在茶学研究中正确应用统计分析方法的思考──对《茶树良种苗期生物产量与光合特性的相关研究》等文统计分析的商榷

用逐步回归、通径分析及主成分方法对《茶树良种苗期生物产量与光合特性的 相关研究》一文（表１）数据进行了分析。结果表明，影响茶树生物产量直接的 主要因素是气孔导度（Ｘ３）和单株叶干重（Ｘ６），茶树单株苗干重（Ｙ）的预测回归 方程是 ＝－０．２８６９＋２６．５８４９Ｘ３＋１．９６４７Ｘ６，并且指出对于多元相关变量采用 简单相关分析剔除变量显然是错误的．

CONSIDERATION ON THE CORRECT APPLICATION OF THE STATISTICAL ANALYSIS METHOD IN THA RESEARCH

Seyeral multivariate statistical methods(stepwise regression, path analysis and principal component analysis) have .been applied to the data of Relationship between biological yield and photosynthetic charateristics of young tea plants (tablel). The result showed that the direct main reason of affecting tea biological yield was stoma conductivity (X2) and leaf dry weight per tea plant (X6), and that tie optimal predicative regression model of biological yield for excellent young tea plants was =-0. 2369+26. 5849X3+l.9647X6, which mhich indicates that it is an obvious mistake for relative multivariate to leave out some variate with simple relative analysis.