45"“ m A STUDY OF WORD VS LETTER GRADING OF TURKEYS Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY WEIIiam Clearcn Mills, Jr. 1960 Isl" LIBRARY Michigan Scam University A STUDY OF WORD VS LETTER GRADING OF TURKEYS By 'William Clearon Mills, Jr. AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Poultry Science 1960 Approved 0%{8W 'William Clearon Mills, Jr. 1 AN ABSTRACT Under the sponsorship of the National Turkey Federation, and the Poultry Science Department, Michigan State University, a study was undertaken to evaluate consumer acceptance of word grades vs letter grades as a system of grade labeling turkeys. Several pre-tests were conducted at'wayne State University, Detroit, Michigan using the Detroit Preference Panel to get consumer preferences regarding word grades "Prime", "Choice", ”Good", and "Commercial" as compared to present U. 5. Grades A, B, and C. The Detroit Preference Panel is selected from a representative group of Detroit residents according to age, education and income. There are from 125 to 150 panel.members who come to'Wayne State University and evaluate products as to their acceptability or rank them in order of preference. About one-half of the panel members make their rankings in the afternoon and the remaining members make their rankings in the evening. The first pre-tests consisted of turkeys labeled Grade A, B and C with each turkey labeled according to its actual quality. The word "Prime" was substituted for Grade A, "Choice" for Grade B, "Good" for Grade B, and "Commercial“ for Grade C. By alternating the various labels on different turkeys it was evident that the most important factor in determining the acceptability of a turkey was its appearance. From these early pre-tests it was also apparent that word labels or grades were preferred over letter grades when the turkeys were of comparable quality. These preferenceS'were significant at the 1 percent level. William.Clearon Mills, Jr. 2 Later preliminary tests comparing groups of turkeys rather than single turkeys revealed similar results. In these tests, when the turkeys labeled with word grade labels were priced two cents per pound higher than the turkeys labeled with letter grades, preference for the turkeys labeled with word grades increased. The results of these tests were significant at the 1 percent level. The main part of the study consisted of actual sales records of turkeys sold in three grocery stores located in different shopping areas of Lansing, Michigan. Thirteen hundred twentybfour Beltsville Small White turkeys ranging in weight from 6 to 8 pounds were sold during the period March 17-19, 1960. The three grade labels tested were "Prime”, "Choice", and Grade A. The "Prime" labels were put on turkeys with less protruding keel bones and which in general had slightly better conformation than the remaining turkeys which were labeled at random as "Choice" or Grade A. Turkeys labeled "Prime” were priced at hl cents per pound and the other two grades were priced at 39 cents per pound. All turkeys were USDA inspected for wholesomeness and were USDA Grade A. Both of these stamps were plainly visible on the breast of each turkey. The only difference in labeling was the colored shipping tag with one of the three grades stamped on each. Each tag was wired to the package at the posterior end of the turkey. Total sales were: "Prime" - 362, “Choice" - 556, Grade A - h06. Statistical analysis on the ekpected 1:1:1 sales ratio revealed that the sales were significantly different from the expected ratio at the 1 percent level. William Clearon Mills, Jr. 3 From the results of this study the author concludes that consumers prefer word grade labels "Prime" and/or "Choice" to Grade A; that consumers prefer the word grade label “Good" to Grade B 3 and that about twentyefive percent of the consumer size turkeys now being sold as Grade A can be selected, labeled as "Prime" and sold at a premium price. It was also found that the consumer will reject turkeys that lack flesh and those with discolorations due to moderate size flesh bruises. Turkeys in the present high B grade can be labeled "Choice" and the consumer will readily purchase them. I I Based on the results of this study, it would be desirable and most profitable to revise the present grade standards of quality as they apply to turkeys to bring them in line with what the consumer considers quality, and to change the present system of ABC grading to a new system embracing the words "Prime", "Choice" and "Good" as grade labels 0 K" \ i; b? 5' A STUDY OF WORD VS LETTER GRADING OF TURKEYS By ‘William Clearon.Mills, Jr. A THESIS Submitted to the School of Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Poultry Science 1960 Approved ACKNOWLEDGEMENTS To my wife and to North Carolina State College, I am most deeply grateful for providing the financial assistance and encouragement to allow me to do graduate study. The author is most appreciative to Dr. L. E. Dawson, Poultry Department, Michigan State University, and to Dr. H. E. Larzelere, Department of Agricultural Economics, Michigan State University, for pre-test data obtained prior to my arrival and for their guidance, their'many suggestions and criticisms during the study and the writing of this thesis. The author is grateful to the Poultry Science Department and the Agricultural Economics Department, Michigan State University, and to the National Turkey Federation for support of this study, and to Mr. Lyle Littlefield, Director of the Division of Foods and Standards of the Michigan Department of Agriculture who gave permission to carry out this study in retail stores. A Special thanks goes to Mr. Bud Schmidt of Schmidt's grocery stores and to his associate, Mr. Mayne Maynard, for their cooperation and assistance in conducting this study in their stores. The author owes much to Mr. J. B. Ward, graduate student, Michigan State University, for making the photographs used in the project. To Mrs. M. J. Bostick and to Mrs. D. Place, Department of Heme Economics, Wayne State University, the author expresses thanks for their assistance in conducting the various panels. ,"(2 re new» n. ~.‘.Q.u§ :r .,.r.\ .n o INTRODUCTION . . . . . REVIEW OF LITERATURE . PROCEDURE . . . . . . RESULTS ....... DISCUSSION . . . . . . SUMMARY AND CONCLUSIONS BIBLIOGRAPHY . . . . . APPENDHQQOOOOO TABLE OF CONTENTS Page 16. 35 39 hi t2 ‘I It“! .I. .uh .,. 2?: m A k . Table l. 2. 3. h. 6. 7. 9. 10. LIST OF TABLES Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoon, May, 1959 o o o o o o o o o o o o o o o o o o o o 0 Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, May, 1959 o o o o o o o o o o o o o o o o o o o o 0 Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoon, May, 1959 O o o o o c o o o o o o o o o o o o o o 0 Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, May, 1959 o o c or. o o o o o o o o o o o o o o o 0 Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoon, May, 1959 OOOOOOOOOOOOOOOOOOOOO Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, may, 1959 O O O O C O O O O O O O O O O C O O O O 0 Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June, 1959 .OOCOOOOOIOOOOOOOO....C Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959 OOOOOOOOOOOOOOOOOOOOOOO Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June, 1959 OOOOOOOOOOOOOOOOOOOOOO. