MUED 6440 Ch. 3

Question 1

The statistical test of homoscedasticity in multivariate situations

Answer 1

Box’s M test for equality of variance

Question 2

Measured on a scale that changes smoothly over possible values rather than in steps; also referred to as interval or quantitative

Answer 2

Continuous variables

Question 3

The application of mathematical procedures to data in order to make them appear more normal

Answer 3

Data transformations

Question 4

The violation of the assumption of homoscedasticity

Answer 4

Heteroscedasticity

Question 5

Assumption that the variability in scores for one continuous variable is roughly the same at all values of another continuous variable

Answer 5

Homoscedasticity

Question 6

Tests the null hypothesis that the population is normally distributed

Answer 6

Kolmogorov-Smirnov statistic

Question 7

Degree of peakedness of a distribution; equal to zero when a distribution is normal

Answer 7

Kurtosis

Question 8

Condition when values for kurtosis are positive, indicating that the distribution is too peaked with long, thin tails

Answer 8

Leptokurtosis

Question 9

A statistical test of the homogeneity of variances

Answer 9

Levene’s test

Question 10

Assumption that there is a straight-line relationship between two variables

Answer 10

Linearity

Question 11

Statistical measure of an outlier; distance of a case from the centroid of the remaining cases where the centroid is the point created by the means of all the variables

Answer 11

Mahalanobis distance

Question 12

Cases with an unusual pattern of scores; values on individual variables may look reasonable, but the combinations of two variables produce values that look unusual or discrepant

Answer 12

Multivariate outliers

Question 13

Assumption that all variables and linear combinations of variables are normally distributed

Answer 13

Normality

Question 14

Cases with extreme values on one variable or on a combination of variables so that it distorts resulting statistics or unduly influence solutions or models; would result in an excessively large residual

Answer 14

Outliers

Question 15

Condition when values for kurtosis are negative, indicating that the distribution is too flat, with many cases in the tails

Answer 15

Platykurtosis

Question 16

Portions of scores not accounted for by the analysis; also a measure of the difference between the obtained and predicted values on the DV, therefore referred to as prediction error

Answer 16

Residuals

Question 17

The degree to which a statistical test is still appropriate to apply when some of its assumptions are not met

Answer 17

Robustness

Question 18

Degree of symmetry of a distribution; equal to zero when distribution is normal;

Answer 18

Skewness

Question 19

Cases with very large standardized scores on a single variable

Answer 19

Univariate outliers