3.2.2. Correlation with Urban Migration Attractiveness

Furthermore, we inquired about the relationship between urban external-MAI in cyber space and urban comprehensive attractiveness for migrants (UAM) in geographical space to further test the validity of the proposed indicators. Based on the push-pull theory which has been widely used in analyzing migration action and willing [48–51], we confirmed the UAM from urban pull perspective. The major migration reasons confirmed by the dynamic monitoring survey of China's migration population have been employed as reference in confirming the objective content of UAM, including work and business, study and training, and relocation. The three aspects separately correspond with the three major migration reasons as job opportunity and income level, living condition, and educational opportunity of children. Based on the data availability principle and integrated analysis of previous studies, eight indicators with respect to three aspects of urban pulling power have been selected as shown in Table 3. From job and income perspectives, Tertiary Industrial Output-Value (TIV) [52] and Urban Residents' Per Capita Disposable Income (IPC) [43] were employed to reflect urban job opportunities and income level; Unemployment Rate (UR), Participant rate of Urban Basic Medical Care System (RBM) [53], and Per Capita Living Area (LPC) [43] were utilized to expose the living condition of local residents; Number of Regular Primary Schools (PSN), Number of Regular Secondary Schools (SSN), and Number of University (UN) were applied to reveal educational opportunity for migrants' children [44].


**Table 3.** Indicator system of urban pulling power.

Note: RMB is the abbreviation of Ren Min Bi (China Yuan), which is the basic monetary unit of China.

Last, we adopted the principal component analysis (PCA) to integrate the index system and to obtain the indicator which reflects urban comprehensive attractiveness for migrants. The components with eigenvalues greater than 1 and the cumulative ratio of total variance greater than 85% are extracted and rotated with the varimax method in SPSS 19.0 (International Business Machines Corporation,

New York, USA), so that each factor has the minimum number of high load variables, which can be expressed as follows: Corporation, New York, USA), so that each factor has the minimum number of high load variables, which can be expressed as follows:

$$
\delta LAM\_k = \sum\_{i=1}^{m} \left[ A\_i \cdot \sum\_{j=1}^{n} C\_{ij} \times X\_{kj}^\* \right] \tag{7}
$$

where *UAM<sup>k</sup>* is urban comprehensive attractiveness for migrants of city *k*; *m* is the number of major components which make the cumulative ratio of the total variance greater than 85%; *A<sup>i</sup>* contributes the major components *i* to UAM of the city; *n* is the number of indexes; *Cij* is the contribution of index *j* to the major components *i*; and *X \* kj* is the standardized value of index *j* in city *k*. where *UAMk* is urban comprehensive attractiveness for migrants of city *k*; *m* is the number of major components which make the cumulative ratio of the total variance greater than 85%; *Ai* contributes the major components *i* to UAM of the city; *n* is the number of indexes; *Cij* is the contribution of index *j* to the major components *i*; and *X\* kj* is the standardized value of index *j* in city *k*.

#### **4. Results 4. Results**

#### *4.1. Correlation between External-MAI and Urban Inflow Migrants 4.1. Correlation between External-MAI and Urban Inflow Migrants*

According to the definition of MAI, the migration tendency of the person from the outside areas can be conveyed through external-MAI. Under the assistance of relevant migration data from the Tencent map, we engaged in exploring the relationship between external-MAI and urban migration population. Pearson correlation coefficient was adopted to reveal the relationship between them; the results have been shown in Table 4 and Figure 3. As we could observe, there are significant positive correlations between external-MAI and population migration in the three urban agglomerations. The Pearson coefficients are 0.948, 0.876, and 0.879 separately in BTH, YRD, and PRD, which has a holistic coefficient of 0.844. All of them have passed the significance test at 99% confidence level. Focused on their spatial heterogeneity, the cities of BTH has the highest correlation. According to the definition of MAI, the migration tendency of the person from the outside areas can be conveyed through external-MAI. Under the assistance of relevant migration data from the Tencent map, we engaged in exploring the relationship between external-MAI and urban migration population. Pearson correlation coefficient was adopted to reveal the relationship between them; the results have been shown in Table 4 and Figure 3. As we could observe, there are significant positive correlations between external-MAI and population migration in the three urban agglomerations. The Pearson coefficients are 0.948, 0.876, and 0.879 separately in BTH, YRD, and PRD, which has a holistic coefficient of 0.844. All of them have passed the significance test at 99% confidence level. Focused on their spatial heterogeneity, the cities of BTH has the highest correlation.

