Ranking for Preference by ronald012589

VIEWS: 4,048 PAGES: 9

									I. Introduction Sensory evaluation is a scientific discipline used to evoke, measure, analyze and interpret sensations as they are perceived by the senses (Gatchalian, 1984). Sensory evaluation utilizes human senses to measure perception and because of this, many opportunities for variability and bias. Thus, sensory evaluation reports require more detailed facts compared to physical or chemical measurements (Meilgaard, 1999). In this current age, the use of sensory evaluation as a tool for measuring product characteristics and acceptability is undeniably mounting. According to Daget (1977), the concept is greatly utilized in the assessment of the product attributes and quality level relative to the degree of consumer preference. Moreover, the method is also important in the selection of best method for efficient use of test procedure. The ultimate providence of manufactured goods is in the hands of its end-users. Thus, a wise producer should want to know more about his potential consumer. In consumer preference or acceptance tests, nature of consumer, gauge of product reaction and importance of consumer test to the manufacturers are important to consider. When conducting sensory evaluation, many variables must be controlled if the object of the sensory test is to measure the true product differences under investigation. According to Meilgaard (1999), it is convenient to group these variables under three major headings, namely: a) test, b) product and c) panel controls. When one of these variables is not properly control, extraneous variables may be introduced and that real variables are obscured and these will result to invalid and unreliable results. In general, there are three major types of sensory tests, namely: a) analytical, b) descriptive, c) acceptance or preference. Under each type are specific methods which are utilized either individually or in combination with each other, depending on the desired information. The information gathered from the test is analyzed using statistical tools. Depending on the extent of inferences and conclusions expected to be derived, a specific statistical method may be selected. In this activity, a simple sensory evaluation test was conducted. The test involved ranking for preference for four different brands of pineapple juice. Twelve panelists were asked to rank the samples based on their preference. The evaluation was conducted properly in such a way no extraneous variables may be introduced. The data gathered was analyzed using Friedman Analysis at 5% level of significance to determine if there was significant difference existed

1

among the four brands of pineapple juice in terms of preference. Since the former test showed significant difference, Fisher’s Least Significant Difference was employed to determine which among the different brands of juices differ significantly from each other.

II. Summary The objective of the sensory evaluation was to use the ranking test as a method of determining preference among the four pineapple juice samples. Four different brands of pineapple juice powdered mix were subjected to a sensory evaluation using a simple ranking test for preference. The samples were coded and presented randomly to the twelve panelists. The subjects were asked to rank the samples according to their preference. Controls for the test, product and panelist were applied and carefully maintained during the sensory evaluation. This was to prevent introduction of extraneous variables that would affect the validity and reliability of the results. The data were analyzed using Friedman Analysis and further multiple comparison procedure using Fisher’s Least Significance Difference (LSD) Test was conducted since the former test showed significant difference. At 5% level of significance, Friedman’s Test showed significant difference in preference among the four different brands of pineapple juice. Brand B (Fresh C) differed significantly from the three other brands; namely, Brands a) A (Kool Aid), b) C (Tang), and c) D (Sunkist). No significant difference existed between Brands C (Tang) and Brand A (Kool Aid); C (Tang) and D (Sunkist); and, A (Kool Aid) and D (Sunkist). Among the four brands of pineapple juice, Brand B (Fresh C) was highly preferred by the 12 panelists as designated by its very small rank total.

III. Objectives Ranking is a simplest way to perform comparisons when the test objective is to compare three or more samples according to a single attribute like sweetness, freshness or preference (Meilgaard, 1999). The test objectives of the sensory evaluation are:   To be able to use the ranking test as a method of comparing the four pineapple juice samples; and, To able to know which of the four pineapple juice samples is the most preferred by the panelists.

2

IV. Experimental To aid product developers in pin-pointing product formulations for consumer testing, different samples are usually presented for preference judgment to untrained panels. This consumer-type test allows for top-ranking products to be selected for final consumer tests. Products ranked highest by the consumer-type panel could be presented for consumer evaluation to determine preference among them or against existing products known to have a considerable market. The genera flow of the activity is given in Figure 1.

