Prove that coefficient of correlation always lies between 1 and -1.
​Please do not give links.

Dear student,

coefficient of correlation = r = cov (x,y)σxσyσx2=Σ(y-x¯)2ncov (x,y)=Σ(x-x¯)(y-y¯)nwe have to prove -1r1Let's consider two terms x-x¯σx  and y-y¯σytaking sum of squares of both termsΣx-x¯σx  ± y-y¯σy2(given no is positive since its square)Σx-x¯σx  ± y-y¯σy2 0opening square we getΣx-x¯σx2  + y-y¯σy2+2x-x¯σx y-y¯σy 0on simplifying,Σx-x¯2σ2x+Σy-y¯2σ2y±2Σx-x¯y-y¯σxσy0diving by n Σx-x¯2nσ2x+Σy-y¯2nσ2y±2Σx-x¯y-y¯nσxnσy0using formula of variance,σ2xσ2x+σ2yσ2y±2covx,yσxσy01+1±2r02+2r01-rr-1or 2-2r022r1r-1r1hence proved
Regards

  • -1
What are you looking for?