Common Mistakes
Statistics · AP · Inference for Categorical Data — Proportions
Focus
Setting Up a Test for a Population Proportion
This topic is assessed on the AP Exam.
Frequent mistakes
### 1. Inference without conditions
What it looks/sounds like: Calculator p-value or CI is reported with no plan step.
Why it happens: Conditions feel optional on technology items.
How to fix: Always state Random, 10%, and Large Counts/Normality as applicable.
### 2. Accept the null
What it looks/sounds like: Conclusion says ‘accept H₀’.
Why it happens: Hypothesis test language is informal.
How to fix: Use ‘fail to reject H₀’ or ‘evidence against H₀’.
### 3. CI misinterpretation
What it looks/sounds like: A 95% CI is described as capturing the sample statistic.
Why it happens: Interval language is about the parameter incorrectly.
How to fix: Say: we are 95% confident the interval captures the population parameter.
### 4. Correct work, wrong question
What it looks/sounds like: The algebra is fine but answers a different quantity than the stem asks for.
Why it happens: Attention locks onto the first operation instead of the final ask.
How to fix: Underline the final question; after solving, verify the boxed answer matches it.
### 5. Notation and justification gaps
What it looks/sounds like: Missing units, limit language, or ‘because’ statements AP graders expect.
Why it happens: Notation is treated as decoration rather than part of the score.
How to fix: Add one justification phrase per major step (definition, theorem, or rule).
How to avoid them
1. State hypotheses in symbols and words. 2. Check Random, 10%, Large Counts. 3. Conclude in context about the parameter.
Quick self-check
- [ ] Can write hypotheses for a proportion
- [ ] Can check inference conditions
- [ ] Can conclude without ‘accept H₀’