Sources of Variation

Variation means differences in our data. For instance, if we were investigating the weight of apples, not every apple would weight the same. But why?

• Was it because we picked the apples at different times of their growth cycle?

• Did the apples get weighed in the same way and accurately? 

• Are all the apples the same species?

• Did some apple trees get more sunlight/fertiliser and others less?

Below are different sources of variation, and how we might manage them

Individual-to-individual variation

When the same variable is measured for different individuals there will be differences in their measurements, simply due to the fact that individuals are different. 

For Example:

A student's travel time to school. Different students are going to take different times to get to school, just because they all happen to live different distances away from school. Also they won't all use the same mode of transport to get to school either.

This can be managed by having a large and representative sample to help account for this variation.

Occassion-to-occasion variation

Repeated measurements of the same individual may vary because of changes in the variable being measured.

For example:

A person's height slightly changes across the day. Generally people are a little taller in the morning after lying down while sleeping and shorter in the evening after standing up all day.

This can be managed by ensuring data is collected at the same time and in the same way.

Measurement variation

Repeated measurements of the same variable may vary because of some unreliability in the tool used to take the measurements. 

For example: 

The weight of rugby players getting measured might not be under the same conditions. Did they all take their shoes off before getting weighed? Were the scales always used on a hard flat surface, or carpet? Were they scales reset to 0kg each time. 

This can be managed by ensuring everyone follows the same plan when their data is collected, and any measurement tools used always start from zero each time.

Induced variation

The difference in measurements of the same variable could be due to the affect of another factor or choices made.

For example:

Length of time a PE student can juggle 3 balls. If some students are given 3 tennis balls and other students are given 3 basket balls to juggle, then this choice to give them different balls may affect the length of time they can juggle.

This can be managed by ensuring all data is collected under the same conditions.