Question

17. Residuals are defined as the portion of scores not accounted for by the multivariate analysis. T or F

18. Homoscedasticity is the assumption that the variability in scores for one continuous variable is roughly the same at all values of another continuous variable. T or F

19. With univariate analyses, homogeneity of variances is assessed statistically with Levene’s test. T or F

20. In multivariate situations, homoscedasticity can be assessed statistically by using Box’s M test for equality of variance-covariance matrices. T or F

21. The main purpose for screening data prior to conducting a multivariate analysis is to deal with the accuracy of the findings. T or F

22. Another purpose for screening data is to enter missing data and assess the effect of and ways to deal with complete data. T or F

Answer 1

17) Residuals are defined as the portion of scores not accounted for by the multivariate analysis.

TRUE

Residuals are portion of variability accounted for by multivariate analysis. They are also referred to as prediction errors because they serve as measures of difference between obtained and predicted values on the given variable.

18) Homoscedasticity is the assumption that the variability in scores for one continuous variable is roughly the same at all values of another continuous variable.

TRUE

The assumption of homoscedasticity is that for any given pair of variables the variability in scores for one continuous variable is roughly the same at all values of another continuous variable. In other words when the data are homoscedastic the accuracy of prediction based on the regression line will be consistent across the full range of both the variables.

19) With univariate analyses, homogeneity of variances is assessed statistically with Levene's test.

TRUE

In the univariate case, the homogeneity of variances is assessed statistically with Levene's test. This statistic provides a test of hypothesis that the samples come from populations with the same variances. If the observed significance level is small (that is, p<0.05) one should reject the null hypothesis that the variances are equal.

20) In multivariate situations, homoscedasticity can be assessed statistically by using Box’s M test for equality of variance-covariance matrices.

TRUE

In multivariate situations, homoscedasticity can be assessed statistically by using Box’s M test for equality of variance-covariance matrices. This test is used to evaluate the hypothesis that the covariance matrices are equal. If the observed significance level is small (that is, p<0.05) one should reject the null hypothesis that the covariance matrices are equal.

21) The main purpose for screening data prior to conducting a multivariate analysis is to deal with the accuracy of the findings.

FALSE

Screening data prior to conducting a multivariate analysis requires univariate screening.

22) Another purpose for screening data is to enter missing data and assess the effect of and ways to deal with complete data.

TRUE