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959 .OOOOOOOOOOOOOOOOOOOOOO iii Page 16 16 18 18 19 20 21 21 22 23 Table ll. 12. 13. 1h. 15. 16. 17. 18. 19. 20. LIST OF TABLES CONTINUED Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June, 1959 O O O O O O O O O O O O O O O O O O O O O O 0 Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959 O O O O O O O O O O O O I O O O O Q 9 O O O 0 Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, October, 1959 C C O O O O I O O O O O O O O O O O O O O O 0 Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, October, 1959 OOOOOOOOOOOOOOOOOOOOOOO Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, Octdber, 1959 OOOOOOOOOOOOOOOOOOOOOCO Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, October, 1959 OOOOOOOOOOOOQOOOOOOOCOO Influence of Price and Grade Label on Preference for Turkeys by the Detroit Preference Panel. February, 1960 COOOOOOOOOOOOOOOOOOO... Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. February, 1960 O O O O O O O O O O O O O O C O O O O O O O 0 Influence of Price and Grade Label on Preference for Turkeys by the Detroit Preference Panel. February, 1960 OOOOOOOOOOOOOOOOOOQOOOO Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. February, 1960 C O O O O O O O O O O O O O O O O O O O O O 0 iv Page 2h 2h 25 26 26 27 28 29 3O 31 Table 21. 22. 23. LIST OF TABLES CONTINUED Page Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. February, 1960 O O O O O O O O O C O O O O O O O O O O O C C O 32 Influence of Grade Label and Price on Turkey Sales, Lansing, Michigan. March, 1960 o o o o o o o o o o o 33 Summary of Turkey Sales, Lansing, Michigan. March,l960..................... 3h Figure I. II. III. IV. VI. VII. VIII. LIST OF FIGURES Illustration of Individual Turkeys Identified for the Detroit Preference Panel . . . . . . . . . . . . . . Illustration of Turkeys Identified by Groups for the Detroit Preference Panel . . . . . . . . . . . . . . Illustration of Turkeys Identified for the Detroit Preference Panel . . . . . . . . . . . . . . . . . . Illustration of Turkeys Identified and Priced for the Detroit Preference Panel . . . . . . . . . . . . . . Display of Experimentally Labeled Turkeys in Okemos Store ogooooooogooooooOoooooo Illustration of Turkeys as Labeled in Sales Test . . Illustration of Price Labeling Used in Sales Test . . Unidentified Shopper Making Purchase During Sales Test................-........ Page 10 10 12 13 37 1. INTRODUCTION Within the turkey industry a debate has arisen as to the desirability of mandatory grade labeling of turkeys as well as the advisability of changing the present grade labeling system. Jessie V. Coles (l95h) discusses two main obstacles to grade labeling. She states "obstacles to grade labeling include the technical problems of measuring and testing qualities to determine conformity to grade. Another Obstacle is that if grades are to be satisfactory they must be based on the qualities that consumers look for in goods". Before an intelligent decision can be made regarding mandatory grade labeling, it is necessary to decide what grade label to use. 'What grade label has the highest psychological significance of quality to the person buying turkeys? To begin to understand consumer preferences as to grade labels, the National Turkey Federation granted financial assistance to Michigan State University to conduct research in this area. The purpose of this study was to evaluate turkey grade labels and their quality'meaning to the consumer. Are the present grade desig- nations A, B and C best; or, would the word labels "Prime", "Choice", "Good", or "Commercial" be more acceptable to the consumer as quality determinants? Another purpose of the study was to determine how and what the consumer considered to be quality. This brings up the question of what is quality, and who is to determine quality and its degree of acceptability? Shall the grades continue to be based on arbitrary technicalities, or shall the grades be built on acceptable attributes as determined by the consumer? There are myriad questions which we could ask that need resolving before we can adequately resolve the question of quality and its designation. For the purpose of this experiment, the author accepts the following definition of quality. Quality is that combination of attributes that have significance in determining the degree of accepta- bility of a product by a consumer. By accepting this definition, the author also accepts that the consumer should be the one to determine what is quality, its degree of acceptability, and how such quality is to be labeled. With these definitions in mind, this eXperiment was designed to test the hypothesis that the consumer will buy a turkey based primarily upon its appearance, and that she will buy the turkey labeled with the grade label which has the highest perceived quality'meaning to her when selecting between turkeys of comparable quality. 3. REVIEW OF LITERATURE The trend in Federal grading of turkeys has been toward total grading. In 1958, 708,9hL,879 pounds of turkey meat was graded by the Federal Government. This represents a 6.7% increase over 1957 and a 551% increase over 19h8, (Kempers, 1959). Effective January 1, 1959, all poultry processing plants con- cerned with interstate commerce were required to operate under compulsory Federal inspection regulations, and all poultry processed therein and packaged for consumers was required to be labeled by age (young, mature, or old). This law has inspired much debate about the future of compulsory grading of turkeys. Wilhelm (1959) summarized the opinions of a National Turkey Federation Committee on Mandatory Grade and Sex Labeling in Turkeys stated, ”The total Objective of the turkey business must be to sell turkeys ... and to obtain this, maximum consumer acceptance is necessary". Wbllney (1960) speaking on the subject of grading stated, "Grading, I believe, is primarily intended as a measure of quality. It affords a yardstick by which these features of the product that bear on its accepta- bility - and in the case of food products, chiefly its eating quality may be indicated." Williamson, Bowen and Genovese (1959) in a report on the eco- nomic effects of beef grades concluded, "It must be concluded, it seems, that the competitive merchandising advantage of "U. 8. Choice" beef has been largely responsible since 1950 for the emphasis given by the food chains to Federal Grades on beef". h. Based on a survey of 25 members of the National Association of Food Chains, Wilhelm (1959) reported that 80% of the Beltsville Small ‘White Turkeys were sold by weight only, 73% of the Broad Breasted Bronze Turkeys were sold by weight and sex, and the remainder were sold in various other ways. This seems to indicate that the larger the turkey the more important identification of sex becomes. Of course this could be merely a preference for smaller turkeys regardless of sex, or an attempt to profit from the consumer's perception of hens as being superior in quality. "All trading, except simple barter, involves price. Price in practically all markets is a common denominator by which are expressed unlike details of size, amount, weight, and quality of products being traded", (Yearbook of Agriculture, 195h). Price then is an American symbol of quality, and any study of grade labeling would be unrealistic without price consideration. Makens and Dawson (1960) reported that personal inspection was a major determinant in the selection of turkeys, and that grade labels "Prime", "Choice", and "Good" were preferred to letter grade labels A, B, and C.