Scribd
Upload
60 views11 pages

PSLP Unit-3 (Regression Lines) - 1

Uploaded by

Pawas Goyal
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
60 views11 pages

PSLP Unit-3 (Regression Lines) - 1

Uploaded by

Pawas Goyal
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/ 11
‘The fitted or estimated regression line is therefore 5=Bo+ Bix (1.9) Note that each pair of observations satisfies the relationship yaRorBarte, FHL 2m where ¢; = y,~ 5) is called the residual. The residual describes the error in the fit of the mode} to the ith observation y;. Later in this chapter, we will use the residuals to-provide-inforsmation about the adequacy of the fitted model. yy — . i Observed valve | Data (y) Estimated regression tine | FIGURE 11-3 Deviations of the data from the estimated | | regression model. : — tel we (& kquations-11-6-are:called-the least squares normal equations. The solution to the non equations results in the least squares estimators Bp and Bs mal ‘The least squares estimates of the intercept and slope in the simple linear regression model are ary | G S&S BI oxygen Purity w. ti en wily We will ft a simple Tre following quantities may be cosipiat lear regression im del to the oxygen purity data in Table 11-1. x 2 n=20 Yx,= 23,92 Lye bees Therefore, the least squares estimates of the slope and intercept are wd a Y= 1.1960 ¥ 609.1605 x Pi Zvi = 170.088.5321 $s «29,2902 7 a x Ea = 2.214.6566 3.92)? ap 0.68088 a 2» = 2,214,6566 22 92)(b83.21) 5 baogy 20 fy So, WOI7744 bi 5.” 68088 14.94748 Bux = 92,1605 ~(14,94748) 1.196 = 74.28331 The fitted simple linear regression model (with the coefficients reported to three decimal places) is This model is plotted in Fig. 11 Hong with the sample di 74.2834 14.9874 lata, Practical Interpretation: Using the regression model, we would predict oxygen purity of j » 89.24% when the hydrocarbon level is x = 1.00% toeryy: that is, 1 is unlikely \y be interpreted as an esti panty when 1 = 1.00%, or as an estimate of a new observation when x = 100%. Th a future observation on purity would be exactly 89.20% when th 11.00%. In subsequent sections, we will see how (0 use confidence intervals and pre exer in estimation from a regression move, taran’a (11-20) where fo is computed from Equation 11-19. The denominator of Equation 11-19 is the stand- ard error of the slope, so we could write the test statistic as r= bizBi0 se(B,) A similar procedure can be used to test hypotheses about the intercept, To test <7: Bo = Boo Hi:Bo #Boo ana) | . er figure 11-5 The oo ote hypothesis Ho:B: = Ois — l | @ . rotrejected. o . @ we would use the statistic : tebe Ee reaps gg _Raeeah Heal bgon SS Tmtereept Hit Bo% Boo Ty=—-Bo=Boo_ _ Bo-Boo 1-22) Efe] 2 Sal af = J wlith Cr-a) dlegaseo of, freedom, and reject the null hypothesis if the computed value of this test statistic, fo, is such that {ro| > taran-a- Note that the denominator of the test statistic in Equation 11-22 is just the stand- ard error of the intercept. A very important special case of the hypotheses of Equation 11-18 is Hy:B, =0 Hy: B, #0 (11-23) ‘These hypotheses relate to the significance of regression. Failure to reject Ho:B, = 0 is equivalent to concluding that there is no linear relationship between x and ¥. This situation is illustrated in Fig. 11-5. Note that this may imply either that i of litle value in explaining the variation in Y and that the best estimator of Y for any x is $= Y (Fig. 11-5(a)] or that the true relationship between x and ¥ is not linear (Fig. 11-5(b)]. Alternatively, if Ho: B; = 0 is rejected, this implies that x is of value in explaining the variability in Y (see, Fig. 11-6). Rejecting Hy: could mean either that the straight-line model is adequate [Fig. 11-6(a)] or that, although there is a linear effect of x, better @ results could be obtained with the addition of higher order polynomial terms in x (Fig. 11-6(b)]. Oxygen Purity Tests of Coefficients | We will test for significance of regression using the model for the oxygen purity data from Example 11-1. The hypotheses are He:B, =0 and we will use 0 = 0.01. From Example 11-1 and Table 11-2 we have f,=14.947 n=20, S_,=0.68088, 6° | Sothe t-statistic in Equation 10-20 becomes oe “Shee | VES se(6,) 1.18 / 0.68088 if i i test statistic is very far Practical Interpretation: Because the reference value of fis foass = 2.88, the value of the fe caeite imo the critical region, implying that Ho, = 0 should be zejeced, Tere i Sirong evidence to suppor th - for this test is P = 1.23107. This was obtained manually with a cael . able 11-2 presents the typical for this problem. Notice that the rs I Ypical computer output for this pro : "puted as 11.35 and that the reported P-value is P = 0.000. The computer also pee 7 Lowe | jyeoltsis Ho: Bo= 0. This statistic is computed from Equation 11-22, with Boo = 0 05 fo |_Spothesis that the intercept is zero is rejected. tatistic value for the slope is istic for testing the ‘Clearly, then, the | Jowing denmition of LUUUL = 64)% ConTIUEILE mies yee ws pe mene nese 1 Confidence Intervals on. ‘Parameters. Under the assumption that the observations are normally and independently distrib. uted, a 100(1 — «)% confidence interval on the slope B; in simple linear regression ig ae F fanuwa fe $BrSBr +a fs M29) ‘Similarly, a 100(1 ~ «)% confidence interval on the intercept Bois Bo —tare-2 |e 2 | stisiotion en |! z| (i130) Oxygen Purity Confidence interval on the Slope We will find 2 95% confidence interval onthe slope of the regression line using the data in Example 11-1. Recall that B, = 14.947, S.,. = 0.68088, and 6? = 1.18 (see Table 11-2). Then, from Equation 11-29, we find 6 : fomasis fg

You might also like