5 Relations between Sensory and Objective Measurements
Flavor Quality: Objective Measurement Downloaded from pubs.acs.org by YORK UNIV on 12/04/18. For personal use only.
for Quality Evaluation of Green Beans JOHN J. POWERS and DAVID R. GODWIN Department of Food Science, University of Georgia, Athens, GA 30602 ROLF E. BARGMANN Department of Statistics and Computer Science, University of Georgia, Athens, GA 30602 For decades, substantial effort has been expended to develop sound objective methods for quality evaluation of foods. Considerable success has been achieved for color measurement; and for texture, moderate success. As was pointed out by Powers and Quinlan (1), part of this success has come about because some of the same forces or properties that cause us humans to respond to the food could be utilized in developing objective tests. Before the origin of gas-liquid chromatography (GLC) two decades ago, objective measurements of the numerous compounds that make up flavor was nigh impossible. Actually, not until a decade ago when Powers and Keith (2) and Dravnieks et al. (3) described practical means of analyzing GLC patterns could GLC measurements be efficiently correlated with flavor (4, 5). Quinlan et al. (6) and Powers (7) have reviewed most of the literature through early 1974. Recent papers are those of Galleto and Bednarczyk (8), Dravnieks et al. (9), Dravnieks (10), Gianturco et al. (11), Jobbágy and Holló (12), Severnants (13), and Powers (14). In spite of much progress, there are still major problems to be solved. Unlike color and texture, the relation between GLC peaks and flavor sensations is peripheral indeed. As Powers (7) pointed out, the properties that enable us to measure a substance chemically may often not be at all related to the properties that cause us to respond sensorially to that compound. In fact, in most cases, we don't know exactly the properties that do make us respond to a compound. Approximately two years ago, we turned to the sensory side to learn if better correlations could be obtained if one dealt with specific flavor or taste descriptions rather than the composite term, flavor. In the intervening years since GLC analysis became practical as a tool for flavor evaluation, others (9, 10, 12, 15, 16, 17, 18, 19, 20) have also endeavored to relate specific odor responses to GLC patterns. * Department of Statistics and Computer Science, University of Georgia. 51
52
FLAVOR
QUALITY:
OBJECTIVE
MEASUREMENT
The problems i n t r y i n g to develop b e t t e r o b j e c t i v e methods for f l a v o r are: (A) We need to l e a r n the p r o p e r t i e s which make us respond s e n s o r i a l l y to t a s t e or odor substances; (B) We probably could use objective/sensory means more e f f i c i e n t l y , even though we are a long way from a t t a i n i n g the f i r s t o b j e c t i v e , i f we could r e l a t e s p e c i f i c terms on both sides of the o b j e c t i v e sensory equation rather than doing as we g e n e r a l l y do now, r e l a t e very s p e c i f i c e n t i t i e s (GLC peaks, f o r example) to a very broad term, f l a v o r ; and (C) No one of the sense m o d a l i t i e s operates i n a vacuum. Each one i s a f f e c t e d by the others. We have to l e a r n more about i n t e r r e l a t i o n s among the senses, e s p e c i a l l y as to sensations which encompass more than one sense modality. We have thus turned to the more general f i e l d of t r y i n g to r e l a t e nuances of sensory response to s p e c i f i c o b j e c t i v e measurements. Bargmann et a l . (21) a p p l i e d component a n a l y s i s to d e s c r i p t o r s f o r blueberry a t t r i b u t e s and q u a l i t y , and Wu et a l . (22) have r e c e n t l y c a r r i e d on a f a c t o r a n a l y s i s of wine d e s c r i p t o r s to i n v e s t i g a t e an o b j e c t i v e means of developing a s u i t a b l e terminology f o r t a s t e and odor sensations. In our laboratory, Hightower (23) has a p p l i e d a minimax approach (24) to component a n a l y s i s of potato chip f l a v o r s . This study i s a part of a continuing e f f o r t to broaden our base of knowledge and methodology so that objective/sensory measurements can be put upon an even f i r m e r foundation. Experimental A p r e l i m i n a r y t r i a l was c a r r i e d on i n mid-1975 f o r the purpose of s e t t i n g up a vocabulary of d e s c r i p t o r terms f o r appearance, c o l o r , mouthfeel, and f l a v o r of canned and frozen beans. Frozen or canned beans were heated f o r s e r v i n g and then sampled by approximately 40 i n d i v i d u a l s who were asked to w r i t e down every sensory response they thought p e r t i n e n t . This l i s t of 53 d e s c r i p t o r s was l a t e r e d i t e d to 27 terms thought to be p e r t i n e n t and not redundant. Sensory e v a l u a t i o n t r i a l s were then made using three commercial brands of canned beans and three of frozen beans, s e l e c t e d from a much l a r g e r group of brands to be sure that the brands a c t u a l l y used d i f f e r e d at l e a s t moderately i n q u a l i t y . The purpose of t h i s phase was to determine whether the terms which the l a r g e group of p a n e l i s t s s a i d were p e r t i n e n t were i n f a c t a c t u a l l y used and of a i d i n d i s c r i m i n a t i n g among the d i f f e r e n t l o t s of beans. A common problem i n a c c e p t a b i l i t y t r i a l s i s to be able to e x p l a i n the r e s u l t s . A p a n e l i s t may r a t e two products as being e q u a l l y acceptable; yet, say the products are d i f f e r e n t , simply because one a t t r i b u t e makes one product d e s i r a b l e whereas a d i f f e r e n t a t t r i b u t e makes the other product e q u a l l y acceptable. T h i s , of course, i s the reason f o r t r y i n g to go behind the general terms, f l a v o r or mouthfeel, f o r example, to seek out the s p e c i f i c sensory q u a l i t i e s which make the food d e s i r a b l e .
5.
POWERS E T A L .
