Lesson 7: Estimating probabilities from data
Estimating conditional probabilities from rescue data
SkillsEstimating probabilities from dataTwo-way tablesIndependent events
Worked example
A Sunshine Coast surf lifesaving club recorded 500 rescues over one summer, noting whether the person rescued was an overseas visitor () and whether they had been swimming outside the flags ().
| Rescues | Outside the flags () | Between the flags () | Total |
|---|---|---|---|
| Overseas visitor () | 112 | 28 | 140 |
| Not an overseas visitor () | 198 | 162 | 360 |
| Total | 310 | 190 | 500 |
(a) Estimate the probability that a rescued swimmer was outside the flags.
(b) Estimate and .
(c) Comment on whether and appear to be independent for these rescues.
Step 1: Estimate the overall probability
Step 2: Estimate the conditional probabilities
and .
Each conditional estimate uses only its own row: the 140 overseas visitors, then the 360 others.
Step 3: Compare the estimates
If were independent of , both would be close to 0.62. Instead, 0.8 and 0.55 are very different.
Step 4: Conclude in context
and do not appear to be independent: rescued overseas visitors were much more likely than other rescued swimmers to have been outside the flags.
With 140 and 360 rescues in the two groups, a difference of 0.25 is far larger than random variation would explain.
Answer
(a) 0.62 (b) and (c) The estimates are very different, so being an overseas visitor and swimming outside the flags do not appear to be independent among rescued swimmers.
Try a similar one
A fresh question of the same kind, checked as you answer, with hints and a full solution.
Method
Relative frequency = count in the event ÷ count in the group.
For a conditional estimate, use only the row or column of the condition.
Compare with : close suggests independence, clearly different suggests not.
Watch out for
Dividing 112 by 500, which estimates , not .
Over-claiming: these are rescues only, so the data can't give the probability that an overseas visitor needs rescuing.