I am trying to see if the data set we are supposed to be analyzing satisfies the

Question

I am trying to see if the data set we are supposed to be analyzing satisfies the assumptions of equal variance and normality of ANOVA. I know you are supposed to perform some type of transformation if there is a pattern in the plots of the residuals. A pattern seems to appear in residuals vs weight (Wt) and the residuals vs the order of data plots, but not in the residuals vs fitted plot. I have already taken the log transformation of the entire data set and plan on removing the point with the large cooks distance and leverage. Would you recommend any other specific transformations to the variables or data?

RStudio Hle dit Code Plots SessRon Build Debug Profile 16R gnif ant at a fig. height 8, fig. width 122 P lisa 180 dua I g. height gray gray k me any a Reeidu vs. Fitted Cook's distance Added-Variable Plots Cook's dist ve Leverage hal(1-ha) 2/8/20

Explanation / Answer

I also see that your residual Vs fitted is not homoscedastic either.

What i would suggest is:

1) Check each variable for skweness. If you find it positively skewed than do log transformation or else if it is negatively skewed than perform square transformation.

2) Also check if you have outliers and than remove the point with the large cooks distance and leverage.

Related Questions

Navigate

• Browse (All)
• Browse I
• Subjects

---

---

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