WhatsApp us
IB DP Maths17 Nov 2026

IB Maths AA SL Statistics: Outliers, Standard Deviation, Correlation and the Two Regression Lines

AA SL statistics mixes quick summary-statistics marks with longer bivariate questions on Pearson's r and regression. What the topic covers, four worked examples (IQR outliers, scaling data, y on x and x on y), the interpretation marks students drop, and a self-check set.

IB Maths AA SL Statistics: Outliers, Standard Deviation, Correlation and the Two Regression Lines

AA SL statistics is the data half of Topic 4, Statistics and probability, which the IB guide gives 27 of the course's 150 hours. It covers sampling, summary statistics (mean, median, quartiles, standard deviation and outliers) and bivariate data: scatter diagrams, Pearson's correlation coefficient r, and regression lines of y on x and of x on y. Most of the calculation is done on the GDC in Paper 2; the marks that separate grades come from choosing the right line and interpreting the answer in context.

This guide works through four examples with every step, then lists the slips our tutors see and gives you a short self-check. For course-matched help, see our AA SL course support.

What the topic includes

AreaWhat you should be able to do
SamplingName and compare simple random, convenience, systematic, quota and stratified sampling; spot bias
Presenting dataHistograms, cumulative frequency graphs, box-and-whisker plots
Summary statisticsMean, median, mode, range, quartiles, IQR, variance and standard deviation (GDC)
OutliersA value more than 1.5 × IQR below Q1 or above Q3
Scaling dataThe effect of adding or multiplying every value by a constant
Bivariate dataScatter diagrams, Pearson's r, regression of y on x and x on y, prediction

Probability is the other half of Topic 4; the sibling guide covers the probability rules that come next in Topic 4. To see how both halves fit with the other four topics, read how Topic 4 fits the rest of the AA SL course.

Worked example 1: is 41 an outlier?

Twelve students scored these marks on a quiz:

12, 15, 17, 18, 18, 20, 21, 22, 24, 25, 27, 41

Median. With 12 values, the median is the mean of the 6th and 7th: (20 + 21)/2 = 20.5.

Quartiles. Split the data into a lower half (first six values) and an upper half (last six). This is the method your GDC uses.

  • Q1 = median of 12, 15, 17, 18, 18, 20 = (17 + 18)/2 = 17.5
  • Q3 = median of 21, 22, 24, 25, 27, 41 = (24 + 25)/2 = 24.5
  • IQR = 24.5 − 17.5 = 7

Fences. Q3 + 1.5 × IQR = 24.5 + 10.5 = 35, and Q1 − 1.5 × IQR = 17.5 − 10.5 = 7.

41 is above 35, so 41 is an outlier. Nothing is below 7.

Why it matters. The mean is 21.7 (3 s.f.) with the outlier and 19.9 without it, while the median only moves from 20.5 to 20. That is why the median and IQR are the safer summary when a data set has an outlier. Questions often ask you to say this in a sentence.

Worked example 2: changing the units

The midday temperature in a city over eight days has mean 24 °C and standard deviation 3.5 °C. Find the mean and standard deviation in °F, using F = 1.8C + 32.

  • Mean: every value is multiplied by 1.8 and then has 32 added, so the mean does the same: 1.8 × 24 + 32 = 75.2 °F.
  • Standard deviation: adding 32 shifts every value equally and does not change the spread. Multiplying by 1.8 stretches it. So the standard deviation is 1.8 × 3.5 = 6.3 °F.

The rule to remember: adding a constant changes the mean but not the standard deviation; multiplying by a constant changes both. Variance is multiplied by the square of the constant.

Worked example 3: correlation and the two regression lines

Seven students recorded the hours x they spent revising and their test score y (%).

x (hours)23568911
y (%)41485260637176

(a) Find r and describe the correlation. Enter the lists in the GDC and run a linear regression: r = 0.988 (3 s.f.). This is a strong positive linear correlation.

(b) Find the regression line of y on x. y = 3.81x + 34.8 (3 s.f.).

(c) Estimate the score of a student who revised for 7 hours. Use the y on x line, because you are predicting y from x: y = 61.4%. Use the unrounded values stored in the GDC; with the rounded coefficients you get 61.5, which can cost an accuracy mark.

(d) Estimate the revision time of a student who scored 45%. Now you are predicting x from y, so you need the x on y line. Run the regression with the lists swapped: x = 0.256y − 8.75, so x = 2.77 hours. Rearranging the y on x line instead would give 2.68 hours. That is the wrong line, and it loses the marks.

