Even babies show signs of preferring certain foods over others.
The market for children's foods is growing, and children have an increasing influence on food purchase decisions. This has lead to greater use of children in the development of new products. Thus, the need for sound methodology for sensory testing with children has increased.

Using children in sensory tests is appropriate when the goal is to better understand their sensory perception or to determine their preferences as consumers. This author is of the opinion that it is better to use adults as judges in sensory evaluations since they have perceptions similar to children, but possess the required cognitive abilities.

Sensory and Cognitive Development

Compared to adults, infants have five times the number of taste buds, and larger and more abundant foliate papillae on the tongue. However, the nervous system supplying taste papillae is not as well developed. Some studies report children as young as five to seven years old have detection thresholds similar to adults, but most research shows they have lower sensitivity.

The gustatory (taste) system, however, is highly functional at birth. In contrast, odor detection shows significant development over even the first few days of life.

The American Society for Testing Materials' (ASTM) Committee 18 is developing guidelines for sensory testing with children. As part of this effort, they have linked children's cognitive skills with their age. For example, the ability to understand scales does not exist until age three. At three to five years, the understanding of simple scales begins, but the use of sorting or identification tasks is more effective for obtaining feedback. Five- to eight-year-olds increasingly understand scales; however, they should be simple. Pre-teens are “capable of understanding scaling concepts with adequate instruction,” and teenagers' scaling cognitive skills are similar to an adults'.

Children can be classified into Piaget's stages of cognitive development. They are “preoperational” between the ages of two and seven, which means they are “perception bound.” A child's ability to pay attention to only one aspect of a situation is particularly limiting. Young children often judge a food based on a sole attribute—such as appearance.

Additional challenges regarding sensory testing include limited verbal skills, short attention spans, and difficulty in task comprehension.

Adults are better able to differentiate among the sensory attributes of, say, ice cream and clearly divide their perception into appearance, taste and texture. Children tend to “smear” attributes, confusing taste and texture (but not appearance), and confusing sweet and tart. Children are influenced more strongly than adults by irrelevant dimensions of complex stimuli.

To deal with potential comprehension problems, one recommendation is to take the child through the test protocol using “visuals” before having him/her taste the actual product. For example, before a paired-comparison test for sweetness intensity, the testers would present a picture of doughnuts and hot dogs, expecting the child to pick the doughnut as the sweeter food. For hedonic scaling, they could present a picture of green beans and chocolate cake, expecting him/her to give a higher rating to the cake.

There is extensive evidence that taste preferences are innate. Olfactory preferences, on the other hand, are mostly learned and develop slowly. In general, repeated exposure drives preference, except when negative peer or parent reactions “teach” young children that an odor is unpleasant. Although the process starts in the womb, children do not show reliable differentiation between the pleasantness of odors until age five.

Testing Protocols

Ways to assess taste or olfactory responses of newborns and infants include lateral tongue movements, automatic reactivity, facial expressions, respiration, heart rate, and differential ingestion and sucking patterns. Except for lateral tongue reflex, all are hedonically motivated.

Children’s limited verbal skills mean special consideration be given to the phrasing of questions and to vocabulary used.
Assessing the responses of older infants and toddlers generally involves non-verbal cues such as body movements, vocalizations and facial expressions. A highly recommended approach is to have the primary caretaker, usually the mother, interpret the behavior of the child as the latter tastes the food. The caregiver then rates acceptance on a traditional hedonic scale. Having adults taste and rate the sample afterwards provides information on adult acceptability of the product, which also matters, since the adult is making purchasing decisions.

Preschoolers to pre-teens may first be asked to perform sensory evaluation tests such as difference and scaling tests. Our studies showed children four and over accurately performed paired-comparison and intensity tests for sweetness, but that two- to three-year-olds were not yet capable of this. Further research has showed that inadequate cognitive ability is the likely issue.

Mothers and other caregivers are good intepreters of a child’s preferences, and also help determine whether adults like the product enough to purchase it.
Published studies are largely inconclusive on the degree to which children can give psychophysical (perception and preference) ratings for complex food stimuli. A study by Zandstra and de Graaf (Food Quality Pref.1998. (9) 5-12) showed that six- to 12-year- olds could scale (rank) the intensity of a beverage's sweet and sour taste, and orange flavor. However, when rating the differences in beverage sweetness, their scale was narrower than what the adults perceived. Some studies have supported this last finding and some have not.

Children also are asked to perform, as consumers, paired-preference, preference-ranking or hedonic scaling tests. Here is a sampling of findings from literature.

Kimmel, et. al. (Food Technology 1994. (48) 92-99) found that children over two reliably performed a paired-preference test, and children as young as four could use a seven-point hedonic scale.

Chen, et. al. (J. Sensory Stud. 1996. (11) 141-163) reported children aged three to six years old could express their degree of food liking using a three-, five-, and seven-point hedonic scale.

Leon, et. al. (Food Quality Pref. 1999. (10) 93-100) used four- to 10-year-olds to investigate three non-verbal methods measuring food liking (paired-comparison, ranking by elimination and hedonic categorization). The results were judged according to three criteria: discrimination of products, response repeatability and validity of the methods. Product discrimination was best with hedonic categorization. Children over five gave reproducible results by all three methods; four- to five-year-olds did not give consistent results by any of the three methods.

It should be noted that children sometimes confuse hedonic and intensity ratings. In one study, two- to five-year-olds had difficulty switching to an intensity test after taking a paired-preference test.

Focus Groups and Observation

Qualitative methods such as focus groups and observation used within a child's own environment can provide useful and reliable information.

Focus groups usually are limited to pre-teens or older. Children should be screened for attributes such as sociability, articulateness, interest in new products and trends and the ability to interact with adults, among other things. For example, they can be used at the ideation stage of product development in kids—only teams with adult facilitators, or in mixed adult-child brainstorming sessions.

Increasingly, observation-based methods are being used. Ideally, they are conducted in a child's own environment. When a leading breakfast cereal manufacturer observed its customers during their daily routine, for example, it was found that the cereal was not primarily used for breakfast, but as a snack—due to its portability.

Other issues need to be considered before beginning sensory work with children. Proper consent needs to be received. Specialized facilities with child-sized furniture and one-way mirrors can be useful. Child-friendly questionnaires with large fonts and writing spaces, and a limited number of pages often are appropriate. The child should be comfortable with the experimenter. And, lastly, the parents' needs should not be neglected. For example, mid- to late-afternoons and weekend testing may be best.

Developing products that please consumers' senses is crucial to new product success. When the consumers are children, sensory testing requires special considerations.