An AI statistics tutor for school: data, graphs and probability

Five friends earn these amounts per week from part-time jobs: 40, 45, 50, 55 and 210. Which average better describes a typical earning?

1. Mean: $\dfrac{40 + 45 + 50 + 55 + 210}{5} = \dfrac{400}{5} = 80$.
2. Median: the middle value in order is 50.
3. Interpret: the mean is pulled up by one large value (an outlier). Four of the five earn less than 80, so the median better describes a typical friend.

The tutor asks you to explain step 3 in your own words, because tests often ask which average is more appropriate, and why.

A bag has 3 red and 2 blue counters. Two are taken out without replacement. What's the probability both are red?

1. First counter red: $\tfrac{3}{5}$.
2. Second counter red, given the first was red: $\tfrac{2}{4}$ (one red has gone).
3. Multiply along the branch: $\tfrac{3}{5} \times \tfrac{2}{4} = \tfrac{6}{20} = \tfrac{3}{10}$.

Drawing the tree on the whiteboard makes "without replacement" visible: the numbers on the second set of branches change.

Tell the tutor which course you're on; we don't claim full coverage of any syllabus.

Many statistics questions end with "what does this tell you?" The tutor practises that habit: compare two data sets using an average and a measure of spread, describe correlation in context ("as hours of revision increase, test scores tend to increase"), and explain why correlation doesn't prove one thing causes another.

The SAT's Problem-Solving and Data Analysis domain includes percentages, rates, data in tables and graphs, probability and inference from samples. If you're preparing for the SAT, see our SAT Math tips and try a free practice test.

The tutor can check your calculations and reasoning, but like any AI it can make mistakes, especially with long lists of data typed in. Double-check data you enter, and follow your teacher's conventions for things like quartile methods, which vary between courses.

A scatter plot shows hours of revision against test score for 20 students, with points rising from left to right.

1. Describe the correlation: positive, because more revision hours tend to go with higher scores.
2. Describe its strength: strong if the points lie close to a line, weak if they're spread out.
3. Use a line of best fit to estimate a score for a given number of hours, but only within the range of the data.
4. Be careful with causation: the plot alone doesn't prove revision causes the higher scores, though it's a reasonable explanation.

The tutor asks you to write each statement in context, as test questions require.