基于门限分位数回归的网上商品销量影响因素探析

网上商品销售与线下商品销售存在较大不同，为探索其消费模式，需要研究各影响因素对网上商品销量的作用机制。文章基于心理抗拒、贝勃定律、卢因人类行为理论等,对网络消费行为进行系统分析；综合运用分位数回归和门限回归方法，建立了门限分位数回归模型，揭示商品价格、商家信誉评分、商家信誉等级、保障标记数量、商品收藏人气、口碑数量和口碑分数等对销量的非线性异质影响。以受众广泛的iPad air2网上销售为研究对象，实证结果表明：提高商家的信誉等级、增加口碑数量能使高销量商家的销量更高，而保障标记数量的增加对热销有阻碍作用；在非热门商品转向热销品的过程中，增加收藏人气、增加口碑数量和一定价格范围内的提价对低销量商家的销量有促进作用。

Analysis of Factors Affecting Online Goods Sales Through Threshold Quantile Regression Approach

There is big difference between selling goods online and offline. In order to explore the consumption patterns, we should study the mechanism of various factors affecting online goods sales. In this paper, we make a systematic study of the behaviors of consumption online through the theory of psychological resistance, Beb's law, and Lewin metal of behavior. We construct a threshold quantile regression model to investigate the effects of price, credit, collection popularity, reputation and guarantee marks, on the sales. We select iPad Air 2 as our research object for its broad customers. The empirical results show that improving the sellers' credit rating and increasing the amount of reputation will lead to the increase of the sales for high-volume sellers. However, increasing the number of guarantee marks is negative to sales promotion. During the process of non-popular goods changing into the popular goods, increasing the amount of collection popularity, reputation and price within a certain range will promote the quantity of sales.