期刊名称：International Journal of Statistics and Applications
印刷版ISSN：2168-5193
电子版ISSN：2168-5215
出版年度：2018
卷号：8
期号：4
页码：211-218
DOI：10.5923/j.statistics.20180804.06
语种：English
出版社：Scientific & Academic Publishing Co.
摘要：The problem of food security is of great concern to all stakeholders for national development. Food shortage in an economy inflicts undesirable costs on citizens as food is a basic human need. Of statistical concern is the estimation of crop yield in a single period called survey period. When data is available over different periods of time, it becomes plausible that the trend of the estimated yields be also investigated. Thus, this study utilized design and model based ratio and regression estimators to provide estimates of total national yields of maize and then fit linear trend to determine the growth in crop yield over time in Nigeria. It further investigated the performance of the estimator using samples of sizes 20 and 30 drawn using simple random sampling without replacement strategy from the 37 survey plots (36 states and FCT, Abuja) and the variables namely, maize yield (Y) and land area (X) were observed for estimation. Similarly, the asymptotic properties of the estimator are also investigated. Results revealed that there was increase in the year by year estimate of total yield of maize in each of the sample size cases except in 2006 where a sudden fall was experienced. It further showed that regression estimator performed better than the ratio estimator when the Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and Coefficient of Variation (CV) are utilized for all sample sizes considered for both design and model based estimators.
