摘要：以广州市为例，采用实证研究方法，运用有序多分类Logistic回归模型研究农民工的个人属性、经济特征及住房特征对农民工住房满意度的影响。研究发现，家庭收入水平、住房面积和人均住房面积对住房满意度有显著影响，但个人特征如年龄、婚姻、文化程度等因素对住房满意度的影响并不显著。提出住房保障产品要结合农民工的工作特点、在选址上要注意与城镇原住居民的融合以及应制定户籍制改革的长远目标等相关建议，以逐步实现住房保障的城乡一体化。
关键词：农民工；住房满意度；Logistic回归模型；影响因素
Abstract：Taking Guangzhou for example，this paper uses empirical methods and orderly multi-classification Logistic regression model to study the impact of personal attributes，economic and housing characteristics on housing satisfaction of migrant workers. The results show that family income，housing area and per capita housing area have significant influence on housing satisfaction，while personal characteristics such as age，marriage，and education level have no significant effect. Therefore，puts forward the recommendations including the housing security products should be combined with the characteristics of migrant workers’ work，the site of housing security products should pay attention to the integration with the urban residents，the government should formulate the long-term goal of household registration reform，in order to gradually realize the integration of urban and rural housing security.
Keywords：migrant workers；housing satisfaction；Logistic regression model；influence factor
参考文献
[1] 丁成日，邱爱军，王瑾.中国快速城市化时期农民工住房类型及其评价[J].城市发展研究，2011（6）.
[2] Malte Lübker. Globalization and perceptions of social inequality[J]. International Labour Review，2004（1）.
[3] 朱嬿，李晨.基于顾客满意度理论的住户满意度模型[J].清华大学学报（自然科学版），2003（10）.
[4] 郭祖军.基于需求视角的住房保障政策选择研究[J].建筑经济，2015（2）.
工程经济，2017（05）：46-50