Quality
Evaluation
of Green
Beans
53
F i r s t sensory t r i a l . The three frozen and canned products were evaluated s i m i l a r l y to the procedure described by Vuataz et a l . (25) and Wu et a l . (22). The p a n e l i s t s evaluated the products h e d o n i c a l l y f o r a c c e p t a b i l i t y , appearance, c o l o r , mouthfeel, and f l a v o r . Hereafter, these f i v e f a c t o r s w i l l be r e f e r r e d to as "general d e s c r i p t o r s " to save having to l i s t them each time. The hedonic terms f o r the general d e s c r i p t o r s were l a t e r t r a n s posed to a 9-point s c a l e . The p a n e l i s t s were a l s o asked to r a t e the degree to which each product was stronger or weaker as compared with a reference sample i n a p a r t i c u l a r a t t r i b u t e , using the 27 d e s c r i p t o r s chosen from the o r i g i n a l l i s t of 53 s p e c i f i c d e s c r i p t o r s . One of the brands of canned beans was the reference sample. The same product was a l s o included among the t e s t products as a coded sample to provide a check upon the p a n e l i s t s . The p a n e l i s t s were thus asked to (A) evaluate the q u a l i t y of the beans, using hedonic terms f o r the f i v e general d e s c r i p t o r s , (B) compare each t e s t product against the reference product f o r each of the 27 s p e c i f i c d e s c r i p t o r s , and then (C) re-evaluate the products, using the f i v e general d e s c r i p t o r s . There were 21 p a n e l i s t s and each product was evaluated four times. Before the t r i a l s s t a r t e d , a t a b l e of random numbers had been used to assign each product to a session so that e v e n t u a l l y each product was examined four times. At any one s e s s i o n , the p a n e l i s t s evaluated four samples plus the reference. Sampling was always between 10:30 to 11:30 AM. S t a t i s t i c a l a n a l y s i s . The sensory data were subjected to a n a l y s i s of variance to e l i m i n a t e i n e f f e c t u a l judges or d e s c r i p t o r s . By u n i v a r i a t e a n a l y s i s (MUDAID program) (26), the treatment/ e r r o r F values were computed. Judges or d e s c r i p t o r s , not s t a t i s t i c a l l y s i g n i f i c a n t , were dropped from f u r t h e r consideration, T h i s procedure i s r o u t i n e i n our l a b o r a t o r y (1, J5, 14, _21, 22, 23). The e d i t e d r e s u l t s were then analyzed by f a c t o r a n a l y s i s to detect d e s c r i p t o r s of major importance and to e f f e c t f u r t h e r e d i t i n g of the d e s c r i p t o r l i s t . The minimax program of Bargmann and Baker (24) was used f o r t h i s purpose, coupled with the f a c t o r a n a l y s i s by stepwise maximum l i k e l i h o o d s o l u t i o n and r o t a t e d (oblique r o t a t i o n ) by Thurstone's A n a l y t i c a l Method of Rotation (Harman) (27). A n a l y s i s of variance and Duncan's m u l t i p l e range t e s t s were used to t e s t f o r s i g n i f i c a n t d i f f e r e n c e s between the products, both f o r the general and 27 s p e c i f i c d e s c r i p t o r s . Main t r i a l s . The main experiment c o n s i s t e d of e v a l u a t i n g eight l o t s of beans, much as above. The t r i a l s were c a r r i e d on i n e a r l y summer 1976. There were 27 p a n e l i s t s , very few of whom were on the 1975 panel; there were 20 s p e c i f i c d e s c r i p t o r s (7 having been dropped as a r e s u l t of the 1975 t r i a l ) , and again the t r i a l s were r e p l i c a t e d four times. The beans consisted of four brands of canned beans, three of f r o z e n beans, and f r e s h beans.
54
F L A V O R Q U A L I T Y : OBJECTIVE
MEASUREMENT
The f r e s h beans were used f o r the reference sample; as before, they were also included as a coded sample. The p a n e l i s t s were asked to evaluate f i v e or s i x products at each s e s s i o n . The products had p r e v i o u s l y been assigned to a p a r t i c u l a r session, using, as before, a t a b l e of random numbers. The judges evaluated the beans h e d o n i c a l l y f o r the f i v e general d e s c r i p t o r s , rated them against the reference sample f o r the 20 s p e c i f i c d e s c r i p t o r s , then re-evaluated the beans again f o r general a t t r i b u t e s . The f r e s h beans were cooked by simmering them f o r 25 minutes, an equal weight of beans with an equal weight of 1% s a l i n e s o l u t i o n . The frozen product was prepared by cooking 510 g of beans i n 700 ml of 1% s a l i n e s o l u t i o n f o r 15 minutes. The canned beans were heated f o r 15 minutes i n t h e i r own packing l i q u o r . For each product, a l l samples f o r sensory and o b j e c t i v e t e s t s were withdrawn from the same cooking v e s s e l at the same time. In other words, the beans withdrawn f o r o b j e c t i v e t e s t i n g were cooked e x a c t l y the same as the beans used f o r the sensory trial. The organoleptic t e s t s were thus s t r i c t l y comparable to the o b j e c t i v e t e s t s i n terms of p r i o r treatment. The p a n e l i s t s were given 15-20 g of sample i n a t r a n s l u c e n t p l a s t i c s o u f l e cup. The cups were c o l o r coded, a given c o l o r was not assigned to the same p o s i t i o n , but randomly changed from session to session. Sensory evaluation was done e i t h e r at 9:30 to 10:30 AM or 3:30 to 4:30 PM. At the end of the t a s t e - t e s t i n g sessions, u n i v a r i a t e a n a l y s i s was used, as described above, to e l i m i n a t e p a n e l i s t s or d e s c r i p t ors not s t a t i s t i c a l l y s i g n i f i c a n t as judged by the treatment/error F value. GLC procedure. The GLC procedure was e s s e n t i a l l y the same as described i n e a r l i e r p u b l i c a t i o n s (28, 29). F i f t y grams of beans were placed i n a 2 l i t e r f l a s k with 1 l i t e r of water f o r the combined s t e a m - d i s t i l l a t i o n - s o l v e n t - e x t r a c t i o n procedure of Likens and Nickerson (30). D i e t h y l ether was the solvent. After e x t r a c t i o n , d i s s o l v e d and e m u l s i f i e d water was frozen out at -28 C. The e x t r a c t was decanted from the i c e c r y s t a l s i n t o a KudernaDanish assembly and the volume reduced to 0.5 ml. For GLC a n a l y s i s , a s i n g l e column chromatograph with a 3.66 m, 6.4mm s t a i n l e s s s t e e l column packed with 5% SP-1000 on Chromosorb W-HP, AW, DMCS, 60/80 mesh, was used. Programming was at 6.4 C/min from 25 to 200 C. A 5 y l sample was i n j e c t e d . Inspection of the chromatograms showed 45 d i s t i n c t peaks. Several other peaks were a l s o d i s c e r n i b l e , but only those which were w e l l resolved were used. The 45 peaks were converted to percent area (4). L i q u i d - s o l i d chromatography. N o n - v o l a t i l e and pigment components were analyzed f o r by a LC method. The products were prepared f o r a n a l y s i s by e x t r a c t i n g 50 g of beans with 100 ml of an e x t r a c t i o n solvent composed of acetone, chloroform, and
5.
POWERS E T A L .
