Relationship Between Typology of Internet Users and Social Networks in Chinese Older Adults

Abstract

Objectives: This study aims to generate a typology of Internet users, examine the association between this typology of Internet users and social network characteristics, and assess the moderating effect of rural–urban residence among older adults in China. Methods: Data were from the 2018 China Longitudinal Aging Social Survey. Latent class analysis and OLS regression with interaction terms were used. Results: Four types of Internet users were identified: Socializers, Social and Leisure Users, Leisure Users, and Advanced Users. Socializers were associated with larger and more supportive social networks, family networks, and friend networks than Leisure and Advanced Users. In addition, compared to their urban counterparts, older Internet users living in rural areas benefitted more from online social communication. Discussion:This study highlights the need to increase awareness among gerontological practitioners with respect to patterns of Internet use among older adults as a potential for enhancing their social well-being.

Keywords

socializers social and leisure users leisure users advanced users social networks

What this paper adds

• This is the first population-based study to understand the typology of Internet users according to diverse use behaviors and its social network implications among older adults in China.

• The typology of Internet users reveals the spread and diversity of Internet use behavior among Chinese older adults and also raises concerns about a new digital divide in terms of “a user typology divide.”

• Different Internet use patterns provided distinctive benefits for social networks of older online users, whereby Socializers had larger and more supportive social networks than other user groups, especially for those living in rural areas.

Applications of study findings

• The Internet use construct identified in this study can be a basis for risk assessment and a means for determining the efficacy of interventions on maintaining or extending social networks.

• Our findings can inform policy efforts to maintain and extend social networks for Chinese older adults by promoting their online social interaction use.

• This research could set the basis for future studies to understand the relationship between Internet use patterns and various health outcomes in Chinese ageing populations.

Introduction

Although technical barriers exist for older adults, the past decades have witnessed a dramatic increase in the number of older adults adopting new technologies. There is an increasing trend of Internet use that may promote active ageing (Boulton-Lewis et al., 2007). In the United States, 75% of people aged 65 years and above used the Internet in 2021 (Pew Research Center, 2021). In China, older adults represent a significant part of the growth in Internet users, with 43.2% of people aged 60 years and above using the Internet (China Internet Network Information Center, 2022). The socioemotional selectivity theory suggested that older adults selectively emphasize emotionally close social partners and disregard more peripheral ones as time in life grows increasingly limited (Carstensen et al., 1999). They are likely to drop some network members and keep those relationships that really matter to them. The decreasing size of family households and the increasing prevalence of nuclear families may lead to a decline in social networks among older adults. The COVID-19 pandemic has further exacerbated social isolation and loneliness among older adults. Internet use may help older adults stay connected to others and provide social support (Chen & Schulz, 2016). The Internet has likely played an increasingly important role in maintaining social well-being during the outbreak of the coronavirus disease (COVID-19) (Gabbiadini et al., 2020).

Internet Use and Social Networks

The major purpose of Internet use for older adults is to maintain connections with social network members (Sims et al., 2017). The Internet is an advanced vehicle of communication that has in recent years become more accessible to various age groups and may provide a means of overcoming barriers to social interaction for the ageing population. The Internet shapes social lives of older adults by providing new ways to communicate with close network members and developing new social ties (Khosravi et al., 2016). Older adults can use the Internet to communicate remotely with family and friends, despite physical impairments or geographic distance. This could help maintain strong ties in their social networks, which provides social support (Zhang et al., 2021).

In addition to social Internet use, there is increasing evidence showing that telecare services play a significant role in the improvement of patients’ self-care and management of chronic illnesses (Or & Karsh, 2009). Mobile health uses mobile devices to provide educational materials, monitor physical changes, and support for medical care (Yu et al., 2021). Telecare services enable healthcare providers to monitor outpatients continuously by receiving and processing critical health information. Older adults view hospital and physician websites to facilitate health care decision-making (Medlock et al., 2015). Having one or more chronic diseases also increases an Internet user’s likelihood to be engaged with online group forums that help people with personal issues or health problems (Fox & Purcell, 2010). Greater Internet use for information seeking has also been shown to be associated with greater perceived social support and community engagement (Choi & Dinitto, 2013; Erickson & Johnson, 2011; Szabo et al., 2019).

