Search

There is increasing interest in methods for the habituation of feral rangeland goats to intensive farming conditions. We tested whether there were production performance and behavioural differences between groups of rangeland goats in an intensive farming system that were either exposed to a high degree of human interaction (HI, n=60) or low degree of human interaction (LI, n=60) over 3 weeks. In the HI group, a stockperson entered the pens twice daily and calmly walked amongst the goats for 20 mins. In the LI group, a stockperson only briefly entered the pens to check water/feed (daily/weekly). At the end of each week the goats were weighed and drafted into 12 subgroups of 10 animals (i.e. 6 sub-groups per treatment). Each sub-group was then tested for agonistic behaviour, avoidance of humans, and flight response. During the flight response test video footage was collected and later used for analysis using Qualitative Behavioural Assessment (QBA). For QBA analysis, the videos of each group, taken each week, were shown in random order to 16 observers who used their own descriptive terms to score the groups' behavioural expression. There was a significant interaction between treatment and time on body mass (F3,174 =5.0; P< 0.01), agonistic behaviour (F3,12 =4.3; P< 0.05) and flight speed (F3,12 =3.9; P< 0.05), with the HI group having significantly higher average body mass (P< 0.05), fewer agonistic events (P< 0.05), and a slower flight speed (P< 0.05) than the LI group after the three weeks. Two main QBA dimensions of behavioural expression were identified by Generalised Procrustes Analysis. QBA dimension 1 scores differed between treatments (P< 0.05); HI goats scored higher on QBA dimension 1 (more 'calm/content') compared to LI goats (more 'agitated/scared'). QBA dimension 1 scores were significantly negatively correlated with the number of agonistic contacts (Rs =-0.62, P< 0.01), and flight speed (Rs =-0.79, P< 0.001), and significantly positively correlated with body mass (Rs =0.68, P

Dogs housed in shelters may experience poor welfare. To ensure these dogs a good quality of life, welfare assessment tools should be sensitive not only to the animals’ physical health but also to their mental state, including the assessment of positive and negative emotions. In this study, we focused on the assessment of shelter dogs’ emotional expression using a Qualitative Behavioural Assessment (QBA) approach. Previous work successfully applied QBA to assess the emotional state of working and rescue dogs, and the observations were carried out on individual dogs in standardised settings with little or no stimulation. Results from such experiments might not be fully representative of the expressive demeanour that a dog could show in shelter conditions, where animals are exposed to a number of social and environmental stimuli. Thus, our aim was to apply QBA to a wider variety of shelter environments and social contexts than has been done so far, giving the animals the opportunity to express a wider repertoire of emotions and allowing for a more comprehensive assessment of dogs’ affective state. A set of descriptive terms was generated using Free-Choice-Profiling methodology by a group of 13 observers. QBA was made by scoring 16 video clips of shelter dogs in very different contexts (e.g. single/pair/group housing, presence/absence of human activity). Generalised Procrustes Analysis showed a high consensus between observers’ scoring patterns (75.7%; p

Qualitative behaviour assessment (QM) was used to validate previous published interpretation whether human stroking has beneficial effects on lambs prematurely separated from their mothers. For a period of 8 weeks a familiar human (FH) subjected 22 Romane lambs to a gentling treatment based on strokes. At the age of 8 weeks each lamb was placed in the experimental pen and tested. Half of them received the gentling treatment (STROKE) for 8 min, and the others were simply exposed to the human presence for the same period of time (PRESENCE). Animals were equipped for heart rate recording and tests were video-recorded. These behavioural and heart rate variability data have been already published. A free choice profiling approach was used to instruct ten observers in qualitative behaviour assessment (QBA). Short video sequences of the tests were randomly presented and scored by the observers unaware of any information about the quantitative data recorded. QBA data were analysed using Generalised Procrustes Analysis (GPA). The GPA consensus profile explained a high percentage of variation among the 10 observers, and differed significantly from the mean randomised profile (P < 0.001). The two main dimensions of the consensus profile explained 42,3% and 18,0% of variation between the tested lambs. All of the interpretation word charts were semantically consistent and oriented to the same meaning despite the use of different descriptors. The most used descriptors to label the first dimension was characterised by descriptors ranging from "curious" (positive end) to "relaxed" (negative end), whereas the second dimension ranged from "indifferent" (positive end) to "sociable" (negative end). Lambs from group PRESENCE received significantly higher scores than group STROKED on the Dimension 2 (chi(2) =4.49, P < 0.05) and were therefore assessed as more indifferent. Conversely, the animals from group STROKED received significantly lower scores on Dimension 1, thus resulting more relaxed than group PRESENCE (chi(2)=4.48, P < 0.05). A principal component analysis showed meaningful relationships between quantitative and qualitative variables with variables indicating relaxation and positive emotions displaying high loadings on the negative end of axis 1, whereas variables showing a low level of arousal but less positive emotions were more related with the positive end of axis 2. In this study QBA confirms previous interpretations from quantitative behavioural and physiological data. In general, we can suggest QBA as a tool to help in the elucidation of quantitative physiological and behavioural data.