Quality
Evaluation
of Green
Beans
55
hexane (1/1/1). The beans and the e x t r a c t i n g s o l v e n t were placed i n a high speed blender f o r 5 min and then separated by two f i l t r a t i o n steps through No. 4 and No. 42 f i l t e r paper, r e s p e c t i v e l y . The e x t r a c t was evaporated at reduced pressure to a dry residue i n a r o t a r y evaporator p a r t i a l l y submerged i n a water bath at 50-60 C. The residue was then r e - e x t r a c t e d with 25 ml of hexane. Drying had a two-fold o b j e c t i v e ; f i r s t , to remove v o l a t i l e compounds which were already being analyzed f o r by GLC a n a l y s i s and, secondly, to remove the p o l a r s o l v e n t s , e s p e c i a l l y water. The p o l a r s o l v e n t s and the h i g h l y - p o l a r water had to be removed p r i o r to the LC a n a l y s i s . The components are c a r r i e d by the i n j e c t e d sample i n i t i a l l y , but when s e p a r a t i o n occurs along the column, p a r t i t i o n i n g between the s t a t i o n a r y phase and the c a r r i e r produces such a high a f f i n i t y f o r the s t a t i o n a r y phase that p o l a r components become non-mobile bands on the column. For LC a n a l y s i s , a Waters A s s o c i a t e s LSC system composed of a Model UGK, U n i v e r s a l L i q u i d Chromatography I n j e c t o r and a Model 6000A solvent d e l i v e r y system with a model 440 absorbance d e t e c t or was used. The flow r a t e was 1.0 ml/min with d e t e c t i o n at 365 nm. The sample s i z e was 10 u l . The s o l v e n t , an adapted v e r s i o n of Pons (31) s o l v e n t , was 750 ml chloroform, 225 ml cyclohexane, 3 ml a c e t o n i t r i l e . and 2 ml 2-propanol. Inspection of the chromatograms y i e l d e d 11 peaks. These peaks were compared as absolute absorbance v a l u e s . 1
Color. The products were analyzed s p e c t r o p h o t o m e t r i c a l l y over the v i s i b l e region of the spectrum. The samples used f o r t h i s a n a l y s i s were a l i q u o t s taken from the hexane-extraction phase of the LC procedure described above. T h i s procedure was q u i t e e f f i c i e n t f o r the e x t r a c t i o n of p l a n t pigments. A Shimadzu Multipurpose Recording Spectrophotometer, Model MPS-50L, equipped with 1 cm c e l l s was used. A l l measurements were made i n the 0-1 absorbance range with necessary d i l u t i o n s to provide on-scale readings. (Just before the t r i a l s s t a r t e d , the instrument went s l i g h t l y out of adjustment so that switching from the 0-1 to the 1-2 range d i d not r e s u l t i n absolute coincidence; r a t h e r than delay the t r i a l s , the d i l u t i o n method was used w i t h backward c a l c u l a t i o n of the values to provide a continuous spectrum.) A f t e r a n a l y s i s , absorbance at 16 d i f f e r e n t wavelength was s e l e c t e d f o r comparisons. Mechanical measurements. Factor a n a l y s i s of the p r e l i m i n a r y t r i a l showed that there was a strong f a c t o r c o n s i s t i n g of "coarse, f i b r o u s , c r i s p , j u i c y , slimy, soggy, and tender." Thought was turned to u t i l i z i n g some mechanical t e s t which might correspond with t h i s sensory f a c t o r . I t was t h e r e f o r e decided to measure the c o e f f i c i e n t of f r i c t i o n between a moveable p l a t e (32) (sled) and the outer surface of the beans. The beans were o r i e n t e d i n two p o s i t i o n s : one was with the beans p a r a l l e d to the f o r c e
56
FLAVOR QUALITY:
OBJECTIVE
MEASUREMENT
and the second was with the beans perpendicular to the d i r e c t i o n of p u l l . The beans were t i g h t l y f i t t e d i n t o an area on the s t a t i o n a r y p l a t e of an Instron, Model 1130, apparatus equipped with a 1 lb. c e l l . The s t a t i c and dynamic surfaces were covered with aluminum f o i l . The s l e d was then p u l l e d on top of the beans u n t i l the |orce became constant. The p l a t e had a surface area of 39.69 cm . I t s dimensions were 6.3 χ 6.3 χ 1.2 cm and i t weighed 169.5 g. The f o r c e was c a l c u l a t e d as Newtons/cm . T e n s i l e strength. The t e n s i l e f o r c e required to p u l l beans apart l o n g i t u d i n a l l y was a l s o measured. The beans were clamped at each end between the jaws of two f i x t u r e s . The ends of the beans were wrapped once with cheesecloth so that the clamps could " b i t e " i n t o the beans s l i g h t l y ; otherwise, the beans tended to s l i p from between the jaws or e l s e the clamping f o r c e caused them to break f i r s t where they were clamped i n s t e a d of i n t e r mediate between the two sets of j a r s . The crosshead speed was 2 in/min. The t e n s i l e strength was taken as the maximum breakingp o i n t f o r c e d i v i d e d by the c r o s s - s e c t i o n a l area of the bean. As before, the t e n s i l e strength was u l t i m a t e l y expressed as Newtons/ cm . Shear f o r c e . The f o r c e required to cut the beans crosswise while i n a Warner-Bratzler type k n i f e assembly was a l s o measured. The crosshead speed was 2 in/min. The shear f o r c e was c a l c u l a t e d from the f o r c e r e s u l t i n g i n f a i l u r e of the bean d i v i d e d by the c r o s s - s e c t j o n a l area of the bean. The values were recorded as Newtons/cm . C l u s t e r a n a l y s i s . The procedure of T r i v e d i (33) was used to carry on " v i r t u a l " c l u s t e r a n a l y s i s . Bargmann and Grainey (34) and T r i v e d i (33) have defined " v i r t u a l . " B a s i c a l l y , the r e l a t i o n may be l i k e n e d to the c l u s t e r of s t a r s one sees when one gazes at the Pleiades against the v a u l t of the heavens and the a c t u a l p o s i t i o n s of the s t a r s . I f one were i n the midst of the P l e i a d e s , the s t a r s would not appear to be c l u s t e r e d at a l l . The c l u s t e r i s i l l u s o r y . We see t h e i r p o s i t i o n s i n three dimensions p r o j e c t e d against the surface of the c e l e s t i a l sphere i n two dimensions. S i m i l a r l y , i t i s w e l l known that random v a r i a b l e s can be regarded as spectors embedded i n the E u c l i d i a n space of i n f i n i t e dimensions. I f the s p h e r i c a l cap which r e s u l t s from the p r o j e c t i o n of the c l u s t e r i s compared with the e n t i r e surface area, one then has a measure of the density of the c l u s t e r . The computer program forms a c l u s t e r when the surface area of a s p h e r i c a l cap i n k dimension extending to the f a r t h e s t p o i n t i n s i d e the c l u s t e r i s compared with the surface area of the k dimension when one a d d i t i o n a l point or a t t r i b u t e i s added to i t . When t h i s r a t i o s u f f e r s a severe drop, a c l u s t e r core i s considered terminated. Those a t t r i b u t e s that have once been included i n a c l u s t e r core are not used again to form another c l u s t e r core. New c l u s t e r s are formed
5.