Internet use changes the way older adults maintain social relationships. Internet use can reconnect, strengthen, and broaden older adults’ social networks by facilitating regular contact with family and friends and developing relationships with new network members through social media (Cotten et al., 2013). Empirical research found that Internet use, including communication, information access, and entertainment, was associated with higher levels of social support (Erickson & Johnson, 2011). One limitation in our understanding about Internet use among older adults is based on research designs where the focus has been on either the frequency of Internet use or a single purpose for Internet use (Antonucci et al., 2017; Heo et al., 2015; Hülür & Macdonald, 2020). It remains unclear how different use patterns are associated with social well-being in later life. The purpose of this study is to close this gap in the scientific literature by examining Internet user types and social network characteristics among Chinese older adults using data from the CLASS, a large nationally representative data source with detailed information on the Internet use behaviors of older adults.

Internet Use Patterns

A few studies have explored Internet use patterns, and some common patterns were identified with data from European countries and Taiwan, China. The identified user types included advanced users, sporadic users, entertainment/leisure users, instrumental users, and non-users (Brandtzæg et al., 2011; Chiu, 2019). Advanced users employ the Internet for many purposes, such as for consuming news, looking for information, playing games, and maintaining contact with family and friends. Sporadic users are characterized by occasional and infrequent use of Internet services, often for a far limited number of purposes. Entertainment/leisure users frequently use Internet radio or TV, downloading games or music and chat with friends and family. Instrumental users are actively engaged in goal-oriented activities, such as searching for information about goods or services and utilizing services related to e-banking, e-commerce, and for making travel plans. Non-users do not use the Internet on a regular basis.

Another limitation of previous research is the focus on Internet use patterns whereby different age groups are combined into a single sample for analysis (Brandtzæg et al., 2011). However, some older adults continue to face technology barriers, and they may also have different levels of access to the Internet than younger adults; thus a focus on older adults is necessary. Currently, 43.2% of people aged 60 years and above use the Internet in China, while the Internet user percentage among those aged 59 year and below is 79.8% (China Internet Network Information Center, 2022). Thus, there is an obvious intergenerational digital use gap. Studies focusing on Internet use patterns among ageing populations could inform technology-assisted behavioral interventions and health delivery options with improved outcomes for this population.

In addition, older adults have distinct structures of social networks compared with younger adults, featured by smaller network size with a focus on close social ties (Hülür & Macdonald, 2020), and the impact of the Internet use on older persons’ social networks remains inconclusive conditional on different Internet use behaviors (Zhang et al., 2021). Online communication helps older adults maintain their close social ties (Heo et al., 2015), but using social media for entertainment and information gathering may exacerbate loneliness due to the lack of social interaction (Shakya & Christakis, 2017). Although previous studies have established the relationship between single Internet use behavior and social networks, indeed, individuals use Internet for multiple purposes. A person-centered, rather than a variable-centered approach, could advance the knowledge on how the underlying Internet use patterns would affect social networks.

Rural Versus Urban Older Online Users

Urban–rural residence makes a substantial difference in Internet use behaviors and social well-being among older adults (Choi et al., 2022). Rural areas have less Internet coverage compared to urban areas in China. Chinese urban and rural Internet users constituted 72.4% and 27.6% of the total Internet users, respectively (China Internet Network Information Center, 2022). In addition, the migration of younger Chinese adults from rural to urban areas in search of employment results in rural empty-nest older adults and increases the risks of social disconnectedness (Wang et al., 2021). Communication through the Internet could be more meaningful for maintaining adequate social ties for rural older adults than their urban counterparts, particularly when in-person intergenerational contact is reduced or absent. Considering the differences in Internet accessibility between rural and urban areas and the prevalence of out migration among adult children from rural to urban areas, further research is needed to explore how different Internet use patterns are associated with social network characteristics in rural and urban areas.

