W351(black) Calculator and no-one... not even our teacher can figure out how to calculate the corellation coefficient....

Does anyone know how to calculate correlation coefficient on THIS calculator as I have been using it for the past 4 or so years and would hate to have to change calculator for such a small detail...

Thanks...



Feel free to move this thread if it is not in the right forum...

LeixlipRed

Can you not calculate it using the formula? Jesus, calculators should be banned, you probably have no idea what the formula represents or how to interpret it. Ugh. Here's a link to calculating the correlation coefficient using the formula: http://easycalculation.com/statistics/learn-correlation.php

TheBody

ChemHickey wrote: »

Hey guys, I'm a Leaving cert student (LC 2012) and, with this new project maths course, we have to do a whole new pile of stats and probablility... Now, I have a Sharp EL W351(black) Calculator and no-one... not even our teacher can figure out how to calculate the corellation coefficient....

Does anyone know how to calculate correlation coefficient on THIS calculator as I have been using it for the past 4 or so years and would hate to have to change calculator for such a small detail...

Thanks...



Feel free to move this thread if it is not in the right forum...

Here is a video on how to do it!!

http://www.wonderhowto.com/how-to-find-r-value-equation-regression-line-w-el531w-303059/

TheBody

LeixlipRed wrote: »

Can you not calculate it using the formula? Jesus, calculators should be banned, you probably have no idea what the formula represents or how to interpret it. Ugh. Here's a link to calculating the correlation coefficient using the formula: http://easycalculation.com/statistics/learn-correlation.php

Steady on there!!! ChenHickey doesn't write the leaving cert syllabus!! They are not required to know the formula for the leaving cert. For what it's worth, I agree with you that it is stupid. Calculator work teaches nothing.

ChemHickey

Thanks guys.... No.. we don't use the formula.... I don't know why but.. if i can't get it working on the calculator i might have to resort to that..

@TheBody, thank you for the video... but It is a different calculator than that one... I know... It is stupid that even thought they have practically the same name, Sharp EL W531 compared to Sharp EL 531W.... I will keep trying on that video nyway to see if it will work eventually... thank you...


Any other tips etc are welcomed!! :L

ChemHickey

TheBody wrote: »

Steady on there!!! ChenHickey doesn't write the leaving cert syllabus!! They are not required to know the formula for the leaving cert. For what it's worth, I agree with you that it is stupid. Calculator work teaches nothing.

Yes, some of the work on the cours eis quite monotonous and is not needing of a calculator.... For certain topics we shouldn't use is it... but for others it is a definite ...

TheBody

Here is the manual!!

http://www.sharp.eu/cps/rde/xbcr/documents/documents/om/30_cal/ELW531-G-H-HA-W535_OM_GB.pdf

ChemHickey

TheBody wrote: »

Here is the manual!!

http://www.sharp.eu/cps/rde/xbcr/documents/documents/om/30_cal/ELW531-G-H-HA-W535_OM_GB.pdf

Thank you!!!!

I shall study it in detail! :L :pac:

MathsManiac

LeixlipRed wrote: »

Can you not calculate it using the formula? Jesus, calculators should be banned, you probably have no idea what the formula represents or how to interpret it. Ugh. Here's a link to calculating the correlation coefficient using the formula: http://easycalculation.com/statistics/learn-correlation.php

I completely disagree. Expecting people to calculate correlation coefficients by hand would be mindless drudgery, and it is something that nobody working in statistics would ever do by hand. Being able to stick numbers into the formula contributes nothing to one's understanding of the concept.

Using the calculator takes the focus off this trivial aspect and allows you to focus on meaning and interpretation.

Under the old syllabus, students wasted wads of time calculating standard deviations, for example, and yet had no idea what this number actually tells you about your data. The kinds of questions that are starting to appear on these new papers are far better, in my view. If they don't understand the concept of correlation and how to interpret it, they won't get very far with these questions.

How many Leaving Cert students do you think would look at the formula for the correlation coefficient and say "Ah, now I have deep insight into the meaning of this quantity - please let me calculate lots of them using an abacus, so I can appreciate its beauty more fully"?

There is a finite (and not very large) amount of time available for each topic in maths at Leaving Cert, and I would far rather see the time available for stats spent interrogating data sets and drawing meaning from them than merrily banging away with formulae.

And nobody's telling them they're not allowed use the formula, if that's what floats their boat.

LeixlipRed

I think you've misintepreted my point. I was complaining about "plug and chug" mathematics. The link was just to let him calculate it if it was urgent. I wasn't suggesting he should learn off some formula blindly, why would I suggest that?

And using the calculator doesn't allow them to shift the focus to understanding it allows them to shift the focus to, "this is how you compute this yoke on the calculator, follow these steps exactly and anytime it's mentioned you should repeat these steps without ever questioning them".

MathsManiac

Fair enough. I took it that you were expressing the view that students should be required to calculate these quantities using the formula rather than by using the relevant feature on a calculator. It has been my experience that, with the best will in the world, this ends up diverting an inappropriate amount of class time to practicing lower-order procedural skills.

But one way or the other, I think we both agree that being able to actually compute the number - by whatever means - is not the important thing that should be dominating the time and effort.

Doing it by calculator is faster, so you have more time for something else. If you choose not to use that time to develop insight and understanding, (as I think you are suggesting teachers might do,) then yes, you'll still be able to answer questions that simply ask you to calculate a correlation coefficient, but you'll be woefully inadequately prepared for questions on the LC project maths papers.

So, my point is that I think it's a good idea that they are letting people use calculators to do what calculators are good at, and demanding greater levels of insight in return.

LeixlipRed

I think it was my belief that the OP seemed to be implying reliance on a calculator to do maths. The comment about buying a new calculator if this one couldn't do the calculation made me focus on the fact he could simply use the definition of the correlation coefficient to work it out.

