Module 10 Knowledge Check
5 questions. Pass mark 4 out of 5. Answer every question before checking the answer key below, then retry after reading the feedback.
1. Fitting smf.ols to a 0/1 outcome can produce…
- A. only values between 0 and 1
- B. predictions outside [0, 1]
- C. an error
- D. identical results to logistic regression
2. Logistic regression models the log-odds because it…
- A. is easier to compute
- B. is unbounded, so a line never predicts an impossible probability
- C. is easier to interpret
- D. requires fewer assumptions
3. Writing smf.ols where you meant smf.logit…
- A. raises an error
- B. silently fits a linear model and returns a full summary
- C. defaults to logistic
- D. produces odds ratios
4. An odds ratio of 2.45 with an observed rate of 0.497 means…
- A. churn is 2.45 times as likely
- B. the odds are multiplied by 2.45, which overstates the risk ratio
- C. the risk rises by 2.45 points
- D. the effect is not significant
5. A model scoring 0.9415 accuracy on a 5.9% positive rate while catching 0 of 117 positives shows that…
- A. the model is excellent
- B. accuracy is nearly meaningless on imbalanced outcomes
- C. the coefficient must be wrong
- D. the threshold should be raised
---
Answer Key and Explanations
Check these only after attempting every question.
1. B — predictions outside [0, 1]
The lesson's linear model predicted −0.088 and −0.510 at longer tenures.
2. B — is unbounded, so a line never predicts an impossible probability
Probability is bounded in [0, 1]; log-odds runs from minus infinity to plus infinity.
3. B — silently fits a linear model and returns a full summary
Nothing warns you, and every coefficient is then on a different scale from the one you think.
4. B — the odds are multiplied by 2.45, which overstates the risk ratio
Odds ratios and risk ratios agree only when the outcome is rare, and 0.497 is not rare.
5. B — accuracy is nearly meaningless on imbalanced outcomes
A rule that always predicts the majority class achieves exactly the same accuracy.
Practical Check
Apply this module to your own work: complete the module activity for *Logistic Regression and Binary Outcomes*, then write one sentence naming what your result shows and one naming what it does not.
Strong Answer Pattern
A strong answer names the task, the evidence used, the check performed, and the remaining limitation. It avoids "proved", "guaranteed", or "always" unless the evidence genuinely supports it.