The Current Study

Understanding Chinese older adults’ Internet use patterns is important both for having a full picture of Internet use in later life and for knowing what type of support digital media can provide to improve the social network characteristics of adults in later life. To fill the research gaps identified above, we use data from the CLASS and address the following study aims: (1) generate a typology of older Internet users in China using LCA; (2) examine the relationship between the Internet user typology and social network characteristics using linear regression models; and (3) assess the moderating effect of rural–urban residence for the relationship between the Internet user typology and social network characteristics.

Method

Data Source and Study Sample

We employed data from the China Longitudinal Aging Social Survey (CLASS), which has detailed information on Internet use among Chinese older adults. CLASS is a nationally representative survey of older adults aged 60 years or above from 28 (out of 31) provinces in China (data and documentation are available at http://class.ruc.edu.cn/). The baseline survey was initiated in 2014. The written informed consent was obtained from each respondent. The follow-up Wave 2 and Wave 3 surveys were conducted in 2016 and 2018, respectively. This study employed Wave 3 survey data only because this wave had rich data on Internet use.

The 2018 CLASS used a multi-stage stratified sampling design with counties as the primary sampling units. A total of 462 villages and neighborhood committees were selected from 134 counties. Households were selected within villages and neighborhood committees. Only one older adult (60 years and above) was randomly selected from each sampled household (Tang et al., 2020). The sample included 11,417 older adults from 28 provinces in China. The age distribution of CLASS is similar to 2010 Chinese census data.

The study sample excluded 9332 respondents who reported never using the Internet. Among the 2085 Internet users, 398 (19%) respondents had missing data on two variables, including monthly household consumption and depressive symptoms. After conducting multiple imputation for these missing data, the final analytic sample included 2085 Internet users aged 60 years and above.

Measures

Social Network Characteristics

Social network characteristics were measured by the Lubben Social Network Scale (LSNS-6) (Lubben et al., 2006). The LSNS-6 consisted of a set of three questions that assessed family ties and a comparable set of three questions that assessed friendship ties. With respect to family ties, we asked participants “How many relatives do you see or hear from at least once a month?”, “How many relatives do you feel at ease with that you can talk about private matters?”, and “How many relatives do you feel close to such that you could call on them for help?” The same items were used to measure friendship ties. The LSNS-6 has good reliability in our study sample (Cronbach’s a = 0.87). The total score ranged from 0 to 30. The total score of LSNS-6 Family subscale (a = 0.81) ranged from 0 to 15. Similarly, the sum score of LSNS-6 Friends subscale (a = 0.82) ranged from 0 to 15. Higher scores in LSNS-6, LSNS-6 Family subscale, and LSNS-6 Friend subscale indicated larger and more supportive social networks, family networks, and friend networks, respectively.

Internet Use

Ten binary indicators of Internet use behaviors were selected to generate a typology of Internet users, including chatting, messaging, watching news, listening to music/radio or watching videos, playing games, information seeking, online shopping, transportation, health management, and finance and economics. For each item, respondents were asked to report whether they participated in each activity when using the Internet (yes = 1).

Control Variables

Demographic characteristics, socioeconomic status, health status, household characteristics, and social activity participation were selected as control variables (Iliffe et al., 2007; Park et al., 2018). Demographic characteristics included age (in years), gender (male = 1), marital status (married = 1), and rural–urban residence (urban residence = 1). Socioeconomic status was measured with education (illiterate = 1, primary school = 2, and junior high school or above = 3), and monthly household consumption (in yuan, transformed by the natural log). Health status was measured by ADL and depressive symptoms. ADL was evaluated by a dichotomous variable (one or more ADL limitations = 1, no ADL limitations = 0). Depressive symptoms were measured by Center for Epidemiologic Studies Depression Scale with a Cronbach’s alpha of 0.78 (Silverstein et al., 2006). Household characteristics included number of living children (zero = 1, only one = 2, two or above = 3) and whether the respondent lived alone (yes = 1). Social activity participation was coded one when respondents reported that they ever participated in at least one of the following activities during the last year: religious activities, educational or training courses, playing cards/chess/mah-jongg, square dance, and community security services.

