D2.2 Determine and compare the theoretical and experimental probabilities of two independent events happening and of two dependent events happening.
Activity 1: Determining Theoretical and Experimental Probabilities
Using a deck with the following cards:

- determine the theoretical and experimental probabilities (10 trials and 20 trials) of drawing two consecutive kings if the first card drawn is put back in the deck (independent events);
- determine the theoretical and experimental probabilities (10 trials and 20 trials) of drawing two consecutive kings if the first card drawn is not put back in the deck (dependent events).
Pose the following questions:
- Is the probability of drawing two consecutive kings higher in independent events or in dependent events? Why is this?
- How can you compare the experimental probabilities to the theoretical probabilities? How does the number of trials affect this comparison?
Activity 2: Compare Theoretical and Experimental Probabilities
Here are the theoretical and experimental probabilities (10 trials and 20 trials) of independent events. The activity consisted of determining the probabilities of drawing 2 consecutive blue tokens if the first token is put back in the bag. The bag contained 2 yellow tokens, 2 green tokens and 6 blue tokens.
Theoretical Probability

Experimental Probability
Relative Frequency Tables
Draw 2 blue tokens in a row and put the first one drawn back into the bag (10 trials).
Results of the experiment: Y, BY, G, BB, BG, G, BB, Y, BY, BB.
Combination of Tokens | Counting | Frequency | Frequency (Fraction) | Frequency (Percentage) |
---|---|---|---|---|
Y |
|| |
2 | \(\frac{2}{10}\) | \(\displaylines{\begin{align} \frac{2}{10} &= 2 \div 10 \\ &= 0.2 \\ &= 20 \% \end{align}}\) |
BY |
|| |
2 | \(\frac{2}{10}\) | \(\displaylines{\begin{align} \frac{2}{10} &= 2 \div 10 \\ &= 0.2 \\ &= 20 \% \end{align}}\) |
G |
|| |
2 | \(\frac{2}{10}\) | \(\displaylines{\begin{align} \frac{2}{10} &= 2 \div 10 \\ &= 0.2 \\ &= 20 \% \end{align}}\) |
BB |
||| |
3 | \(\frac{3}{10}\) | \(\displaylines{\begin{align} \frac{3}{10} &= 3 \div 10 \\ &= 0.3 \\ &= 30 \% \end{align}}\) |
BG |
| |
1 | \(\frac{1}{10}\) | \(\displaylines{\begin{align} \frac{1}{10} &= 1 \div 10 \\ &= 0.1 \\ &= 10 \% \end{align}}\) |
Total | 10 | 10 | \(\frac{10}{10}\) | 100 % |
Note: In the tables, you must write the colour of the token drawn each time a first draw is made. If, on the first draw, a blue token is obtained, a second draw must be made. If a token of another colour is drawn, there is no need to draw another token, because it is already lost.
Draw 2 blue tokens in a row and put the first one drawn back into the bag (20 trials).
Results of the experiment: BB, Y, G, BB, BY, G, BG, Y, BB, Y, BG, BY, G, BB, G, BY, BB, Y.
Combination of Tokens | Counting | Frequency | Frequency (Fraction) | Frequency (Percentage) |
---|---|---|---|---|
BB |
|||| ||| |
7 | \(\frac{7}{20}\) | \(\displaylines{\begin{align} \frac{7}{20} &= 7 \div 20 \\ &= 0.35 \\ &= 35 \% \end{align}}\) |
Y |
|||| |
4 | \(\frac{4}{20}\) | \(\displaylines{\begin{align} \frac{4}{20} &= 4 \div 20 \\ &= 0.2 \\ &= 20 \% \end{align}}\) |
BG |
|| |
2 | \(\frac{2}{20}\) | \(\displaylines{\begin{align} \frac{2}{20} &= 2 \div 20 \\ &= 0.1 \\ &= 10 \% \end{align}}\) |
G |
|||| |
4 | \(\frac{4}{20}\) | \(\displaylines{\begin{align} \frac{4}{20} &= 4 \div 20 \\ &= 0.2 \\ &= 20 \% \end{align}}\) |
BY |
||| |
3 | \(\frac{3}{20}\) | \(\displaylines{\begin{align} \frac{3}{20} &= 3 \div 20 \\ &= 0.15 \\ &= 15 \% \end{align}}\) |
Total | 20 | 20 | \(\frac{20}{20}\) | 100 % |
Source: translated from En avant, les maths!, 7e année, ML, Données, p. 5-8.
Ask students to analyze and compare the theoretical and experimental probabilities of getting two consecutive blue tokens. Here are some sample questions to ask them to extend their reasoning:
- If you compare the experimental probability of getting two consecutive blue tokens after 10 trials with the probability after 20 trials, what do you notice?
- What do you notice when you compare the experimental probability of 10 trials and 20 trials with the theoretical probability?
- What do you notice if you add or compare the sum of the probabilities of the outcomes for each of the relative frequency tables with the tree diagram?
Encourage students to determine the theoretical and experimental probabilities (10 trials and 20 trials) of drawing two blue tokens if the first token drawn is not put back into the bag (dependent events). Ask them to compare the probabilities obtained with those previously calculated for the independent events.