Page 197 - DECO504_STATISTICAL_METHODS_IN_ECONOMICS_HINDI

bdkbZ—12% js[kh; izrhixeu fo'ys"k.k % ifjp; ,oa izrhixeu dh js[kk,¡

vk/qfud lkaf[;dh esa izrhixeu dh /kj.kk osQoy bldh okLrfod fo'ks"krkvksa osQ vè;;u rd gh lhfer ugha uksV
gSA bldk iz;ksx mu lHkh {ks=kksa esa fd;k tkrk gS ftlesa nks ;k nks ls vf/d lEcfU/r leadekykvksa osQ in&ewY;ksa
esa lkekU; ekè; dh vksj okil vkus dh izo`fÙk gksrh gSA izrhixeu osQ vk/kj ij lkekftd] vk£Fkd o
O;kolkf;d {ks=kksa esa fofHkUu rF;ksa osQ eè; lEcU/ksa dk fo'ys"k.k djosQ ,d inewY; ls lEcfU/r nwljh
(vkfJr) Js.kh osQ loksZi;qDr ewY; dk vuqeku fd;k tk ldrk gSA
ftu {ks=kksa ls lEcfU/r leadekykvksa osQ fofo/ py&ewY;ksa esa lkekU; ekè; dh vksj okil vkus dh izo`fÙk ikbZ
tkrh gSA ogka izrhixeu fo'ys"k.k dh O;kogkfjd mi;ksfxrk gSA

lglEcU/ fo'ys"k.k tgk¡ ?kVukvksa osQ lgifjorZu osQ lEcU/ dks crkrk gS ogk¡
izrhixeu fo'ys"k.k lEcU/ dh izo`fÙk vkSj ek=kk osQ vk/kj ij Hkkoh vuqeku iznku
djus esa lgk;rk djrk gSA

bldh lgk;rk ls ewY; osQ vk/kj ij ekax dk] o"kkZ dh ek=kk] cht] [kkn] vkfn osQ vk/kj ij Ñf"k mit rFkk
iwath osQ c<+kus ;k ?kVkus ij ykHk] bR;kfn esa gksus okys ifjorZu dk vuqeku yxk;k tk ldrk gSA izcU/ vf/
dkfj;ksa }kjk O;olk; osQ fu;U=k.k midj.k (Control tool) osQ :i esa izrhixeu fo'ys"k.k dk iz;ksx fd;k tkrk
gSA bl izfof/ (Process) osQ vk/kj ij mfpr O;kolkf;d fu.kZ; ysrk gh ljy ugha gks tkrk gSA oju~ fu.kZ;
dks O;kogkfjdrk dh dlkSVh ij ij[kk Hkh tk ldrk gSA tc nks pyksa osQ chp dkj.k o izHkko (Cause and
Effect) dk lEcU/ gksrk gS rks izrhixeu lehdj.k (Regression Equation) dh lgk;rk ls ,d ewY; ij vk/
kfjr nwljk ewY; cM+h ljyrk ls fudkyk tk ldrk gSA

x dk loksZÙke ekè; ewY; Kkr djuk gks] rks x dk y ij izrhixeu lehdj.k (x on y) dk iz;ksx
djuk gksxk vkSj ;fn y dk loksZÙke ekè; ewY; Kkr djuk gks] rks y dk x ij izrhixeu lehdj.k
(y on x) dk iz;ksx fd;k tk,xkA

lglEcU/ vkSj izrhixeu esa vUrj (Difference Between Correlation and Regression)

¶tcfd lglEcU/ nks ;k vf/d ?kVukvksa esa lgifjorZu dh ?kfu"Brk dk ijh{k.k djrk gS] izrhixeu
fo'ys"k.k (Regression Analysis) bl lEcU/ dh izÑfr o ek=kk dk eki djosQ gesa Hkkoh vuqeku yxkus
dh {kerk iznku djrk gSA¸ 1
lglEcU/ ,oa izrhixeu esa vUrj fuEu izdkj gSμ
(i) dkj.k&ifj.kke lEcU/ (Cause-Effect Relation)μpy&ewY;ksa osQ dkj.k ifj.kke lEcU/ksa dks lglEcU/
fo'ys"k.k vf/d Li"V djrk gS] ijUrq nksuksa esa dkSu&lk dkj.k vkSj dkSu&lk ifj.kke gS ;g lglEcU/
ls ekywe ugha iM+rkA blosQ foijhr] izrhixeu fo'ys"k.k esa ,d py dks Lora=k ekudj nwljs vkfJr
dk ewY; Kkr fd;k tkrk gSA bl izdkj Lora=k py dkj.k gksrk gS vkSj vkfJr ifj.kkeA
(ii) lglEcU/ dh ek=kk ,oa izÑfr (Degree and Nature of Correlation)μlglEcU/ fo'ys"k.k osQ
}kjk nks pjksa osQ eè; lEcU/ dh ek=kk dk vkHkkl gksrk gS] ijUrq izrhixeu fo'ys"k.k ls lEcU/ dh

1. “While correlation analysis tests the closeness with which two (or more) phenomena covary,
regression analysis measures the nature and extent of the relation, thus enabling us to make
predictions.” —W.Z. Hirsch

LOVELY PROFESSIONAL UNIVERSITY 191