Elliott Book 5

Front 1

In discriminant analysis, when would we not interpret the analysis

Back 1

if the discriminate functions estimated were not statistically significant.

Front 2

how do you calculate Wilk's lambda

Back 2

done is SPSS

Front 3

Wiilk’s lambda statistic

Back 3

is the product of the univaraite lambda for each function

Front 4

Wiilk’s lambda statistic: how do you determine the significance level

Back 4

. The significance level is estimated based on a chi square transformation of the statistic.

Front 5

If you reject Ho in Wilk's lambda it means

Back 5

If the null hypothesis is rejected, it indicates significant discrimination and you can proceed to interpret the results

Front 6

o The value of the coefficient for a particular predictor depends on

Back 6

the other predictors included in the discriminate function

Front 7

o The value of the coefficient for a particular predictor indicates

Back 7

which variable values result in large and small function values and associate them with particular groups

Front 8

o Generally, predictors with relatively large standardized coefficients are

Back 8

contribute more the discriminating power of the function, and are therefore more important

Front 9

o Examining the standardized discriminate function coefficients- given low intercorrelations between the predictors,

Back 9

one might cautiously use the magnitudes of the standardized coefficients to suggest which variable is most important predictor in discriminating between the groups. And then make the rest of the list of importance accordingly.

Front 10

o Unstandardized discriminate function coefficients-

Back 10

can be applied to the raw values of the variables in the holdout set for classification purposes.

Front 11

structure correlations is also called

Back 11

canonical or discriminat loadings

Front 12

structure correlations

Back 12

these simple correlations between each predictor and the discriminate function represent that variance that the predictor shares with the function. The greater the magnitude of structure correlation, the more important is the corresponding predictor

Front 13

develop a characteristic profile for each group by

Back 13

describing each group in terms of the group means for the predictor variables. If the important predictors have been identified, then a comparison of the group means on these variables can assist in understanding the intergroup differences.

Front 14

characteristic profile

Back 14

o It would be reasonable to develop a profile of the two groups in terms of the three predictors that seem to be the most important

Front 15

 The analysis sample is used for what?
the validation sample is used for what?

Back 15

estimating the discriminate function,
developing the classification matrix

Front 16

 Discriminate weights-

Back 16

estimated by using the analysis sample, are multiplied by the values of the predictor variables in the holdout sample to generate discriminate scores for the cases in the holdout sample. The cases are then assigned to groups based on their scores and an appropriate decision rule.

Front 17

 Classification accuracy achieved by discriminate analysis vs. accuracy obtained by chance

Back 17

 Classification accuracy achieved by discriminate analysis should be at least 25 percent greater than the obtained by chance

Front 18

two-group discriminate analysis

Back 18

When the criterion variable has two categories,

Front 19

examples of discriminate analysis

Back 19

• Examples of discriminate analysis:
o In terms of demographic characteristics, how do customers who exhibit store loyalty differ from those who do not?
o Do heavy, medium, and light users of soft drinks differ in terms of their consumption of frozen food?
o What psychographic characteristics help differentiate between price-sensitive and non price sensitive buyers of groceries?
o Do the various market segments differ in their media consumption habits?
o In terms of lifestyles, what are the differences between heavy patrons of regional department store chains and patrons of national claims?
o What are the distinguishing characteristics of consumers who respond to direct mail solicitations?

Front 20

• The coefficients, or weights (b) are estimated so that

Back 20

groups differ as much as possible on the values of the discriminate function. This occurs when the ratio of between group sum of squares to within group sum of squares is at its maximum