‘ This work was based on results obtained from a consumer preference panel where the consumer was placed in a buying situation without influence of brand names. They also conducted a mail survey of consumers to determine the opinions of a larger segment of the consuming public. Results of the survey were partially contradictory to the results of the preference panel. One important aspect of the mail survey was that the consumers did not have an opportunity to inspect turkeys as the questionnaire was answered in the home. A majority of the consumers S. surveyed did not remember if the turkeys where they shopped were labeled as to grade. These consumers probably were more grade conscious than the average since they have received a number of publications regarding foods and food products from the Michigan State University Cooperative Extension Service. Makens (1960) reported that grade labeling results in discrimi- nation against birds marked B or Choice by members of the Detroit Preference Panel. Earlier work did not test the preference for word grades when different prices were attached to the various grade labels, and only individual turkeys were used for comparison. A search of the literature revealed no other work pertaining to word vs letter grades on turkeys. PROCEDURE The design of this study was to use the Detroit Preference Panel for preliminary studies or pre-tests in order to gather as much information as possible about consumer attitudes and opinions as well as preferences regarding the use of word grades vs letter grades on turkeys. The Detroit Consumer Preference Panel was selected in the following manner. "Mail questionnaires were sent to about 6500 names obtained from the latest Detroit telephone directory. About 5% of these were returned by the post office for nondelivery. About 20% of the remainder were filled out and returned either from the first or follow up questionnaire. "The basic questions asked concerned age group, education group, income group, and willingness to come to a display room in Detroit to rate samples of the different products displayed. A majority of those selected for the panel were in the middle income group, and had received a high school education, and were in the 31 to h5 age group", (Larzelere and Gibb, 1956). i There were from 125 to 150 panel members (male and female) who came to Wayne State University and evaluated the turkeys used in this experiment as to their acceptability or ranked them in order of prefer- ence. About one-half of the panel members made their rankings in the afternoon, and the remaining members made their rankings in the evening. Several pre-tests were made at different times to determine the reliability of rankings. The initial pre-tests consisted of Broad Breasted Bronze Turkeys labeled "Prime”, "Choice”, and "Commercial" to 7. correspond with Grades A, B, and C, respectively. The various labels were switched between individual turkeys and groups of turkeys in an attempt to arrive at the perceived quality connotation of each word and the type of turkeys the consumer would associate with each word. Both hens and toms were used. In later pre—tests, the word "Good" was used to designate a Grade B turkey for comparisons. In this phase the consumer merely ranked the turkeys in order of his or her preference. Typewriter symbols %, &, #, *, and () were used to identify the turkeys to eliminate any possible meaning which might have been associated with letters or numbers had they been used in identifying the turkeys. These symbols together with the word or letter grade were placed on cards in front of each turkey. See figures 1 and 2 for examples. The consumer ranked the turkeys in order of preference from.most to least acceptable by merely putting the number 1 by the symbol on the panel report card that corresponded to his or her first preference and the number 2 by the symbol that corresponded to his or her second choice, etc. See appendix page h? for sample panel report card. It is obvious that large price differences would result in a shifting of selections. However, from the theoretical economic stand- point, when two turkeys are of the same sex, of equal size, fleshing and finish and apparent quality, any price increase of one over the other should result in the lower priced turkey being selected. This should not hold true if one grade label has a perceived higher quality meaning than another label. The author was primarily concerned with the significance of grade labels and not price. 0 . _ - 7135- ' u. 5. ' ‘ '71. 5. P l" I me r-.- C [,0 ice Co MMERIRI Figure I. Illustration of Individual Turkeys Identified for the Detroit Preference Panel “9" .. Pr imc. ~: ;._ Figure II. Illustration of Turkeys Identified by Groups for . . the Detroit Preference Panel 9. Price differences of three cents per pound between grades were used in the May 1959 pre-tests to observe the possible effect of price on selection. Three cents per pound was used since at that time the price difference between Grade A and Grade B was three cents per pound and different grades were being used for comparisons. The Detroit retail price of forty-five cents per pound for Grade A turkeys was used as a base price in the final pre-tests in February 1960 where in.most cases turkeys of comparable quality were being used to evaluate significance of grade labels. Price differences were used to see if price would alter the consumers' previously expressed. preference for word grade labels where no price had been used. See figures III and IV. The final phase of the study was a sales experiment designed to test the validity of the preferences indicated by the Detroit Prefer- ence Panel. To determine this validity, actual sales records of turkeys from three grocery stores located in different shopping areas of Lansing, Michigan, were obtained and analyzed. Since space was a limiting factor (space was not available to test all combinations previously used), turkeys labeled Grade A, "Prime", and "Choice" were used in this test. The timing of the sales test was excellent because we were able to take advantage of a planned pre-Easter turkey'sale. In this way many more observations were obtained in a shorter time than would have been possible otherwise. The "Prime" label was placed on the turkeys with less protrud- ing keel bones and which in general were of slightly better conformation 10. 