The Effects of Water and Heat Variation on the Volume of Long-Grained and Short-Grained Rice - Research Paper Example

A Screening Design to Assess the Effects of Water and Heat Variation on the Volume of Long-grained and Short-grained Rice Obtained after Cooking : Institution: Abstract The results of a screening experimental design were used to obtain the optimum temperatures and water needed to separately prepare a long-grained and short-grained rice meal to the highest volume possible. Temperatures were alternated at 50 and 60 degrees Celsius, while water volumes were alternated between one and two cups respectively. It was found that the data obtained followed a close pattern between the long-grain and short-grain versions, indicating that the conditions needed to achieve highest volumes for each variant were suggestively similar. Increasing the temperatures from 50 to 60 degrees Celsius resulted in a reduction in the total volume of rice obtained, while increasing the water volume resulted in an increase in volume of cooked rice. Introduction Background At various stages of testing the usefulness of an industrial entity, the management is faced with situations that require concrete experimental evidence to enable them make informed decisions and, in the meantime, set a basis for scientific reasoning behind major decisions. There are several types of designs of experiments that can be adopted in such situations. However, going deeper into the natures of individual situations, it is apparent that individual, specialized designs of experiments suit different experimental situations. One such design is the screening design. Screening tests usually require a limited number of runs which makes them resource/ cost-effective and time-saving. At times, further analysis is needed, which means that screening designs only serve as a yardstick for further research. However, this happens only when it is absolutely necessary to generate further data based on the components already tested. Screening designs are an integral component of various product launching activities. Their importance stems from the fact that through the screening tests, experimenters are able to locate certain compounds/ elements/ components that have higher effect on the overall effectiveness of the end product, therefore suggesting more attention to them unlike those that show limited effect on the end product (Stout & Hardwick, 2005). Furthermore, screening designs are highly relevant when the situation requires that the experimenter locates compounds or components that are suitable for the manufacture of a certain product. Through an accept-reject criterion, the experimenter is able to establish whether some of the components are really worth retaining (Stout & Hardwick, 2005). For those that appear effective to the final product, it is always a question of whether their effectiveness differs across the different components. Statement of the Problem and the Research Question Numerous factors are thought to influence all simple and complex processes. The researcher was confronting the question: do variations in the amounts of water, heat, and grain-type affect the amount of rice yield after cooking? The rationale behind carrying out this experiment is to establish a domestic and industrial standard by which cooking each type of rice (short or long-grained) would be based. Although guiding notes have always been provided to guide such activities as cooking of rice, including the amounts of each ingredient to use, such manuscripts could easily be rendered irrelevant by the varying produce qualities over time, as technology keeps perfecting every day. This screening experiment was designed to examine some of the factors that affect cooking of rice. In particular, variation in the amount of cooked rice that results from regulation of the amount of heat and water and varying between long and short-grain rice types was investigated. In the experiment, the amount of heat is factor A, with the levels of interest being 50 and 60 degrees Celsius at low and high respectively. The amount of water is factor B, with the high level showing the utilization of 2 cups of water and the low level showing the utilization of 1 cup of water. The experiment was carried out at two levels: once using long-grained rice and at the second level using a short-grained variant. Three cups of rice were used in each case. The factor levels were intended to help bring out the optimum amounts of each substance to use. The Variables As already described, four variables were identified for the study. The dependent variable (DV) is the amount of rice produced through cooking, bearing the conditionality