Aim To investigate the differences in the head shapes of scaphocephaly patients during the pre- and postoperative phases. Material and methods Cranium shape data were collected from the two-dimensional digital images. The generalised Procrustes analysis was used to obtain mean shapes in the pre- and postoperative phases. The shape deformation of the cranium from the pre- to postoperative phases was evaluated using the thin plate spline method. Results There was no significant cranial shape difference in the pre and postoperative phases. The high-level deformations from the pre- to postoperative phase were seen in thin plate spline graphs. The highest deformation was observed at the biparietal dimension. Conclusion In this study, we showed that the structural deformation of the cranium was correlated with sagittal craniosynostosis. The present study also shows that pre- and postoperative head shapes of patients with scaphocephaly can be compared using head shape with the landmark-based geometric morphometric method by taking into consideration the topographic distribution.

Qualitative Behaviour Assessment (QBA) was utilised to examine the behavioural expression of dogs in different housing environments and the results were compared to measurements of quantitative behaviour and physiology. Firstly, quantitative behavioural and physiological differences were investigated between dogs in 3 housing environments (short-term shelter confinement, ≤4 days, n = 10; long-term shelter confinement, >30 days, n = 9; and domestic living situations, n = 10). Each dog’s behaviour was recorded over a 4 h period using an ethogram consisting of 21 behaviour categories. Dogs in both short (SD) and long (LD) term confinement displayed higher frequencies of paw-lifting ( P P P P H = 0.55, P = 0.76). Secondly, quantitative behaviour and QBA were combined to investigate differences among these same 29 dogs when filmed for 1 min in both their Home Environment and a standardised Novel Environment. QBA of these video clips was made by 10 observers utilising Free-Choice-Profiling methodology. Generalised Procrustes Analysis was used to calculate a consensus profile and three main dimensions of dog expression in both Environments. The observers repeated dog scores on these dimensions with high accuracy ( P H = 17.86, P H = 13.58, P H = 5.97, P H = 6.82, P P

The effect of cross-modal sensory interactions between cheese aroma and cheese taste on both cheese flavour intensity and cheese flavour character were investigated. Cheese aroma consisted of a mixture of ten aroma compounds, whilst cheese taste was constructed with the five basic tastes in water solution. Interactions were investigated using a combination of a high resolution design (HRD) and central composite design (CCD). The HRD was 2 aroma × 5 NaCl × 5 lactic acid levels. The CCD was constructed based on a 23 factorial design that covered the HRD space. Both HRD and CCD gave a total of 57 samples. A panel of experienced assessors (n = 8–10) evaluated cheese flavour intensity relative to reference. In addition, samples were sorted based on similarity of flavour character. Fourteen samples were selected for free choice profiling and analysed using generalised procrustes analysis (GPA). Cheese flavour intensity was non-linear and dependent on both taste level and aroma level. The status of flavour character being cheese-like was dependent on taste levels, where either NaCl or lactic acid at a high or low level altered the flavour character to being atypical of cheese. A cheese-like flavour character was maintained across a wide range of NaCl concentrations, but only across a narrow range of lactic acid concentrations. Aroma level changed the character of cheese flavour. Overall, the balance in concentration of both tastes and aroma levels was important in maintaining cheese flavour character as cheese flavour intensity was modified.

