Impact of digital literacy on aging attitudes of the elderly from the perspective of social cognitive theory | BMC Public Health
Benchmark regression
In order to verify the impact of digital literacy on aging attitudes, based on the formula (1) model, this paper uses Ordinal Logistic Regression method to analysis the impact, and the results are shown in Table 3.
In Table 3, most independent variables are significant at the level of 1%, and the log likelihood test is passed, indicating that the model is set reasonably and effectively, and the results had a certain stability. From the perspective of the sub-dimension of digital literacy, the regression coefficients of other variables are significantly negative except that the digital security indicator has no significant effect on the aging attitude, indicating that the improvement of digital cognition and digital skills will significantly inhibit the pessimistic aging attitude of the elderly. Hypothesis 1 passes the test. From the perspective of the comprehensive dimension of digital literacy, the influence of digital literacy on aging attitude is significantly negative, that is, every 1% increase in digital literacy, the occurrence rate of the elderly’s aging attitude decreasing by at least one level increases by 90.60%. The results show that the improvement of digital cognition and digital skills can effectively reduce the pessimistic aging attitude of the elderly. This is mainly because in the context of the rapid development of digital information and technology, once a large amount of digital information exceeds the receptivity of the elderly, the elderly will appear abnormal symptoms such as confusion, frustration, and impaired judgment. Therefore, the improvement of digital cognition and skills can effectively alleviate the elderly’s pessimistic aging attitude. The research conclusion of Charness [28] and Wu et al. [29] also confirm this view. However, the digital security indicator is not significant, mainly because digital security reflects the elderly’s prevention awareness of information leakage and economic fraud during the use of digital tools. Digital security has more impact on the consumption behavior, and has little impact on the elderly’s digital social networking, digital entertainment and other activities, so the impact on aging attitude is not significant.
From the perspective of control variables, the impact of age on aging attitude is significantly positive, indicating that with the increase of age, the pessimistic aging attitude of the elderly is more serious. This is mainly because with the increase of age, the elderly’s behavioral ability, cognitive function and social participation decline, so the aging attitude becomes more pessimistic. The impact of gender on aging attitude is significantly negative, indicating that the aging attitude of female elderly is more negative. This may be due to the fact that elderly women are more sensitive to changes in physical appearance and cognitive function, and are more likely to have bad emotions such as low self-esteem and depression, which will affect their aging attitude.
Robustness test
In order to enhance the robustness of the conclusion, this paper adopts the following three methods for testing.
Replacing the explained variable. In this paper, the question “As I get older, I feel that life has no direction” is used as a measure of aging attitude, and its mean, median and Standard deviation values are 2.01, 2 and 0.87 respectively, which indicates that there is no deviation value. As shown in Table 4 (3), the effect of digital literacy on aging attitude is still significantly negative.
Adjusting the weight. In this paper, the equal weight method is adopted to measure digital literacy in benchmark regression. Considering the subjectivity of equal weight method, we further adopt the entropy method to measure digital literacy. As shown in Table 4 (2), the regression coefficient is still significantly negative.
Principal component analysis. In order to reduce the correlation between digital literacy indicators, we further uses the principal component analysis method to assign weights to indicators. As shown in Table 4 (1), the regression coefficient is still significantly negative.
The above robustness test results are consistent and significantly negative, indicating that digital literacy has a certain negative impact on aging attitudes, and the benchmark regression results have strong robustness.
Endogeneity test
Although the benchmark regression in this paper controls the individual characteristics of the elderly, there are still unobserved missing variables. Considering that there may be a certain reverse causal relationship between the digital literacy and the aging attitude, the model may have endogeneity problems. Based on the reason, combined with the availability of questionnaire data, this paper selects “network access”, that is, whether the house where the elderly live has network signals, as an instrumental variable [30]. On the one hand, whether the house has network signals reflects the basis of the digital life, which can directly reflect the digital literacy level of the elderly. On the other hand, whether the house has Internet signal does not directly affect the aging attitude of the elderly. Table 5 shows the hausman test and the regression results for the instrumental variables. The P value of hausman test is 0.023, indicating the rejection of the null hypothesis and the existence of endogeneity problems. When other variables are controlled, the influence of digital literacy on aging attitude is still significant, with a coefficient of -0.504. At the same time, the underidentified null hypothesis and the null hypothesis of weak instrumental variables are rejected, indicating that the benchmark regression results are reliable, and digital literacy has an inhibitory effect on aging attitude.
