| Objective Sex ratio at birth(SRB) is an important index affecting the future overall marriage and childbirth level. The phenomenon of high SRB has appeared in many countries of Asia, and China is one of counties with the highest SRB in the world. Higher SRB has caused wide public concern of experts and researchers. Many researches in China and abroad focused on the changing trend, the reason of imbalance and the influence. This study analyzed the distribution characteristic and change of SRB with time and among regions in China since 1990, and explored the influence factors of SRB by spatial regression model, so as to provide reference basis for the control of SRB in China.Methods The data of SRB and corresponding economy, education and health care were collected from the fourth(1990), fifth(2000) and sixth(2010) population censuses for the 31 provinces, municipalities and autonomous regions(provinces) in China. And the data were analyzed by the methods of spatial correlation, hot spot analysis and spatial regression model. The application softwares were Arc GIS10.2, Excel2010 and Open Geoda.Results 1. The profile of SRB in China. The SRB of the fourth census was 111.3, which was fairly high, and six provinces, municipalities and autonomous regions were in the normal range. The SRB of the fifth census was 117.8, which was moderately high, and only Xinjiang and Tibet were in the normal range. The SRB of the sixth census was 118.0, which was moderately high, and only Xinjiang and Tibet were in the normal range. 2. The analysis of spatial clusters. The Global Moran’s I of SRB in the fourth census was Z=0.252, P=0.801, which showed no statistically significant area clustering. The Global Moran’s I of SRB in the fifth and sixth censuses was Z=2.968, P=0.003 in 2000 and Z=4.093, P<0.001 in 2010, which showed statistically significant area clustering. The low-low clusters were always distributed in the regions of Xinjiang and Tibet in the three censuses. But the high-high clusters were located in the central and southern China, as well as the number of the high-high clusters increased. 3. The detection of hot spots. Five hot spots and two cold spots were detected in the fourth census. Nine hot spots and two cold spots were detected in the fifth census. Ten hot spots and four cold spots were detected in the sixth census. Hunan, Jiangxi and Zhejiang and so on were always the hot spots and the range of hot spots increased gradually since 1990. 4. The analysis of influence factors. The General Linear Model was fitted in the study of the relation between SRB and its influence factors in 1990 and the determination coefficient was 0.948. In the model, 32?(10)(10)(28)x076.1x435.1774.16 y was the expression. The spatial lag model was fitted in 2000 and the determination coefficient of this model was 0.615. In the model, the expression was 5421?(10)(28)14.093x-0.152x-1.335x0.001x-142.578 y. The spatial lag model was fitted in 2010, which the determination coefficient was 0.816. In the model, the expression was 7654?-(10)--(28)x504.0x451.2x314.6x199.0091.133 y [x1 represented per capita disposable income, x2 represented the education rate of more than senior high school Degree, x3 represented the rate of agriculture account, x4 represented population rate of minorities, x5 represented average beds per one thousand people, x6 represented birth rate and x7 represented the rate of the second child or above].Conclusion 1. In the three censuses since 1990, the SRB in China increased gradually, and the number of provinces in which SRBs were higher than the normal range increased gradually. 2. The areas of SRB in China had a significant spatial clustering, which located in the central and southern China and the scope of spatial clustering increased gradually. 3. The spatial regression model was better than the traditional linear regression model when spatial clustering existed. Birth rate, the rate of the second child or above and population rate of minorities had influenced the SRB in China. |
PDF Full Text Request