Abstract:
Rotational sampling procedures are extensively used in national surveys to reduce administrative and
respondent burdens and response bias introduced by sampling the same unit in consecutive periods. In
this paper, three stage diminishing sample size design is introduced to estimate mono-crop yield made in
the same location over consecutive time periods. It is suggested that the approach proposed in this paper
reduces administrative and respondent burdens and the response bias. The correlation structure of
observations in consecutive time periods is used without sacrificing the precession of estimates in order
to reduce the sample size in three consecutive sampling stages. The best com binedestim ate of average
mono-crop yield at each consecutive period is. obtained by weighing the independent estimates inversely
as their variances. In this paper the complete estimation procedure is presented on the proposed rotation
sampling design taking the case of paddy yield estimation which is recommended for estimating the
national average production of other seasonal mono-crop yields.
