What's the issue?
In any research with children, including that relating to media and the internet, age differences are consistently amongst the most important background factors. Reporting findings by age, charting age trends, or comparing age groups is expected by most readers, and it would be the absence of age differences, not their discovery, that would be counterintuitive, if and when it occurred. A useful principle, therefore, is to assume that each child is capable of providing valid and insightful information, provided that s/he is approached appropriately and that the data are interpreted carefully.
A range of principled or commonsense rules of thumb are evident in published accounts of research. In general two major turning points can be assumed with key adjustments in methods being made for respondents older or younger than 7-8 years, and older or younger than 11-12 years. It is worth noting that these age transitions tend to mirror the transition points in Piaget's stage mode (Piaget & Inhelder, 1969).
For younger children it is common to rely on proxy respondents such as parents or teachers but it is possible to use other methods such as drawing, role play and observational methods over interviews for children under 6.
Besides these general rules, one should also consider questions related to the use of each research method with children from particular age groups and its adequacy in regard to the problem at hand:
Interviews: Children are not used to interviews, so it is important to create a familiar and fearless atmosphere. A hand puppet can be a perfect medium to engage with the child and to adjust to her/his language. Individual and group semi-structured interviews are possible with children older than 7 years old. However, less structured methods are needed for younger children (Christensen & James, 2000). As for individual interviews with young children, even children as young as 4 and 5 years olds are effective in referential communication (i.e. describing an object to a listener). This is only true on the condition that they have to describe familiar objects in a face to face interaction in a familiar, naturalistic setting (Bukatko & Daehler, 2001).
Participative observation: This method enables the observation of how children interact with each other while using media. Thus, emphasis can be placed on how a single child deals with the media, or on the exposure of a social system, in which children are growing up (e.g. family, nursery school, school), to the media.
Children's drawings: The advantage of children's drawings is the possibility of revealing aspects which cannot be verbalised. They provide an insight into the visual and intellectual capabilities of children, the emotions experienced while they are drawing, as well as their level of development. But children's explanations of their drawings are needed in order to interpret them adequately.
Experiments: Experiments are often favoured when dealing with very young children, who aren't yet able to verbalise their experiences and mindsets. However, young children, even preschoolers, have the language skills to describe what they remember. Young children remember familiar (repeated) events in terms of scripts. It is remarkable that all children recall older items better (recency effect) whereas a good recall for early items (primacy effect) is more apparent with children aged 7 years and older.
Questions to consider
When researching with children, particularly in the case of very young ones, combined approaches and alternative methods should be tried, as well as different perspectives on media and internet uses. Otherwise, research could be partially compromised at best, or completely beyond reach at worst. Sometimes solutions to problems rely on methodological imagination.
A common flaw in research with children is addressing the child as more mature, or more competent, than they are - overestimating their linguistic skills, for example, or underestimating the gap between competence (what they can really do) and performance (what the researcher has been able to observe them doing).
Pitfalls to avoid
Don't assume that children under 8-9 years can give accurate time estimates.
Focus groups with teens can be especially marked by social desirability biases.
Example of good practice: a study of very young children and their media use
In a study concerning the relevance of media (especially media figures and heroes) in children's friendships, peer groups and nursery schools, I worked with children aged three and above in order to discover their own perspectives. The aim of this study was to find out which television series, and especially their heroes, are meaningful to children of preschool age in the contexts of their daily lives (including nursery school, peer groups, friends and family). The issue of how children deal with the stories and symbols carried by their favourite television series in their daily lives was addressed. Data collection took place in a nursery school in West-Germany (75 children) and one in East-Germany (43 children). The social-ecological approach used by Baacke and Bronfenbrenner (Paus-Haase, 1998: 61) was the basis for the theoretical foundation (differentiation concerning the concepts of media and education, media equipment and the equipment in the rooms, as well as the social-ecological environment). The children who formed the focus group were chosen by theoretical sampling (concerning knowledge of media products, social relationships and peer groups, cognitive and linguistic development, as well as by age and gender). (Ingrid Paus-Hasebrink, Austria)
Example of good practice: experimental method with very young children
The 'this-or-that' method, which is used in experiments, is found to be useful with preschoolers between the age of 4 and 6 years old to do likeability research (Zaman & Abeele, 2007). At the beginning of this experiment, each child is asked to play with 2 objects, e.g. games (the order in which the games are presented are counterbalanced). The researcher tries to obtrude as little as possible and undirected play is supported (no tasks, since these conflicts with the explorative nature of games). After both conditions are finished, a likeability questionnaire is administered. Likeability was measured with five questions: 1) Which game did you find most fun (most fun), 2) Which game would you want to receive as a gift (wanted gift), 3) which game would you like to take home with you (take home), 4) which game would you like to play again (play again) and 5) which game did you find the most stupid (most stupid - this question was reversed in the final likeability measure)? These answers were triangulated with free play at the end of the test: as a 'reward' for participating the child could choose one of the two interfaces and play the game again. Besides quantitative measurements, qualitative material was also gathered. We video-recorded interaction styles and comments uttered by the toddler when playing the game. Only after the complete test was finalized (playing the two conditions and answering the likeability questions) did the facilitator follow up on this qualitative information and ask the toddler to explain a little more on exactly why one condition was chosen over one another according to the contextual laddering method (Zaman, 2008) (Bieke Zaman, Belgium)
Studies (see Annex C)
UK Children Go Online, Eurobarometer 250, SAFT, Mediappro
Berk, L. E. (2007). Physical and Cognitive Development in Early Childhood. In Exploring Lifespan Development. Allyn & Bacon.
Bukatko, D., & Daehler, M. (2001). Child Development. A thematic approach (4th ed.). Boston: Houghton Mifflin Company.
Rice, F. P. (1998). Human Development. A life-span Approach (3rd ed.). New Jersey: Prentice Hall.