Analyses

We used descriptive statistics to describe the characteristics of the study sample. To explore distinctive types of Internet users, we conducted LCA with the 10 indicators described above. LCA is a statistical method that identifies unobserved classes within a population based on responses to a set of observed categorical variables. The number of latent classes is determined by a variety of statistical indices, including the likelihood-ratio G² statistic, Akaike’s information criterion (AIC; Akaike, 1974), Bayesian information criterion (BIC; Raftery, 1986), and entropy. Better model fit is determined by lower values of the first three indices and higher entropy values (Guo et al., 2020).

To address the second research question, linear regression models were estimated to test the association between the Internet user typology and social network characteristics, as well as family and friend network characteristics, while controlling for confounding variables. Subsequently, we included interaction terms involving Internet user typology and rural–urban residence to assess the moderating effect of rural–urban residence on social network characteristics, family network characteristics, and friend network characteristics. LCA models were estimated with Mplus, version 8.3. Regression analyses were performed using Stata, version 15.1.

Results

Sample Characteristics

The characteristics of study sample are presented in Table 1. Regarding social network characteristics, the mean score of LSNS-6 among older people was 14.84 (SD = 5.28). They had an average family subscale score of 7.77 (SD = 2.70) and friend subscale score of 7.07 (SD = 3.08). The average age of the sample individuals was 66.95 years (SD = 5.29). The proportions of female and male respondents were almost the same. Respondents who were currently married accounted for more than four-fifths of the sample. About 80% of respondents lived in urban areas. About two-thirds of the sample graduated from junior high school or above. The average monthly household consumption was 3723.72 yuan (SD = 140.54). Nearly 4% of respondents required assistance with at least one of the six ADLs. The average score of depressive symptoms among respondents was 7.63 (SD = 0.06). Nearly 60% of respondents had two or more living children, 6% of them lived alone, and 70% of them ever participated in social activities during the last year.

Table 1.

Sample Characteristics (Mean/Proportions) (N = 2085).

Variable	Mean (SD)/N (%)	Variable	Mean (SD)/N (%)
Social network score	14.84 (5.28)	Household characteristics
Family network score	7.77 (2.70)	Number of living children
Friend network score	7.07 (3.08)	0	61 (2.93)
Demographic characteristics		1	828 (39.71)
Age (in years)	66.95 (5.29)	2 or above	1196 (57.36)
Female	994 (47.67)	Lived alone	135 (6.47)
Married	1730 (82.97)	Social activity participation	1474 (70.70)
Urban	1660 (79.62)	Internet use behaviors
Socioeconomic status		Chatting	1782 (85.47)
Education		Messaging	1193 (57.22)
Illiterate	148 (7.10)	Watching news	1275 (61.15)
Primary school	549 (26.33)	Information seeking	542 (26.00)
Junior high school or above	1388 (66.57)	Listening to music/radio or watching videos	991 (47.53)
Monthly household consumption	3723.72 (140.54)	Playing games	423 (20.29)
Health status		Online shopping	273 (13.09)
Had one or more ADL limitations	83 (3.98)	Transportation	170 (8.15)
Depressive symptoms	7.63 (0.06)	Health management	85 (4.08)
		Finance and economics	112 (5.37)

Internet Use Behaviors and Patterns

Table 1 also summarizes the distribution of Internet use behaviors. Most of the older adults used the Internet for chatting (85.47%), followed by watching news (61.15%), messaging (57.22%), listening to music/radio or watching videos (47.53%), information seeking (26%), playing games (20.29%), online shopping (13.09%), transportation (8.15%), finance and economics (5.37%), and health management (4.08%).

The model-fit statistics for the LCA models with one to five classes are presented in Table 2. The lowest value of BIC existed in the 4-class model, indicating the best model fit. Although models with a higher number of classes yielded lower values of AIC and G², a closer inspection of the 2-parameter estimates showed the drop in AIC and G² was more substantial with each additional class, up to the 4-class model. In addition, the Lo–Mendell–Rubin (LMR) test showed that the 5-class model was not significant, indicating no statistically significant improvement of fit for the 5-class model. Given these 4 indices, we selected the model containing four latent classes.

Table 2.

Comparison of Fit Statistics for LCA Models With One to Five Classes (N = 2085).