**Table 4.** The Pearson coefficient between population migration and external-migration attention indexe (MAI). **Table 4.** The Pearson coefficient between population migration and external-migration attention indexe (MAI).


Note: UA: urban agglomeration; BTH: Beijing-Tianjin-105 Hebei metropolitan region; YRD: the Yangtze River Delta; PRD: the Pearl River Delta. Note: UA: urban agglomeration; BTH: Beijing-Tianjin-105 Hebei metropolitan region; YRD: the Yangtze River Delta; PRD: the Pearl River Delta.

**Figure 3.** Scatter plot of external-MAI and migration population. **Figure 3.** Scatter plot of external-MAI and migration population.

Applying the principal component analysis, we obtained UAM of target cities based on the statistical data; the correlation study was deployed between the comprehensive UAM and external-MAI. As shown in Table 5 and Figure 4, we could observe a significant correlation between the UAM and external-MAI in the study areas. The coefficients of the whole area, BTH, YRD, and PRD are separately 0.829, 0.924, 0.984, and 0.789. The high correlation between them illustrated that urban received external-MAI is highly correlated to the attractiveness of urban itself. The relationship between such a cyber-based index and a traditional statistic-based index can be implied. Applying the principal component analysis, we obtained UAM of target cities based on the statistical data; the correlation study was deployed between the comprehensive UAM and external-MAI. As shown in Table 5 and Figure 4, we could observe a significant correlation between the UAM and external-MAI in the study areas. The coefficients of the whole area, BTH, YRD, and PRD are separately 0.829, 0.924, 0.984, and 0.789. The high correlation between them illustrated that urban received external-MAI is highly correlated to the attractiveness of urban itself. The relationship between such a cyber-based index and a traditional statistic-based index can be implied.

**Table 5.** The Pearson coefficient between urban comprehensive attractiveness for migrants (UAM) and external-MAI. **Table 5.** The Pearson coefficient between urban comprehensive attractiveness for migrants (UAM) and external-MAI.


Note: UA: urban agglomeration; BTH: Beijing-Tianjin-105 Hebei metropolitan region; YRD: the Yangtze River Delta; PRD: the Pearl River Delta. Yangtze River Delta; PRD: the Pearl River Delta.

**Figure 4.** The scatter plot of external-MAI and UAM.

**Figure 4.** The scatter plot of external-MAI and UAM. Furthermore, the Pearson correlation coefficients between the selected indexes and external-MAI have been calculated, as shown in Table 6. We can see that all the two indexes for job opportunities and income levels have the highest correlation with external-MAI in the study area. For the living condition perspective, a positive correlation can be observed between the Participant Rate of Urban Basic Medical Care System and external-MAI in BTH and YRP. However, significant correlations cannot be observed between the unemployment rate per capita living area with external-MAI. Paying attention to the education opportunities, significant correlations can be found in BTH and YRD between the three educational indexes and population attention index. In PRD, only the number of primary schools significantly correlates with external-MAI. In the three urban agglomerations, the strongest correlations are depicted between the Tertiary Industrial Output-Value and external-MAI, which reflect job opportunities in the areas being conventionally attractive for the potential migrants. Insignificant low correlation between the unemployment rate per capita living Furthermore, the Pearson correlation coefficients between the selected indexes and external-MAI have been calculated, as shown in Table 6. We can see that all the two indexes for job opportunities and income levels have the highest correlation with external-MAI in the study area. For the living condition perspective, a positive correlation can be observed between the Participant Rate of Urban Basic Medical Care System and external-MAI in BTH and YRP. However, significant correlations cannot be observed between the unemployment rate per capita living area with external-MAI. Paying attention to the education opportunities, significant correlations can be found in BTH and YRD between the three educational indexes and population attention index. In PRD, only the number of primary schools significantly correlates with external-MAI. In the three urban agglomerations, the strongest correlations are depicted between the Tertiary Industrial Output-Value and external-MAI, which reflect job opportunities in the areas being conventionally attractive for the potential migrants. Insignificant low correlation between the unemployment rate per capita living area with external-MAI can be detected.