difference in the various setups. This variable is measured in terms of cubic centimetres. The independent variables (IV) include the amount of heat applied at a specific stage (measured in degrees – Celsius), the amount of water measured in cups (each cup has approximately 250 ml of water when full), and the type of rice (either long-grained or short-grained). The amount of rice inputted was not considered as a viable variable since a constant volume was used across all experiments. Hypotheses Three hypotheses were formulated to help explain the intended results. They are: H1: Variation in the amount of heat does not impact the final volume of rice cooked; against: varying the amount of heat results in different volumes of rice yields. H2: varying the volumes of water used to cook rice does not affect the final volume of rice obtained; against: varying the volumes of water impacts the volume of rice produced. H3: Long-grained and short-grained rice yield equal amounts of rice at similar conditions of water and heat regulation; against: varying volumes of rice are obtained for long-grained and short-grained rice types cooked at similar heat and water conditions. Limitations and delimitations of this study The study was limited by the fact that it did not incorporate other ingredients used in the preparation of a rice meal, despite there being many such ingredients. While the implications of these unconsidered ingredients cannot be underestimated, it is worthwhile to note that in practice, many people do not take rice with any other ingredients apart from water and the rice grains. The researcher is of the opinion that medical reasons, personal likes, and extreme poverty could be some of the reasons behind this tendency. That in itself proves that basing the study on these two most common ingredients ensures the results cater for group of rice consumers. Owing to financial and time constraints, it was difficult to extend this study to other types of rice. In fact, different manufacturers have different preferences for production mechanisms, as it is these mechanisms that set them apart from their peers, and help retain a specific market end to themselves. However, long-grained and short-grained rice types are common for every producer. Therefore, the long-grain/ short-grain criterion is the most general way to approach a question on rice types. Terms and Definitions Long-grained rice: rice that is considered thin and long. Short-grained rice: rice that is generally short from end-to-end, but mainly appears stout. This report contains a summary of the research aims and findings in the form of an abstract, followed by an introductory section, a review of literature, methods section, data analysis and findings section, and a conclusion based on the findings. Methods The findings of this experimental study are based on the findings of two related experiments carried out at varying conditions of the independent variables. The amount of heat is factor A, with the temperatures being controlled at 50 and 60 degrees Celsius at low and high respectively. The amount of water is factor B, with the high level showing the utilization of 2 cups of water and the low level showing the utilization of 1 cup of water. The experiment was carried out at two levels: once using long-grained rice and at the second level using a short-grained variant. Experimental designs (of which screening designs are part) offer real data necessary for situational analysis. They are simple to carry out at many times, and they offer cheap alternatives that knowledgeable individuals can adopt for personal trials. Screening designs are among the cheapest types of experimental designs to setup, and they consume relatively little time to complete (Stout & Hardwick, 2005). Technically, this study involves establishing whether water, rice type, and variation in temperatures are significant determinants of the final volume of rice obtained after cooking. That, in itself, forms the technical grounds on which the screening design was chosen. The arrangement adopted (by order of replications) assumes a pre-test post-test control group design. By arranging each factor as a block, the effects of environmental factors are equally distributed among the replicates. That way, the burden of accounting for the error arising from such factors was remarkably diminished. Consequently, the study reels on the assumption that the error term is small enough not to influence the outcome of the analysis. Using statistical software such as SPSS to generate the results would enable easy determination of the existence of an error term. However, using the manual format used, the exact value of the error term