Number of Classes	G²	AIC	BIC	Entropy	LMR
1 class	1876.216	19,139.131	19,195.557	N/A	N/A
2 classes	1276.156	18,276.884	18,395.377	0.623	0.000
3 classes	956.033	17,958.995	18,139.556	0.766	0.000
4 classes	741.031	17,775.285	18,017.914	0.707	0.000
5 classes	672.285	17,719.246	18,023.942	0.704	0.224

The parameters for the 4-class model are presented in Table 3. The prevalence of the four latent classes was reported under each class. Item-response probabilities indicated the estimated likelihood that an individual in a given latent class answers “Yes” to each of the 10 indicators. The item-response probabilities in bold defined the classes and guided the labelling of the classes.

Table 3.

Four-Class Model of Typology of Internet Users Among Chinese Older Adults (N = 2085).

	Socializers	Social and Leisure Users	Leisure Users	Advanced Users
Latent class prevalence	47.67%	33.67%	12.71%	5.95%
Item-response probabilities
Chatting	0.943***	0.991***	0.000	0.979***
Messaging	0.465***	0.827***	0.000	0.781***
Watching news	0.351***	0.813***	0.713***	0.945***
Listening to music/radio or watching videos	0.264***	0.606***	0.585***	0.881***
Playing games	0.026*	0.365***	0.135***	0.517***
Information seeking	0.109***	0.389***	0.071***	0.797***
Online shopping	0.052***	0.148***	0.067***	0.626***
Transportation	0.048***	0.032*	0.000	0.694***
Health management	0.032***	0.013	0.018*	0.286***
Finance and economics	0.028***	0.032**	0.067***	0.309***
Defining characteristics	Relatively high frequency of chatting and lowest recreation	Highest ratings on chatting and messaging and relatively high frequency of watching news and listening to music/radio or watching videos	Relatively high frequency of watching news and listening to music/radio or watching videos and lowest likelihoods endorsing chatting and messaging	Highest ratings on all the leisure and instrumental activities and relatively high ratings on chatting and messaging

Notes. *p < .05, **p < .01, and ***p < .001. Item-response probabilities in bold indicate defining characteristics of each latent class.

The most common class of Internet users was Socializers (47.67%), characterized by a relatively high probability of chatting, together with the lowest recreation use (e.g., watching news, listening to music/radio or watching videos, and playing games) and the lowest use of the Internet for some instrumental activities (e.g., information seeking, online shopping, and finance and economics). The second most common class was Social and Leisure Users (33.67%), characterized by the highest social communication use (e.g., chatting and messaging), together with a relatively high probability of watching news and listening to music/radio or watching videos. The third class was Leisure Users (12.71%), which manifested a relatively high probability of watching news and listening to music/radio or watching videos, as well as the lowest social communication use (e.g., chatting and messaging), and instrumental use (e.g., transportation and information seeking). The least common class was Advanced Users (5.95%), characterized by the highest ratings on all the leisure activities and instrumental activities, as well as relatively high likelihoods endorsing social communication use.

Typology of Internet Users and Social Networks

The regression model findings for the Internet user typology and social network characteristics are summarized in Table 4. We used the “Socializers,” the most common typology of Internet users, as the reference group in the regression analysis. In Model 1, older Internet users in the Leisure Users group were associated with smaller social networks (b = −2.05, SE = 0.39, p < 0.001) as well as family (b = −0.88, SE = 0.20, p < 0.001) and friend (b = −1.17, SE = 0.23, p < 0.001) networks than those in the Socializers group. Older Internet users in the Advanced Users group were associated with smaller social networks (b = −1.53, SE = 0.50, p < 0.01) and family (b = −0.80, SE = 0.26, p < 0.01) and friend (b = −0.73, SE = 0.30, p < 0.05) networks than their counterparts in the Socializers group. In Model 2, urban/rural residence could moderate the relationship between Leisure Users and social networks (b = 1.69, SE = 0.79, p < 0.05), family networks (b = 0.77, SE = 0.40, p < 0.05), and friend networks (b = 0.92, SE = 0.46, p < 0.05). Figure 1 shows that the urban Leisure Users had stronger social networks, family networks, and friend networks than their rural counterparts.

Table 4.

Summary of Results from Regression Analyses of Typology of Internet Users and Social Networks Among Chinese Older Adults (N = 2085).

