Commentary: Conducting Sensory Research With Children

Children today are more involved than ever in choosing what they buy, use and eat. With an increasing number of parents, guardians and older siblings working outside the home, many children return home from school each day to an empty house. Often it is the children's responsibility to decide what to do, what to play with, what to eat and how to prepare it. A Gallup Survey of fourth through eighth graders revealed that 83% make their own snacks, 80% make their own breakfasts and 57% are involved in food purchases. This in itself is a prime reason that children's responses to products must be assessed. The article by Popper and Kroll gives us an excellent feel for the general state of the art in children's testing. Unfortunately, as the article clearly shows in the reference list, that state of the art often is unpublished, untested opinion rather than peer-reviewed, published science. That is not a reflection on Popper and Kroll, nor on the authors cited in the reference list – it is just simple fact. Howard Moskowitz's review, which I also have read, focused extensively on the business sense of using children for testing. I will not take that path because I agree with his comments and have little to add. Instead, I suggest that a major problem in the field of sensory testing with children (and to some degree with adults) is that the science often gets lost in the milieu of personal experiences, professional disagreements and belief systems – things that are not necessarily based on good science. I am pleased that Popper and Kroll specifically commented that more research is needed, because that certainly is true. It is especially the case for children, because the age at which children learn various concepts, are exposed to products and ideas and interact with measurement tools and technology changes with each succeeding generation. Much of the original research on Piaget's cognitive development theory and the ages associated with various aspects of that theory are now decades old. Although many of the innate psychophysiologic development issues have not changed a great deal, children are maturing earlier, and the cognitive demands and processing skills to meet those demands have changed. It is essential that we continually test the methods that we use to determine at what age they can be used, how they need to be adapted for certain age groups and how societal influences may have affected the techniques. Personal experiences and anecdotal evidence are insufficient to meet our needs. Popper and Kroll provide an excellent example of adaptation of a sensory technique (ranking) to make it more useful with children. Instead of being asked to "rank all the samples," the child is asked only to pick the best product. That product is removed, and then the child is asked to pick the best sample of those remaining. We also use a similar method in our work with children, except that we ask them to pick the best and worst, remove those and proceed accordingly. Such adaptations are addressed in the ASTM International E2299-03 Standard Guide for Sensory Evaluation of Products by Children (ASTM 2003), but I do not know of anyone who has actually published research showing that the technique is appropriate. We would not dream of believing the data from a new technique for measuring the chemical or physical composition of a product without validation of that method, but we do not hesitate to make decisions that will influence millions of dollars of consumer spending based on a new sensory test, as long as the test: (1) makes logical sense; (2) got the OK from a consultant who said his or her experience shows that it works; (3) costs less to do than what we are doing now; and (4) seems to give us the answer we expect when we use it on our product. Popper and Kroll mention another example, this time an example that has been researched, of how changing societal influences on language can allow adaptation of sensory techniques in the so-called Peryam and Kroll (P & K) scale for children. I doubt that, a generation ago, a scale using words such as "super good" or "super bad" would have been possible because the connotation of "super" was "best, bigger or greater" (e.g., supermarket or superman), not necessarily "extreme," which is its connotation in the P & K scale. However, in the late 1980s and early 1990s, when the scale was in the development and testing stage, the word "super" was very much in use by children to mean "extreme." Changing language, especially the vernacular of children, and a more global focus in many product categories means that we will always need to revalidate and update sensory methods and techniques that depend heavily on vocabulary. One area of particular importance that Popper and Kroll address is that of differentiating testing abilities among children. Certainly preschoolers are different in their testing abilities from 12-year-olds. But what about children in between those ages – the kindergarten to 5th grade crowd? Are they more like adults or young children in their ability to conduct sensory tests? Spaeth et al. (1992) showed that 8- to 10-year-olds could use traditional 9-point hedonic scaling techniques as well as any of the other scales tested, but those authors pointed out that they used children clearly in the age group associated with Piaget's concrete operational stage. Popper and Kroll discuss some issues of facial scales but do not provide concrete information about the age groups for which they are necessary and appropriate. In most published research studies, researchers have been careful to steer clear of the "cusp years"– those years that typically serve as the end of one developmental stage and the beginning of another. For example, age 6 is the end of Piaget's preoperational stage and 7 is the beginning of the concrete operational stage. Thus, most research is done with 4- to 5-year-olds or 8- to 10-year-olds. That occurs because there is too much chance that the cognitive development of any given child is slightly different from the "norms." Thus, it is easier to exclude those children who might be too old or too young developmentally (e.g., a bright 6-year-old who is already in the concrete operational stage) for a given task. However, we know that children have different abilities to do various tasks and we know that all the age groups are involved in decision making for a variety of products. Instead of excluding groups of children, would not it make more sense to develop a carefully researched screening tool to determine whether a given child had the ability to do a certain task or use a certain procedure than to arbitrarily set age limits that make no scientific sense and ultimately no business sense either? Popper and Kroll have done an admirable job of touching on many key points in the discussion of testing with children. It is now up to us to move ahead with logic and sound science into a new era of evaluating children's needs and wants.