If the assumption of homogeneity of variances is violated in a one-way ANOVA, which method might you consider to address the issue? 

A) Conduct a Chi-squared test instead 

B) Use a Welch's ANOVA 

C) Remove the smallest group from the analysis 

D) Use a paired t-test instead