Free Sorting Research Articles

Sample Size Requirements for Stable Clustering of Free Partition Sorting Data

Convergence rates of cluster analysis results are studied for six sets of partition sorting data. Hierarchical cluster analysis, fuzzy cluster analysis, k-medoids clustering and four variants of the consensus clustering method are covered. The six data sets are kinship terms, colour plates, drug data, consultancy Web site items, social groups and prenatal tests. Convergence rates depend on the number of clusters extracted, on the clustering method and on the domain. For domains of 25 items or more, cluster analysis results identical to the final result were most consistently obtained using the third method of Gordon and Vichi (2001). Example evaluations are given for the most stable numbers of clusters for five sets of partition sorting using the method of Gordon and Vichi. The interpretations resemble previously published interpretations of results obtained from non-partitioning statistical methods.

Odour quality of spray-dried hens’ egg powders: The influence of composition, processing and storage conditions

This study aimed to determine whether compositional or processing parameters have an influence on the odour quality of egg powders. The parameters tested were: whole egg vs. egg yolk, polyunsaturated fatty acid (PUFA) enrichment, spray-drying temperature (160°C vs. 180°C), production scale (industrial vs. pilot plant), storage temperature (15°C vs. 30°C) and time (1, 2, 4 and 8months). The quality of egg powders was evaluated by sensory analysis using free sorting, and by gas chromatography coupled to mass spectrometry and olfactometry. PUFA enrichment and spray-drying temperature do not affect the odour of egg yolk powders. There are significant differences between the odour of whole-egg and egg-yolk powders as well as between powders produced on an industrial scale or in a pilot plant. An increase in the odour intensity of egg powders was observed during storage, while unpleasant odours were perceived when the egg powders were stored at 30°C.

Assessing and taking into account the subjects’ experience and knowledge in consumer studies. Application to the free sorting of wine glasses

The degree of knowledge of a panel of consumers about the stimuli under study is assessed by means of a questionnaire. From the scores obtained by means of a Rasch measurement model, the consumers were segmented into three groups. In order to assess the impact of the consumers’ knowledge and experience onto their perception of wine glasses, a free sorting experiment was undergone. The group of consumers identified as connoisseurs sorted the glasses into six categories based on usage and technical properties. The group of consumers identified as non connoisseurs showed less agreement in their sorting of the glasses and mainly based their assessment on descriptive properties related to the design and the shape of the glasses. Thus, the perceptual representations of the glasses and their descriptions are different between the groups of consumers and are impacted by the previous knowledge and the experience of the subjects and their involvement toward the products.

Understanding how consumers categorise nutritional labels: A consumer derived typology for front-of-pack nutrition labelling

Significant ongoing debate exists amongst stakeholders as to the best front-of-pack labelling approach and emerging evidence suggests that the plethora of schemes may cause confusion for the consumer. To gain a better understanding of the relevant psychological phenomena and consumer perspectives surrounding FoP labelling schemes and their optimal development a Multiple Sort Procedure study involving free sorting of a range of nutritional labels presented on cards was performed in four countries (n=60). The underlying structure of the qualitative data generated was explored using Multiple Scalogram Analysis. Elicitation of categorisations from consumers has the potential to provide a very important perspective in this arena and results demonstrated that the amount of information contained within a nutrition label has high salience for consumers, as does the health utility of the label although a dichotomy exists in the affective evaluation of the labels containing varying degrees of information aggregation. Classification of exiting front-of-pack labelling systems on a proposed dimension of ‘directiveness’ leads to a better understanding of why some schemes may be more effective than others in particular situations or for particular consumers. Based on this research an enhanced hypothetical front-of-pack labelling scheme which combines both directive and non-directive elements is proposed.