1’ Choice. Figure III. Illustration of Turkeys Identified for the Detroit Preference Panel Figure IV. Illustration of Turkeys Identified and Priced for the Detroit Preference Panel than the turkeys labeled Grade A or ”Choice". The grade A and "Choice" labels were put on the remaining turkeys at random as needed to fill the sales area. See figure V for illustration of sales display. The actual labeling was done by the store personnel who knew nothing about the study except that the labels were to be wired onto the turkeys as needed to restock the sales area. All turkeys used in the sales trial were Commercial Beltsville Small White Turkeys weighing from 6 to 8 pounds. All were USDA inspected for wholesomeness and were USDA Grade A. Both the wholesomeness and the USDA.Grade A labels were plainly visible on the breast of each turkey. The only difference in labeling was the colored shipping tags (see appendix page ho for tags used) which were wired to the posterior end of the package. See figure VI for illustration. In previous studies at Michigan State University, MacNeil, Larzelere, and Dawson, (1958) consumers have shown a preference for poultry meat packaged in green cartons or containers. For this reason the green tag and arrow on point of sale poster was used for the Grade A label in order to give every advantage to these turkeys. This tag carried the grade label, store imprint, the net weight, price per pound and the total price of the turkey. See figure VI. The "Prime" turkeys were priced at hl¢ per pound since some selection based on general conformation had been made. The Grade A and "Choice" turkeys were priced at 39¢ per pound. Price being a symr bol and a nexus of quality it was thought that these turkeys would sell in a 1:1:1 ratio. The sales data were analyzed by the Chi Square test using this as the expected sales ratio. 12. SCHMIDT'S PRE-EASTER BELTSVILLE TURKEY SALE CVVOVCEF (fiflflDEFA ilk" lln- ~11:me fur... I TURKEY Figure V. Display of Ekperimentally Labeled Turkeys in Okenos Store 13. Figure VI. Illustration of Turkeys as Labeled in Sales Test 1h. egos node» an soup mqaamnaq conga mo noapaasaaaau .HH> unease 15. Fifteen hundred turkeys were available for the study during the period March 17 - 19, 1960. At the end of the sales period the unused labels of the different types were counted to determine the number of each grade designation sold. This number was compared with the number of unsold turkeys as a double check. The statistical analysis of the results of the Detroit Preference Panel where more than two samples were compared was by computing the coefficient of concordance. The levels of the coefficients were tested for significance by the Chi Square statistic, (Kendall, 1955). See appendix page h3 for computation. The statistical analysis of the results of the February 1960 Panel where only two samples were compared was by use of the Chi Square statistic, (Snedecor, 1956). See appendix page hh for compu- tation. The sales trial was also analyzed by the Chi Square statistic. See appendix page hS for computation. 16. RESULTS In order to evaluate the relative quality meaning of various words as grade labels for turkeys a series of tests was conducted to observe the relative preference for the words U. S. Prime, U. 3. Choice, and U. S. Commercial as grade labels. The turkeys in this class were labeled as to quality represented. The turkey labeled U. S. Prime was high A quality; U. S. Choice was B quality'due to a breast bruise; U. 5. Commercial was B quality due to pin feathers, a small breast bruise, and torn package, otherwise it was A quality. Results are shown in Tables 1 and 2. Table 1. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoong_flay,#l959. Cents % Grade per lst Rank Symbol label spound Aplace 1 2 3 Persons % U.S. Prime 39 68.8 33 8 7 & U.S. Choice 36 h.2 2 33 13 () U.S. Commercial 33 27.1 13 8 27 Table 2. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, May, 1959., Cents 3 Grade per lst Rank Symbol labe1* ,pound pplace l 2 pg} Persons % 11.5. Prime 39 10.6 39 28 15 & U.S. Choice 36 h6.3 38 37 6 () U.S. Commercial 33 6.1 S 16 S9 *The labels on the U. S. Choice and the U. S. Commercial turkeys were reversed in the evening; no other changes were made. 17. The data show a marked increase in preference for the label "U. S. Choice" over the label "U. 3. Commercial" as evidenced by the increased numbers who selected this label over the number who selected the ”U. 3. Commercial" label in the afternoon, and also by the wide differences in the second and third preferences of the two birds. The data also support the original hypothesis that the consumer will buy a turkey based primarily on appearance, and that she will select the turkey labeled with the grade label that has the highest quality meaning to her. The panel rejected the bird with the breast bruise (U. 5. Choice in the afternoon and U. 5. Commercial in the evening). This rejection because of moderate size breast bruise was evidenced through- ‘out the project. The data were subjected to a statistical analysis (coefficient of concordance) which showed the panel members were in significant agreement at the 1 percent level. This means that the panel members had a community of preference in their evaluation of the series of samples. This does not mean that there is necessarily a significant difference between any two samples within a series. In another series of tests turkey hens were labeled with a variety of grade labels to observe indicated preferences. The turkeys labeled U. 3. Grade A and U. S. Prime were high A quality; those labeled U. 5. Grade B and U. 5. Choice were high B quality. The results of this series are shown in Tables 3 and h. 18. Table 3. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoon, May, 1959. Cents % per lst Rank Symbol Grade label “pound pplace l 2 3 3h Persons % U.S. Grade A Hen 39 8.3 h 18 18 8 & U.S. Grade B Hen 36 6.3 3 5 S 35 () U.S. Prime Hen 39 58.3 28 9 6 S % U.S. Choice Hen 36 27.1 13 l6 l9 - These data show that the turkey labeled U. S. Prime was pre- ferred over the turkey labeled U. 3. Grade A. It also shows that U. 3. Choice was preferred over U. S. Grade B as a grade label for turkeys, and that a higher percentage of the panel members preferred a high B quality turkey labeled U. 5. Choice than preferred a high A quality turkey labeled U. 8. Grade A when the price difference was 3 cents per pound. It is interesting to note that no panel member ranked the turkey labeled U. 8. Choice as fourth on the preference rank. Agreement was significant at the 1 percent level using the coefficient of concordance statistic and Chi Square test. Table h. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, May, 1959. -—_‘ J‘— _#7 Cents % per lst Rank Symbol Grade label pound __p1ace 1 2 3 h Persons % U.S. Grade A Hen 39 h.9 h 20 37 21 & U.S. Grade B Hen 36 9.8 8 7 1h 53 () U.S. Prime Hen 39 h8.8 LC 17 18 7 e' U.S. Choice Hen 36 36.6 30 38 13 1 19. All labels remained on the same birds in the evening. It is interesting to note the selection of the turkey labeled U. S. Choice as second preference in both afternoon and evening sections of the panel. Again this turkey was of high B quality and only one panel member ranked it in fourth place. This indicates that the perceived.meaning of the U. S. Choice label is higher than the U. 3. Grade A label. Such selections by the panel members support the original hypothesis. Rankings in Table h are in significant agreement at the 1 percent level. In order to determine the acceptance of similar grade labels on both hens and toms, a series of tests was set up using both sexes. The hens used in this test were of comparable quality. The tom labeled U. 5. Grade A was slightly'red, and the tom labeled U. 5. Prime was in a bag which was torn at the wing and tail. Other factors such as fleshing, finish, and conformation were comparable. Results are shown in Tables 5 and 6. Table 5. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Afternoon,_May,_l9S9. _-'_ Cents % per lst Rank Symbol Grade label _pound _place II 2 3 h Persons a U.S. Grade A Tom 36 12.5 6 5 15 22 % U.S. Grade A Hen 39 27.1 13 21 6 7 & U.S. Prime Tom 36 lO.U S 8 20 15 () U.S. Prime Hen 39 35h.2 26 1h 7 1 The data show that the preference for toms is less than for hens, and that the percent of first preference for the toms was in 20. favor of the U. 3. Grade A. However, when the total rankings for the toms are considered, preference was for the U. 5. Prime Tom. Table 6. Influence of Grade label and Price on Preference for Turkeys by the Detroit Preference Panel. Evening, May, 1959. Cents g per lst Rank Symbol Grade label ,pound ,place 1 2 3 3H— Persons * U.S. Grade A Tom 36 8.5 7 6 26 h2 % U.S. Grade A Hen 39 17.1 1h h2 13 23 & U.S. Prime Tom 36 15.9 l3 l3 32 23 () U.S. Prime Hen 39 63.h 52' 19 9 l Labels remained on same birds in the evening. Total rankings as well as the percentages of first preference were in agreement during the evening section of the panel. When results from both afternoon and evening panels are averaged together, the first preference is for the turkeys labeled U. S. Prime. The slight redness may have had some effect on the appearance of the U. 5. Grade A tom due to the light differences between the afternoon and night rankings. The data also show that hens are more preferred by consumers than are toms. Rankings were in significant agreement at the 1 percent level. In June, 1959 another series of tests was conducted with the same Preference Panel. A different group of turkeys was used to test the same grade labels and to determine the reliability of the results of the first panel tests. During this series of tests some of the labels were switched from one turkey to another to further test the general hypothesis that 21. the consumer will buy a turkey based primarily upon appearance and that she will buy the turkey with the grade label which has the highest perceived quality meaning to her provided the turkeys are of the same quality. The results of the first series are shown in Tables 7 and 8. Table 7. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June, 1959. % lst Rank Symbol Grade label* ,place 1 2 3 Persons % U. 8. Prime 53.3 2h 12 9 & U. S. Choice h2.2 19 18 8 () U. 5. Commercial h.h 2 16 27 E-‘The turkey labeled U. 3. Prime was a top A quality turkey; the turkey labeled U. 5. Choice was of A quality except for small bruises; the turkey labeled U. 5. Commercial was lacking in flesh. Table 8. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959. % lst Rank Symbol Grade label _place 1 2 3 Persons % U. 5. Prime 59.7 to 17 10 & U. S. Choice 32.8 22 31 1h () U. 5. Commercial 7.5 5 19 1.3 No changes were made in labeling from the afternoon section of the panel. Rankings, both afternoon and evening, were in significant agreement at the 1 percent level and support the general hypothesis. To determine if preference changes when labels are switched to different turkeys, a class of turkey hens was set up using birds of 22. different quality. Turkeys labeled U. 3. Prime and U. 5. Grade A were high A quality. Turkeys labeled U. S. Grade B and U. 5. Choice were low A or high B quality. The results are shown in Tables 9 and 10. Table 9. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June, 1959. %77 lst Rank Symbol Grade label place 1 2 3 U Persons % U. S. Grade A 51.1 23 8 12 2 () U. 3. Prime 35.6 16 19 6 h & U. 8. Choice 6.7 3 17 2h 1 * U. S. Grade B 6.7 - 3 2 2 38 While the percentage of first preference between the Grade A and Prime is large, when the total rankings of these two turkeys are analyzed there is no significant difference between them. There were no differences in the percentage of first preference between the Choice and Grade B, but when these two rankings are analyzed there is a significant difference in favor of the turkey labeled Choice. Rankings as a group are in significant agreement at the 1 percent level. Table 10 shows the results when the labels were switched. The turkey listed in Table 9 as Grade A is now'labeled Prime and the turkey listed in Table 9 as refine; is now labeled Grade A. Likewise the labels between the Chtice aud Grade B turkeys were reversed. 23. Table 10. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959. % lst Rank Symbol Grade label ,place 1 2 3 3h Persons () U. 3. Prime 55.2 37 16 7 7 % U. 3. Grade A 26.9 18 23 22 h i U. 8. Grade B 11.9 8 10 19 3O & U. S. Choice 9.0 6 18 2O 23 Results show strong preference for the turkey labeled Prime. This indicates that the panel members perceived the first turkey listed in Table 9 and 10 to be more desirable. This perceived superior quality was not significant in Table 9, but is significant in Table 10 when the turkey was labeled U. 5. Prime which is in agreement with the original hypothesis. The same was true for the turkey labeled U. S. Choice as shown in Table 9 and Grade B in Table 10. This selection of the turkey perceived to be of superior quality strongly supports the general hypothesis. Another series of tests using both hens and toms was tested. In this class the labels were reversed between the afternoon and evening sections of the panel. The hens were comparable A quality, and the toms were comparable A quality. The results are shown in Tables 11 and 12. 2h. Table 11. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, June,,l959. lst Rank Symbol Grade label _place 1 2 3 h Persons % U. 5. Grade A Hen 22.2 10 23 S 7 () U. S. Prime Hen h2.2 19 9 12 S & U. S. Prime Tom 26.7 12 7 l7 9 a U. 5. Grade A Tom 6.7 3 6 12 2h The turkeys labeled U. S. Prime ranked high in percentage of first preference. The rankings of this class are in significant agree- ment at the 1 percent level. For the evening evaluation of this same class the labels were reversed between the hens and between the toms. No other changes were made. The results of this change are seen in Table 12. Table 12. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, June, 1959. 3P lst Rank Symbol Grade label place 1 2 3 3H Persons () U0 5. Prime Hen 5307 36 18 8 S i U. 5. Grade A Hen 22.h 15 31 1h 7 i U. 5. Grade A Tom l6.h 11 h 21 31 & U. 8. Prime Tom 9.0 6 16 2h 21 In the afternoon (Table 11) one hen was preferred most often. In the evening, with the labels being reversed, another hen was preferred and in each case it was the hen labeled U. S. Prime. 25. The difference between the rankings of the toms in the after- noon (Table 11) was significant, however, in the evening section the difference was not significant. This indicates that one of the toms was perceived to be of better quality and that the label U. 8. Prime was most preferred as evidenced by the change in significance. The results of both of these sections of the panel support the original hypothesis. Rankings are in significant agreement at the 1 percent level. Two classes of turkeys were retested in October. In this series the turkeys labeled U. 3. Prime and U. S. Grade A were of compar- able A quality, and the turkeys labeled U. S. Grade B and U. 5. Choice were of comparable B.quality. The results are shown in Tables 13 and 1h. Table 13. