Scribd
Upload
25 views

Statistics Chap-7

Uploaded by

Md Mahmudur Rahman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
25 views

Statistics Chap-7

Uploaded by

Md Mahmudur Rahman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 14
“« The statistical tool with the between vo variables is studic of Correlati he degree such fhat the the chang another. |, numerical index that tells related and to what exter .s with the variations in) t ibolized stl y 6) variables. The values ra + Correlati different i tars eet tive and Zero ‘Types of Correlation: Positive, Negi (C0) is followed by Positive correlation: ble (Y); the cprrelation +» When the increase oF decre corresponding increase oF js said to be positive correlation ive correlations range from 0 to +1; the up, 1 of correlation ation specifies that, for | increase in the other. ” have a perfect positive cqrrelatjon ease in one variable ‘decrease in the other varial sper limit i.e. +1 is the + The posit perfect positive cocfficien + The perfect positive correl. variable, there is proportional + For example “Heat” and “Temperature every unit increase in one Negative correlation: + Ion the other hand, the increase in one variable (X) results in a corresponding decrease in the other variable (Y), the correlation is be negative correlation. The negative correlation ranges from 0 to ~ 1; the lower limit givjng the perfect negative correlation. a The perfect negativt.correlation indicates that for every unit incr¢ase in one variable, there is proportional unit decrease in the other. 2 ‘Types of Correlation: (+, -, 0 rel tions) positive Relation | Negative Relation | Zero Relation, ieee: Keay nets X10 Io 15 go 50 yi 5 12 70 55 ‘a 14 60 60 * Zero corre ns no relationship between the two varidbles X and Y: ie. the change in one variable (X) is not assoc: ated with the change in the other variable (Y). * The zero correlation is the mid-point (neutral position) of th. Jto+L te partial and multiple ¢ aes ariables are studied ‘o When only two variabl correlation. When three or more yariables ultiple or partial correlation. relation three or mor: Yin partial correlation. we but consider only two variabl rand Non-Linear Correlation | | Line: relation: : ariables is Cons! Linear ‘Ta the linear, relation, the chang= of two ¥: , the (vo ¥ oid «= Lincar correlation 18 the ratio of o BS rection or opposite dir and the graphical other variable is either in the same di ble with respect (0 Sipresentation of the one Varia } straight line Non-linear relation: ables does jnot bear = In the Non-linear relation , the change of two vari ‘a constant ratio. = Consider the situation. Fi Variable increases proportionately up to eetease in the first variable the second v + The graphical reprepentation of the two variables will be a curved Fes ins relationship between the two Variables is termed as the | curvilinear correlation. srst. with increase of one varinble, tye second some point: after thak with an ariable starts decreaking. + Merely computation of col enificanee until ai the coefficient be in order correlation tell us about tr = What do we mean by the dO rr correlation? Its simp! eee aa nd unless ly. like: the foll Interpretation Moderate Postive/Nea elation does not have any wwe determine to be significant. he data? — ‘obtained value Oo} lowing table: oe r and what d f coeflicient ortect Posilive/Negative Correlation ery High PositiverNegative Correia h Positive/Nogative Correlation jative Correlation how large 9} ust bes: pt mmm Diagram Method ethods The Scatter Dia plest method fe rudy the correlation BE values for each pair of 8 form of dots thereby obtain eraph ip the ‘as the nymber jable is plotted oP ‘as many points of observations, st the scatter of several points the degree of Then by looki correlation is ascertained ams tell about the degree utter dil The following types of § of correlation between variable X and variable Y- Method: * The Spear™ Gorretation CoelMiclens deed to study the strengtl OF association between F yethod is applicd 10 te ordinal set of num sj one afer the other so {het ranks are Tity or quantity, such 85 ranking the one ranked last in thr B he two ran! ich can bf pers, Whil + Thism nrranged in order, t relation coefficie the ott + Inthe rank co! nt method, the iven 10 individual on the basis of position 1* till Nth position for ts qual and gocs jent (in normpal rank Jation coeffic! culate the rank cor pisctos =1-$ rank (tie in ranks hap] + The formula to ca ja|for *r? is situation) is: pen) is given the formul + When same ames ee ame sro" +p (mi = mi) + i HN + The Method of Least method that tells th variables by using the squat regression coefficient that of x on «The numerical notation of the formu fficient oe correlation by the coe’ [ou XP e root of mee Squares js another e degree of correlation y and y on X. method of Jeast squares j $57, first it is to calculate Dxy and byx using the mathematica} between the duct of tw thé’pro' Ja to calculate the b given, fe Determination tion is the squat 1 of determi alue of the correlation. rete ¥: plains the level of variance in the dey hip with the independent varia 5 ‘ion explains the proportion of tbe explained var ding to the regression equation orrelatioh? Which Coefficient 0! re of the coefficient ofc + The Coefficient js calculated to interpr «+ itis useful because it ex explained by its relations! § © ~ Thecocfficient ofdetermina the relative reduction in variance correspon Shoutihe mean of the dependent variable. pendent variable caused o7 ation oF will be 0.64, which means thal 61” he value of r= 0.8, then < jiained by the independent variable }vhile 307 + For example, ifth ‘variation in the dependent variable is exP! remains unexplained. 5 + Thus, the coefficientof determination is the ra of explained variance to thp total ras ret tells about he strength of Lincar association between the var les, say X nd ¥- The value of lics between and I and observes the following relatjonship Servi be value ofr" will alwaysbe greater than. ‘r* unless the | =0 oF a “The coefficientof determination also explains that how well he repr ¢ fits the pascal Sa P! ell the regression fine fits the - csi {0 the points plotted on a scatter diagram, the ore likely igmrand the farther the ine from the points the Iessey is the abi} 5 of Coefficient of Correlation jont of correlation G 1) always lies of the coclli (el to +1) 41. Such as 3 correlation e correlation } between 3 2 r+. perf “1, perfect negati / 0, no correlation |e dpeificient of correlation is indepen - scale. By origin. it means subtracting any non-ZeTo the given value of Nand Y the vale of * i seal itmeans, there no effect on the value of ~ vided or multiplied by. any constant, ITicient of correlation is a geometri cient of correlation is a geometric mean of two ct positiv ndent of the ori constant from cS cally itis represented as: ye re erse is not mie, he follov jor last 5 year Karl Pea: Sales (X): Profit (¥): ae Profits of an Or ent of and yiculate co-

You might also like