Quick and dirty but still pretty good: a review of new descriptive methods in food science

SummaryFor food scientists and industrials, descriptive profiling is an essential tool that involves the evaluation of both the qualitative and quantitative sensory characteristics of a product by a panel. Recently, in response to industrial demands to develop faster and more cost‐effective methods of descriptive analysis, several methods have been offered as alternatives to conventional profiling. These methods can be classified in three families: (i) verbal‐based methods (flash profile and check‐all‐that‐apply), (ii) similarity‐based methods (free sorting task and projective mapping aka Napping®) and (iii) reference‐based methods (polarised sensory positioning and pivot profile). We successively present these three classes of methods in terms of origin, principles, statistical analysis, applications to food products, variations of the methods and the Pros and Cons.

About consensus in free sorting task

A partir de jugements individuels sous forme de categories (un profil de partitions sur un ensemble X), on cherche a etablir des categories collectives, ici appelees concepts. Nous comparons deux approches combinatoires. La premiere consiste a calculer une partition consensus, la mediane du profil, c’est-a-dire la partition de X dont la somme des distances aux jugements individuels est minimum ; les concepts sont alors les classes de cette partition consensus. La seconde commence par calculer une distance D sur X, basee sur le profil, et a construire un Xarbre associe a D ; les concepts sont alors certains sous-arbres de cet X-arbre. Nous cherchons a comparer ces deux approches, a mesurer leur congruence, en particulier, dans quelle mesure les classes de la partition consensus, sont des sous-arbres du X-arbre et reciproquement.

Inverse MDS: Inferring Dissimilarity Structure from Multiple Item Arrangements

The pairwise dissimilarities of a set of items can be intuitively visualized by a 2D arrangement of the items, in which the distances reflect the dissimilarities. Such an arrangement can be obtained by multidimensional scaling (MDS). We propose a method for the inverse process: inferring the pairwise dissimilarities from multiple 2D arrangements of items. Perceptual dissimilarities are classically measured using pairwise dissimilarity judgments. However, alternative methods including free sorting and 2D arrangements have previously been proposed. The present proposal is novel (a) in that the dissimilarity matrix is estimated by “inverse MDS” based on multiple arrangements of item subsets, and (b) in that the subsets are designed by an adaptive algorithm that aims to provide optimal evidence for the dissimilarity estimates. The subject arranges the items (represented as icons on a computer screen) by means of mouse drag-and-drop operations. The multi-arrangement method can be construed as a generalization of simpler methods: It reduces to pairwise dissimilarity judgments if each arrangement contains only two items, and to free sorting if the items are categorically arranged into discrete piles. Multi-arrangement combines the advantages of these methods. It is efficient (because the subject communicates many dissimilarity judgments with each mouse drag), psychologically attractive (because dissimilarities are judged in context), and can characterize continuous high-dimensional dissimilarity structures. We present two procedures for estimating the dissimilarity matrix: a simple weighted-aligned-average of the partial dissimilarity matrices and a computationally intensive algorithm, which estimates the dissimilarity matrix by iteratively minimizing the error of MDS-predictions of the subject’s arrangements. The Matlab code for interactive arrangement and dissimilarity estimation is available from the authors upon request.

Executive function performance on the children's kitchen task assessment with children with sickle cell disease and matched controls

Objective: The aim of this study was to compare executive function abilities of 8- to 12-year-old children with sickle cell disease (SCD) with a matched control sample. The measures included the parent and teacher Behavior Rating Inventory of Executive Function (BRIEF); the Delis-Kaplan Executive Function System (D-KEFS), Free Sorting Test and Color Word Interference Test; and the Children's Kitchen Task Assessment (CKTA), a new performance measure. Methods: Twenty-two children with SCD were recruited from one hematology clinic and 22 community children, matched on characteristics of age, gender, and race, were selected from a larger sample of controls for comparison. Parents and teachers completed rating scales. Results:As hypothesized, children with SCD scored significantly lower than matched controls on Digit Span Forward; on 5 of the 9 D-KFES Color Word Interference and Sorting tasks; on CKTA organization, initiation, and task completion; and on the BRIEF's parent and teacher Metacognitive Index (MI) and Global Executive Composite (GEC) scores. Conclusion:Cognitive and performance evaluations indicate lower executive function among children with SCD. Results substantiate the need for evaluative triangulation for children with SCD: Neurocognitive testing supported by performance testing, and adult reflection of a child's daily performance compared to other children. These elements will provide rich data to create educational support for children with SCD who have frequent hospitalizations, school absences, and the potential presence of cerebral vascular accident symptomology.