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon,_0ctober, 1959. if lst Rank Symbol Grade label place 1 2 3 U Persons () U. 5. Prime 59.3 35 17 h 3 r ' U. 5. Grade A 28.8 17 22 19 1 % U. 5. Grade B 3.U 2 6 13 38 & U. 5. Choice 8.5 S 16 23 15 The preference for the turkey labeled U. 3. Prime is Obvious at first glance. However, the preference for the turkey labeled U. 5. Choice when compared to the Grade B becomes most obvious when you note the number of people ranking the turkey labeled U. 5. Choice as second or third. 26. The results of this test support the original hypothesis, and the rankings are in significant agreement at the 1 percent level. Table 1h. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, October, 1959. %' lst Rank Symbol Grade label place 1 2 3 AU Persons () U. 8. Prime 69.0 to lo 1; I. * U. S. Grade A 13.8 8 29 17 h 1 U. 5. Grade B 3.h 2 7 21 28 & U. S. Choice 13.8 8 12 16 21 No changes were made in labeling the turkeys from those in Table 13. Rankings by this group of panel members are the same as the rankings in Table 13. There is significant agreement of rankings at the 1 percent level. Another test using both hens and toms was conducted at the same time. The hens were comparable A quality and thetpms were comparable A quality. The results are shown in Tables 15 and 16. Table 15. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Afternoon, OctOber,_l959. -;_ ‘;m %V lst Rank Symbol Grade label ,place 1 2 3 HP _fii Persons () U. 5. Grade A Hen 22.0 13 23 13 10 * U. 5. Prime Hen hh.l 26 16 9 8 i U. S. Grade A Tom 28.8 l7 13 13 16 8‘ U0 so Prime Tom 1001 6 9 23 2]. Rankings are in significant agreement at the 1 percent level. 27. Table 16. Influence of Grade Label on Preference for Turkeys by the Detroit Preference Panel. Evening, October, 1959. _—% lst Rank Symbol Grade label place 1 2 3 4h Persons () U. S. Grade A Hen 2h.6 1h l9 l7 7 * U. S. Prime Hen 52.6 30 7 ll 9 % U. S. Grade A Tom 15.8 9 17 17 1h & U. 5. Prime Tom 7.0 h 1h 12 27 Labels remained on same birds in the evening. The panel members failed to rank the turkey labeled U. S. Prime Tom over the turkey labeled U. S. Grade A. This indicates that the panel members perceived the U. S. Grade A Tom to be of superior quality as compared to the turkey labeled U. S. Prime Tom. This action supports the original hypothesis. Again rankings were in significant agreement at the 1 percent level. In previous tests the panel members had been ranking individual turkeys in order of preference. One seldom finds in grocery stores a selection of turkeys that do not vary in some way. The panel members were now asked to select their preference from turkeys where some variation existed similar to what one finds in grocery stores. Each class of turkeys consisted of two groups of two turkeys each, and each group contained some variation; but, the variation within one group (in the author's Opinion) did not exceed the variation within the other group in the same class. 28. The first class was set up to evaluate preference of turkeys of the same quality using two different grade labels, and to get some idea of the quality connotation of the grade labels Prime and Grade A when applied to groups of turkeys. The turkeys in this series were of comparable A quality with the only difference being the grade label and price per pound. Table 17. Influence of Price and Grade Label on Preference for Turkeys by the Detroit Preference Panel. February, 1960. Cents % Cents 3%- Z per lst per lst no Class Label pound ,place Label ,pound ,place choice* I A1 Prime ... 8h.7 Grade A ... 11.8 3.h I E2 Prime 15 81.0 Grade A ts 15.2 3.8 III A Prime 71.2 Grade A 23.7 5.1 III E Prime U7 8h.8 Grade A h5 12.7 2.5 I A - Afternoon - 59 Panel members 2 E - Evening - 79 panel members *The cards of two panel members in both the afternoon and evening sections were unusable. At times one or two panel members were unable to arrive at a decision of first choice which explains the variation in percent no choice. In each class the turkeys labeled Prime were preferred over those labeled Grade A. This preference for Prime increased when the price was increased 2¢ per pound over the Grade A price. The results are significantly different at the 1 percent level using the Chi Square statistic. See Appendix page h3 for computation. 29. A class of turkeys was set up to test the theory that the best turkeys now being offered as A quality can.be selected, labeled as ”Prime" and sold at a premium price when the remaining A quality turkeys are labeled as "Choice". The results of this test are shown in Table 18. Table 18. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. February, 1960. "II Cents % Gents i— g per lst per lst no Class Label pound place Label ,pound place choice VA]. Prime ... hug]. ChOice so. 5205 301‘» VE2 Prime h? h5.6 Choice US 51.9 2.5 1A - Afternoon - 59 panel members 2E = Evening - 79 panel members The data indicate that the consumer perceives the word "Prime" and ”Choice" as being practically equal in their quality connotation. The differences are not significant at the 5 percent level. Over h5 percent of the panel members preferred the turkeys labeled Prime at 2¢ per pound higher than the turkeys labeled Choice. The data indicate that a selected portion of turkeys now being marketed as Grade A can be sold at higher prices when they are labeled Prime and the remaining A quality turkeys are labeled Choice. Price studies would be needed to further test these results and to determine the amount of premium one could expect. Two classes were set up to determine whether consumers would purchase turkeys labeled Choice that were B quality'because of small bruises, small areas of missing skin, processing defects, and a few scattered pin feathers. These turkeys were A quality flesh and finish. The turkeys within these two classes were comparable B quality. 30. Table 19. Influence of Price and Grade Label on Preference for Turkeys by the Detroit Preference Panel. February, 1960. Cents % Cents $7 $ per lst per lst no Class Label gpound Aplace label pound place choice II A1 Choice 89.8 Grade B 6.7 3.14 II E2 Choice 141 89.9 Grade B hi 5.1 5.0 IV A Choice 89.8 Grade B 5.1 5.1 IV E Choice h3 96.2 Grade B bl 1.3 2.5 1A - Afternoon - 59 panel members 2E = Evening - 79 panel members The results indicate that consumers will accept Grade B turkeys of good flesh and finish but with minor defects when such turkeys are labeled Choice. This acceptance is conclusively shown by the vast majority selecting the turkeys labeled Choice as compared to similar turkeys labeled Grade B. Preference for the turkeys labeled Choice increased by S.h percent when priced 2¢ per pound higher than those labeled Grade B. The differences between Choice and Grade B labeled turkeys are significant at the 1 percent level. Previous tests showed that consumers prefer the grade label Choice to Grade B. The following classes were set up to determine the relative quality'meaning of the grade label Good vs Grade B when they are compared to the grade label Choice. In this test the turkeys labeled Choice were A quality. The turkeys labeled Good and Grade B were comparable B quality'due to small breast bruises, pin feathers and damaged package materials. Results are shown in Table 20. 