POWERS E T A L .
Quality
Evaluation
of Green
Beans
57
from the r e s i d u a l a t t r i b u t e s of prior-formed c l u s t e r s . In e f f e c t , the v e c t o r s are p r o j e c t e d onto a Unisphere where the p o i n t s on the Unisphere are connected by great c i r c l e s . The cosine of each of the great c i r c l e s represents the c o r r e l a t i o n between each p a i r of random v a r i a b l e s . The r e p r e s e n t a t i o n of c o r r e l a t i o n s as cosines of angles between v e c t o r s was f i r s t introduced by K a r l Pearson (35) i n 1901. To organize the data f o r c l u s t e r a n a l y s i s , the 27 sensory values that e x i s t e d f o r each product and r e p l i c a t i o n had to be reduced to one experimental u n i t to correspond with the one u n i t which e x i s t e d f o r each of the o b j e c t i v e measurements. The gen e r a l d e s c r i p t o r s were weighted by the f o l l o w i n g formula
- 1 +
w
|x - x 1
k
I
u
1
where χ was the score the judge gave knowingly to the reference sample and χ was the score he gave when he d i d not know he was judging the reference product. I f x^ equals χ , the weight attached to t h i s judgment on t h i s p a r t i c u l a r a t t r i b u t e w i l l be one. I f he makes a r a t h e r extreme misjudgment, the d i f f e r e n c e x^ minus could be, say, 4 p o i n t s . In t h i s case, h i s weight would be only 1/5 of that attached to a judge whose r a t i n g of the reference product was c o n s i s t e n t r e g a r d l e s s of whether he evaluated i t knowingly or unknowingly. For the 20 s p e c i f i c d e s c r i p t o r s , the weighting a t t r i b u t e d to each judge was somewhat d i f f e r e n t .
w
-
1 + jx -5l * u 1
In t h i s formula, χ i s the score the judge gave to the coded reference sample. Since he was comparing i t against the known reference sample and the "no d i f f e r e n c e " score was 5, he should have assigned a 5. I f he assigned any other score, he was being i n c o n s i s t e n t and the weight of h i s score was decreased accord i n g l y . By weighting the general and s p e c i f i c d e s c r i p t o r scores i n the f a s h i o n above, we then had 32 experimental u n i t s (8 products χ 4 r e p l i c a t i o n s ) on which we had a l l the o b j e c t i v e determinations and the composite judgment on each of the 30 d e s c r i p t o r s . A c o r r e l a t i o n matrix was then formed between the 58 v a r i a b l e s on the b a s i s of the 32 experimental u n i t s . The 28 o b j e c t i v e a t t r i b u t e s included 17 GLC peaks, 3 s p e c t r a l r a t i o s , 4 LC peaks, and 4 mechanical measurements. Results
1975
Factor a n a l y s i s . E d i t i n g of the 27 d e s c r i p t o r s used i n the p r e l i m i n a r y t r i a l s r e s u l t e d i n seven being dropped because
58
FLAVOR QUALITY:
OBJECTIVE
MEASUREMENT
they e i t h e r were not used or they d i d not d i s c r i m i n a t e among the products. Factor a n a l y s i s y i e l d e d the eight f a c t o r s l i s t e d below: Factor 1
Factor 4
Off-flavor 0.52 Persistent 0.42 aftertaste Bland f l a v o r 0.31 Pleasant -0.44 aftertaste
Coarse Fibrous Crisp Slimy Juicy Soggy Tender
Factor 7 0.57 0.57 0.53 -0.53 -0.55 -0.63 -0.66
Factor 2 Color-1 0.69 Color-2 0.64 Appearance-1 0.61 Appearance-2 0.58 Bright c o l o r 0.33 Pale c o l o r -0.38 Color o f f -0.43 shade Factor 3 Fibrous Coarse Buttery
0.47 0.46 0.34
Factor 5 Off-flavor Persistent aftertaste Hay-like taste Coarse Fibrous Bland Juicy
0.65 0.55 0.42 0.41 0.38 0.36 0.29
Factor 6 Coarse Fibrous Off-flavor Persistent aftertaste Hay-like taste Slimy
0.50 0.49 0.44 0.43
Soggy Slimy Juicy Buttery Tender
0.50 0.47 0.43 0.41 0.34
Factor 8 Color-1 Appearance-•1 Appearance--2 Color-2 Coarseness Fibrous Crisp Bright color Juicy Slimy Pale c o l o r Color o f f shade Tender
0.55 0.51 0.49 0.49 0.36 0.36 0.33 0.31 -0.32 -0.34 -0.39 -0.46 -0.50
0.33 0.31
Comment w i l l be reserved u n t i l we come to the c l u s t e r a n a l y s i s of the main experiment because the f a c t o r a n a l y s i s and the c l u s t e r a n a l y s i s corroborated each other, except to point out some of the major observations. Factor 7, f o r example, r e l a t e s e s s e n t i a l l y to mouthfeel except the f l a v o r note of "buttery" was a part of that f a c t o r . The same thing may be seen f o r f a c t o r 3; buttery again accompanies the mouthfeel sensations. Factor 4 shows the two sets of terms combined except b u t t e r y i s not s u f f i c i e n t l y w e l l c o r r e l a t e d to come w i t h i n the f a c t o r . Sensory c o n t r i b u t i o n s to a c c e p t a b i l i t y . One of the problems i n d e v i s i n g o b j e c t i v e t e s t s to s u b s t i t u t e f o r or to complement sensory evaluations i s to know how to weight the o b j e c t i v e t e s t so that i t w i l l c o n t r i b u t e i n the same manner to a c c e p t a b i l i t y as does each sense modality. L i s t e d below are the simple and
5.
POWERS E T AL.
Quality
Evaluation
of Green
59
Beans
m u l t i p l e c o r r e l a t i o n c o e f f i c i e n t s f o r the 1975 and 1976 e x p e r i ments . Multiple
correlations
Acceptability Acceptability
Acceptability Simple
1975 T r i a l s
1976 T r i a l s
f l a v o r , mouthfeel, appearance & c o l o r vs. f l a v o r , mouthfeel & appearance
0.917
0.917
0.915
0.916
vs.
