中文摘要 |
目前一般研究者在進行產業分析時,大部分忽略一個問題。在進行代表性個體分析時,以算術平均來設定聯合密度函數,視樣本中每一家廠商的重要性均相同,而影響數學期望值的計算結果,卻未加以注意,而陷入林國雄所稱「統計默契」的陷阱中而自身不知警惕。任何人在從事產業研究分析時,必須瞭解研究對象的性質,若是屬於代表性個體分析,必須注意聯合分配密度函數的設定具不唯一的特性,而不可在未評估其前提合理性之前,直接利用算術平均的概念來設定代表性個體的聯合分配密度函數。
Most researchers will generally neglect one key problem when doing the industrialanalysis. In the analysis of the representative individual analysis, one may usually usearithmetic mean to set up the joint distribution density function. The contribution from eachmanufacturer in the sample is deemed as the same. This will affect the calculation ofmathematical expectation without being noticed. And ends up, it falls into the so-called“statistical connivance” trap named by Kuo-Hsiung Lin unconsciously.Anyone who isworking on the analyses of industry should understand the attributes of research subjects. Forthose representative individual analyses, researcher must bear in mind that there’s nouniqueness in setting up the joint distribution density function, and not directly using theconcept of arithmetic mean before assessing its reasonableness. |