31. Table 20. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. Februarngl960. Cents % Cents % % per lst per lst no Class Label pound _place Label pound place choice VI A1 Choice 000 7907 600d .00 17.0 I 3011‘» VI E2 Choice 15 65.8 Good 83 31.6 2.5 VII A Choice ... 78.0 Grade B ... 15.2 6.8 VII E Choice 1.5 ' ]8.5 Grade B 1.3 18.9 2.5 1 A. - Afternoon - 59 panel members E = Evening - 79 panel members The data show that with the use of the grade label Good 31.6 percent of the panel members ranked these turkeys as first preference when the price difference was 2¢ per pound; whereas, only 18.9 percent of the panel members selected the turkeys labeled Grade B as their first preference with the same price difference. Results show that the grade label Good is more acceptable than the label Grade B. The data also show that when there was a price difference, the percentage selecting turkeys labeled Choice decreased when compared with those turkeys labeled Good; whereas, when comparable turkeys labeled Choice are compared to turkeys labeled Grade B, the percentage preferring Choice increased slightly when the price difference was 2¢ per pound. There was a significant difference at the 1 percent level between the number of persons selecting Choice and those selecting either Good or Grade B. 32. It was established by this time that the panel members preferred the word grade labels to the letter grade labels. To determine to some extent if the use of these word grade labels could narrow the price spread between Grade A and Grade B turkeys, two classes were set up where the turkeys labeled Grade A and Choice were comparable A quality, and the turkeys labeled Good and Grade B were comparable B quality with moderate size flesh bruises on the posterior end of the keel bone and with B quality finish. Results of this test are shown in Table 21. Table 21. Influence of Grade Label and Price on Preference for Turkeys by the Detroit Preference Panel. February, 1960 Cents 37_~ Cents 8% % per lst per lst no Class Label «ppund place Label _pound __place choice VIII :11 Grade A 81.1; Grade B 13.6 5.1 VIII E2 Grade A LS 75.9 Grade B 111 21.5 2.5 IX A Choice ... 79.7 Good ... 17.0 3ch IX E Choice hS 88.6 Good h3 8.9 2.5 1A - Afternoon - 59 Panel members 2E - Evening - 79 panel members The data show that only when the price spread is about h¢ per pound will the consumer accept birds with moderate flesh bruises, and then only about 21 percent will select them. In this test narrowing the price Spread resulted in an increased preference for the turkeys labeled Choice. This was probably due to the price spread of the turkeys labeled Choice and Good, and to the objection of the breast bruises. This was not totally uneXpected since price is a symbol and nexus of quality to most consumers. 33. The main part of the study consisted of actual sales records of turkeys from three grocery stores located in different shopping areas of Lansing, Michigan. Space was not available to test all the comparisons previously made with the Detroit Preference Panel. Since only 1,500 Beltsville Small White Turkeys were available for the sales study, only grade labels Prime, Choice and Grade A were used. All turkeys in the sales trial ranged in weight from 6 to 8 pounds. All turkeys were USDA inspected for wholesomeness and were USDA Grade A. Both the wholesomeness and the USDA Grade A labels were plainly visible on the breast of each turkey. The only difference in labeling was the colored shipping tag which was wired to the package at the posterior of the turkey. The tag carried the store imprint, grade label, net weight, price per pound and total price per turkey. The turkeys labeled Prime were priced at forty-one cents per pound and the turkeys labeled Choice and Grade A were priced at thirty-nine cents per pound. The results of this sales experiment are shown in Tables 22 and 23. Table 22. Influence of Grade Label and Price on Turkey Sales, Lansing, Michigan. March, 1960. % of 1 Number sold _r total Store Prime Choice Grade A Total sales Cedar 120 178 12h h22 31.9 Saginaw lhh 2h8 158 550 hl.5 Okemos 98 130 12h 35 2 26.6 lSchmidt's independent retail grocery chain in Lansing, Michigan 3b. This table shows the distribution of sales in each of the three stores used in the study. Table 23. Summary of Turkey Sales, Lansing,,Michigan. March, 1960. irof Number total label sold sales Prime 362 27.3 Choice 556 h2.0 Grade A 1:06 30. Total l32h 100.0 This table shows the number of turkeys of each grade label sold. An analysis of these results by the Chi Square test for significance revealed significant differences in sales between the grade labels at the 1 percent level. This analysis was based on the hypothesis that the turkeys would be purchased in a 1:121 ratio. This hypothesis was rejected and it was concluded that grade labels had significance in influencing purchases. 35. DISCUSSION OF RESULTS Grades were originally developed to facilitate trade by producers and distributors for trading in wholesale quantities and terminal market transactions. More recently they have become increas- ingly useful to retailers in buying and merchandising products and to consumers in buying. One of the main problems of establishing grade standards and using grades is that such standards must be based on the qualities consumers look for and recognize as being quality determinants of a particular product. This study was an attempt to determine what consumers look for in determining quality of turkeys and how such quality should be labeled. Value and quality of a product are what the individual perceives as value and quality, nothing more. 'Who is to say that a turkey that has been classified as B.quality'because of some missing skin, a broken bone, or one that is lacking in finish is not a high quality turkey? There is no known difference in the nutritive value of turkeys that do not conform to the present arbitrary standards of A quality as compared to those turkeys that do conform to these standards. Certainly the consumer did not reject such turkeys in this study when these turkeys were labeled with word grade labels, but they did reject them when they were labeled with Grade B - a term that by its definition implies something of inferior quality. 