Application of free sorting tasks to sound quality experiments

In many studies devoted to the sound quality of industrial products, a perceptual space is determined through dissimilarity judgements on pairs of stimuli. A drawback of this procedure is that it can be very time consuming if the number of stimuli is large. An alternative procedure consists in a free sorting of sounds: averaging individual results provides a set of data which are considered as indicators of dissimilarities and analyzed using a multi-dimensional scaling method. The validity of this alternative can be discussed, as the psychological processes involved in the two procedures are different. This study compared these two approaches in a particular case (door closure sounds). In this specific case, it was observed that dissimilarities obtained from the two procedures can be different, the more so as sounds are dissimilar and these differences can lead to slightly different perceptual spaces. Nevertheless, a free sorting experiment is a reliable way of reducing the number of stimuli in a large set of sounds. It allows selecting some representative sounds and narrowing the set of sounds while keeping in the subset most of the timbre features. This provides a useful preliminary step to a paired-comparison experiment.

Taxonomic free sorting

Free sorting is a quick and reliable means to assess similarities among a set of stimuli by a panel of subjects. In this procedure, subjects form as many groups of stimuli as they consider necessary, with the understanding that each group is formed with stimuli that are perceived as similar. However, no information is available about the structure of the groups formed by each subject. In practice, these groups are considered as equally distant from each other. This is a questionable assumption as it may not reflect the perception of the subjects towards the stimuli. Taxonomic free sorting is designed to cope with this problem. Once the subjects have achieved the free sorting of the stimuli, they are instructed to organize, step by step, the stimuli into a hierarchical structure thus giving more insight into the similarities among products. Statistical treatments of the data thus obtained are discussed and the procedure of evaluation is illustrated on the basis of a case study.

Automatic stress-free sorting of sturgeons inside culture tanks using image processing

Summary A periodic sorting of fish is important to achieve optimal growth rates and efficient fish farm operations. Conventional methods, however, are labour intensive, cause stress to the fish and increase the risk of injuries. To overcome these problems a device has been developed for automatic and stress free sorting of fish by means of digital image processing. The main part of the device consists of an underwater monitoring channel, which is able to direct fish through different exits according to their weight. The decisions on weight class sorting are induced by basic computer algorithms. Experimental results show that a linear regression model based on a correlation between silhouette area and absolute body mass is able to estimate body mass with a relative average error of 5.5% (n = 50). The rates at which the fish swim through the device are found to be seven and 30 fish per hour, suggesting a dependence on the fish density in the sorting tank. It is hypothesized that the relative position and the behaviour of the fish within the sorting channel, among other factors, influence the model accuracy and hence the sorting decision. For future work, an improved sorting device is planned by adding commercially relevant functions for individual fish identification and monitoring that are based upon finger-print-like morphological features of individual fish.

Identification of new food alternatives: How do consumers categorize meat and meat substitutes?

New meat substitutes need to be recognized as alternatives to meat. We therefore investigated which category representations consumers have of meat and meat substitutes. Thirty-four non-vegetarian participants performed a free sorting task with 17 meat products and 19 commercially available meat substitutes, followed by similarity and typicality ratings. Results indicated that categorization was largely influenced by the taxonomic classification of meat, so by categories that refer to the animal source like ‘pork’, ‘beef’ etc. Hence, meat substitutes were grouped separately from non-processed meat products. However, there were categories (e.g. ‘pieces’ and ‘sausages’) that contained both meat substitutes and processed meat products, as these products were perceived to be very similar. New meat substitutes should have a certain resemblance to meat in order to replace meat on the plate. This can be achieved by either similarity in appearance or by referring to shared scripts/goals, such as a similar application in meals.