Qualitative behavioural assessment (QBA) was applied to investigate the expressive behaviour of sheep with varying intestinal parasite burdens over two experiments. The expressive behaviour of sheep naturally infected with intestinal parasites was assessed from video footage collected in the paddock, and assessments were compared pre- and post-treatment with anthelmintic drench. The first experiment assessed sheep with a range of parasite burdens (n = 28), and the second compared sheep that expressed clinical symptoms of parasitism (Anaemic, n = 5) with those that did not (Non-anaemic, n = 5). Behavioural expression scores were validated against individual clinical evaluations (faecal egg counts – FEC, faecal consistency and anaemia scores), production parameters (body mass and body condition score), and quantitative locomotive measures (walking speed and return order to paddock). Twenty-two observers scored 28 video clips using QBA in experiment 1, and in the second experiment, 35 observers scored 20 video clips that depicted the 10 focal sheep pre- and post-treatment. QBA scores were analysed using Generalised Procrustes Analysis (GPA), and sheep scores on the main GPA dimensions were evaluated in relation to parasite burden using Spearman Rank correlations and repeated-measures ANOVA for experiment 1 and 2, respectively. In both experiments, observers reached significant (P

Qualitative behavioural assessment (QBA) is based on observers' ability to capture the dynamic complexity of an animal's demeanour as it interacts with the environment, in terms such as tense, anxious or relaxed. Sensitivity to context is part of QBA's integrative capacity and discriminatory power; however, when not properly managed it can also be a source of undesirable variability and bias. This study investigated the sensitivity of QBA to variations in the visual or verbal information provided to observers, using free-choice profiling (FCP) methodology. FCP allows observers to generate their own descriptive terms for animal demeanour, against which each animal's expressions are quantified on a visual analogue scale. The resulting scores were analysed with Generalised Procrustes Analysis (GPA), generating two or more multi-variate dimensions of animal expression. Study 1 examined how 63 observers rated the same video clips of individual sheep during land transport, when these clips were interspersed with two different sets of video footage. Scores attributed to the sheep in the two viewing sessions correlated significantly (GPA dimension 1: r s =0.95, P0.001, GPA dimension 2: r s =0.66, P=0.037) indicating that comparative rankings of animals on expressive dimensions were highly similar, however, their mean numerical scores on these dimensions had shifted (RM-ANOVA: Dim1: P0.001, Dim2: P0.001). Study 2 investigated the effect of being given different amounts of background information on two separate groups of observers assessing footage of 22 individual sheep in a behavioural demand facility. One group was given no contextual information regarding this facility, whereas the second group was told that animals were moving towards and away from a feeder (in view) to access feed. Scores attributed to individual sheep by the two observer groups correlated significantly (Dim1: r s =0.92, P0.001, Dim2: r s =0.52, P=0.013). A number of descriptive terms were generated by both observer groups and used in similar ways, other terms were unique to each group. The group given additional information about the experimental facility scored the sheep's behaviour as more 'directed' and 'focused' than observers who had not been told. Thus, in neither of the two studies did experimentally imposed variations in context alter the characterisations of animals relative to each other, but in Study 1 this did affect the mean numerical values underlying these characterisations, indicating a need for careful attention to the use of visual analogue scales.

In this paper we explore a relational understanding of landscape qualities. We asked three independent groups of human observers to assess the expressive qualities of a range of landscapes in the UK and in Spain, either by means of personal visits or from a projected digital image. We employed a Free Choice Profiling (FCP) methodology, in which observers generated their own descriptive terminologies and then used these to quantify perceived landscape qualities on visual analogue scales. Data were analysed using Generalised Procrustes Analysis, a multivariate statistical technique that does not rely on fixed variables to identify underlying dimensions of assessment. The three observer groups each showed significant agreement, and generated two main consensus dimensions that suggested landscape 'health' and 'development in time' as common perceived themes of landscape expressivity. We critically discuss these outcomes in context of the landscape assessment literature, and suggest ways forward for further development and research.

