Carrying out a reliable survey
Example situation
You have been asked to carry out a survey comparing the number of people living in Scotland who prefer countryside holidays to those who prefer beach ones.
Consider the following questions on your own and then click the buttons for answers.
Question
Would it be suitable to ask ten people in total?
No. Ten would be too small a The number of participants, responses or results that are looked at in a survey, experiment or piece of research.. This is not a big enough number of people to give an accurate representation of all people in Scotland.
Question
Would it be suitable to go to a seaside town to ask people?
Possibly, but if we only went there, our data would likely be biased. Ideally there should be a balance between people living in cities, the countryside and seaside towns.
Question
Would it be suitable to go to a school and ask a selection of pupils?
Yes, but if you were wanting to find out about all people living in Scotland you would need to also interview people of other ages and in different locations.
Outcome
If you were going to carry out the survey above you would need to interview:
- a large enough number of people
- a mix of ages and genders
- people from varied locations across Scotland
The above would avoid bias of one group over another and ensure the data collected gave a true representation of the population of Scotland.
Titles
When presenting data it is important to ensure you have a title that gives an acurate indication of what is being presented.
In this example, it might include a sub-heading saying, 鈥1000 people surveyed in towns and cities across Scotland 鈥 included a balance of age and gender.'