In conducting the initial assessment of the variables for a simple linear regression analysis, it was noticed that one of the variables is not normally distributed. The variable has a high level of positive skewness
Should the regression analysis be continued? Why or why not?
The linear regression will not necessarily be valid as the assumption of normality is violated. One could try to transform the variable, most likely applying the natural log of the variable, which might 'normalize' it. If the variable cannot be transformed or the variable is not normal after transformation, one could apply non-parametric methods to assess the relationship. Spearman's correlation would be a viable option.
You might also like to view...
In the MySQL Table Editor, the abbreviation AI stands for AUTO_INCREMENT
Indicate whether the statement is true or false
A publisher is interested in estimating the proportion of textbooks that students resell at the end of the semester. He is interested in making this estimate using a confidence level of 95 percent and a margin of error of ±0.02
Based upon his prior experience, he believes that ? is somewhere around 0.60. Given this information, the required sample size is over 2,300 students. Indicate whether the statement is true or false