This study examined whether observers could distinguish between cattle that were exposed to various road transport conditions: Experiment 1 compared a manipulated flooring treatment (non-grip flooring, NG) with a control transport event (grip flooring, G) and Experiment 2 compared a manipulated driving style (stop-start driving, SS) with a control transport event of smooth, continuous (C) driving. The behavioural expression of cattle was assessed through the process of Qualitative Behavioural Assessment (QBA), and these assessments were tested for correlation with various physiological parameters. Fourteen Angus steers were assessed. Blood samples were collected immediately before and after transport, and heart rate and core body temperature were measured continuously throughout each transport event. Continuous video footage recorded during each transport event was edited to isolate short clips of individual animals which were randomly ordered and shown to observers for QBA and analysis by Generalised Procrustes Analysis (GPA) which identifies common dimensions of behavioural expression. There was significant consensus amongst 39 observers in their assessment of behavioural expression of the cattle (P < 0.001). In Experiment 1, observers scored cattle exposed to NG flooring during road transport relatively higher (on visual analogue scales) for terms such as 'agitated', 'restless' and 'anxious' compared with cattle exposed to G flooring, which were scored higher for 'calm', 'comfortable' and 'relaxed' (GPA dimension 1, P < 0.001). In Experiment 2, cattle exposed to SS driving received higher average GPA dimension 1 scores (P < 0.01) and higher GPA dimension 2 scores (P < 0.05). These cattle were therefore scored relatively higher for terms such as 'restless', 'agitated' and 'scared' (GPA dimension 1) or 'curious', 'interested' and 'inquisitive' (GPA dimension 2) compared with cattle exposed to C driving, which were scored higher for 'calm', 'relaxed' and 'comfortable' (GPA dimension 1), or 'stressed', 'tense' and 'alert' (GPA dimension 2). There were some significant correlations between physiological responses and behavioural expression of animals for both experiments, with informative correlations between the different dimensions of behavioural expression and white blood cell counts, red blood cell parameters and heart rate. For example, the neutrophil: lymphocyte ratio, a typical marker of stress in ruminants, was elevated in cattle that were described as more 'agitated', 'restless' and 'anxious' (G-NG flooring: GPA dimension 1) or 'stressed', 'tense' and 'alert' (C-SS driving: GPA dimension 2). These results suggest that the QBA process captured behavioural manifestations of stress in cattle. We conclude that QBA is a valuable method of assessing cattle welfare under the conditions tested since there was significant consensus in the ability of human observers to interpret behavioural expression of cattle during these experimental conditions (i.e. QBA is repeatable), observers could distinguish between transport treatments on the basis of the animals' QBA scores, and these scores were correlated with meaningful physiological measures.

This study examined the behavioural expression of cattle immediately prior to slaughter through the process of Qualitative Behavioural Assessment (QBA), and compared these results to measurements of physiology and temperament. Twenty-eight Angus steers were filmed while in a funnel chute as they were being moved towards an abattoir killing box. Footage of cattle was shown in random order to 15 observers. Observers assessed the cattle using a qualitative approach based on Free Choice Profiling (FCP) methodology, which gives observers complete freedom to choose their own descriptive terms. Data were analysed with Generalised Procrustes Analysis (GPA). There was significant consensus (P 5.7 (classified as a 'dark cutter') had a low value for GPA dimension 2 (scored as relatively more 'annoyed'/'frightened'). The findings from this study suggest that QBA could contribute to assessing pre-slaughter animal handling, highlighting potential issues to be followed up with additional measures. The significant correlations between GPA dimension 1 with slaughter order and plasma lactate warrant further investigation, comparing behavioural expression with aspects of meat quality.

Extreme weather events are expected to increase in frequency and/or severity under climate change. Recent examples of these types of events, such as the heat wave in Europe in 2003, have caused con...

Sea urchin roe is a high value product; however, previous attempts to market the roe from the New Zealand sea urchin (Evechinus chloroticus) have been unsuccessful due to its inconsistent sensory quality. The current study investigated the odour active volatile profiles of roe from male and female sea urchins, harvested from two locations in New Zealand, using solvent assisted flavour evaporation (SAFE) extracts and a gas chromatography–olfactometry–finger span cross modality (GC–O–FSCM) method. Panellists detected 81 odour active compounds, 18 of which were identified using GC–mass spectrometry (MS). Generalised procrustes analysis (GPA) of the data revealed that there were differences in the volatile profile that were location and gender specific. These differences may have contributed to the inconsistent sensory properties of sea urchin roe. The differences in volatile profiles between sea urchin populations and genders must be appreciated before developing effective strategies to produce consistent sea urchin roe quality.