AN INVESTIGATION USING THREE APPROACHES TO UNDERSTAND THE INFLUENCE OF EXTRINSIC PRODUCT CUES ON CONSUMER BEHAVIOR: AN EXAMPLE OF AUSTRALIAN WINES

ABSTRACTThe aim of this study was to demonstrate the value of and differences between three approaches to measuring extrinsic product attribute influences on consumer acceptance of Australian wines. The approaches included three complimentary tasks: conjoint assessment of wine product concepts (derived from a free sorting task), liking ratings for commercial wine labels and a real‐choice study where the same wine bottles were presented to the consumer to choose their preferred wine based on the label. Results from the conjoint assessment indicated that price and wine region were of greatest utility. Findings from the real choice task mirrored the rating of wine labels, indicating that respondents were consistent in their liking and choice behavior across the two tasks. The choice study facilitated the validation of the experimental findings and suggested that label liking is a strong predictor of choice behavior.PRACTICAL APPLICATIONSThe study provides methodological guidelines for studying consumer responses to extrinsic product attributes. The labeling study provides important information for wine marketers and label designers that could be used to optimize the presentation of product information to meet consumers' acceptance.

Perceptive maps of dishes varying in glutamate content with professional and naive subjects

The present study was undertaken to examine how western subjects perceive umami taste and how they can sensory discriminate among dish variants having different natural umami taste (glutamate) levels. The food matrix used in this study was prepared following a traditional Italian recipe of asparagus soup with ingredients having natural high endogenous content of umami compounds (i.e. Parmigiano, commercial meat broth, and fresh asparagus). The ingredients ratio was modified in order to modulate the flavour and, in particular, the umami taste. The study enrolled two groups of subjects: one group of panellists tested for normal taste perception but with no sensory training, and a second group of professional panellists trained in the recognition and quantification of umami taste. Panellists in both groups participated in a free sorting and verbalisation tasks. Additionally, a descriptive profiling was performed by the trained group. Both the groups distinguished the samples according to their umami compounds levels, although only the trained panellists recognised the taste increase as umami and provided a classification of the samples closer to formulation’s differences. The untrained subjects distinguished the soup samples for flavour intensification, although they did not recognise the ingredients responsible for enhancing the flavour.

Sensory methodologies and the taste of water

Describing the taste of water is a challenge since drinking water is supposed to have almost no taste. In this study, different classical sensory methodologies have been applied in order to assess sensory characteristics of water and have been compared: sensory profiling, Temporal Dominance of Sensations and free sorting task. These methodologies present drawbacks: sensory profile and TDS do not provide an effective discrimination of the taste of water and the free sorting task is efficient but does not enable data aggregation. A new methodology based on comparison with a set of references and named “Polarized Sensory Positioning” (PSP) has been developed enabling to easily define the sensory characteristics of water without presenting too many samples. Finally, this method provides a new type of sensory data requesting dedicated data analysis.

The Odor of Colors: Can Wine Experts and Novices Distinguish the Odors of White, Red, and Rosé Wines?

Recently, several papers have investigated color-induced olfactory biases in wine tasting. In particular, Morrot et al. (Brain and Language, 79, 309–320, 2001) reported that visual information mostly drove wine description and that odor information was relatively unimportant in wine tasting. The present paper aims to study the relationship between the color of wine and its odor through a different approach. We hypothesize that people have stable mental representations of the aroma of the three wine color categories (red, white, and rose) and that visual information is not a necessary clue to correctly categorize wines by color. In order to explore this issue, we adopted two complementary approaches. In the first one, we presented 18 wines (six reds, six whites, and six roses) in dark glasses to our participants who were asked to smell the wines and categorize them into three categories: “red wine,” “white wine,” or “rose wine.” Because we expected categorization performance to be affected by participants’ expertise, we used two groups of participants corresponding to wine experts and wine novices. The second approach was designed in order to verify whether the most salient perceptual differences among samples were correlated with the output of the ternary sorting task. Using the same 18 wines, we asked a third panel composed of trained assessors to perform a wine description, a free sorting task based on wines’ odor similarity, and finally, the same ternary sorting task carried out by experts and novices. We found that experts and novices were able to correctly identify red and white wines but not rose wines. Contrary to our expectations, experts and novices performed at the same level. Trained panelists also categorized accurately white wines and red wines but not rose wines. From a more perceptual point of view, the free sorting task yielded virtually the same result. Finally, in terms of wine description, again, a clear segmentation was obtained between white and red wines. White wines were described by yellow or orange odorant sources, while the red wines were described by dark odorant sources. In the light of our results, cognitive mechanisms potentially involved in the organization of sensory knowledge and wine categorization are also discussed.