area with external-MAI can be detected.


**Table 6.** Correlation coefficient between external-MAI and urban pulling indicators.

Note: \*: Pearson correlation is significant at the 0.01 level. TIV: Tertiary Industrial Output-Value; IPC: Urban Residents' Per Capita Disposable Income; UR: Unemployment Rate; RBM: Participant rate of Urban Basic Medical Care System; LPC: Per Capita Living Area; SSN: Number of Regular Secondary Schools; PSN: Number of Regular Primary Schools; UN: Number of University.

### *4.2. Correlation between Local-MAI and Floating Population Inner City*

The results of correlation analysis between local-MAI and local floating population have been shown in Table 7 and Figure 5. We can see that, no matter in the whole study area or the individual urban agglomeration, high correlation coefficients were gained. Especially in the YRD, the relevant coefficient has arrived at 0.950. PRD has a relatively lower value but is still higher than 0.75. Divided by the median value of local-MAI and local floating population, the cities in the study area can be divided into four types. Thereinto, 78.95% of them has high-high or low-low features. For the cities with higher-than-average floating population and higher-than-average local-MAI, there are three located in the BTH (Beijing, Tianjin, and Baoding), two in YRD (Shanghai and Suzhou), and two in PRD (Shenzhen and Guangzhou).

**Table 7.** The Pearson coefficient between local-MAI and floating population.


Note: UA: urban agglomeration; BTH: Beijing-Tianjin-105 Hebei metropolitan region; YRD: the Yangtze River Delta; PRD: the Pearl River Delta. *Int. J. Environ. Res. Public Health* **2020**, *17*, x 12 of 18

**Figure 5.** The scatter plot of local-MAI and floating population. **Figure 5.** The scatter plot of local-MAI and floating population.

**Figure 6.** Scatter plot of external-MAI and Local-MAI.

To explore the relationship between intercity-MAI and intercity population flow, the results have been shown in Table 8 and Figure 7. As we could notice, the average value of intercity-MAI is 1.00, Guangzhou-Shenzhen has the highest intercity-MAI at 5.19; and Shenzhen-Chengde has the lowest index of 0.04. For the individual urban agglomeration, the intercity-MAI among Beijing, Tianjin, and Shijiazhuang has the highest top three values in BTH. The same level of intercity-MAI

*4.3. Correlation between Intercity-MAI and Intercity Population Flow* 

local-MAI also has a higher external-MAI

To further excavate information from MAI, the relationship between local-MAI and external-

To further excavate information from MAI, the relationship between local-MAI and external-MAI has been explored; the results are shown in Figure 6. There is a highly positive correlation between the two indexes, of which the r is 0.7538 and *p* is 0.01. It is shown that the city with higher local-MAI also has a higher external-MAI To further excavate information from MAI, the relationship between local-MAI and external-MAI has been explored; the results are shown in Figure 6. There is a highly positive correlation between the two indexes, of which the r is 0.7538 and *p* is 0.01. It is shown that the city with higher local-MAI also has a higher external-MAI

**Figure 5.** The scatter plot of local-MAI and floating population.

*Int. J. Environ. Res. Public Health* **2020**, *17*, x 12 of 18

**Figure 6.** Scatter plot of external-MAI and Local-MAI. **Figure 6.** Scatter plot of external-MAI and Local-MAI.