Product/Sample Preparation

Short Briefing of the Panelist

Distribution of Samples and Questionnaires to Panelists

Collection of Questionnaires

Analysis and Interpretation of Results A. The Design Physical setting must be designed so as to minimize the subjects’ biases, maximize their sensitivity, and eliminate variables which do not come from the products themselves. In the activity, to avoid subjects from interacting, and so to prevent introduction of biases and distraction, the concept of the booth was used to aid in more or less objective evaluation of the product. The panel booth area was composed of 6 individual booths. The booths were arranged all side-by-side, with two sets of three facing each other. Each booth was exactly the same as the other booth in all aspects. The evaluation was divided into two sets wherein only six panels were allowed to evaluate first then followed by the other six. The room was maintained in a temperature favorable and comfortable to the panelist. Light and other physical conditions did not introduce any destruction to the panel members.

3

The four samples of different brands of pineapple juice were tested to determine the most preferred juice. The preference for pineapple juice was evaluated by 12 panel members. The test set was assessed once only.

B. The Sensory Evaluation Test In ranking test, panelists receive three or more coded samples and are asked to rank samples for intensity of some specific characteristic. Ranking tests are often used to screen one or two "best samples' from a group of samples rather than to thoroughly test all samples. However, no indication of the magnitude of difference between samples is obtained because samples are only evaluated in relationship to each other. The panelists were given four coded pineapple juice samples. Then, each of them was asked to rank the samples based on their preferences. The sample scoresheet used in the activity is given in Appendix 1.

C. The Panel Members A large number of respondents selected through adequate sampling techniques and training are utilized in a consumer test. For consumer panels, the judges are neither given any training nor tests expected to be replicated. Each respondent is expected to be a replication of another. In the activity, 12 panel members were asked to evaluate the samples. All 12 panels tested the four different brands of pineapple juice samples.

D. The Sample Preparation and Presentation The samples originated from four different brands of packed powdered juice. About 45 g powdered juice for each brand was dissolved in 500 ml water to attain a very concentrated pineapple juice. The samples were contained in more or less equal amount in a uniform plastic cups to avoid another variable that will affect the performance of the panelists. Before the session started, the panel members were informed that they will be going to evaluate four different brands of pineapple juices. They were notified that they were going to rank the coded samples based on their preferences. Members worked individually in booths and no discussion took place after session. The panel members were also asked to write the sample code of the product in their respective scoresheet (see Appendix 1 for the sample scoresheet used in the

4

evaluation) at their own convenience. The four samples were presented simultaneously in balanced and random order. There was no instruction to spit the sample tasted but the group provided spitting cups in case the panel decided to dispose tasted sample.

E. Statistical Analysis The results were analyzed and interpreted using Friedman Analysis, a simple ranking test. Analysis by Friedman’s Test is preferred to the use of Kramer’s tables, as the latter provides inaccurate evaluation of samples of intermediate rank (Meilgaard, 1999). In the analysis, the scores were tabulated and the rank sums for each brand were calculated. The value of the test statistic T was then computed (see Appendix 2). Using the Table of Critical Values of the ChiSquare Distribution, the upper 5% critical value of the χ2 with 3 degrees of freedom was determined. Then, the value of the test statistic was compared with the χ2 value. If the value of T exceeds the upper 5% critical value of the χ2 with 3 degrees of freedom, then conclude that significant differences exist among samples. Then, multiple comparison procedure was used to determine which samples are significantly different from each other. In the activity, the Fisher’s Least Significance Difference (LSD) Test was used.

V. Results and Discussion The sum of all ranks per brand is shown in Table 1 and the corresponding statistical analyses in Tables 2 and 3. Table 2 shows the Friedman’s Test and Fisher’s Least Significance Difference (LSD) Test in Table 3.

Table 1. Rank Totals for Each Brand of Pineapple Juice.
Brand A “Kool Aid” 1 3 2 4 3 3 4 2 5 3 6 2 7 3 8 4 9 3 10 1 11 3 12 4 Rank Totals 35 Note: Rank of 1 = best; rank of 4 = worst Subject No. Brand B “Fresh C” 2 1 1 1 1 1 1 2 2 3 1 2 18 Brand C “Tang” 1 3 2 4 2 4 2 3 1 4 4 1 31 Brand D “Sunkist” 4 2 4 3 4 3 4 1 4 2 2 3 36

