Why students lose marks in statistics and it is always the conclusion that is wrong not the calculation
Statistics is the only topic in IB AA HL where a student can execute every calculation perfectly and still receive zero marks — because the conclusion contradicts the working. A student computes a p-value of 0.03, compares it to a significance level of 0.05, and concludes do not reject the null hypothesis. The arithmetic is correct. The statistical reasoning is backwards. In 25 years of teaching, I have seen this error across every level of student, every year. The problem is not mathematics. It is that students learn the mechanical steps of hypothesis testing without internalising what the p-value actually measures. A small p-value is evidence against the null hypothesis. These worksheets connect the calculation to the meaning.
Recognition Training
Probability & Discrete Distributions
Conditional probability, Bayes' theorem, and the binomial distribution. The most reliable source of errors: confusing P(A given B) with P(B given A). These are different quantities related by Bayes' theorem and treating them as equal is one of the most common probabilistic errors in IB examinations.
Normal Distribution & Continuous Probability
Standardising to the Z-distribution and working backwards from a probability to find an unknown mean or standard deviation. The most commonly lost mark: using the wrong tail when computing inverse normal values.
Hypothesis Testing & Confidence Intervals
Setting up hypotheses, computing a test statistic, comparing the p-value to the significance level, and stating the conclusion in context. The non-negotiable requirement: the conclusion must reference the original real-world claim, not abstract statistical language.
The 4 Patterns Behind Every Lost Mark
Reversing the p-value comparison
A p-value smaller than the significance level means reject the null hypothesis. Students who see a small p-value and conclude do not reject have reversed the logic. Write the comparison explicitly every time and never skip the comparison step.
Confusing conditional probability directions
P(disease given positive test) and P(positive test given disease) are different probabilities. Confusing them -- the base rate fallacy -- produces numerically plausible but logically wrong answers. Bayes' theorem is the tool for converting between them.
Stating the conclusion without context
Reject the null hypothesis is not a complete conclusion. The conclusion must reference the original claim in context. Every word in the contextual conclusion earns marks that a bare statistical statement does not.
Using the wrong tail in inverse normal calculations
The normal distribution is symmetric about its mean. Students who look up a probability and take the result directly without considering which tail the question asks for produce inverse normal values in the wrong half of the distribution.
The Full Diagnostic Path
- 50 original exam-style questions across 4 sections
- Full worked solutions with M1/A1/R1 IB mark scheme
- Mistake analysis on every question
- Sections: probability and conditional probability, binomial and normal distributions, hypothesis testing, confidence intervals
- IB Examiner commentary per section
Need one-to-one help with probability and statistics?
Book a session with Abdul Wadood and work through the techniques that are costing you marks.