The responsiveness of 10 horses and 10 ponies to environmental challenge (represented by an open field test) was assessed using a qualitative approach based on free choice profiling methodology (FCP), which gives observers complete freedom to choose their own descriptive terms. Data were analysed with generalised Procrustes analysis (GPA), a multivariate statistical technique associated with FCP. A cross-validation of the outcomes of this approach to data recorded through quantitative behaviour analysis, and through a questionnaire given to the animals’ owner/riding instructor, was also performed using principal component analysis (PCA). Twelve undergraduate students generated their own descriptive vocabularies, by watching 20 horse/pony video clips lasting 2.5 min each. GPA showed that the consensus profile explained a high percentage of variation among the 12 observers, and differed significantly from the mean randomised profile ( p < 0.001). Two main dimensions of the consensus profile were identified, explaining 60% and 5.2% of the variation between animals, respectively. The 12 observer word charts interpreting these dimensions were semantically consistent, as they all converged towards the same meaning, albeit using different terms. The most used term to describe the positive end of axis 1 was ‘‘quiet’’, whereas ‘‘attentive’’ was the best positive descriptor of axis 2. The most frequently used descriptors for the negative ends of axes 1 and 2 were ‘‘nervous’’ and ‘‘bored’’, respectively. Thus, axis 1 was labelled as ‘‘quiet/nervous’’ and axis 2 was named as ‘‘attentive/bored’’. A marked effect of animal category was observed on the scores of the animals on the first dimension ( p < 0.001). Horses received significantly higher scores, and were thus assessed as more quiet and calm, than ponies. Conversely, ponies tended to receive lower scores on the second dimension ( p < 0.12), therefore they appeared less curious and attentive. The results of the PCA showed that the variables from different types of measurement clearly had meaningful relationships. For instance, the variables with the highest loading on the positive end of axis 1 were all indicative of tractable and docile animals, whereas axis 2 showed high loadings on the positive end for variables indicating attentive animals. Qualitative behaviour assessment proved to be an appropriate methodology for the study of horse behavioural responsiveness, in that it provided a multifaceted characterisation of horse behavioural expression that was in agreement with other quantitative and subjective assessments of the animals’ behaviour.

Summary Rheological and composition measurements used as industrial indices of textural quality in Cheddar cheeses were related to sensory data, obtained by free choice profiling using an untrained panel. Procrustes analysis of compositional and instrumental data on a group of 19 Cheddars displaying a range of textural characteristics showed certain important Instron variables and compositional variables that discriminated between samples in terms of textural characteristics. However, neither analysis discriminated between the samples in the same way as the consumers perceived texture. the second dimension of the sensory data discriminated between samples in terms of mouthfeel, or tertiary textural properties, correlating with moisture content but not easily quantifiable using force-deformation analysis. Sensory dimension 3 related to perceived textural maturity. Composition parameters did not change significantly with maturity of texture, but measures of elasticity and cohesiveness did correlate significantly.

Generalised procrustes analysis and multiple factor analysis are multivariate statistical methods that belong to the family of multiblock methods. Both methods are often used for analysis of data from projective mapping (a.k.a. Napping). In this study, generalised procrustes analysis and multiple factor analysis are compared for a number of simulated and real data sets. The type of data used in this study were (I) random data from Monte Carlo simulations; (II) constructed data that were manipulated according to some specific criteria; (III) real data from nine Napping experiments. Focus will be on similarities of the consensus solutions. In addition we considered interpretation of the RV coefficient and individual differences between assessors.

Canonical variate analysis (CVA) is concerned with the analysis of J classes of samples, all described by the same variables. Generalised canonical correlation analysis (GCCA) is concerned with the analysis of K sets of variables, all describing the same samples. A generalised procrustes analysis context is used for data partitioned into J classes of samples and K sets of variables to explore the links between GCCA and CVA. Biplot methodology is used to exploit the visualisation properties of these techniques. This methodology is illustrated by an example of 1425 samples described by three sets of variables (K = 3), the initial analysis of which suggests a grouping of the samples into four classes (J = 4), followed by subsequent more detailed analyses.