36. It was found that lack of flesh and discoloration from bruises were the two most important defects that were unacceptable by the consumer. It was also found that word grades Prime, Choice, and Good were grade labels that were perceived to represent high quality and these labels were proved conclusively to be more acceptable than labels Grade A and Grade B. Tests using the Detroit Preference Panel were in agreement that word grade labels were a more acceptable means of identifying turkey quality. This was also proved in the sales test by 132h consumers tho bought turkeys with their own money. The-sales test is even more conclusive when it is remembered that 69% of the purchasers during this test actually bought turkeys labeled either Prime or Choice when the USDA Grade A label was plainly visible on the package of each turkey offered for sale. These purchases proved that the hypothesis that the turkeys would be purchased in a 1:1:1 ratio was incorrect. This means that the grade labels used signifi- cantly influenced purchases of turkeys. See Figure VIII. The series of tests conducted with the Detroit Preference Panel also showed that the consumer purchases turkey based mainly on the appearance of the turkey as evidenced by the continued selection of a particular turkey even when the label was changed. In each case where this happened it was noted that the word labels Prime and Choice influenced the rankings, in some cases changing the significance of the differences between the two turkeys being compared. Such action on the part of the panel members and in the subsequent sales test prove that the original hypothesis that the 37. Figure VIII. Unidentified.Shopper Making Purchase During Sales Test 38. consumer will purchase a turkey based primarily on its appearance and that the consumer will purchase the turkey labeled with the grade label that has the highest perceived quality connotation when the turkeys are of comparable quality is correct. While it is not the purpose of the author to editoralize, it does seem inconsistent not to mention the potential value of this study to the turkey'industry. - The turkey industry cannot afford to let their opinions regarding grade labeling stagnate and become resistant to change in face of a changing food.marketing era. To do so invites competing products to continually prosper at the eXpense of the entire turkey industry. There seems to be some "magic" associated with the terms 8 "Prime" and "Choice" as symbols of quality, and those engaged in the production, merchandising and retailing of turkeys would profit substantially by taking advantage of these silent salesmen. The results of this study would also apply to the marketing of cooked items as well as parts. The economic potential which would result from the revision of the present grade standards of quality and adopting the word grade labels used in this study is tremendous. 39. SUMMARY AND CONCLUSIONS Under the sponsorship of the National Turkey Federation a study was undertaken to evaluate consumer preference of word grades vs letter grades of turkeys. Several pro-tests were conducted at wayne State University in Detroit using the Detroit Preference Panel to get consumer reaction to word grades "Prime", "Choice", "Good", and "Commercial" in comparison to Grades A, B, and C. Later actual sales of turkeys were conducted in three grocery stores located in different shopping areas of Lansing, Michigan. From the results of this study several conclusions can be drawn. The consumer prefers word grades "Prime", and/or "Choice" to Grade A. Sixtybnine percent of the sales were of the "Prime" and ”Choice” grades while only thirty-one percent was of Grade A. The data further show that about twenty-five percent of the con- sumer size (h-lh pounds) with the best conformation can be sold at a premium price when labeled "Prime". Since the sales trial gave validity to the results of the Preference Panel, it is concluded that the consumer prefers the word grade "Good" to letter Grade B. It was found that the consumer will reject turkeys that lack flesh and those with moderate discolorations due to bruises. ho. This rejection of bruised turkeys held true with the use of the preferred ”Choice" label. This study revealed that a major revision of the turkey standards of quality is needed to take advantage of the more liberal opinions which consumers have as to what represents acceptable quality and how such quality should be labeled. Consumers demonstrated during this study that they do not like and will not buy turkeys labeled Grade B. They do like and will buy the same turkeys if they are labeled Choice or Good. Based on the results of this study, it would be desirable and profitable to change from letter grading of turkeys to the use of word grades, and that this change would be readily acceptable by the consumers. Such a revision of the present grade standards would be profitable to producers, distributors, and retailers and at the same time give to the consumers a turkey labeled in such a.manner that is more acceptable and has more meaning. bl. BIBLIOGRAPHY Anonymous, Yearbook gfggriculture, U. S. Government Printing Office, Washington 29, D. C. 195h. p. lh2 Coles, Jessie V., Yearbook 2: Agriculture, U. S. Government Printing Office, Washington 2;, D. C. 195h. p. 169 Dawson, L. E., Annual Report to National Turkey Federation on project "Grade and Sex Labeling of Market Turkeys”. 1960 Kempers, B. W;, Volume of Products Graded by the Grading Service, USDA, Agriculture Marketing Service, Poultry Division, 1959. p. l Kendall, Maurice G., Rank Correlation Methods, Hafner Publishing Co., New York, 1955. pp. 93-10; Larzelere, H. E. and Gibb, R. D., Consumers Opinions 2: Quality in Pork Chops. Michigan Agricultural Experiment Station Quarterly Bulletin, vol. 39, No. 2, pp. 327-333. 1956 MacNeil, J..R., Larzelere, H. E., Dawson, I“ E., The Influence 2f Certain Packaging Materials and Processing'Tagspgn the Consumer Preference 22 Chicken Fryers. Michigan Agriculture Experiment Station'efiarteriy'Bfiiletin, vol. 80, No. 3, p. 593, 1958 Makens, J. 0., Unpublished Masters Thesis, Michigan State University, East Lansing, Michigan. 1960 Snedecor, George W., Statistical Methods, The Iowa State College Press, Ames, Iowa. Fifth Edition 1956, pp. 18-30. 'Wilhelm, L. A., Report to the Directors of the National Turkey Federation, 1960. Williams, W. F., Bowen, E. K., Genovese, Frank, Economic Effects 33 U.S. Grades for Beef, U.S. Government Printing Office, washington 29, D.C. 1959. p. lhO Wollney, Frank G., Speech presented before the National Turkey Federation, January 5, 1960 142. APPENDIX h3. Computation of the Coefficient of Concordance Table 1. Influence of grade label and price on preference for Turkeys by the Detroit Preference Panel. Afternoon, N 3'1959 Cents % per lst Rank Symbol Grade label ,pound ,place 1 2 3 1 U. 5. Prime 39 68.8 33 8 7 a U. 5. Choice 36 h.2 2 33 13 () U. S. Commercial 33 27.1 13 8 27 Symbol Rank times number within each rank totaled % 133 E“. ranks of all samples 110 number of ranks E .. .397. _ 3 X ', 9506667 M - persons making rankings h8 N - samples 3 Symbol ..Sc' - d d2 - 95.667 ~25.667 658.795 - 95.667 11.333 l28.h37 992.667 sum of actual deviations w . 12 (d2) . 11912 . .2151; M2(N3-N) 33293 12 - M(n-1)w h8(2)(.215h) 20.678h* *Significant at the 1% level 2. Computation of Chi Square for class 1A table 17 number % Class Label selecting* lst choice I A Prime 50 8b.? I A Grade A 7 11.8 * 2 panel members made no selection 2 x2 _ _gE (f1 ' F1) X2 - Chi Square F _ f1 - Observed frequency Fi - Expected frequency X2 _ ._550—29.5)2 + (7-29.5)2 59 926.5 15.70* *Significant at the 1% level hS. 3. Computation of Chi Square for sales test % of number total Label sold sales Prime 362 2703 Choice 556 h2.0 Grade A h06 30.2 132h 100.0 2 x2 _ £(fi - Fi) F Expected sales ratio 1:1:1 (hh7.33-362)2 + (55541.7.33)? + (hh7.33-h06)2 hh7.33 20787 hh7.33 h6050* *Significant at the 1% level 946. peep moamm mafiozo mermnw amapcora ov rows manned 0.3. bf]. Undo phonon Hogan £95m .m 35530 .x. . t. n .fi a e .95 ..IIIIIA fins ....III... .9 a a 5 e .E C H 3230 deacon «o 838 o .8558... - -7 w, E552: Ease 28:.on is. {If— MICHIGAN STATE UNIVERSITY LIBRARIE IIL ILILIILIILLII II I 48