Why Do Regression Models Never Fit Perfectly in IB Maths?
Many IB Mathematics: Applications & Interpretation students feel uneasy when a regression curve does not pass through all data points. After learning exact functions earlier in their education, seeing scatter points spread around a line or curve can feel like failure — as if the model is “wrong.”
IB includes regression precisely to challenge this idea. Regression models are approximations, not exact rules. A perfect fit is usually a warning sign, not a success.
What a Regression Model Is Actually Doing
A regression model finds a function that best represents the overall trend in data.
It does not aim to match every point. Instead, it balances deviations above and below the model to minimise overall error. IB expects students to understand that regression describes typical behaviour, not exact outcomes.
This distinction is central to Applications & Interpretation.
Why Real-World Data Is Never Perfect
Real data includes variability.
This comes from:
- Measurement error
- Natural variation
- Uncontrolled external factors
- Simplifying assumptions
IB expects students to recognise that no realistic dataset will align perfectly with a single function. A model that fits too well may be overfitted or unrealistic.
Why Students Expect a Perfect Fit
Earlier maths education often rewards exact relationships.
Students become used to functions that pass through every point. Regression breaks this expectation deliberately. IB wants students to move from certainty to statistical reasoning, where variability is normal and informative.