Generalised Procrustes analysis (GPA) is a method for producing a group average from rotated versions of a set of individual data matrices followed by bi-linear approximation of this group average for graphical inspection. Partial Least Squares Regression (PLSR) is a method for relating one data matrix to another data matrix, via bi-linear low-rank regression modelling. The merger of these methods proposed aims to produce an average (e.g. a sensory group panel average), which balances an “intersubjective”, internal consensus between the individual assessors’ data against an “objective” external correspondence between the sensory data and other types of data on the same samples (e.g. design information, chemical or physical measurements or consumer data). Several ways of merging GPA with PLSR are possible, of which one is selected and applied. The proposed “GP–PLSR” method is compared to a conventional GPA followed by an independent PLSR, using a data set about milk samples assessed by a group of sensory judges with respect to a set of sensory descriptor terms, and also characterised by experimental design information about the samples. The GP–PLSR gave a more design-relevant group average than traditional GPA. The proposed algorithm was tested under artificially increased noise levels.

Flash profile is a quick sensory profiling technique designed to meet industrial needs. It is based on the combination of free choice profiling and a comparative evaluation of the whole product set. This study aimed at comparing Flash profile with conventional profiling. This has been done with two product spaces: strawberry blended yoghurts and apricot “fromages frais”, both from the French market. Multidimensional data treatment such as Generalised Procrustes Analysis and Canonical Variate Analysis (CVA) allowed us to assess the correspondence between the sensory maps obtained with both methods. In the case of strawberry yoghurts, very similar results were obtained with the two methods. On the other hand the sensory positioning of the apricot fresh cheeses somewhat differed between the methods. For both product sets, Flash profile was slightly more discriminating than the conventional profile. Advantages and limits of Flash profile are discussed.

The present paper describes and compares three methods to analyse volatile compounds emitted by seven car seat foam samples: “electronic nose” with metal oxide semiconductor (MOS) and metal oxide semiconductor field effect transistors (MOSFET), ATD/GC/MS and sensory analysis. Principal component analysis (PCA) and generalised procrustes analysis (GPA) were used to investigate the discrimination between foam samples obtained separately by the three methods. Multiple factorial analysis (MFA) was applied to explore the relationships between “electronic nose” measurements and sensory assessments, which were discussed according to the identified volatile compounds. The established relationships highlighted the relevance of the “electronic nose” system to partly transcribe the olfactory information given by human panel. However, it brought a lower quantity of information than sensory analysis and an interesting correlation with sensory attributes and chemical compounds, but not as high as we could expect. Consequently, further work needs to be done to improve the correlation.

Proficiency testing in sensory analysis is an important step towards demonstrating that results from one sensory panel are consistent with the results of other sensory panels. The uniqueness of sensory analysis poses some specific problems for measuring the proficiency of the human instrument (panel). Individuals within a panel and people from different cultures may have different thresholds of perception, and product experience of the panel may lead to differences in the ability to discriminate among samples. Such factors make the job of the sensory scientist and statistician more challenging, because of the difficulties in defining the expected level of performance. As part of an EU supported project, ProfiSens, 12 panels undertook descriptive profiling on six samples of red wine. Four were designated 'validation' panels, whose data were used firstly to establish the expected profile results, and secondly to set the performance criteria that a trained sensory panel would be expected to achieve. Four key measures of a panel's performance were investigated: (1) the number of significant sensory dimensions identified by using generalised Procrustes analysis; (2) the number of pairs of samples that a panel found to be different at a specified level of significance; (3) how well the 'sensory map' of a panel agreed with the 'expected sensory map'; and (4) how well the assessors agreed with each other within a panel, and with the consensus configuration. For each of these criteria, an 'expected result' was considered, as was an overall measure of performance. The data from the remaining eight panels were analysed, and the level of performance recorded for each of the stated criteria. Results indicated differing levels of performance. The study also revealed a number of key issues that need to be addressed to successfully run proficiency testing for descriptive profile analysis. A simpler performance scheme is proposed to address issues related to attaching arbitrary weightings to each of the performance criteria and to address potential problems associated with combining different measurement criteria into a single performance score. While some issues still need addressing, this project has made significant contributions to proficiency testing for sensory analysis. © 2002 Elsevier Science Ltd. All rights reserved.

