New research from the Centre for Longitudinal Studies (CLS) has tested how children’s responses to surveys are affected by the way the questions are asked.
CLS researchers Kate Smith and Lucinda Platt conducted two experiments with a sample of almost 3,000 10 to 15-year-olds in advance of the age 11 survey of the Millennium Cohort Study (MCS). They assessed how the children’s answers to potentially sensitive questions varied when they were given different response categories.
In the first experiment, the children were asked how often they were bullied at school. The researchers found that children gave quite different answers when they were asked to choose from frequency response categories (for example, ‘1-3 times a week’, ‘more than 4 times a week’ and ‘a few times every week’) than when they were given descriptive categories, such as ‘not much’, ‘quite a lot’ and ‘a lot’.
The researchers suggest that children may struggle to answer questions about frequencies without clear definitions, and could take into account the intensity of their experience as well as the actual number of times an event has occurred.
In the second experiment, the children were asked to estimate how many times they had drank alcohol. They were asked to respond to an open-ended question, as well as to choose from set categories ranging from ‘0 occasions’ to ‘40 or more occasions’.
Although the responses to both types of questions were largely consistent, there was a significantly lower response rate to the open-ended question, suggesting that children find it easier to select an answer from a list and that response categories give them a better idea of what information is being sought.
The researchers emphasise the need to test survey questions to ensure the best quality data can be collected. They note that this is particularly important for surveys with children, as previous research has indicated they are more likely to misinterpret questions or respond without fully thinking through their answers.
‘How do children answer questions about frequencies and quantities? Evidence from a large-scale field test’, by Kate Smith and Lucinda Platt, was published by the Centre for Longitudinal Studies in December 2013.