Clearly, as someone who works in maths education I realise that the emphasis must be on understanding. And it never was under the old syllabus and I wonder will it ever be under project maths.

MathsManiac

Every time a syllabus changes, there is always an ambition to get greater understanding and less reliance on routine procedures. It doesn't always work out that way, judging by the fact that same complaints get made the next time change rolls around!

Still worth the effort, though.

Ostrom

Pearson, Spearmann or Kendall....?

LeixlipRed

Pearson's but don't tell the teachers there's other types, I done that last year teaching them some basic stats and it confused em more :pac:

**Timbuk2**

I think that students should be thought the formula first, shown how to calculate the correlation coefficient from scratch with an easy example (e.g. only 5 or 6 data points, calculated Sxy on it's own, then calculate Sxx on it's own, then calculate Syy on it's own and put it all into the formula). Then the teacher should go through what the coefficient means, is it close to what you would have guesstimated from looking at your scatter plot sketch.

After they know the formula, then it should be allowed to use the calculator. In real life statistics, it's all done by computer. But I think it's important that people realise what the computer (or calculator) is actually doing when people hit enter and get the result back.

**Timbuk2**

Sorry about the double post, I have the Sharp EL-W531 (Black, it says WriteView on the top right), and this is how I calculate the correlation coefficient.

Press MODE
Press 1: Stat
Press 1: Line (in the case that you have two variables, usually x and y)
Press 2nd F then MODE to clear anything that might be stored as data memory.
Enter your first data point, separating the x and y variables with a comma (the comma button is located above the 'Change' button, it says (x,y)).
So if I wanted to enter 3,4, I would type [3] [(x,y)] [4]
After you type in the data point, press the Change button (in STAT mode, this is actually the 'DATA' button, as indicated by the writing on top).
At the bottom of the screen, it will say DATA SET= 1, which means it has been entered.
Now type in data point number 2, and do the same until you reach the last data point.

Once you have all your data points in, that's the hard bit finished.
The Pearson correlation coefficient for samples is usually denoted by a small r, which you can find above the Divide button on the calculator. To display it, press the Alpha button (below 2nd F), then the divide button, then press equals. And it should display r (the correlation coefficient).

This sounds very complicated, but once you get used to it, you'll have it done in no time! The stat mode is very useful on the calculator, because once you input the data, you can display all sorts of numbers in seconds.
You could find the standard deviation of the x data by pressing the sx button (Alpha + 5), or the sum of the x terms by pressing the Sum of x button (Alpha + decimal point button), or the mean of the x terms by pressing the x bar button (alpha + 4), and similar for the y variables!

I did the old course for the leaving cert (in 2010), so I'm not sure how much statistics you actually do, but if you do any regression, the regression coefficients can easily be displayed. Note the a and b (directly above where r is). A regression line is usually given in the form Y = a + bx, where a is the y-intercept, b is the slope and Y is the estimated value of y. If you haven't done that yet, just ignore that bit! Calculating regression coefficients a and b involves similar techniques to calculating the correlation coefficient!

And if you do use your calculator in the leaving cert, make some indication on the paper that's how you got your answer - otherwise it might look as if you are pulling numbers from thin air :P

ChemHickey

**Timbuk2** wrote: »

Sorry about the double post, I have the Sharp EL-W531 (Black, it says WriteView on the top right), and this is how I calculate the correlation coefficient.

Press MODE
Press 1: Stat
Press 1: Line (in the case that you have two variables, usually x and y)
Press 2nd F then MODE to clear anything that might be stored as data memory.
Enter your first data point, separating the x and y variables with a comma (the comma button is located above the 'Change' button, it says (x,y)).
So if I wanted to enter 3,4, I would type [3] [(x,y)] [4]
After you type in the data point, press the Change button (in STAT mode, this is actually the 'DATA' button, as indicated by the writing on top).
At the bottom of the screen, it will say DATA SET= 1, which means it has been entered.
Now type in data point number 2, and do the same until you reach the last data point.

Once you have all your data points in, that's the hard bit finished.
The Pearson correlation coefficient for samples is usually denoted by a small r, which you can find above the Divide button on the calculator. To display it, press the Alpha button (below 2nd F), then the divide button, then press equals. And it should display r (the correlation coefficient).

This sounds very complicated, but once you get used to it, you'll have it done in no time! The stat mode is very useful on the calculator, because once you input the data, you can display all sorts of numbers in seconds.
You could find the standard deviation of the x data by pressing the sx button (Alpha + 5), or the sum of the x terms by pressing the Sum of x button (Alpha + decimal point button), or the mean of the x terms by pressing the x bar button (alpha + 4), and similar for the y variables!

I did the old course for the leaving cert (in 2010), so I'm not sure how much statistics you actually do, but if you do any regression, the regression coefficients can easily be displayed. Note the a and b (directly above where r is). A regression line is usually given in the form Y = a + bx, where a is the y-intercept, b is the slope and Y is the estimated value of y. If you haven't done that yet, just ignore that bit! Calculating regression coefficients a and b involves similar techniques to calculating the correlation coefficient!

And if you do use your calculator in the leaving cert, make some indication on the paper that's how you got your answer - otherwise it might look as if you are pulling numbers from thin air :P

Thank you!!! I must try that out!!! My calcultaor is outside in our extension and its too dark too go out at the moment so... I'll update you in the morning!! It seems like it'll work and if so... YOU'RE A SAVIOUR!!

ChemHickey

It works!!! Thank you!!! =]

monkeybhands

hi, i think this is an old convo, but i have the exact same problem, only i finally was able to decipher the instructions and did exactly what chemhickey said and it still comes up as error? i have no idea what im doing wrong.