(e) Comment on predicting the score for 25 hours of revision. The y on x line gives about 130%, which is impossible. 25 hours is far outside the data (2 to 11 hours), so this is extrapolation and is not reliable. Predictions inside the data range (interpolation) are reasonable when r is strong.

Worked example 4: where the two lines meet

For a data set, the regression line of y on x is y = 1.5x + 4 and the regression line of x on y is x = 0.5y + 1. Find the mean of x and the mean of y.

Both regression lines pass through the mean point (x̄, ȳ). So solve them together:

x = 0.5(1.5x + 4) + 1 = 0.75x + 3, so 0.25x = 3 and x̄ = 12. Then ȳ = 1.5 × 12 + 4 = 22.

Check in the second line: 0.5 × 22 + 1 = 12. A quick extra check, not a syllabus formula: the product of the two gradients, 1.5 × 0.5 = 0.75, equals r², so r = √0.75 ≈ 0.866. That is positive, as both gradients are.

Using the GDC well

  • Clear old lists before you start, and check the number of data points the calculator reports.
  • For standard deviation, use the value labelled σx (which divides by n). In IB exams, a given data set is treated as the population.
  • Store the regression coefficients so predictions use full accuracy.
  • Turn on the display of r and r² if your model hides them.

Our guide to GDC statistics menus and other calculator skills shows these steps for the common calculator families.

Mistakes our tutors see

  1. Using the y on x line to predict x. Swap the lists and find x on y, as in example 3(d).
  2. Saying "no correlation" when r is near 0. r measures linear correlation only. The data could still follow a curve.
  3. Claiming cause. A strong r shows association, not that revising more causes higher marks. Other factors may be involved.
  4. Extrapolating without comment. If x is outside the data range, say the estimate is unreliable and why.
  5. Rounding coefficients early, then losing the final accuracy mark.
  6. Changing the standard deviation when adding a constant. Only multiplying changes the spread.
  7. Leaving the description incomplete. "Strong positive linear" earns more than "positive".

Self-check

  1. For 4, 7, 7, 9, 10, 12, 30, find Q1, Q3 and the IQR, and decide whether 30 is an outlier.
  2. A data set has mean 12 and standard deviation 3. Each value is doubled and then 5 is added. Find the new mean and standard deviation.
  3. Describe r = −0.92.
  4. The y on x line is y = 2.4x + 13 and x̄ = 6. Find ȳ.

Answers. 1: Q1 = 7, Q3 = 12, IQR = 5; the upper fence is 19.5, so 30 is an outlier. 2: mean 29, standard deviation 6. 3: strong negative linear correlation. 4: ȳ = 2.4 × 6 + 13 = 27.4, because the line passes through the mean point.

Exam technique

  • Paper 1 (no calculator) tests the ideas: reading a box plot, the 1.5 × IQR rule, scaling data, or using the mean point.
  • Paper 2 (GDC) gives a table and asks for r, a line and a prediction. Write the equation you used before giving the number.
  • Interpretation lines are marks. End each answer with a sentence in the context of the question, for example "a student who revises one more hour scores about 3.8 percentage points more, on average."

A note on the new course

The IB has announced a new AA course for first teaching in August 2027, with first assessment in May 2029. Bivariate data is one of the areas being removed from AA SL in that course. Students sitting exams up to November 2028 follow the current guide, so they still need everything above.

Where this leads

Regression is also a natural tool for the exploration, and our explainer on how the exploration is marked shows how to use it with care. AI SL students meet Spearman's rank and non-linear models too; our sibling post on Spearman's rank and regression for AI SL covers that course. If you are moving up a level, our guide to what AA HL adds to the syllabus lists the extra content.

How Maths Bodhi helps

In the free demo, our tutor gives a short data question and listens to how the student explains r and the choice of line. Most lost marks in this topic are in the words, so we practise written interpretation as much as the GDC steps. Progress is tracked through past-paper-style questions with written method. See our statistics tutoring for how sessions run.

Home tuition is available in Gurgaon (Gurugram). Students elsewhere in India and abroad learn with us online. You can book a free demo class.

Written and checked by the Maths Bodhi team. Facts checked against official IB sources, November 2026.

