Sample size for principal component analysis in corn

Abstract The objective of this work was to determine the number of plants required to estimate the eigenvalues of the principal components analysis in corn (Zea mays) traits. Twelve traits were measured in 361, 373, and 416 plants of single-, three-way, and double-cross hybrids, respectively, in the 2008/2009 crop year; and in 1,777, 1,693, and 1,720 plants of single-, three-way, and double-cross hybrids, respectively, in the 2009/2010 crop year (six cases), totaling 6,340 plants. Principal component analysis was performed for the six cases. Sample size (number of plants) for the eigenvalue estimations of the principal components was determined by resampling with replacement and application of the model linear response and plateau model. The measurement of 267 plants is sufficient to estimate the eigenvalues of the principal components in corn traits.

摘要 本研究旨在确定估算玉米（Zea mays）性状主成分分析特征值所需的植株样本量。2008/2009生长季，分别针对单交、三交及双交杂交种，测定了361、373、416株植株的12个性状；2009/2010生长季则分别对单交、三交及双交杂交种的1777、1693、1720株植株开展性状测定，共计6组试验，总样本量达6340株。针对上述6组试验分别实施主成分分析。采用有放回重抽样法，并结合线性响应-平台模型，确定主成分特征值估算所需的样本量（即植株数量）。研究结果表明，测定267株植株即可满足玉米性状主成分特征值的估算需求。