Analysis on the influence factors of college students about the willingness to work in rural e-commerce

Among the samples participating in the survey, a larger number of people are more extroverted, which is further in line with the personality characteristics of Chongqing locals from the perspective of personality analysis. From the analysis of internship experience, nearly two-thirds of Chongqing e-commerce majors have not participated in the practice of rural e-commerce. From the analysis of academic planning, nearly 70% of college students choose direct employment. From the analysis of job location, students willing to stay in Chongqing for employment are very optimistic. The salary of e-commerce majors in Chongqing is expected to be concentrated in and around 4001–5000 yuan, which is a reasonable level for the income of fresh graduates in Chongqing. The specific data is shown in Table 6.

Overall, the average value of the measurement questions of the professional attitude dimension is above 3 points, which shows that the professional attitude of Chongqing e-commerce majors is relatively optimistic. The average value of the two variables of employment situation perception and occupation expectation is higher, reflecting that the e-commerce industry is more optimistic about its employment situation and the current development of rural e-commerce. However, the understanding of quotient is not high and needs to be strengthened. Specific data are shown in Table 7.

From the perspective of family members, the number of only-child families is 125, which accounted for 43.25%. Further, the number of non-one-child families is 164, which accounted for 56.75%. From the perspective of family attitudes, 100 people support the rural e-commerce industry, which accounted for 34.60%. Further, there are 189 people who do not support it, and this number accounted for 65.4%. From the perspective of the family’s network, the average level is 3.03, which indicates that the ability to help children in their career is slight. Most of the parents’ education level is below high school or technical secondary school. The annual family income level is mainly concentrated between 50,000 and 150,000 yuan. The specific data of parents’ education level and family annual income level are shown in Figures 2 and 3.

Fig. 2

Fig. 3

The mean values of school guidance and public opinion are all above 3 points, which is more optimistic. The higher scores of social public opinion indicate that the students and friends around the respondent have a positive view and practice of the rural e-commerce industry. The specific data is shown in Table 8 given below.

Through the cross-analysis of the factors affecting employment willingness, only the P-value of the three indicators of ‘internship experience’, ‘salary expectations’ and ‘family attitudes’ is less than 0.05, which indicates the practical experience, salary expectations and family members of Chongqing e-commerce college students. Attitudes are significantly related to their willingness to work in the rural e-commerce industry. The specific data is shown in Table 9 given below.

Correlation analysis of factors affecting employment willingness shows that the significance of the variables is 0.000. There are two asterisks (**) in the upper right corner, which indicate these variables are significantly correlated at the 0.01 level. Further, the correlation coefficient is positive, which indicates a positive correlation verifying the speculation. The specific data is shown in Table 10 given below.

‘Employment willingness’ is a binary variable and a binary logistic regression model that is used to carry out further exploration. There are 17 independent variables in the above four dimensions, and the dependent variable belongs to the employment willingness, which is determined within the range of [0,1]. The specific probability model is shown in the following Eq. (1): (1) Ln(pi1−pi)=β0+∑iβiXi+ε Ln\left( {{{{p_i}} \over {1 - {p_i}}}} \right) = {\beta _0} + \sum\limits_i {\beta _i}{X_i} + \varepsilon

In Eq. (1), the probability of Chongqing e-commerce majors who are willing to work in rural e-commerce can be represented by p, the explanatory variable that affects the willingness to work can be represented by Xi, β represents the coefficient, and the random error term can be represented by ɛ. When p=1, it means willing, and when 1-p=0, it means unwilling.

According to the regression model, the predicted value is 76.8%, which is greater than 50%. Also, the comprehensive test chi-square value of the model coefficient is 123.97, which is significant at the 1% level, indicating that the model effect is good. The specific logistic regression analysis results are shown in Table 11 given below.

In the dimension of individual characteristics, the variables of gender, personality, household registration and job location are not significant at the 5% level, which indicates that these variables have no significant impact on the willingness to engage in rural e-commerce work. The academic planning (reference: employment) is significant at the 5% level, and B=−0.797<0, which indicates that people who choose direct employment will prefer to engage in rural e-commerce work. The internship experience (reference: yes) is significant at the 1% level, and B=−0.899<0, which indicates that the people who have participated in the practice are inclined to engage in rural e-commerce work. Salary expectations (reference: 3000 yuan or less) are significant at the 1% level, and B=−0.598<0, which indicates that people with lower salary expectations will prefer to engage in rural e-commerce jobs. Hypotheses 1 and 2 are true.

In the dimension of occupational attitude, situation perception and occupational expectations are not significant in the model (p>0.05), which means that these variables did not significantly affect their rural e-commerce work willingness. Occupation perception is significant in the model (p<0.05, B=0.216), and it has a positive effect, indicating that college students with better occupation perception tend to engage in rural e-commerce work. Hypothesis 3 is true.

In the family background dimension, family members, family relationship, education level and income level are not significant in the model (p>0.05), which indicates that these variables have no effect on the willingness to engage in rural e-commerce. Family attitudes (reference: yes) are significant in the model (p<0.01, B=−2.404), which indicates that family attitudes have a significant and negative impact on e-commerce college students’ willingness to engage in rural e-commerce. These prove that hypothesis 4 is true and that hypotheses 5 and hypothesis 6 are not true.

In the external environment dimension, school guidance and public opinion are significant in the model (p>0.05), which indicates that school guidance and social public opinion have an impact on e-commerce college students’ willingness to engage in rural e-commerce. The B value is greater than 0, which means that the better the school guidance and public opinion are, the higher the willingness of college students will be to engage in rural e-commerce. Hypotheses 7 and 8 are established.

After the regression model test, the three categorical variables of internship experience, salary expectations, and family attitudes all have an extraordinary impact on the willingness of e-commerce majors to engage in rural e-commerce, which is consistent with the results of the chi-square test. The academic planning variables are inconsistent with the chi-square test results. The variables of career perception, school guidance, and public opinion have an important impact on the willingness to work, but other variables are not significant, and the correlation analysis has not been verified. This situation is probably because the chi-square test only considers the relationship between a single variable and the explained variable. However, the regression analysis has carried out a multi-factor analysis, and the various factors in the regression model have mutual influence, which results in different test results. The results of the regression analysis shall prevail. Therefore, the seven variables of internship experience, academic planning, salary expectation, career perception, family attitude, school guidance and public opinion all have a significant impact on the willingness to work, which means that the remaining assumptions are all valid, except for Hypotheses 5 and 6.