is significantly more complex to establish, and, therefore, it was considered to have no bearing on the analysis-results. The main threat to validity is the possibility of regression between pre-test and post-test values. A correlation analysis on the data allows the researcher to establish whether there is high correlation that would indicate existence of regression. However, this threat to internal validity was addressed by introduction of two parallel tests (for each type of rice), which allows assessment of values at different points, and closely monitor the trends in the output. The budget for the experiments’ supplies was initially estimated at 10 Kg of rice at for $20, uncontaminated 10 liters of water for $10, two thermometers at $5 each, and a volumetric flask at $25. The test was carried out at the 5% level of significance (α). The power of the test was set at 80%. The power of the test is the probability that the test is able to detect the null hypothesis as being false when it is indeed false. Naturally, the higher the power of a statistical test, the lower the threshold of rejecting the null hypothesis. By maintaining 80% as the power and 5% as the statistical significance of the tests, a fair balance that would not overstretch either the power or the p-value was established. The tests were replicated three times, as shown in the data tables 1 and 2. In this design, which is a 22 design, the high and the low levels of both A and B are depicted by (+) and (–) respectively on the A and B axes. Hence, (+) on the A axis stands for the high level of heat and (–) stands for the high level of heat. This representation is the same in the B axis but stands for the number of cups of water. The reason for using the above factor levels is because they will effectively help us determine the amount of heat and the amount of water needed to achieve the highest possible volume of cooked rice. This particular experiment is replicated 3 times which means there is 12 runs in number. The order of the runs is totally random and not preset, hence; this is an experiment that is completely randomized. The data obtained were recorded in tables 1 and 2 below. Table 1: Factor, Treatment Combination, Replicate, Total for Short-grained Rice Factor Treatment Combination Replicate Total 1 2 3 A B - - A low, B low 29 25 26 80 + - A high, B low 38 34 28 100 - + A low, B high 16 21 23 60 + + A high, B high 35 25 30 90 (Source of data: Experiment 1) Table 2: Factor, Treatment Combination, Replicate, Total for Long-grained Rice Factor Treatment Combination Replicate Total 1 2 3 A B - - A low, B low 30 23 24 77 + - A high, B low 37 35 29 101 - + A low, B high 15 21 23 59 + + A high, B high 36 26 29 91 (Source of data: Experiment 2) Data Analysis and Findings Analysis of the data obtained for the tables was done concurrently for the sake of comparisons. Remarkably, the results obtained across the two tables were closely related for the different points they were obtained for. To get the effect of A at the high and low levels of B, the following calculations are done. (Experiment 1): A = 1÷ [(2×3) (90+100-60-80)] = 8.33 (Experiment 2): A = 1÷ [(2×3) (91+101-59-77)] = 7.4375 To get the effect of B at the high and low levels of A, the following calculations are done. (Experiment 1): B = 1 ÷ [(2×3) (90 + 60 – 100 – 80)] = - 5.00 (Experiment 2): B = 1 ÷ [(2×3) (91 + 59 – 101 – 77)] = - 5.357 To get the interaction effect AB, the following calculations are done. (Experiment 1): AB = 1 ÷ [(2×3) (90 + 80 – 100 -60)] = 1.67 (Experiment 2): AB = 1 ÷ [(2×3) (91 + 77 – 101 -59)] = 2.0875 From both calculations, we find that A, which is the amount of heat, has a positive, giving the notion that increasing the amount from of 50 to 60 degrees Celsius will ensure the rice increases in volume by increasing temperature. The effect of B which is the amount of water has a negative, which hints that, an increase in the amount of water used will result in a reduction in the output volume. As for the interaction effect of the two, the effect remains much like the increase in heat alone. Clearly, the values for both experiments appear to follow a closeness that would allow us to determine the trends on a set of needed values of one experiment from the other. At this level, we opt to adopt the results of the first experiment and determine further trend values for the overall experimentation. To determine the variables that are going to be important for the experiment, the direction and magnitude of the effects of the factors are established through examination of variances. The sums of the squares of A, B and AB are established. As the total effect of A, B, and AB is found, the sum of