Sorry, this activity is currently hidden
Topic outline
-
-
Video transcript
Welcome to this short ‘insights video’ where we are going to look at some of the misunderstanding that learners have around the Poisson distribution. And whether, or not, they can use the Poisson distribution as a method to predict outcomes in a situation, or particular events happening over a period of time.
There are two sets of criteria learners need to understand and apply to see if events can be predicted using this method.
The first is looks at the situation itself and asks questions about the kind of events that are occurring.
And the second is a mathematical test.
If both sets of conditions are true then the situation, or events, can be predicted using this method.
First, looking at the kind of events…
Events occur at a constant rate’. What are constant rate events?
Learners sometimes struggle to understand what this means and real world examples are a good way of helping them.
Looking at the number of traffic accidents on a road over a period of time, or the number of defects in a material roll, per metre, are both examples of constant rate events.
Whereas predicting the number of customers going into a coffee shop might seem to go in at constant rate, but when you think about it, the results would be very different at say coffee time, or lunchtime compared to the middle of the afternoon, so this would not be suitable events.
Then, are events ‘singular’ ?
A good example is predicting the number of light bulbs that will fail in a factory, each bulb failure can only happen once. These are therefore singular events.
How about, are the events ‘independent’ of each other? For this, maybe think about an insurance company and the claims the office receives over a period of time. The claims from different customers at different times are going to be independent of each one another.
Finally learners need to mathematically test the situation by looking at the mean and variance of the data they are given. For the Poisson distribution to be valid the mean and the variance of the data set have to be the same.
Learners don’t always realise the importance of this. If the mean is not the same as the variance in the data, then the formula will not give accurate outcomes.
I hope this short insights video on helping learners decide if the Poisson is the right method for modelling outcomes for a given situation over a period of time.
Thank you.
-