Influence mechanism analysis
In order to further clarify the influence mechanism of digital literacy on aging attitude, this paper, from the perspectives of enhancement effect and empowerment effect, takes autonomous ability and adaptability as mediator variables respectively to represent self-efficacy, and discusses the influence path of digital literacy on aging attitude. The results are shown in Table 6.
In Table 6, the mediating effects of autonomous ability and adaptability on aging attitude are both significantly negative, and hypothesis 2 is verified through significance test. From the perspective of empowerment effect, when the autonomous ability of the elderly increases by 1%, the incidence rate of at least one level of reduction in aging attitude increases by 65.37%, indicating that the autonomous ability greatly affects the behavioral choices and emotional experiences of the elderly in the life process. Therefore, the improvement of digital literacy can effectively reduce the dependence of the elderly on family and society. The elderly can choose different types of elderly care services and activities by themselves, which can meet the needs of the elderly for self-esteem and independence, and restrain the negative aging attitude. From the perspective of enhancement effect, when the adaptability of the elderly increased by 1%, the incidence of aging attitude decreased by at least one level increased by 76.82%. This is mainly because the improvement of digital literacy means the improvement of the elderly’s ability to participate in the digital society and use digital products, which not only eliminates the gap between the elderly and the digital society, but also is conducive to the elderly to choose diversified contents and forms of elderly care services. Digital literacy can effectively meet the subjective initiative and social value needs of the elderly, and improve aging attitude, which is consistent with the research conclusion of Hasan [31] and Castilla [32].
Heterogeneity analysis
The above research results show that digital literacy can significantly inhibit aging attitudes. However, differences in individual characteristics, social characteristics and regional characteristics may directly affect the digital literacy level of the elderly, resulting in certain heterogeneity in aging attitudes of different elderly groups. Therefore, considering the particularity of the elderly group, this paper carries out heterogeneity analysis from three aspects of urban and rural attributes, income categories, age and gender.distribution. Among them, we classify the elderly with a monthly income of less than 3,000 as the low income group and sample size is 412, and those with a monthly income of more than 3,000 as the high income group and sample size is 400. Those under 75 years old are classified as the low age group and sample size is 455, and those over 75 years old are classified as the high age group and sample size is 357. Rural sample size is 388 and urban sample size is 424. Female sample size is 379 and male sample size is 433. The results are shown in Table 7.
In Table 7, under different samples of elderly people, the influence of digital literacy on aging attitude is significantly negative, which further proves that the influence of digital literacy on aging attitude is stable.
From the perspective of urban and rural attributes, the digital literacy of the urban elderly has a greater impact on aging attitude, mainly because in urban, elderly care, medical care, shopping, entertainment and other activities are more dependent on digital platforms, and the forms and products of elderly care services are mostly online services or online consumption. The improvement of digital literacy can effectively increase the adaptability and autonomous ability of the elderly and inhibit negative attitudes. The life of the elderly in rural areas is mainly based on mutual assistance in farming, chatting, playing cards, and family care and so on, and less dependent on digital tools. Moreover, the rural digital infrastructure is relatively imperfect, which also reduces the dependence of the elderly on digital products, so the impact is relatively small.
From the perspective of income level, the digital literacy of high-income elderly people has a greater impact on the aging attitude, which may be due to the fact that high-income elderly people have higher requirements for the form and content of elderly care services, and pay more attention to the integration of digital technology and elderly care services. And the improvement of digital literacy of elderly people is conducive to the elderly to seek and use different levels of elderly care services. For low-income elderly people, digital economic activities have a certain economic threshold, and the lack of economic sources limits the elderly’s use and dependence on digital products, so aging attitudes of the elderly are less affected by digital literacy.
From the perspective of age, the digital literacy of the younger elderly has a greater impact on the aging attitude, which is consistent with the research results of Győrffy et al. [33]. This is mainly due to the decline of physical function, cognitive ability and adaptive ability with the aging of the elderly, which seriously hinders the elderly’s use of digital products and digital services. Older people’s needs are more health care and less dependent on digital life, so the older the age is, the smaller the effect of digital literacy on aging attitudes is.
From the perspective of gender, the digital literacy of elderly men has a greater impact on their aging attitudes. This might be because elderly men attach more importance to the use of digital products than elderly women, and the proficiency of digital skills directly affects the self-efficacy of the elderly. Furthermore, for elderly Chinese women, most of their time is spent taking care of grandchildren and managing family affairs, and their exposure to digital products is relatively low. Therefore, their digital literacy has a relatively small impact on their attitudes towards aging, which is consistent with the research conclusion of Getenet S et al. [34].