A.R.S. BhatSanthosh N. Shanbhag2016-11-102016-11-102011http://krishikosh.egranth.ac.in/handle/1/85010In the present study the Coastal Area under Arabian Sea is considered as broad area and villages in the Coastal Area are considered as small area. The sampling units (non-farm households) are selected by the method of two stage samplings. Villages are considered as first stage units and farming households are considered as second stage units. The coastal area was separated into two parts namely Udupi and Dakshina Kannada. Small area estimates were obtained for the villages by fitting regression based models considering each of the above as broad area. Small area estimates were also obtained considering all these areas together as one broad area. The estimates were obtained by fitting Poisson regression model and Logistic regression models. For fitting Poisson and Logistic regression models, the explanatory variables used are family size, education level of household (number of years of schooling), total land holding (in acres), total annual non-farm income (in thousand rupees). The fitted models are then used to estimate the number of households engaged in non-farm activities for non-sampled villages in the broad area. The present investigation indicates that Logistic estimator and Composite estimator are the best small area estimators. The information on household’s farm activities at village level has a greater importance to the government and NGO’s in policy formulation, fund allocation for socio-economic development of a region and for establishment of small scale industries at the village level.Thesis