Raske sensoriske metoder har blitt veldig populære i matvitenskap og spesielt i internasjonal matindustri. De er appelerende fordi de er kostnadseffektive og raskere å gjennomføre enn noen av de tradisjonelle sensoriske metoder som vanligvis anvendes for å få tak i informasjon om forbrukere. En metode som har blitt spesielt populær er projective mapping hvor hver forbruker plasserer et antall produkter på et ark basert på deres likheter og forskjeller. Hver forbruker bruker sine egne kriterier for å avgjøre hva som gjør at produkter er lignende eller ikke og hvor de skal plasseres på arket. For å få nyttig informasjon fra disse individuelle produktplasseringene på projective mapping arket trengs det en passende statistisk metode som kan håndtere denne type data. To metoder som har etablert seg for analyse av projective mapping data er generalised procrustes analysis og multiple factor analysis. Begge benytter seg av veldig forskjellige tilnærminger for å håndtere og analysere dataene, noe som gjør at man stiller spørsmål om resultatene fra disse to metodene vil være forskjellige eller ikke. I tilllegg ble det testet en kombinasjon av disse to metodene for aa undersøke om dette kan gi bedre resultater enn generalised procrustes analysis og multiple factor analysis hver for seg. Denne masteroppgaven prøver å gi noe insikt i hvilke forskjeller i resultatene det kan oppstå ved å teste ut generalised procrustes analysis, multiple factor analysis og kombinasjonen av disse to på tre type data: tilfeldige data i Monte Carlo simuleringer; konstruerte eller desginete data som ble manipulert på kontrollerte måter for å kunne undersøke hva slags isolerte situasjoner metodene kan håndtere; ni reelle datasett hvor forskjellige type produkter ble testet av et varierende antall individer. Resultatene av analysene ga ingen kart svar om hvilken metode skulle foretrekkes framfor den andre siden i noen tilfeller generalised procrustes analysis fungerte bedre enn multiple factor analysis og omvendt. Kombinasjonen av de to metodene gaminst tilfredstillende resultater.

Sensory characterisation of virgin olive oil is completely standardised in the European Union. The sensory profile used to evaluate samples is made up of 16 descriptors, three of them being generic. These three attributes [Other ripe fruit(s), Other positive attribute(s) and Other negative attribute(s)] could be of great additional interest in the characterisation of different samples. However, as they are not specifically defined it is difficult to reach a descriptor consensus among assessors evaluating the samples. This situation is similar to that produced when using a Free Choice Profile and so the Generalised Procrustes Analysis (GPA) could be extremely useful in obtaining all the information contained in these three generic descriptors. In this study GPA was applied to the descriptors generated by a 25 member expert panel under the headings Other ripe fruits and Other positive attributes when evaluating 11 samples of virgin olive oil from the variety ‘Arbequina’ grown in three different regions. The Generalised Procrustes Analysis was efficient in the sensory characterisation of the 11 samples studied using the information provided by the two positive generic descriptors included in the official sensory profile. The overall result was equivalent to that obtained using Correspondence Analysis, although the GPA gave more information since it also included the intensity of each of the descriptors and was less affected by odd descriptors. This technique was as useful in the characterisation of the samples as the other six positive descriptors defined in the official sensory profile (Olive fruity, Apple, Green, Sweet, Bitter, and Pungent), providing additional information of great scientific and commercial interest.

There is a growing need to monitor the health and welfare of farm animals, and to develop methods which do so efficiently and reliably. A crucial problem with current methods is the integration of separate measurements taken during farm visits into accurate judgements of an animal's overall welfare state. This paper proposes that the qualitative assessment of animal behaviour may serve as an integrative methodology which could guide the interpretation of other, more detailed welfare measurements. Recent research has demonstrated qualitative behaviour assessment to be reliable and repeatable under controlled experimental conditions. The paper concludes with a discussion of ways in which this method may be adapted for practical on-farm use.

A sensory vocabulary of 20 terms each with a corresponding reference material was developed over 7 sessions using pork patties derived from the meat of carriers and non-carriers of the RN − gene. Patties were oven-cooked at 150 and 170°C and chill-stored for up to 5 days to facilitate warmed-over flavour development. Generalised Procrustes Analysis (GPA) was used to investigate sensory terms and their individual use by panellists over the sessions. GPA explained variance indicated that the final vocabulary displayed a similar amount of information to that of the initial vocabulary of 42 terms. Individual panellists scale use was found to converge over the sessions. Panel agreement on many odour and flavour terms appeared to be enhanced as term synonyms were removed in vocabulary development. Sample discriminability decreased from sessions 1–4, where term concepts were verbally communicated to the panel. Term reference introduction in session 5 caused a levelling in sample discriminability and a reduction in agreement, most likely related to perceptual confusion. Subsequently, references enhanced both discriminability and agreement. Thus, it may be more useful to introduce reference materials earlier, if not in the first session, of the vocabulary development process.