0.906
0.912
0.884 0.801 0.626 0.597
0.891 0.794 0.607 0.555
vs.
f l a v o r and mouthfeel
correlations
Acceptability Acceptability Acceptability Acceptability
vs. f l a v o r vs. mouthfeel vs. appearance vs. c o l o r
It i s q u i t e obvious that f l a v o r and mouthfeel e s s e n t i a l l y determine a c c e p t a b i l i t y and that appearance and c o l o r are of l e s s e r importance. This does not mean that appearance and c o l o r are unimportant. Rather, w i t h i n the commercial range, remaining v a r i a t i o n s i n c o l o r or appearance are apparently of l e s s e r importance i n determining a c c e p t a b i l i t y than the v a r i a t i o n s that occur i n t e x t u r a l and f l a v o r q u a l i t i e s . Emotional vs. a n a l y t i c a l judgments. One of the aspects we have been i n t e r e s t e d i n i s the e f f e c t of a n a l y t i c a l thought such as has to go i n t o the r a t i n g of each d e s c r i p t o r r e l a t i v e to the reference sample versus the rather low-key thought o r emotion i n v o l v e d i n r a t i n g foods h e d o n i c a l l y . Should the samples be judged f o r general c h a r a c t e r i s t i c s p r i o r to or a f t e r the a n a l y t i c a l phase? They cannot be judged apart without a great d e a l of e x t r a r e p l i c a t i o n to overcome the v a r i a t i o n r e s u l t i n g from using d i f f e r e n t l o t s and cooking batches. Immediately below are the c o r r e l a t i o n c o e f f i c i e n t s between corresponding general d e s c r i p t ors evaluated before and a f t e r the a n a l y t i c a l comparison phase. Factor Acceptability Appearance Color Mouthfeel Flavor
1 1 1 1 1
vs. vs. vs. vs. vs.
2 2 2 2 2
1975 Experiment
1976 Experiment
0.841 0.833 0.849 0.798 0.859
0.863 0.837 0.850 0.847 0.834
There was some evidence that a f t e r a n a l y t i c a l thought the paneli s t s rated the general q u a l i t i e s of the beans somewhat d i f f e r e n t l y than they had when they were merely expressing l i k i n g - d i s l i k i n g without a l o t of thought as to why. Cluster
analysis.
The three strongest c l u s t e r s
included
60
FLAVOR QUALITY:
f a c t o r s f o r texture, below:
f l a v o r , and
appearance,
OBJECTIVE
MEASUREMENT
They are
Cluster 1—Texture
Cluster 2 — F l a v o r
Cluster
Coarse Fibrous Tender Juicy Crisp Buttery Hay-like f l a v o r Bright c o l o r Slimy Soggy
Flavor-1 Flavor-2 Acceptability-1 Acceptability-2 Mouthfeel-2 Mouthfeel-1 Pleasant a f t e r t a s t e Off-flavor Sweet
Color-1 Color-2 Appearance-1 Appearance-2
listed
3—Appearance
The program was set to place terms i n the core (form a c l u s t e r ) i f the c o r r e l a t i o n c o e f f i c i e n t was above 0.70. Terms not used then went i n t o a r e s i d u a l group of f a c t o r s , from which a secon dary core set could be extracted and i n turn a t h i r d core and subsequent cores, denominated " v i r t u a l c l u s t e r s . " A c t u a l l y , 17 d i f f e r e n t c l u s t e r s were generated. The f i r s t three c l u s t e r s included only sensory f a c t o r s ; the l a t e r c l u s t e r s consisted predominantly of o b j e c t i v e measurements. The 58 χ 58 c o r r e l a t i o n matrix i s too massive to reproduce. From Tables I, I I , and I I I , one can observe some of the sensory-objective c o r r e l a t i o n s . As was true f o r the f a c t o r a n a l y s i s , a t t r i b u t e s which we tend to think of as being texture, f l a v o r , or appearance f a c t o r s are so w e l l c o r r e l a t e d with each other that they sometimes appear i n a d i f f e r e n t c l u s t e r than one would expect. Note that "buttery," " h a y - l i k e f l a v o r , " and " b r i g h t c o l o r " show up i n a c l u s t e r otherwise r e l a t i n g to texture. To i l l u s t r a t e the numerous sensory-objective c o r r e l a t i o n s that d i d appear i n the 58 χ 58 matrix, three l i s t s are given as f o l l o w s : C o r r e l a t i o n of absorbance r a t i o 525/610 nm Color-1 Color-2 Color, off-shade Bright c o l o r GC 24 GC 32 Ratio 467/525 Ratio 525/665
-0.67 -0.64 0.69 -0.77 0.64 0.67 -0.74 0.66
Green veg. t a s t e -0.87 0.76 Buttery f l a v o r Hay-like t a s t e -0.65 0.71 Process f l a v o r
with -0.81 Crisp Coarse -0.75 0.70 Juicy Slimy 0.64 0.71 Soggy Fibrous -0.75 0.78 Tender
Coarse Fibrous Tender Juicy Crisp Buttery Hay-like taste Green veg. t a s t e Bright c o l o r Slimy Soggy Process f l a v o r Color, o f f shade 525/610 Shear f o r c e Sweet Flavor Mouthfeel Acceptability Pleasant a f t e r t a s t e j T e n s i l e strength LC09 Off-flavor GC-24 GC-32 Color Pale c o l o r 525/665
Table I
Buttery -.90 -.94 .94 .89 -.88
Crisp .94 .92 .93 .91
Juicy .95 .92 .95
Tender
-.96 -.97
Fibrous
.97
C l u s t e r No. 1 Bright Color .82 .87 .87 .74 .83 .82 .82 .91
Green Veg. taste .90 .91 .90 .83 .92 .86 .84
Hay-like taste .91 .93 -.94 -.87 .89 -.92
.87 .86 .84 .88 .81 .77 .76 .78 .65
Slimy
Coarse Fibrous Tender Juicy Crisp Buttery Hay«like t a s t e Green Veg. t a s t e Bright c o l o r Slimy Soggy Process f l a v o r Color, o f f shade 525/610 Shear f o r c e Sweet Flavor Mouthfeel Acceptability Pleasant a f t e r t a s t e T e n s i l e strength LC09 Off-flavor GC-24 GC-32 Color Pale c o l o r 525/664
Table I (continued)
.77 .79 .80 .75 .79 .67 .67 .83 .77 .71 .75
Process flavor t
.74 .78 .81 .77 .73 .71 .71 .75 .76 .75 .70 .87
Color off-shade
C l u s t e r No. 1
.75 .75 .78 .70 .81 .76 .65 .87 .77 .64 .71 .71 .69
;
Sweet
.77 j -.74 .79 ! -.79 .76 -.76 .67 -.71 -.61 .68 .84 -.76 -.82 .78 -.60 .68 -.61 .75 .58 -.58 .49 -.53 .44 -.58 .46 -.62 .43 -.57 -.73
j Shear 525/610 f o r c e -.70 -.74 .76 .67 -.61 .84 -.85 -.67 -.71 .50 .43 .40 .48 .50 -.62 .76
Flavor .64 .65 .68 .62 .55 .73 .73 .60 .60 .39 .37 .38 .36 .51 .55 .65 .88
Mouthfeel .64 .65 .65 .59 .51 .74 .77 .59 .59 .42 .36 .31 .31 .42 .57 .69 .95 .90
Acceptability -.60 -.64 .65 .56 -.56 .78 -.74 -.57 -.60 .39 .35 .24 .30 .46 -.49 .74 .84 .74 .82
Pleasant aftertaste
Coarse Fibrous Tender Juicy Crisp Buttery Hay-like taste Green Veg. t a s t e Bright c o l o r Slimy Soggy Process f l a v o r Color, o f f shade 525/607 Shear f o r c e Sweet Flavor Mouthfeel Acceptability Pleasant a f t e r t a s t e T e n s i l e strength LC09 Off-flavor GC-24 GC-32 Color Pale c o l o r 525/665
Table I (continued)
.68 .70 -.66 -.71 .58 -.64 .66 .52 .51 -.58 -.54 -.47 -.45 -.37 .71 -.67 -.51 -.56 -.47 -.31
Tensile strength
1
! ; J ι
!