Frequently asked questions

What statistics is in IB Maths AA SL?
Sampling methods, data presentation, summary statistics including quartiles, standard deviation and outliers, the effect of scaling data, and bivariate data: scatter diagrams, Pearson's r and regression lines of y on x and x on y.
When do I use the regression line of x on y?
Use it when you are estimating x from a given value of y. The y on x line is only for estimating y from x, and rearranging it gives a different, incorrect answer.
How do I find an outlier in IB maths?
Find Q1, Q3 and the IQR. A value is an outlier if it is more than 1.5 × IQR below Q1 or more than 1.5 × IQR above Q3.
What does Pearson's r tell me?
It measures the strength and direction of a linear relationship, from −1 to 1. A value near 0 means no linear correlation, but there could still be a curved relationship, and a strong r never proves cause.
Do I calculate r by hand in AA SL?
No. In AA SL you find r and the regression lines with the GDC, so these calculations sit in Paper 2. Paper 1 can still ask you to interpret r or use the mean point.
Is bivariate data still in AA SL?
Yes for the current course, which applies to exams up to November 2028. The IB has announced that the new AA course, first assessed in May 2029, removes bivariate data from AA SL.
IB maths AA SL statisticsregression and correlation IBPearson's rIB maths GDC statisticsIB maths tutor Gurgaon

Related guides

Keep exploring this maths topic

IB DP Maths17 Nov 2026

IB Maths AA SL Topics and Syllabus: The Five Topics, Teaching Hours, What Each Paper Tests and a Tick-Off Checklist

Every IB Maths Analysis and Approaches SL topic for exams in 2027 and 2028, with the IB's suggested teaching hours, what Paper 1 and Paper 2 expect, a sub-topic checklist to tick, and what changes for students who start DP in August 2027.

IB maths AA SL topicsIB maths AA SL syllabusIB maths AA SL papersAA SL checklist

By Maths Bodhi Team

Read guide

IB DP Maths17 Nov 2026

IB Maths AI SL Normal Distribution and Descriptive Statistics: Quartiles, Outliers, Standard Deviation and Normal Probabilities

How to summarise a data set the way AI SL markschemes expect, then move on to normal distribution probabilities and inverse normal questions on the GDC. Four worked examples, the usual traps and a self-check.

IB maths AI SL normal distributionIB maths AI SL descriptive statisticsinverse normalbox plots and outliers

By Maths Bodhi Team

Read guide

IB DP Maths17 Nov 2026

IB Maths AI SL Regression and Spearman's Rank: Pearson's r, the Regression Line and When to Use Each Coefficient

Pearson's r measures how close data is to a straight line; Spearman's rank measures whether one variable consistently rises or falls with the other. How to calculate both on the GDC, handle tied ranks, use the regression line safely and interpret everything in context.

IB maths AI SL regressionSpearman's rank correlationPearson's correlation coefficientIB maths AI SL bivariate data

By Maths Bodhi Team

Read guide

IB DP Maths17 Nov 2026

IB Maths AA SL Probability Questions: Conditional Probability, Independent Events, Venn and Tree Diagrams

AA SL probability questions turn on two ideas, conditional probability and independence, shown through Venn diagrams, tables and tree diagrams. The formulae, four original worked examples including a "without replacement" tree and a test-result question, the traps, and a self-check set.

IB maths AA SL probabilityconditional probability IBindependent eventstree diagrams IB maths

By Maths Bodhi Team

Read guide

IB DP Maths17 Nov 2026

IB Maths SL Geometry and Trigonometry: 3D Solids, Bearings, and the Sine and Cosine Rules (AA SL and AI SL)

The geometry and trigonometry that AA SL and AI SL share: 3D distance and solids, angles in pyramids, the sine and cosine rules, bearings and angles of elevation. Which parts belong to which course, four worked examples, the ambiguous case for AA, and a self-check set.

IB maths SL bearingssine rule and cosine rule IB3D geometry IB mathsIB maths AA SL and AI SL

By Maths Bodhi Team

Read guide

IB DP Maths10 Nov 2026

How to Improve Your IB Maths AA SL Grade from a 4 to a 6: A 12-Week Plan Built on the Syllabus Hours and Paper Weights

A practical plan for IB Maths Analysis and Approaches SL students stuck at a 4. Where the marks are, which topics to fix first using the IB's own teaching hours, how to treat Paper 1 (no calculator) and Paper 2 differently, and a week-by-week 12-week plan.

IB maths AA SLimprove IB maths gradeIB maths study planIB maths Paper 1 and Paper 2

By Maths Bodhi Team

Read guide