【摘要】：針對非線性模型中調(diào)節(jié)作用檢驗(yàn)不能只依賴交叉項(xiàng)系數(shù)的問題,目前存在兩個(gè)解決辦法:構(gòu)造更好的統(tǒng)計(jì)量和通過圖形直觀表現(xiàn)。但是已有研究對于非線性模型中的離散型和連續(xù)型調(diào)節(jié)變量沒有作進(jìn)一步區(qū)分,尤其是非線性模型中連續(xù)型調(diào)節(jié)變量的檢驗(yàn)一直缺乏簡便有效的方法。文章從模型和數(shù)據(jù)兩個(gè)維度劃分了四種調(diào)節(jié)變量的類型,梳理了不同類型調(diào)節(jié)變量的檢驗(yàn)方法和存在的問題,針對連續(xù)型調(diào)節(jié)變量檢驗(yàn)的問題,提出了對連續(xù)變量進(jìn)行多點(diǎn)分割的處理方式,并采用三維作圖的方法,直觀展現(xiàn)了調(diào)節(jié)作用的連續(xù)變化。
[Abstract]:In order to solve the problem that the adjustment function test in nonlinear model can not only depend on the cross-term coefficient, there are two solutions: constructing better statistics and visualizing representation by graphics. However, there is no further distinction between discrete and continuous variable in nonlinear model, especially the test of continuous variable in nonlinear model is lack of simple and effective method. This paper divides four types of adjustment variables from two dimensions of model and data, combs the test methods and existing problems of different types of adjustment variables, aiming at the problems of continuous adjustment variable testing. A method of multi-point segmentation for continuous variables is proposed, and a three-dimensional mapping method is used to visualize the continuous variation of the adjusting action.
【作者單位】： 山東大學(xué)縣域發(fā)展研究院;北京郵電大學(xué)經(jīng)濟(jì)管理學(xué)院;中國人民大學(xué)信息學(xué)院;
【基金】：中央高�；究蒲袠I(yè)務(wù)費(fèi)專項(xiàng)基金資助項(xiàng)目 山東大學(xué)基本科研業(yè)務(wù)費(fèi)專項(xiàng)基金資助項(xiàng)目(2017GN006)
【分類號】：C81

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 盧謝峰;韓立敏;;中介變量、調(diào)節(jié)變量與協(xié)變量——概念、統(tǒng)計(jì)檢驗(yàn)及其比較[J];心理科學(xué);2007年04期

，

本文編號：2155221