分层抽样方法在渔业生产统计中的应用--以松花江干流捕捞抽样调查为例

摘要

This article deals with stratified sampling method, error analysis, and the determination of sampling number in stratified sampling. The process of determining the sample size of the survey of small-scale fisheries was described in an example of sampling survey of capture fisheries in main stream of Songhua River. The location stratified sampling was conducted in CPUE of gillnet fishing in Fangzheng, Yilan and Huachuan, and the capture pattern stratified sampling was carried out in stow net in Yilan. The stratified sam-pling results and the corresponding simple random sampling errors revealed that there was 3.4 times more error in simple random sam-pling than that in location stratified sampling, 5.2 times more than capture pattern stratified sampling, indicating the stratified sampling was characterized by less error and higher accuracy. Within the funding the stratified sampling method should be preferred in a fishery production survey.%本文介绍分层抽样方法、误差分析以及分层抽样样本量的确定方法。以松花江干流捕捞抽样调查为例，介绍小规模渔业抽样调查的样本量的确定过程。选取方正、依兰、桦川境内渔船刺网作业进行地点分层，选取依兰境内刺网、张网作业进行作业方式分层抽样，同时将分层抽样结果和相应的简单随机抽样进行误差对比分析。结果表明：简单随机抽样误差是地点分层抽样的3.4倍，是作业方式分层抽样的5.2倍，说明分层抽样误差小、精度高。在经费允许条件下，分层抽样是渔业生产统计调查的首选方法。