www.iaset.us editor@iaset.us MISSING PLOT TECHNIQUES USING REGRESSION ANALYSIS AND ITS COMPARISON WITH RANDOMIZED BLOCK DESIGN MISSING PLOT TECHNIQUES RICHA SETH 1 , D. K. GHOSH 2 & N. D. SHAH 3 1 K. S. School of Business Management, Gujarat University, Ahmedabad, Gujarat, India 2 Department of Statistics, Saurashtra University, Rajkot, Gujarat, India 3 Principal, M. C. Shah Commerce College, Ahmedabad, Gujarat, India ABSTRACT In this paper, we have estimated the values of one and two missing observations using the regression method of analysis. Again, we estimated the value of same missing observations (one and two) by expressing the regression data in the format of randomized block design data. In this investigation, we observed that the estimated value of the missing observation(s) come out to be more or less same. This procedure has been shown by taking two suitable examples. Further, we analyzed the data to obtain ANOVA table using both the methods, where we found the same significant result. KEYWORDS: Analysis of Variance, Blocking, Missing Observation, Regression Analysis, Response Variable 1. INTRODUCTION For the statistical procedure in Design of Experiments, the matrix presentation of the data of a randomized complete block design is similar to that of the exhibit of a factorial experiment with two factors being studied at different levels. When there is a missing value in the dataset of a randomized block design, we obtain the estimate of missing value using least square method of estimation and then carry out the analysis of variance using the estimated value of the missing observation. Regression analysis provides a mathematical relationship between the response variable and the factors affecting it. The expression of relationship obtained by regression can also give an estimate of the missing value, so that the further analysis can be performed. A study of the relation between the Analysis of Variance and Regression Analysis techniques has been carried out by Arner [1] and Karen [5] in the context of data analysis. For the statistical analysis procedure, regression analysis, requirement of the data is in the form of response variable values and the values of the factors affecting it, without considering the levels of the factors involved. In a variation of this, the statistical procedure design of experiments requires the data in the form of response variable values being affected by different levels of the factors involved. 1.1. Analysis of Variance The statistical tool Analysis of Variance is applied to analyze the data of a Design of experiment. Fisher [4] introduced the term ‘Analysis of Variance’ and defined it as the separation of variance ascribable to one group of causes from the variance ascribable to the other group. Under this technique, the total variation in the sampled data is divided into components of variation due to different independent factors. Each of these estimates of the variations due to assignable factors is compared with the estimate of the variation due to chance factor and identified whether the variation due to the assignable cause is significant or not. International Journal of Applied Mathematics & Statistical Sciences (IJAMSS) ISSN(P): 2319-3972; ISSN(E): 2319-3980 Vol. 6, Issue 4, Jun – Jul 2017; 55-66 © IASET