5

Rank totals ranged from 18 to 35. Brand B got the lowest total rank (18) and Brand D got the highest (36). Brands A, C, and D with nearly identical rank totals were significantly preferred by the panelists in almost the same degree, but significantly less preferred than the Brand B, as indicated by the Brand B’s very small total rank. Using Friedman’s Test, the differences in preference among the four different brands of pineapple juice was determined. Table 2. Friedman’s Test: Simple Ranking Test Parameter Value Test Statistic (T) 10.30 Level of Significance (α) 0.05 χ2 with 3 degrees of freedom 7.815^ Note: If T is greater than χ2, there is significant difference. ^ The chi-square value was sourced out from Walpole (1982). Table 1 shows how the results were compiled and the rank sums were calculated. The value of the test statistic T was 10.30 (see Appendix 2 for the calculation of T). Using the Table of Critical Values of the Chi-Square Distribution, the upper 5% critical value of the χ2 with 3 degrees of freedom was 7.812. Hence, the four different brands of pineapple juice were significantly different at the 5% level in preference since the value of T (10.30) was greater than 7.812. To determine which brand/s was/were significantly different, multiple comparison procedure using Fisher’s Least Significance Difference (LSD) Test was used. Table 3. Multiple Comparison Test Using Fisher’s LSD for Rank Sum Brand C Brand A Brand D Brand B 13* 17* 18* Brand C 4 5 Brand A 1 Note: The value represents the difference between the rank totals of the two brands. * means significant difference (the value is greater than the LSDrank = 12.40) The calculated critical value for the multiple comparison was 12.40 (see Appendix 3 for the computation). Any two samples whose rank sum difference is greater than the LSDrank = 12.40 are significantly different at the 5% level. Therefore, Brand B was significantly different in terms of preference from Brands C, A and D. It was highly preferred than the other three Brands as indicated by its small rank total. Brand C was not significantly different from Brands A and D, nor A from D in terms of preference. 6

VI. Conclusions At 5% level of significance, Friedman’s Test showed significant difference among the four different brands of pineapple juice in terms of preference. The value of the test statistic T = 10.30 was greater than the upper 5% critical value of the χ2 with 3 degrees of freedom = 7.812. To determine which brands were significantly different, Fisher’s LSD Test was used. At 5% level of significance, Fisher’s LSD Test showed significant difference in terms of preference between Brand B (Fresh C) and Brands A (Kool Aid), C (Tang) or D (Sunkist). The rank total difference between any of the two brands is greater than the critical value of the multiple comparison. No significant difference existed between Brands C (Tang) and Brand A (Kool Aid); C (Tang) and D (Sunkist); and, A (Kool Aid) and D (Sunkist). Among the four brands of pineapple juice, Brand B (Fresh C) was highly preferred by the 12 panelists as designated by its very small rank total.

VII. References Daget, N. (1977). Sensory Evaluation or sensory measurements. Nestle Research News. Nestle Products Technical Assisitance Co. Ltd. Switzerland. pp. 43-56. Gatchlian, M. M. (1989). Sensory Evaluation Method for Quality Assessment and Development. Diliman, Quezon City: UP College of Home Economics. Meilgaard, M., Civille, G. V., and Carr, B. T. (1999). Sensory Evaluation Techniques. 3rd ed. USA: CRC Press. Walpole, R. E. (1982). Introduction to Statistics. 3rd ed. Singapore: Pearson Education, Asia Pte Ltd.

7

Appendix 1 Scoresheet for Ranking for Preference for Pineapple Juice

Ranking for Preference Test for Pineapple Juice

Name: _____________________________________________ Set No. ___________________ Date: ______________________________________________ Instructions: You are presented with four coded pineapple juice samples. Please assess them for appearance, flavor, and over-all qualities, and then rank the samples according to your preference, by writing the corresponding sample code in the appropriate column. A rank First is best. Rank First Second Third Fourth Sample Code

Comments: ____________________________________________________________________ ____________________________________________________________________ Thank You!!!

8

Appendix 2 Calculation of the Test Statistic T in Friedman Analysis T = {[12/bt(t+1)]Σrank total2 }-3b(t+1) T = {[12/(12)(4)(5)][352 + 182 + 312 + 362] }-3(12)(5) T = 10.30 Where: b is the number of panelists t is the number of samples or treatments

Appendix 3 Calculation for the Critical Value of the Multiple Comparisons LSDrank = z/2bt(t+1)/6 LSDrank = 1.96(12)(4)(5)/6 LSDrank = 12.40 Where: b is the number of panelists t is the number of samples or treatments Z/2 is used since n > 30 (n is 48)

9


								
To top