Given the following output from the LOGISTIC procedure: Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate Intercept 1 0.1119 0.0304 24.8969 <.0001 - MBA 1 0.9699 0.0385 633.6092 <.0001 0.2074

The correct answer is A. Greatest: MBA Least: DOWN_AMT. Option A is correct because relative importance is evaluated using the standardized estimates (last column) for variables that are statistically significant (p < 0.05). HOME is excluded immediately - its p-value of 0.2035 exceeds α = 0.05, leaving only MBA, DOWN_AMT, and TERM. Am

Logistic Regression

Question

Given the following output from the LOGISTIC procedure:
Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate
Intercept 1 0.1119 0.0304 24.8969 <.0001 -
MBA 1 0.9699 0.0385 633.6092 <.0001 0.2074
DOWN_AMT 1 0.000072 1.39e-5 448.5986 <.0001 -0.2861
TERM 1 -0.5229 0.1143 24.2343 <.0001 -0.0308
HOME 1 -0.00402 0.00317 1.6168 0.2035 -0.0114
Which variables, among those that are statistically significant at an alpha of 0.05, have the greatest and least relative importance in the fitted model?

Options

AGreatest: MBA Least: DOWN_AMT
BGreatest: MBA

How the community answered

(33 responses)

A
70% (23)
B
30% (10)

Explanation

Option A is correct because relative importance is evaluated using the standardized estimates (last column) for variables that are statistically significant (p < 0.05). HOME is excluded immediately - its p-value of 0.2035 exceeds α = 0.05, leaving only MBA, DOWN_AMT, and TERM. Among these, MBA has the highest standardized estimate (+0.2074), making it the most positively influential predictor, while DOWN_AMT has the most negative standardized estimate (−0.2861), placing it at the opposite extreme - the least (in the direction of reducing the outcome). TERM falls in between (−0.0308), so it is neither greatest nor least.

Why distractors fail: Option B is incomplete - naming only the greatest without identifying the least doesn't fully answer the question, and any choice that includes HOME is invalid since HOME doesn't meet the significance threshold.

Memory tip: Think "signed ranking matters" - the standardized estimate column tells you both direction and magnitude. The variable closest to the top of the signed scale is "greatest," the one closest to the bottom is "least." Always eliminate non-significant variables first before comparing.

Topics

#statistical significance#parameter estimation#model interpretation#logistic regression output

Community Discussion

No community discussion yet for this question.

Full A00-240 Practice