·
.61 .63 -.72 -.61 .57 -.67 .64 .63 .69 -.56 -.45 -.55 -.67 -.63 .47 -.46 -.56 -.39 -.43 -.48 .33
LC09
C l u s t e r No. 1
1
.54 .55 -.57 -.49 .50 -.67 .67 .50 .48 -.33 -.31 -.10 -.17 -.41 .35 -.63 -.78 -.65 -.76 -.90 .26 .48
Off-flavor
1
-.55 -.54 .62 .60 -.54 .57 -.44 -.55 -.57 .53 .47 .52 .60 .64 -.44 .37 .38 .41 .28 .33 -.34 -.55 -.23
GC-24 -.58 -.54 .62 .61 -.54 .52 -.48 -.60 -.48 .61 .59 .55 .58 .67 -.33 .32 .34 .35 .32 .32 -.25 -.57 -.30 .74
GC-32 Pale color -.54 -.57 .53 .47 -.50 .44 -.52 -.57 -.60 .59 .55 .56 .53 .40 -.55 .37 .31 .08 .27 .19 -.33 -.42 -.17 .25 .49 -.50
Color .57 .60 -.65 -.56 .60 -.50 .55 .63 .67 -.55 -.53 -.74 -.82 -.67 .53 -.28 -.24 -.17 -.06 -.06 .40 .65 .01 -.51 -.49 -.52 -.53 .51 .49 -.51 .55 -.39 -.56 -.43 .50 .50 .48 .45 .66 -.44 .33 .34 .29 .30 .24 -.43 -.40 -.25 .27 .42 -.31 .32
525/665
j
!
.37 .45 -.43 -.26 .29 -.50 .54 .48 .64 -.26 -.16 -.19 -.29 -.35 .53 -.46 -.60 -.37 -.54 -.62 .22 .56 .60 -.24 -.14 .22 -.44 -.24
GC-20
Color Appearance Color, o f f shade Process f l a v o r 525/610 LC 09 GC-24 GC-32 Shear f o r c e Pale c o l o r GC-37
Table I I I
.83
Appearance -.82 -.52
-.74 -.51 .87
-.67 -.48 .69 .71
.65 .43 -.67 -.55 -.63
-.51 -.39 .60 .52 .64 -.55 -.49 -.31 .58 .55 .67 -.57 .74
.53 .18 -.62 -.58 -.57 .47 -.44 -.33
-.50 -.21 .53 .56 .40 -.42 .25 .49 -.55
.38 .27 -.50 -.58 -.40 .49 -.38 -.37 .34 -.33
.52 .42 -.42 -.48 -.41 .32 -.30 -.07 .47 -.27 .34
Color, Process Shear P a l e 525/610 LC 09 GC 24 GC 32 f o r c e c o l o r GC 37 GC 47 o f f shade f l a v o r
C l u s t e r No. 3
!
.88 .90
.84 .82 .74
-.78 -.76 -.65 -.90
.76 .69 .65 .74 -.63
-.62 -.57 -.54 -.49 .35 -.73
-.60 -.54 -.37 -.62 .60 -.46 .53
-.61 -.47 -.57 -.46 .43 -.69 .42 .25
-.51 -.47 -.56 -.31 .26 -.67 .71 .22 .54
C l u s t e r No. 2. i Pleas Ac Off ant ceptTensile Shear a b i l Mouth a f t e r f l a strength Bland GC-20 force Sweet vor t a s t e feel ity
.95 Flavor Acceptability Mouthfeel Pleasant a f t e r t a s t e Off-flavor Sweet Shear f o r c e GC-20 Bland T e n s i l e strength GC-33 Uniform LC09 LC02
Table I I
-.60 -.59 -.46 -.51 .54 -.32 .22 .63 .39 .25
ι
-
525/ 610
.50 ! .50 -.56 -.50 1 .43 -.43 .44 .42 ! .36 -.39 .55 .51 1 .33 -.48! .42 .46 i .48 -.38 -.41 S-.30 .45 .43 S .64 -.46 57 1-.44 .47 -.18 — .32 .56 -.20!-.35 -.72 .30 -.58 j-.22 -.42 .33 -.42 ,-.37 -.30 ! .27 ί-.41 -.12 -.25 ! .24 .04 I-.41 -.63 ! .26
1
1 'UniGC-33 form;LC09 LC02
1
66
FLAVOR QUALITY:
OBJECTIVE
MEASUREMENT
C o r r e l a t i o n s of shear f o r c e w i t h Soggy Coarse Crisp Mouthfeel-1 Slimy Juicy Tender
0.79 0.77 0.68 -0.55 -0.58 -0.71 -0.76
Hay-like f l a v o r Green veg. t a s t e Acceptability-1 Process t a s t e Flavor-1 Sweet Buttery f l a v o r
0.78 0.68 -0.57 -0.58 -0.62 -0.73 -0.76
Bright color 0.75 Pale d o l o r -0.55 Ratio 525/610 -0.57 Color, off-shade -0.62
C o r r e l a t i o n s with LC Peak No. Color-1 0.65 Color, o f f -shade --0.67 Bright c o l o r 0.69 525/610 nm r a t i o --0.63
-0.67 Buttery Green veg. t a s t e 0.63 0.64 Hay-like taste
9 0. 61 Coarse Juicy -0. 61 F i b r o u s 0. 63 Tender -0. 72
Discussion Some of the observations and r e l a t i o n s have been commented on above. One of the s t r i k i n g things with regard to the GLC measurements was that most of the GLC peaks were n e g a t i v e l y r e l a t e d to f l a v o r i n d i c a t i n g that the compounds being measured were detrimental to f l a v o r . Some of the r e l a t i o n s with regard to a b r i g h t c o l o r were a l s o noteworthy. The c o r r e l a t i o n s are l i s t e d below: C o r r e l a t i o n s of b r i g h t c o l o r w i t h Color-1 Green veg. taste Hay-like flavor Crisp Coarse Fibrous GC 20
0.67 0.91 0.82 0.83 0.82 0.87 0.64
Acceptability-1 Flavor-1 Sweet Buttery Process f l a v o r Pleasant a f t e r taste Juicy
-0.59 -0.71 -0.61 -0.82 -0.77 -0.60
Slimy Soggy Tender Pale c o l o r Color, off-shade
-0.65 -0.67 -0.87 -0.60 -0.76
-0.74
The c o r r e l a t i o n s i n d i c a t e the r i s k s involved i f one confines oneself to one sense modality (as many of us have done i n the past). The i n t e r r e l a t i o n s between the sense m o d a l i t i e s i s so strong that p r e f e r a b l y a l l should be evaluated at the same time, e s p e c i a l l y i f one hopes to make informed judgments as to the o b j e c t i v e t e s t s which w i l l n e a r l y r e f l e c t o v e r a l l a c c e p t a b i l i t y . Within the past few years, increased emphasis has been placed on r e l a t i n g d e s c r i p t o r terms to q u a l i t y or s p e c i f i c chemicals (9-13, 15-23, 36-48). The procedures depend on a r t i c u l a t i n g a l i s t of d e s c r i p t o r s based upon " s n i f f i n g " or sampling the food with considerable deep and mature thought as to the sensations being perceived. Objective means of e v a l u a t i n g the
5.
