Observations on three different topics have been made and suitable modifications have been suggested. First related to linear growth rates. The linear growth rates computed from the indices data depend too much on the base period. The situation can be remedied by computing linear growth rates relative to the estimated responses of the items in the initial year. Second related to estimates of co-variance of half-sibs and full-sibs which did not utilize the full available information. Estimates of co-variance of half-sibs and full-sibs have been obtained while considering all the relevant components in the MANOVA of sire × dam experiment. The last related to the estimation of contributions of area and average yield towards increasing the production of a crop. The suggested method is based on the natural multiplicative model in contrast to the rather less suited additive models in vogue.
