Participants
Participants were recruited via a primary school in Wageningen, The Netherlands. Three age groups were included in the study, each age group representing a different stage of cognitive development: 4–5-year-old children (pre-operational stage: group A), 7–8-year-old children (concrete operational stage: group B) and 11–12-year-old children (formal operational stage: group C). The age groups correspond to the first, fourth and last grade of primary school in The Netherlands.
In group A, the two youngest and two oldest boys and the two youngest and two oldest girls who had permission to participate joined the study. In group B, four boys and four girls were selected by lottery. In group C, three groups were formed: one group of four girls, one group of four boys and a mixed group of two boys and two girls. The principal of the school and the parents of the children provided their informed consent.
Qualitative research
Because little is known about the relation between cognitive development and children's preferences, a qualitative approach was used to gain more insight and understanding about this relationship [30,31]. Qualitative research provides the opportunity to explore new topics and directions; thereby using the respondents' own words to give meaning to their world [30].
In our study we used focus group interviews for group C children. The purpose of a focus group is to elicit respondents' personal perceptions of a defined area of interest through carefully planned, semi-structured discussions [32]. Respondents can react and build upon other members' responses, and this in turn leads to more thoughtful and in-depth information [31,33,34]. In addition, it is possible to ask follow-up questions for clarification [31]. Focus groups are especially valuable for obtaining data from children [31-33].
Duo-interviews were held with the younger children in groups A and B. Children of this age need more assistance and have fewer communication skills. That is why duo-interviews are often used for this age group in market research [35]. Two children of the same sex and same age are interviewed together. Advantages are that the situation for the children is more natural and less scary than if they were alone, they speak more freely when they are with a child of the same age and same sex, and the attention of the interviewer is not all the time on one child. When the children know each other, there is another advantage: the children can point each other to untrue statements. The children can build upon each other's responses, but do not have to process responses of many other group members. Although the children can interfere with each other, we expected that the advantages of this method would compensate for the disadvantage of interference. Besides, care was taken to prevent dominant children from biasing the results: the moderator used different questioning strategies to include all participants' opinions and, if there was a dominant child in the group, the moderator took care to invite the less dominant child to give his or her opinion before the dominant child.
Procedure
In May and June 2005, three focus group sessions (four children in each) were held with group C. Each session took approximately 90 minutes. In group B, four duo-interviews were held, each lasting approximately 75 minutes. All four duo-interviews in group A took about 60 minutes. Because young children have a short attention span, the 4-year-old children were interviewed twice, for half an hour each time, on two different days. The interviews were held in a separate, quiet room at school during lessons with no parents or teacher present. All interviews were recorded and video taped. A research assistant was present to take notes, keep track of the time, record non-verbal information and control the video and sound recording. On finishing the conversation, each participant received a small present.
Question route
As advised by Morrison-Beedy et al. [34] for multiple group comparison, a structured interview guide was devised to ensure consistency in data collection. This guide was applied in the focus group discussions as well as in the duo-interviews and ensured a proper introduction, which is very important to make the children feel at ease. The rules of conversation were explained to them, as well as confidentiality, anonymity, recording of the sessions, and the fact that there are no wrong answers [21,34].
In addition to questions, game tasks and fruit tasting were included, to get richer information and to keep the children concentrating. Health was the final topic introduced during the conversations to prevent the children from focussing on health during the whole interview. An expert in child interviewing checked the interview guide and made suggestions for improvement. The guide was pilot tested on three children aged 5–6 years. Improvements were made to ensure that the questions and game tasks were clear and understandable.
To get a better understanding of children's preferences and their perceptions, thoughts, learning and communication about fruit and vegetables, diverse topics were addressed during the conversations: spontaneously probed preferences and dislikes, attributes leading to liking and disliking, categorization of fruit and vegetables, tasting fruits, healthy eating strategies, appropriate eating situations for fruit and vegetables, free associations and the concept of health. To make it easier and concrete for the children, various real fruits and vegetables were brought to the sessions. Picture cards were used to assist the children in pointing out appropriate eating times and occasions (six eating times: breakfast, morning break, lunch, afternoon, dinner and evening + six occasions: home, school, party, sport, being with friends and TV/computer). Seven fruits were chosen for tasting. We included fruits that varied widely in their taste, appearance, frequency of use and familiarity: strawberry, apple, mango, papaya, kiwi, grapefruit and lemon. Six vegetables were chosen based on the same arguments, but these were not tasted: carrot, cauliflower, egg plant, red peppers, French beans and chicory. The question route can be found in Table 2.
Data analysis
The recorded interviews were transcribed by the interviewer and the assistants. The interviewer checked the transcript with the video records, in order to add non-verbal information. We developed a coding framework based on the research aims, the interview guide and findings in the literature. The qualitative data analysis package N6 from QRS International (version 2002) was used to code and organise the data systematically. Significant statements were coded with a label and corresponding statements were coded with the same label. Based on the children's statements, we chose an appropriate term for each label to summarize the statements within a category. This organisation of data into different categories and sub-categories assisted a more effective comparison of the groups.
Initially, data analysis was carried out by the first author, who was physically present in the room when the duo-interviews and focus groups were conducted, as is advised by Krueger and Casey [32]. First, the data within an age category were analysed. Subsequently, the data of the three age groups were compared thoroughly to detect patterns and find similarities and differences. In discussion sessions with all authors of the manuscript, the results of the analyses were repeatedly and thoroughly discussed. In addition, the results were presented and discussed with other researchers and with external experts in the field of research with children and taste, to check interpretations and conclusions. Key concepts and patterns are discussed below.
Answers to all your Biology Questions