Categorization skills and recall in brain damaged children: a multiple case study

During development, children become capable of categorically associating stimuli and of using these relationships for memory recall. Brain damage in childhood can interfere with this development. This study investigated categorical association of stimuli and recall in four children with brain damages. The etiology, topography and timing of the lesions were diverse. Tasks included naming and immediate recall of 30 perceptually and semantically related figures, free sorting, delayed recall, and cued recall of the same material. Traditional neuropsychological tests were also employed. Two children with brain damage sustained in middle childhood relied on perceptual rather than on categorical associations in making associations between figures and showed deficits in delayed or cued recall, in contrast to those with perinatal lesions. One child exhibited normal performance in recall despite categorical association deficits. The present results suggest that brain damaged children show deficits in categorization and recall that are not usually identified in traditional neuropsychological tests.

SORT-CC: A procedure for the statistical treatment of free sorting data

A statistical approach for the analysis of free sorting data is discussed. In a first stage, the sorting data from each subject are arranged into a dataset consisting of indicator variables which reflect the memberships of the stimuli to the groups formed by the subject under consideration. Thereafter, an appropriate standardization is applied on these data and a three way statistical method, namely Common Components and Specific Weights Analysis, is performed on the datasets thus obtained. This makes it possible to take account of the individual differences among the subjects and to depict graphical displays showing the relationships among the stimuli on the one hand and among subjects on the other hand. The general strategy of analysis is illustrated on the basis of a dataset and the outcomes are compared to those of other strategies of analysis of sorting data.

Classification of everyday sounds: Influence of the degree of sound source identification

Based on a review of everyday sound classification studies, we define three types of similarity that might underlie the formation of categories: acoustical similarity, related to the perception of the sound signal properties; causal similarity, related to the perception of the cause of the sound; semantic similarity, related to the meaning of the sounds. This study focuses on the influence of the degree of sound source identification on the type of similarity used during a free sorting task. Our hypothesis is that well-identified sounds will lead to categories based on causal or semantic similarities, while non-identified sounds will lead to categories based on acoustical similarity. First we measure the causal uncertainty of each sound of a set of kitchen sounds. We ask to participants to describe the cause of the sounds in order to measure the degree of identification for each sound. The results of a free sorting task on this sound corpus indicate that participants groups together sounds according to our assumptions. The classification of the well-identified sounds shows a clear distinction between classes of sound events (gas, liquid, solids, motor) confirming a taxonomy proposed in literature. [Work supported by the CLOSED project of the NEST program]

Is there a perceptive signature of vehicles vibrations

The vibro‐acoustic comfort of vehicles running at low speed has been studied for several years by car manufacturers. Even if car passengers are exposed to a complex environment involving sight, hearing and touch, it is generally agreed that vibrations transmitted through the seat is a very significant parameter in such situations. Previous laboratory experimentations revealed that vehicles were strongly discriminated even when subjects were submitted to seat vibrations only. The ranking of vehicles was also identical for different tested roads. All these observations raised the question of the existence of an identifiable signature of vehicles, independent of the road type. A perceptive experiment designed to evaluate the influence of such phenomenon was conducted. Subjects were submitted to a free sorting test. The categorisation task consisted in grouping vibrations stimuli recorded in 9 cars running on 3 differents roads according to their similarity. The RMS level of stimuli was normalised so that its effect was not predominant during the experiment. Results showed that stimuli groups corresponded to the 3 tested roads, and were correlated to time envelopes of vibrations. Vehicles were not significantly discriminated.