1. Predict before you test
Start with a simple question: if a fair coin is tossed 20 times, how many heads do you expect? Ask learners to make a prediction before running the trials. The important discussion comes afterward: did the result match exactly, and if not, does that mean the coin was unfair?
Repeat with 100 tosses and compare how the proportion changes. This introduces the idea that expected probability describes a long-run tendency, not a guarantee for every short run.
2. Build a dice frequency table
Roll a six-sided die 30 or 60 times and tally how often each face appears. Before starting, note that each face has the same theoretical probability of 1/6 on a fair die. After the trial, convert the tallies into fractions, decimals or percentages depending on the learners' level.
Compare results between groups. Different groups will usually get different patterns even though they used the same probability model.
3. Compare with and without replacement
Put the numbers 1 to 10 into a random number picker. First, allow repeats and generate five numbers. Then switch to no-repeat mode and generate five more. Ask what changed about the possible sequences.
With replacement, the same number can appear more than once and every draw starts from the same pool. Without replacement, the pool changes after each selection, so later probabilities are different.
4. Design an unequal spinner on paper
Draw a circle divided into unequal sections, such as one half red, one quarter blue and two eighths green and yellow. Ask which colour should appear most often and why. Then compare the physical spinner with a digital equal-choice picker. This highlights an important idea: random does not automatically mean every outcome has equal probability.
5. Investigate streaks
Many people expect random results to alternate more than they really do. Toss a coin repeatedly and look for runs such as three heads in a row. Discuss why a streak can occur even when each toss is independent. The previous result does not force the next toss to “balance things out”.
6. Ask better probability questions
- What outcomes are possible?
- Are they equally likely?
- What do you predict before testing?
- How many trials are enough to see a useful pattern?
- Why might two groups get different experimental results?
- What changes if selected items are removed from the pool?
Keep the activity age-appropriate
For younger learners, focus on language such as impossible, unlikely, even chance, likely and certain. Older learners can calculate theoretical probability, experimental frequency and relative frequency. A digital tool is useful for running many trials quickly, but physical coins, dice and cards can make the mechanics easier to see.