Welcome to the main Logistic Regression Module quiz. We will try to hurt your poor brain with a set of 11 questions. Actually, it is not so bad… Click VIEW FEEDBACK when you’re ready to see the answers. When you want to move on just click EXIT QUIZ. Enjoy!

1. Assuming a continuous explanatory variable on the x-axis and a binary proportion on the y-axis, which of the two curves would be most likely to reflect the relationship?

A B

2. Match the statistical term with the definition.

Definition

Nagelkerke R-squared

Wald

Exp(B)

-2LL

The odds ratio - tells us how more or less likely an outcome is to occur when there is a unit change in the associated explanatory variable.

The model 'deviance' - the lower this value is the better your model is at predicting the binary outcome.

A statistic used for judging whether or not a given explanatory variable is substantially contributing to the overall model.

A psuedo r-squared statistic for estimating the proportion of variance explained by a logistic regression model.

3. What does the log-likelihood represent?

The probability of the outcome occurring for a given case A measure of the explanatory power of the model The odds ratio for a given explanatory variable

4. The following table of coefficients is the output of a logistic regression analysis which explores the factors underlying whether or not a student is deemed eligible for free school meals:How many explanatory variables are included in this model?

3 5 9 10 11

5. Using the table in the previous question, how much more likely is it that a student from a single parent family will be eligible for free school meals compared to a student not from a single parent family (assuming that the baseline category is ‘not a single parent family’)? Please type your answer to the nearest whole number.

6. It is approximately 8 times more likely that someone from the baseline SEC category ‘Routine, semi-routine or unemployed’ will be eligible for free school meals than someone from the category ‘Managerial & professional’. True or false (using the same table as before)?

True False

7. Below is a classification plot:Please tick the boxes for all statements which are true.

The data suggests that the outcome is as likely to occur as it is not to. For most cases the predicted probability of the outcome occurring is zero. Where the model states that probability of the outcome occuring is between .7 and .8, the outcome does not occur in approximately one quarter of cases. Where the probability of the outcome occurring is less than .5, the outcome never occurs.

8. The model was adjusted to explore a possible interaction between SEC and single parent family. The following output was generated:Is the interaction effect statistically significant?

Yes No Not enough data

9. The line graph below shows the mean proportion of pupils eligible for free school meal (FSM) separately by maternal education and SEC.Please tick all statements which are true.

Those from the ‘Routine, semi-routine and unemployed’ SEC category are most likely to be eligible for free school meals regardless of maternal education Of all those from the ‘Routine, semi-routine and unemployed’ SEC category, those who have a mother with a degree are least likely to be eligible for free school meals There appears to be an interaction between SEC and maternal education in relation to free school meal eligibility There does not appear to be an interaction between SEC and maternal education in relation to free school meal eligibility

10. The following output related to how well the ‘free school meals’ model fits the data (how accurately it classifies cases)Please tick all statements that apply.

The model is a perfect fit for the data The model fits the data better than the ‘baseline’ model The model classifies 86% of cases correctly. The model is more accurate when classifying those who are eligible for a free school meal.

11. When might the assumption of ‘Independent Errors’ be at particular risk of being violated?

When residuals are unevenly distributed across the range of values for an explanatory variable When all explanatory variables are categorical When cases are clustered in hierarchies

Privacy statement

This quiz is anonymous. No data which personally identifies you is collected on the quiz, and the data you provide is used solely to help us improve the delivery of our courses.

VIEW FEEDBACKwhen you’re ready to see the answers. When you want to move on just clickEXIT QUIZ. Enjoy!