Descriptive profiling is known to require long and expensive training of assessors to provide reliable and consistent results. However, alternative methods for obtaining this type of information are limited. Two experiments were conducted to compare the quality of results obtained by conventional profiling and an alternative solution, a ranking test. In the first experiment, a trained panel of eight assessors evaluated 10 attributes of sweet corn whereas in the second experiment, an untrained panel of 20 assessors performed the same task on the same samples but using a ranking procedure. The results from both experiments are coherent as far as overall product discrimination is concerned. Slight discrepancies were found in the discriminating attributes from one panel to another. These experiments also showed that methods used to measure consensus for profile data such as generalised Procrustes analysis can be applied to ranking data to provide the same kind of results. The findings suggest that whenever one has insufficient time to train a panel, one should consider using an untrained panel and conduct a ranking test. Further studies, however, are needed to verify if ranking tests can be used fruitfully or adapted to a wider range of products and a greater number of attributes.

This study aimed to examine in detail the underlying factors which affect visual texture perception based on existing sensory methodology. The visual properties of dairy dessert samples were manipulated via a full factorial design, with four factors, screened from a fractional design of eight factors. Fifteen assessors, trained solely in visual profiling based on the Quantitative Descriptive Analysis (QDA)1 method, analysed 16 formulations of mousse dessert products over ten visual texture attributes. The visual profile data were examined using analysis of variance (ANOVA), principal component analysis (PCA) and generalised Procrustes analysis (GPA) methods. ANOVA results showed that cream levels significantly affected all the visual textures (p < 0.05). It was also found that levels of blue colourant significantly affected the perception of creaminess (p < 0.05). The results suggest the possibilities of food formulators producing detectable variation in visual texture parameters which may subsequently affect quality perception in chilled dairy dessert products.

Comparisons of volatile compounds released during consumption by different assessors with individual differences in the assessors’chewing patterns, saliva production rates and ultimately their expressions of perceived flavor have received little research attention to date, although such comparisons are fundamental to the understanding of flavor. To address this, eight untrained assessors were chosen and each consumed six Cheddar cheeses during Buccal Headspace Analysis of the volatile compounds released, while in parallel measures of each assessor's mastication behavior using Electromyography, their stimulated saliva production during consumption and their sensory perceptions of the cheeses flavor during Free Choice Profiling were determined. Relationships between the volatile compounds released and the sensory and physiological measures were investigated using Principal Components Analysis, Generalised Procrustes Analysis and Partial Least Squares regression. It was found that although there were differences between assessors’mastication behavior and saliva production rates, the assessors’individual volatile profiles obtained by Buccal Headspace Analysis were similar for each cheese examined. Also, Partial Least Squares was successful in predicting the most important flavor differences between cheeses from the volatile compounds released during their consumption by different assessors.

The results of the analysis of data from a ring trial, carried out by sensory laboratories in seven countries (Denmark, France, Germany, Italy, Norway, Switzerland and the United Kingdom) are reported. Samples of the same batches of twelve varieties of cheese were tested. The protocol for this ring trial was developed from that used in an earlier ring trial in which each laboratory used its normal procedures for descriptive profiling ( Nielsen and Zannoni, 1998 ). The major element of the revised methodology was the consensus vocabulary. All laboratories were able to differentiate between the cheese varieties using the new vocabulary, hence showing a high degree of agreement between assessors within a laboratory. There was a very high degree of agreement between laboratories on the use of attributes for `Texture', with a high degree of agreement for `Aroma, Taste' and for `Smell'. A Generalised Procrustes Analysis (GPA) of the laboratory means showed that the samples lay in the same underlying sensory space and that the laboratories agreed about the location of the samples in the space. There were some differences in the interpretation of the underlying sensory dimensions consistent with differences in the use of the attributes by the different laboratories.