This paper examines the main elements of Indian food policy in the context of recent economic and agricultural developments. Specifically, implications of recent global food crisis, climate change and shortfall in the agricultural growth are discussed. Given the rapid increase in demand for food grains globally and likely adverse impact of climate change on yield; the policy of self-sufficiency and government interventions to ensure physical and economic access should be strengthened. The country need to pay more attention to nutritional issues as 1/3rd of our people suffer from malnutrition. Concerted efforts should also be made to raise food grain yields not only to meet food requirement but also release some area for cultivation of high value crops like fruits and vegetables. Application of available stock of knowledge and technology for bridging the yield gap, particularly in the eastern region, can have immediate impacts. A long term strategy should also target developing plant varieties and crop management practices for adaptation and mitigation of climate change.Key words : Trends in foodgrain production, Temporal and spatial trends, Demand projections, Food and nutritional security.

  N.A

  In the context of design of experiments, reference is drawn to well-known examples involving the use of covariates. Recent advances in the Theory of Optimal Covariates? Designs suggest possibility of significant improvements through ?optimal? choice of the covariate values. A blend of theory and applications is discussed. Key words : Optimal designs of covariates models, A-optimality, Efficiency.

  The status of development of different states of eastern region of the country was estimated with the help of composite index based on optimum combination of a number of socio-economic indicators. Five major states and seven smaller states of the region are included in the study. The data on various indicators for the year 2001-02 are used in the analysis. The level of development is examined separately for agricultural sector, infrastructural facilities and overall socio-economic field. West Bengal is ranked first among the major states and Mizoram obtains the first position among smaller states in socio-economic development. Wide disparities in the level of development are found among different states. Infrastructural facilities are found influencing the socio-economic development in the positive direction both for major and smaller states. For bringing out uniform regional development, potential targets of different indicators are estimated for low developed states. Key words : Developmental indicators, Composite index, Model states, Potential targets.

  N.A

  N.A

  N.A

  Appropriate methodology has been developed for estimation of production of important flowers in Delhi on the basis of market arrivals data in flower mandis. The estimated market arrivals figures were in agreement with the estimated figures obtained through village survey. Key words : Flower mandi, Market arrivals, Village survey approach.

  Generalized linear mixed models (GLMMs) containing fixed and random area-specific effects are often used for small area estimation (SAE) of discrete variables (McGilchrst 1994 and Rao 2003). In GLMM, the random area effects take account for between areas variation beyond that is explained by auxiliary variables included in the model. These area effects are generally assumed to be independent in SAE. However, in practice area effects are correlated with neighbouring areas and the correlation decays to zero as distance increases. In this paper we investigate SAE based on GLMM with spatially correlated random area effects where the neighbourhood structure is described by a contiguity matrix. We use simulation studies to compare the performances of empirical best predictor for small area proportions under such models with and without spatially correlated area effects. The simulation studies are based on two real data sets. Our empirical results show only marginal gains when spatial dependence between small areas is incorporated into the SAE model. Key words : Cost function, Domain estimation, Optimal sample design, Probability of item response.

  A generalization of diagonal systematic sampling scheme for estimation of a finite population mean is introduced. The assumption of n ? k is relaxed here and hence the proposed method is applicable for all values of n provided N = kn. The relative performance of proposed diagonal systematic sample mean along with those of the simple random and systematic sample means is assessed for a natural population. Key words : Diagonal systematic sampling, Systematic sampling, Natural population, Linear trend, Trend free sampling.

  In this paper, we consider the estimation problem of population proportion, say, ?A bearing a stigmatizing characteristic in a community. We take into account the unrelated question model with the assumption that the population proportion of unrelated innocuous attribute is not known. We modify that model on combining the direct and randomized responses and an unbiased estimator of ?A based on randomized responses obtained from persons chosen by simple random sampling with replacement is obtained for this modified model. In this work, a new attempt has been made to construct shrinkage estimator on this modified model based on an adequate prior value ?A0 of ?A. The efficiency properties of the new shrinkage estimator over the usual modified estimator are discussed here theoretically along with some numerical illustrations. In addition the unbiased estimator of the mean squared error is also derived. Key words : Direct-cum-Randomized responses, Bias, Mean squared error, Improved efficiency, Prior value.

  This paper considers estimation of both domain and population totals for an item of interest using a two-phase sampling where the domain identity is realized, but the item response is not necessarily available from a phase I sampled unit. The optimality of sample design is studied considering the probability of item response, the cost of phase I vs. phase II sampling, and the item variability in the domains. Numerical evaluations made using a simulation study show that the proposed sampling and estimation method is more efficient than an alternative method given by Agrawal and Midha (2007). Key words : Cost function, Domain estimation, Optimal sample design, Probability of item response.

  In a stress-strength model, the reliability of a system is measured by the probability of the strength of the system exceeding the environmental stress. As the strength of the system depends on a number of controllable factors, the accuracy in estimating the reliability can be enhanced by a proper choice of these factors. In this paper, we have assumed stress and strength to be independent exponential variables, with mean strength being a function of the controllable factors. The optimum design for the estimation of the system reliability has been proposed using suitable optimality criterion. Key words : Stress-strength reliability, Exponential distribution, Optimum design, Central composite design.

  This article describes some methods of construction of mixed level ?NOD-optimal supersaturated designs. The methods of construction exploit the layout and the property of Uniform designs and Hadamard matrices. Mathematical expression for X(?NOD) and E(32) criteria have been obtained for many designs constructed in this paper. Some examples are given to illustrate the methods of construction. A catalogue of 67 optimal mixed level supersaturated designs with at most 60 runs and 60 factors is prepared. Some other important features of the designs are also given in the catalogue. All designs are ?NOD- optimal while some designs are 32- optimal too. Key words : Uniform design, Hadamard matrices, Efficiency criteria.

  N.A

  N.A

  N.A

  N.A

  N.A

  N.A

  N.A