Desempenho de crianças em testes sensoriais discriminativos e afetivos com escalas hibridas ilustradas. / The performance of children in discriminative and affective sensory tests with illustrated hybrid scales.
AUTOR(ES)
Rosamaria da Re
DATA DE PUBLICAÇÃO
2006
RESUMO
In both developed and emerging countries, children represent a significant part of the consumer market, a market which, due to globalization, has become extremely competitive and avid for novelties. As the number of products destined for the child food market increases and the role of the child in the decision to buy also increases, sensory evaluations carried out with children become more important for the food industries. Unfortunately, little systematic investigation into the validity and effectiveness of the application of sensory techniques to children has been carried out. Sensory methods applied to the infantile public are usually limited to paired comparison, ranking and category scale tests. Linear scales are rarely used with children, despite showing the following advantages over the category scales: i) they offer more liberty to the consumers to express their sensory perceptions; ii) they reduce the numerical and contextual effects; iii) they provide continuous data, which, by their very nature, are more adequate and increase the discriminative power of the parametric and non-parametric statistical tests. Some researchers have used hybrid hedonic scales in consumer tests, reporting that they are easy to use and favour sample segmentation in comparison to the category hedonic scale. However, the use of hybrid scales with the infantile population has been little explored. Thus the general objective of the present study was to propose and evaluate the performance of illustrated hybrid scales with 8 to 11 year old children for both the evaluation of stimulus intensity and the degree of acceptance of the food. With this objective, two experiments were carried out as follows. In the first experiment, two illustrated hybrid intensity scales were proposed and evaluated. A panel of ninety-six children evaluated the intensities of sweetness and sourness in eight orange drink formulations in duplicate, using illustrated and ranking hybrid intensity scales. The drinks contained different levels of sucrose (8.17g/100ml; 9.67g/100ml; 11.17g/100ml and 13.17g/100ml) and citric acid (0.10g/100ml; 0.15g/100ml; 0.20g/100ml and 0.25g/100ml). A second panel consisting of ninety-six young adults also evaluated the sweetness and sourness of the samples in duplicate, using the hybrid intensity scale. The Friedman variance analysis was applied to the data to compare the discriminative power of the children when using the ranking and illustrated hybrid scales, computing the values for Fr and pFr generated by each scale for each stimulus in both replications. In addition, the same data were transformed into ranking values and analysed by the Kramer non-parametric test, associated with the Basker tables. For each panel, the results of the hybrid scales were evaluated using ANOVA with the following source variations: sample, panellist, sample presentation order and replication. The values for Fsample and pFsample were used to compare the discriminative power of the individuals with each scale and the values for pFpanellist, to evaluate their expansiveness. Differences between the means were evaluated using the Tukey test (p=5%). Random variability of the panellists with each scale was assessed from the values for QMerror and standard deviation. Finally, in order to evaluate if the data obtained attended the statistical presuppositions of normality required by ANOVA, the asymmetry and kurtosis indexes were calculated. The results suggested that the children presented equally high discriminative power for both the ranking scale (Frsamplesweetness = 163.59, pFrsamplesweetness <0.0001; Frsamplesourness = 143.67, pFrsamplesourness <0.0001), and the illustrated hybrid intensity scales (Frsamplesweetness = 110.42, pFrsamplesweetness <0.0001; Frsamplesourness = 85.08, pFrsamplesourness <0.0001) in the evaluation of the intensity of the sweet and sour stimuli in the drinks. The discriminative power of the children (Fsamplesweetness = 84.30, pFsamplesweetness <0.0001; Fsamplesourness = 76.88, pFsamplesourness <0.0001) was comparable to that of the adults (Fsamplesweetness = 368.45, pFsamplesweetness <0.0001; Fsamplesourness = 278.25, pFsamplesourness <0.0001) for both the stimuli. For the sour stimulus, the data generated by the child panel presented less restriction with respect to the presuppositions of normality than those generated by the adult panel. The contrary was observed for the sweet stimulus. The problem of expansiveness between the children was small, being non-significant for the sour stimulus (pFpanellist = 0.2972) and only significant for the sweet stimulus for the 10 year olds (pFpanellist <0.0001). The results generated by the children using the illustrated hybrid scales were highly reliable, presenting good reproducibility in the judgements (pFreplicationsweetness = 0.6514, pFreplicationsourness = 0.1229), better than that of the adult panel (pFreplicationsweetness = 0.0323, pFreplicationsourness = 0.2566). However, the younger children, between 8 and 9 years old, presented greater random variability in their responses for both stimuli, (QMerrorsweetness @ 6.70; QMerrorsourness @ 7.26) as compared to the older children between 10 and 11 years old, (QMerrorsweetness @ 5.36; QMerrorsourness @ 5.51) and the adults (QMerrorsweetness = 2.19; QMerrorsourness = 2.59). These differences possibly reflect variations in the cognitive development of the children. In general one can conclude that the illustrated hybrid intensity scales proposed in the present study represented an efficient and reliable method to evaluate segmentation of drinks by children as a function of the intensity of sweetness and sourness. The objective of the second experiment was to propose an illustrated facial hybrid hedonic scale for use in affective tests with children, and evaluate its performance as compared to the 9-point mixed facial hedonic scale. Two consumer panels, one composed of ninety-six 8 to 11 year old children, and the other of ninety-six young adults, evaluated six orange drink formulations containing different concentrations of sucrose and citric acid. The children evaluated the samples using both the illustrated facial hybrid hedonic scale and the 9-point mixed facial hedonic scale, whilst the adults only used a 10 cm hybrid hedonic scale. For the children, the effects of scale and sample presentation orders, age and sex were balanced. Finally, tests for the conservation of discreet and continuous amounts, for class inclusion and for seriation were applied to the children and, based on the responses obtained, they were classified into one of the three cognitive development stages: i) Concrete Operational; ii) Transition and, iii) Pre-operational. The data were evaluated by ANOVA, Tukey?s means test and Internal Preference Mapping (MDPREF). The values for Fsample and pFsample were used to evaluate the discriminative power of the individuals, whilst pFpanellist and QMerror indicated, respectively, differences in expansiveness of the children when using the scales and random data variability. The children presented good sample segmentation capacity using both the illustrated facial hybrid hedonic scale (Fsample = 61.58, pFsample <0.0001) and the structured hedonic scale (Fsample = 48.40, pFsample <0.0001), showing performance similar to that of the adults (Fsample = 61.51, pFsample <0.0001). In generating the MDPREFs, the hybrid hedonic scale generated slightly greater segmentation both for the samples and for the children. Expansiveness was significant for the children, both for the illustrated facial hybrid hedonic scale (pFpanellist = 0.0075) and for the structured hedonic scale (pFpanellist = 0.0368), similar to that observed for the adults (pFpanellist <0.0001). Sex only showed a significant effect [F(1.445) = 11.53, p = 0.0007] amongst the children when using the 9-point mixed facial hedonic scale. The stage of cognitive development of the children used in this study showed no significant effect (p?0.05) with any of the scales. Children and adults diverged greatly with respect to the acceptance and preference of the six drink samples: whilst the children preferred the sweeter, less sour samples, the adults greatly disliked these samples, preferring the more sour samples with intermediate sweetness. The results of the present study show that the illustrated facial hybrid hedonic scale can be used as efficiently as the category facial hedonic scale in the segmentation of drinks according to acceptance by 8 to 11 year old children, whether they be in a Concrete Operational or Transition stage of cognitive development.
ASSUNTO(S)
escalas de intensidade analise sensorial rating scales consumer preferencia do consumidor - aceitação children consumidor crianças consumer sensoring analysis
