Why students lose marks in Statistics and Probability and it is almost always the wrong conditional probability
Statistics and Probability is the topic where students know every formula and still answer the wrong question, because P(A given B) and P(B given A) use the same two events and look almost identical on the page. A student calculates the probability of testing positive given the disease when the question asked for the probability of having the disease given a positive test, and the arithmetic that follows is flawless for a question that was never asked. The same reversal shows up in two way tables, where dividing by the total number of students instead of the relevant row or column gives an answer that looks reasonable but answers a different question entirely. In 25 years of teaching, this single reversal accounts for more lost marks in this topic than any error in the binomial or normal distribution formulas themselves. These worksheets train students to state in words which event is given before a single number is substituted.
Recognition Training
Statistics and Probability - Mistake Analysis - Set I
Six questions on basic probability and complements, counting outcomes for two dice, conditional probability, the binomial distribution, the normal distribution, and expected value for a discrete random variable. Mistake analysis on every question, focused on identifying which quantity each formula actually produces.
Statistics and Probability - Mistake Analysis - Set II
Six questions on the addition law, testing for independence, the mean and variance of a binomial distribution, the upper tail of a normal distribution, an inverse normal problem for an unknown mean, and the variance formula for a discrete random variable. Mistake analysis on every question.
Statistics and Probability - Mistake Analysis - Set III
Six questions built around harder Paper 2 applications, including conditional probability from a two way table, cumulative binomial probability, inverse normal problems for both an unknown mean and an unknown standard deviation, Bayes theorem applied to a medical testing scenario, and the geometric distribution. Mistake analysis on every question.
The 4 Patterns Behind Every Lost Mark
Inverting Which Event Is Given in a Conditional Probability
P(A given B) and P(B given A) involve the same two events but are rarely equal, since the formula divides by the probability of the event that is given, not the event being found. Students who calculate P(positive test given disease) when the question asks for P(disease given positive test) have answered a different, related question. Writing out in words which event is given before selecting a formula prevents this reversal.
Dividing by the Grand Total Instead of the Relevant Row or Column
In a two way table, a conditional probability restricts to the row or column matching the given condition before dividing, not the total number of items in the table. Students who divide by the grand total compute an unconditional probability, which answers a different question than the one asked. Identify the given condition first, find its row or column total, then use that as the denominator.
Confusing the Upper Tail and Lower Tail in Normal Distribution Problems
A normal distribution question asking for the probability above a value requires the upper tail, found as 1 minus the cumulative probability below that value, not the cumulative probability itself. Students who read P(X > value) directly from the standard normal table without subtracting from 1 report the lower tail probability instead, which is the complement of the answer required.
Sign Errors When Solving Inverse Normal Problems for an Unknown Mean or Standard Deviation
When a probability is given as less than 0.5, the corresponding z value is negative, and when finding an unknown standard deviation, the result must always be positive regardless of the sign that appears during the algebra. Students who use a positive z value when the probability indicates a lower tail, or who report a negative value for standard deviation, have not checked their answer against what the situation actually requires.
The Full Diagnostic Path
- 75 original exam-style questions covering probability, the binomial and normal distributions, and Bayesian applications
- Full worked solutions with M1/A1/R1 IB mark scheme annotations
- Mistake analysis on every question targeting conditional probability reversals, tail confusion, and sign errors
- Sections: probability rules and conditional probability, binomial distribution, normal distribution, expected value and variance, Bayes theorem and advanced applications
- IB Examiner commentary per section on where marks are most commonly lost
Need one-to-one help with Statistics and Probability?
Book a session with Abdul Wadood and work through the techniques that are costing you marks.