Introduction
In many real-world problems, the number of trials in a binomial distribution is so large that calculating exact binomial probabilities becomes impractical. The normal distribution provides a powerful and examinable approximation in these cases. The 9709 syllabus requires you to know precisely when the approximation is valid, which normal distribution to use, and how to apply the all-important continuity correction that adjusts for the fact that you are approximating a discrete distribution with a continuous one.
Core Concept
Suppose , where . As grows large, the binomial distribution becomes increasingly bell-shaped and symmetric, and begins to resemble a normal distribution. This happens provided both tails of the distribution have enough probability mass — which the conditions and ensure.
The approximation: When the conditions are met, we use
with and .
The continuity correction is essential. Because is discrete (it takes integer values), the event corresponds to a unit-wide interval. When we switch to the continuous variable , we replace each integer with the interval . This is not optional — failing to apply it in an exam will cost marks.
| Binomial event | Approximating normal event |
|---|---|
A useful mental check: inequalities that include move the boundary towards by 0.5; inequalities that exclude move the boundary away from by 0.5.
Unlock the full The Normal Distribution note with Nova
You're reading the preview. Unlock the complete note — every worked example, examiner pitfall and practice question — plus 24/7 AI tutoring from Nova that teaches directly from these notes.