Verizon 2:41 PM 54%@ O. + 10 of 10 150 175 200 the standardined residuals from t

Question

Verizon 2:41 PM 54%@ O. + 10 of 10 150 175 200 the standardined residuals from this regression Figure 656 shows the plot of the standardined residuals versus the fimed values Figures 657 and 658 show, respectively, plots of the DFITS and Cook's D sistics. These data are available ina file named PETROS on the CD. FIGURE 6.57 Pet Time-Series Plat of DFIT 18 16 2 32 56 72 B0 03 24324048 56647280 Are there any unusaal observations thar should be checked before accepting these negression esuls Iso, which observations? Can you determine what might be causing centain ob- ervatioes to appear sl? Justify your

Explanation / Answer

Unsual observations are those observations whose values are either too high or too low ina given set of observations.These observations are also called as outlier.It is essential to check unsual observations before working on regression analysis as these observation makes the regression equation less reliablle as it greatly affects the slope of the regression equation.From Fig 6.57,unusal observation occur at index 17 and 18 respectively where value of DFITS was observed too low and too high respectively.From Fig 18,outlier exists at 17 and 18th indices.Presence of outlier or extreme observation is the cause of such graph in both Cook's D and DFITS.

Related Questions

Navigate

• Browse (All)
• Browse V
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.