POWERS E T A L .
Quality
Evaluation
of
Green
Beans
67
d e s c r i p t o r s a r r i v e d at should be used to seek out main d i s c r i m i nators (49, 50, 51) because our s u b j e c t i v e impressions do not always c o i n c i d e with the way we a c t u a l l y use words (45), as demonstrated by Lehrer (45), Wu et a l . (22) and i n t h i s study. The u l t i m a t e goal, of course, i s to be able to develop a general equation s u i t a b l e f o r p r e d i c t i n g the q u a l i t y of the food from o b j e c t i v e measurements or a mixture of sensory-objective terms. There are s e v e r a l s t u d i e s where covariance or m u l t i r e g r e s s i o n procedures have been used (13, 52, 53, 54) to e s t a b l i s h r e l a t i o n s between response to one or a few sense m o d a l i t i e s and o b j e c t i v e measurements. This study shows that a l l of the senses should be taken i n t o account because one component o f t e n i n f l u e n c e s one's response to the food i n ways not suspected (or at l e a s t f u l l y understood) unless the data are evaluated objectively. The road to a general p r e d i c t i n g equation f o r q u a l i t y f o r any given commodity w i l l be long and tortuous. Cluster analysis i s but one of the methods a p p l i c a b l e i n moving along that road. I t permits r e l a t i o n s among components to be e s t a b l i s h e d . Component a n a l y s i s should have value, too, i n demonstrating the strength of the a s s o c i a t i o n between p a r t i c u l a r o b j e c t i v e measurements and sensory components. From these analyses, one should be able to reduce the number of v a r i a b l e s which have to be t e s t e d i n the f u t u r e . Sometimes a t e s t can be eliminated because another t e s t measures the same thing or e l s e the t e s t i s of so l i t t l e value i t should be e l i m i n a t e d . Once one has s e t t l e d upon appropriate and p r e c i s e t e s t s , then d e c i s i o n s w i l l have to be made as to how the o b j e c t i v e t e s t s are to be weighed so as to make the p r e d i c t i o n match that which would have been s a i d about the a c c e p t a b i l i t y of the food i f i t had been evaluated s e n s o r i a l l y f o r a l l of the major sense f a c t o r s . M u l t i p l e and simple c o r r e l a t i o n s c a l c u l a t e d between the d i f f e r e n t sense m o d a l i t i e s and a c c e p t a b i l i t y , as we d i d i n t h i s study and i n p r i o r s t u d i e s ÇL, 6^, _7. 14), provides information h e l p f u l f o r that phase of the task. To c a l c u l a t e these c o r r e l a t i o n s i s q u i t e simple. As was pointed out i n the beginning, we w i l l be handicapped as long as we have to use words to describe sensations, because d e s c r i p t o r s are so s u b j e c t i v e , r a t h e r than being able to measure the p a r t i c u l a r property that generates the sensation i n the f i r s t p l a c e . The day when we w i l l be able to measure p r o p e r t i e s i n s t e a d of having to use words f o r most sense responses i s a long way o f f . In the meantime, we can be o b j e c t i v e i n determining the r e l a t i o n s among d e s c r i p t o r s f o r sensations by such methods as f a c t o r a n a l y s i s , which procedure should permit us to p i c k those d e s c r i p t o r s most appropriate and p e r t i n e n t . In turn c l u s t e r a n a l y s i s , component and m u l t i - r e g r e s s i o n techniques should enable us to q u a n t i f y the r e l a t i o n s between sensory and o b j e c t i v e measurements.
68
FLAVOR QUALITY: OBJECTIVE MEASUREMENT
Acknowledgements The authors thank Messrs, Tommy Mundy and Davis Staser, NSF Highschool Summer Participants, who conducted the 1975 sensory t r i a l s and carried on the preliminary s t a t i s t i c a l analyses. Mrs, Louise Wu made the factor analysis. Appreciation i s expressed to Mrs. Wu and Messrs. José Novo, Randall Evans, and Ronald Cox whose efforts permitted the carrying on of the sensory and objective tests simultaneously for the 1976 t r i a l s , Literature Cited 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
11. 12. 13. 14.
15. 16.
