Help with stats coursework

-K- · 21 Jan 2009 at 21:58

Hi guys, I am doing a stats coursework at the moment, and I have no idea on some parts of the coursework which I am required to do.

The coursework wants me to find if there is a relationship between 2 set of data, so I assume I use a scatter diagram for this to find out. Then they asked if it is possible to obtain a formula for predicting a change in data A from change in data B?

How do I obtain this formula from the diagrams or data? Thanks for any help given.

RemainNameless · 21 Jan 2009 at 22:04

Fit a line to your data points and then try and find an equation which approximately matches your line of best fit.

HeadlessChicken · 21 Jan 2009 at 22:23

2 sets of data could indicate a t-test, or correlation/regression. What are the 2 data sets? And what level are you studying?

aslco · 21 Jan 2009 at 22:27

What level is this? If i think back my GCSE stats involved lots of spearmans rank and scatter graphcs with lines of best fit.

Nitrojan · 21 Jan 2009 at 22:31

Good old Spearmans Rank Correlation Co-Efficient!

Red Porsche · 21 Jan 2009 at 22:31

Ask your teacher? they are supposed to teach you what you need to know so you can apply it.

pingwing · 21 Jan 2009 at 22:38

Nitrojan said:
Good old Spearmans Rank Correlation Co-Efficient!

Ah i remember this, GSCE i ranked teh actuall data opposed to the co-efficent i think....

I got an E in the course work *egg on my face* redeemed my self massivly in the exam with a B!

Didnt make that mistake at A-level thankfully :cool:

-K- · 21 Jan 2009 at 22:49

Yea my teachers been out of class for almost a month due to some hip nerve twisting....

This is the data

Im suppose to use the 2 right end columns.

dante6491 · 21 Jan 2009 at 22:54

Product moment correlation co-efficient. You can then do y upon x regression lines.

HeadlessChicken · 21 Jan 2009 at 23:13

Simple regression/correlation.

Plot a scatter.

Formula to predict any change:

Y=MX+C (IIRC!!!)

X + Y = obvious
M = gradient (rate of change) Y1-Y2 / X1-X2
C = intercept

aslco · 21 Jan 2009 at 23:15

I could send you my project i did and see if its of the right level....i'm not sure if i still have it or not but its worht a shot. i will only be able to do it tomorrow though. NO COPYING for your own sake!

-K- · 22 Jan 2009 at 11:37

dante6491 said:
Product moment correlation co-efficient. You can then do y upon x regression lines.

After finding the PMCC, what do you do with the answer?

And thanks all for the help, appreciate it