the squares of the total effects is gotten by: SSA = [ab + a – b – (1)]2 , 4n SSB = [ab + b – a – (1)]2 4n Mathematically, the 3 total effects are orthogonal. This means that they are all statistically independent of each other. As per the experiment, the variables that are natural constitute two levels. Again, for natural variables that constitute two levels, the coding will give rise to the well known +, - 1 notation. This notation is for the levels of the variables that are coded. Hence, from the experiments we get: a1 = Amount of heat – ( 50 + 60) / 2 (60 – 50)/ 2 = Amount of heat – 55 55 This means that if the amount of heat needed to cook the rice to an optimum volume is at a low level which in the experiment is 50 degrees Celsius, then a1 = -1. It also implies that is the amount of heat needed to cook the rice to an optimal volume is at a high level which in the experiment is 60 degrees Celsius, then a1 = +1. These same calculations can also be used with the amount of water needed to cook the rice properly by: a2 = Amount of water – (1 + 2)/ 2 (2 – 1)/ 2 = Amount of water – 1.5 0.5 Hence, if the amount of water needed to achieve a maximum volume is at a high level which in the experiment is 2 cups, then, a2 = + 1. Also, if the amount of the water required to cook the rice to the highest volume is at a low level which in the experiment is 1, then a2 = -1. From these, we obtain the regression model which is: у = 27.5 + (8.33 ÷ 2) (a1) + (-5.00 ÷) (a2) The regression model can be utilized to get the fitted or expected value of у at the four points of the experimental design. Further, residuals are the differences between the fitted and observed values of у. When the amount of heat needed to achieve the highest volume of the rice is at the low level and the amount of water needed to achieve the same is at the low level, meaning that a1 = -1 and the a2 = -1 respectively, then the predicted amount of both the heat and water needed is: y = 27.5 + (8.33 ÷ 2) (-1) + (-5.00 ÷ 2) (-1) = 25.84 At this treatment combination there are 3 observations, the residuals for these observations are: е1= 29 – 25.84 = 3.16, е2 = 25 – 25.84 = - 0.84 e3 = 26 – 25.84 = 0.16 The next residuals and predicted values are also calculated in the same way for the experiment. For the low and high level of the amount of water and the amount of heat respectively: у = 27.5 + (8.33÷2) (+1) + (-5.00 ÷ 2) (-1) = 34.17. The residuals are: e4 = 38 – 34.17 = +3.83 e5 = 34 – 34.17 = -0.17 e6 = 28 – 34.17 = -6.17 For the high level of the amount of water needed to cook the rice and for the low level of the amount heat required to cook the rice, the predicted amount of both is: У = 27.5 + (8.33 ÷ 2) (-1) + (5.00 ÷ 2) (+1) = 20.84. The residuals are: e7 = 16 – 20.84 = -4.84 e8 = 21 – 20.84 = 0.16 e9 = 23 – 20.84 = 2.16 Lastly, the high level of both factors of the experiment gives the predicted amount of: y = 27.5 + (8.33 ÷ 2) (+1) + (-5.00 ÷ 2) (+1) = 29.17. For the residuals, e10 = 35 – 29.17 = 5.83 e11 = 25 – 29.17 = -4.17 e12 = 30 – 29.17 = 0.83 These results are used to plot a normal probability plot and a plot of the residuals against the predicted amount. If they satisfy the experiment then the conclusions are valid. Conclusion According to the results, it is obvious that increasing the amount of water to the rice results in an increase in the volume of rice obtained. However, doing the same for the heat component resulted in a decrease in the final volume of rice obtained. References Box, G. E. P. & Meyer, R. D. (1993). Finding the active factors in fractionated screening experiments. Journal of Quality Technology. 25 (2): 94 – 105. Kesharwani, R. K., Singh, D. V. & Misra, K. (2013). Computation-based virtual screening for designing novel antimalarial drugs by targeting falcipain-III: A structure-based drug designing approach. Journal of Vector Borne Diseases. 50: 93 – 102. Li, R. & Lin, D. K. J. (2009). Variable selection for screening experiments. Quality Technology and Quantitative Management. 6 (3): 271 – 280. [PubMed]. Montes, A., Tenorio, A., Gordillo, M. D., Pereyra, C. & Ossa, E. J. M. (2009). Screening design of experiment applied to supercritical antisolvent precipitation of amoxicillin: Exploring new miscible conditions. The Journal of Supercritical Fluids. 51: 399 – 403. Elsevier. doi:10.1016/j.supflu.2009.10.007. Stout, Q. F. & Hardwick, J. (2005). Optimal screening designs with flexible cost and constraint structures. Journal of Sequential Planning and Inference. 132: 149 – 162. Xiao, L., Lin, D. K. J. & Bai, F. (2012). Constructing definitive screening designs using conference matrices. Journal of Quality Technology. 44 (1): 1 – 8. ProQuest. Read More