Powers, J. J. and Quinlan, M. C., J. Agr. Fd. Chem. (1974) 22, 744-749. Powers, J. J. and Keith, E. S., Abstracts of Papers, 2nd Intern'l. Congress of Food Science and Technol. (1966), p. 439, Warsaw, Poland. Dravnieks, A . , Krotoszynski, B. K . , Abstracts, 154th meeting, Am. Chem. Soc. (1967). Powers, J. J. and Keith, E. S., J. Food Sci. (1968) 33, 207213. Dravnieks, A. and Krotoszynski, Β. K . , J. Gas Chromatography (1968) 6, 144-149. Quinlan, M. C., Bargmann, R. E., E l - G a l a l l i , Υ. Μ., and Powers, J. J., J. Food Sci. (1974) 39, 794-799. Powers, J. J., Proceedings, IV Int. Congress of Food Science and Technology (1974) vol. I I , 173-182, Madrid, Spain. Galleto, W. G. and Bednarczyk, A. A . , J. Food Sci. (1975) 40, 1165-1167. Dravnieks, Α . , Krotoszynski, B. K . , and Shah, J., J. Pharmaceutical Sci. (1974) 63, 36-40. Dravnieks, Α . , i n , Correlating Sensory/Objective Measurements --New Methods for Answering Old Problems, edited by Powers, J. J. and Moskowitz, H. W., Pub. STP 594, Am. Soc. for Testing and Materials (1976), 5-25, Philadelphia. Gianturco, Μ. Α . , Biggers, R. E., and Ridling, Β. H., J. Agr. Fd. Chem. (1974) 22, 758-764. Jobbágy, A. and Holló, J., Nahrung (1976) 20, 295-305. Sevenants, M. R., Abstracts, 168th meeting, Am. Chem. Soc. (1974). Powers, J. J., in, Correlating Sensory/Objective Measurements --New Methods for Answering Old Problems, edited by Powers, J. J. and Moskowitz, H. W., Pub. STP 594, Am. Soc, for Testing and Materials (1976), 111-122, Philadelphia. von Sydow, E. and Karlsson, G . , Lebens.-Wiss. u. Technol. (1974) 4, 152-155. Karlsson-Ekström, G. and von Sydow, E., Lebens.-Wiss. u. Technol. (1973) 6, 86-89.
5. powers et al. 17. 18. 19. 20. 21.
22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
Quality Evaluation of Green Beans
69
Persson, T . , von Sydow, E., and Åkesson, C . , J. Food Sci. (1973) 38, 386-392. Persson, T . , von Sydow, E., and Åkesson, C . , J. Food Sci. (1973) 38, 682-689. Dravnieks, Α., Reilich, H. G., Whitfield, J., and Watson, C. Α., J. Food Sci. (1973) 38, 34-39. Tanaka, T . , Saito, N., and Yokotsuda, T . , J. Ferment. Technol. (1970) 48, 56-62. Bargmann, R. E., Wu, L., and Powers, J. J., in, Correlating Sensory/Objective Measurements--New Methods for Answering Old Problems, edited by Powers, J. J. and Moskowitz, H. W., Pub. STP 594, Am. Soc. for Testing and Materials (1976), 56-72, Philadelphia. Wu, L., Bargmann, R. E., and Powers, J. J., J. Food Sci. (submitted) Hightower, J. Α., M.S. thesis, University of Georgia, Athens, Ga. Bargmann, R. E. and Baker, F . , Proc., Symp. Applications of Statistics (1976), Dayton, Ohio. Vuataz, L., Sotek, J., and Rahim, Η. Μ., Proceedings Fourth Intnl. Congress of Food Sci, and Technol., Madrid, Spain, Sept. 1974. Applebaum, M. and Bargmann, R. E., Tech. Rept. NONR 1834(39), Univ. of Illinois, Urbana, Ill. (1976). Harman, J. J., Modern Factor Analysis, 2nd Ed., (1968), University of Chicago Press. Young, L. L., Bargmann, R. E., and Powers, J. J., J. Food Sci. (1970) 35, 219-223. Milutinovic, L., Bargmann, R. Ε . , Chang, Κ. Y . , Chastain, M., and Powers, J. J., J. Food Sci. (1970) 35, 224-228. Likens, S. T. and Nickerson, G. B., Proc., Am. Soc. Brewing Chemists, Annual Meeting (1964), 5-13. Pons, W. Α., J r . , Assoc. Off. Anal. Chemists (1976) 59, 101-105. Rao, V. Ν. Μ., Unpublished Rept, Dept. of Food Science, Univ. of Ga., Athens, Ga. 30602. Trivedi, S. J., Themis Report No. 17, Technical Report No. 75 Department of Statistics and Computer Science, Univ. of Georgia, Athens, Ga. 30602. Bargmann, R. E. and Grainey, R. W., Themis Report No. 1, University of Georgia, Oct. 1969. Pearson, Κ., Philosophical Magazine (1901) 6, 559-572. Dravnieks, Α., Krotoszynski, Β. Κ., Keith, L., and Bush, I. M., J. Pharmaceut. Sci. (1970) 59, 495-501. von Sydow, Ε . , Andersson, J., Anjou, Κ., Karlsson, G., Land, D., and Griffiths, N., Lebensmittel-Wissenschaft u. Technol. (1970) 3, 11Meilgaard, Μ., Elizondo, Α., and Mackinney, Α., Wallerstein Laboratories Communications XXXIV (1971) 114, 95-110.
70 39. 40. 41. 42. 43. 44.
45. 46. 47. 48. 49. 50. 51. 52. 53. 54.
FLAVOR QUALITY: OBJECTIVE MEASUREMENT Clapperton, J. F . , J. Inst. Brew. (1973) 79, 495-508. Persson, T. and von Sydow, Ε., J. Food Sci. (1973) 38, 377385. Mccredy, J. Μ., Sonnemann, J. C., and Lehmann, S. J., Food Technol. (1974) 28, 36-41. Palmer, D. Η., J. Sci. Fd. Agr. (1974) 25, 153-164. Schiffman, S. S., Science (1974) 185, 112-117. McDaniel, M. R. and Sawyer, F. Μ., Paper presented at the annual meeting, Canadian Institute of Food Science and Technology, Halifax, Nova Scotia, June 1975 (Abstract in Newsletter of Sensory Evaluation Div., Institute of Food Technologists). Lehrer, Α., Language (1975) 51, 901-923. Williams, Α. Α., J. Sci. Fd. Agr. (1975) 26, 567-580. Dravnieks, Α., J. Soc. Cosmet. Chem. (1975) 26, 551-557. Brower, K. R. and Schafer, R., J. Chem. Educ. (1975) 52, 538-540. Harper, R., Applied Statistics (1956) 5, 32-48. Harries, J. Μ., J. Sci. Fd. Agr. (1973) 24, 1571-1581. Frijters, J. E. R., Poultry Sci. (1976) 55, 229-234. Kosaric, Ν., Duong, T. B. and Svrcek, W. Υ., J. Food Sci. (1973) 38, 369-373. Biswas, Α. Κ., Biswas, Α. Κ., and Sarkar, A. R., J. Sci. Fd. Agr. (1971) 22, 196-204. Andersson, Y., Drake, B., Granquist, Α., Halldin, L., Johansson, B., Pangborn, R. Μ., and Åkesson, C., J. Texture Studies (1973) 4, 119-144.
