Home > Department of Media and Communications > Research > EU Kids Online > Best Practice Guide > FAQ 20: How do you adjust data collection methods for different age groups?


FAQ 20: How do you adjust data collection methods for different age groups?

What's the issue?

It is important to separate what you want to ask children from how you ask it. How you ask it must depend on how old they are - on their competence to understand what you are asking, and to express themselves in their reply.
If children think a question is too old for them, they will still try to give an answer, but the answer may be meaningless. If they think a question is too young for them, they will become bored or may give a silly answer deliberately.
This issue is important in ensuring your data are reliable and valid. But it is also important to ensure that child participants are treated with respect.

Common practice

  • Researchers working with children generally conduct a pilot study, putting their questions to the age group they wish to research. Only with careful piloting of questions (for a survey or interview) can you be sure that the children understand your question and can express their answer.
  • If qualitative methods are used, the language and approach can be adjusted in the research situation, provided the researcher is experienced in working with children.
  • For surveys, younger children will require a simpler version of the questions and a shorter questionnaire - 20 minutes is a long time to sit still for a seven year old. Most researchers don't ask questions about time use to children younger than about nine years old.
  • Many researchers would use the age categories used by schools as a guide to maturity (elementary children, primary children, secondary/high school, etc), as these categories inform social norms of in/dependence.
  • Straightforward questioning may be supplemented with prompts and stimulus material (Bragg, 2007). Some examples include:

    o controversial or representative statements to initiate reactions: e.g. McCallum et al used four 'statements cards' about learning as prompts with children aged 6 and 11 (McCallum, Hargreaves, & Gipps, 2000).

    o focus groups may use colour cards to access feelings where (Bragg, 2007) different colours evoke different emotions, and can be interesting as a way to talk about positive and negative aspects of, for instance, a project (De Bono, 2000).

    o timelines - children draw or get a timeline and mark on it the ups and downs of a project, a period of time, their own lives, etc. These could take the form of 'confidence lines' that show how a person's confidence has changed over the course of a project, or what they can do afterwards that they could not do before (Bragg, 2007).

    o ranking exercises: children may be given a set of cards or photographs of activities or issues to rank in order of importance (ibid).

Pitfalls to avoid

  • It is tempting to treat children of different ages in the same group - for example, conducting a focus group with children aged 7-10. But this is an inappropriate group - the younger children will be intimidated and the older ones may feel insulted.
  • The cognitive capacity of younger and older children must be considered carefully - don't ask younger children questions containing double negatives (e.g. 'do you agree or disagree that it is a problem that some children can't access the internet?'); ethically, it is important not to introduce 'adult' ideas (e.g. of images of sexual violence) to children who have not already experienced these in their daily lives.
  • Remember that, while teenagers will probably understand your questions easily, they are very sensitive to the presence of peers - it may be better to interview them alone.

Further resources

Bragg, S. (2007). Consulting Young People: A Review of the Literature Creative Partnerships. London.

The Methodological Issues Report D4.1 contains an extended discussion of when methods developed for adults can be adapted for children and when distinct methods should be developed (Lobe et al., 2007).
See the Portuguese (Ponte & Malho, 2008) questionnaire on the EU Kids Online website for an example of a child-friendly questionnaire with pictures.

Examples of good practice

In my research, prior meetings with teachers indicated that questionnaires for the younger children should be very simple and not include open questions, since at this stage they were able to read but still struggled with their writing skills. For the questionnaires for the younger children, each page included one question visually aided by drawings. Open questions included on the 4th year questionnaires were excluded from the 1st year's. In the case of 1st year children, the teacher explained the task and read the questions out loud, waiting for everyone to answer. It was difficult to keep them from shouting their answers and making side comments, but overall they seemed to be concerned that their answers were not copied - they warned each other not to do that. For group interviews, the children, particularly the younger ones, were very enthusiastic and keen to talk about cartoons, but some fourth-grade groups were slightly reluctant and made clear attempts to distance themselves from a genre they said was for younger children - stressing that they did not see cartoons, and that they watched other programmes like sports or soap-operas; or even refusing to comment at all on the clips shown, like the afternoon group of girls. From these reactions it seemed that to ask them to expose themselves in front of an adult and their peers by displaying any knowledge of, and thus admitting watching, the genre raised questions of status. Thus, I opted to use different task-oriented research techniques that would allow different ways of contributing. By asking the children to create and criticise a cartoon the emphasis was not on their viewing habits but on their creative competencies. (Sofia Leitão, Portugal)
While for young children, we asked simple questions about the importance of the internet in their lives (what do you like or not like about it? would you miss it?), older children should not be underestimated. In focus groups with 15-19 year olds, the UK Children Go Online project asked questions such as, "Now that the internet is here and part of your life, what difference would it make if you no longer had access yourself?", "What difference would it make if the internet disappeared altogether? Would things be better or worse?", "Do you think we pay too much attention to computers in our society? Do we overrate the internet and how it can change things?", "What about those left out, those people who don't have internet access? Why might they not have or not want to have internet access?, Do you think they're missing out on something? What consequences does it have for them?"
(Sonia Livingstone, UK)

A researcher's mistake

A common mistake in research where children are involved is to take something which is designed for adults and use it, sometimes in a modified way, sometimes unchanged, to research children. An example of how this can lead to serious errors comes from the long-term research project, 'Children and Television in Iceland' (Broddason, 2003). To measure TV viewing, the research used a diary listing the programmes of the three biggest TV stations in Iceland from Monday to Sunday in the week before the survey. Respondents could indicate whether they had watched a particular programme. Beforehand, the researchers' main concern was whether the kids would have difficulties in remembering what they had watched and understand the format of the diary. As it turned out though, this set-up proved itself very well. An unexpected problem turned up, however, in the form of a painful mistake. The diary was modelled on a diary intended for adults, and so the children's programme (starting at 09:00 on the two biggest channels) on Saturday and Sunday mornings was simply omitted and the diary started at 12:00 noon. Despite countless preparatory meetings where at least a dozen individuals looked at the diary and a small pilot study with young children, this problem was not discovered until after the data had been collected.
(Kjartan Olafsson, Iceland)