	Model 1						Model 2
	Social networks		Family networks		Friend networks		Social networks		Family networks		Friend networks
	b	SE	b	SE	b	SE	b	SE	b	SE	b	SE
Typology of Internet users (ref = Socializers)
Leisure users	−2.05***	0.39	−0.88***	0.20	−1.17***	0.23	−2.97***	0.63	−1.31***	0.32	−1.66***	0.37
Advanced users	−1.53**	0.50	−0.80**	0.26	−0.73*	0.30	0.10	1.36	−0.13	0.70	0.24	0.80
Social and Leisure users	−0.25	0.26	−0.02	0.13	−0.24	0.15	0.25	0.64	0.17	0.33	0.08	0.37
Typology of Internet usersurban residence (ref = Socializersurban residence)
Leisure users*urban residence							1.69*	0.79	0.77*	0.40	0.92*	0.46
Advanced users*urban residence							−1.86	1.46	−0.76	0.74	−1.10	0.85
Social and Leisure users*urban residence							−0.59	0.69	−0.22	0.35	−0.37	0.41
Age	0.00	0.02	0.01	0.01	−0.01	0.01	0.00	0.02	0.01	0.01	−0.01	0.01
Male	−0.10	0.23	−0.01	0.12	−0.08	0.14	−0.07	0.23	0.00	0.12	−0.07	0.14
Married	0.65	0.35	0.22	0.18	0.44*	0.21	0.66	0.35	0.22	0.18	0.44*	0.21
Urban residence	0.03	0.34	−0.13	0.17	0.16	0.20	−0.05	0.48	−0.19	0.24	0.14	0.28
Education (ref = illiterate)
Primary school	1.42**	0.50	0.49	0.26	0.93**	0.29	1.23*	0.51	0.41	0.26	0.82**	0.30
Junior high school or above	1.76***	0.50	0.72**	0.26	1.04***	0.29	1.59**	0.50	0.64*	0.26	0.94**	0.30
Logarithm of monthly household expenditure	−0.12	0.17	−0.10	0.09	−0.01	0.10	−0.11	0.18	−0.10	0.09	−0.01	0.10
Had one or more ADL limitations	−1.86**	0.59	−1.11***	0.30	−0.75*	0.35	−1.75**	0.59	−1.06***	0.30	−0.69*	0.35
Depressive symptoms	−0.10	0.05	−0.07*	0.03	−0.03	0.03	−0.09	0.05	−0.06*	0.03	−0.03	0.03
Number of living children (ref = 0)
1	1.39*	0.71	0.55	0.36	0.84*	0.41	1.30	0.70	0.51	0.36	0.78	0.41
2 or above	2.20**	0.70	1.15**	0.36	1.05*	0.41	2.06**	0.70	1.09**	0.36	0.97*	0.41
Lived alone	−0.58	0.53	−0.81**	0.27	0.23	0.31	−0.60	0.53	−0.81**	0.27	0.22	0.31
Social activity participation	−0.25	0.25	0.03	0.13	−0.28	0.15	−0.19	0.26	0.06	0.13	−0.25	0.15
Intercept	13.23***	2.16	6.98***	1.10	6.25***	1.26	13.27***	2.19	6.98***	1.12	6.29***	1.28

Notes. (1) b = unstandardized coefficient; SE = standard error. (2) *p < .05; **p < .01; ***p < .001.

Figure 1.

Interaction effects of typology of Internet users and rural–urban residence for social network characteristics, including family and friend networks.

Discussion

To our knowledge, this is the first study based on a nationally representative sample to identify a typology of Internet users among older adults in China and examine the relationship between the typology of Internet users and social networks, including family and friend networks. We highlight three major findings from this study. First, the typology of Internet users was characterized as representing four classes, including Socializers (47.67%), Social and Leisure Users (33.67%), Leisure Users (12.71%), and Advanced Users (5.95%). Second, Socializers were significantly associated with larger and more supportive social networks, family networks, and friend networks than Leisure Users and Advanced Users. Third, rural–urban residence moderated the relationship between the typology of Internet users, social network characteristics, and family and friend network characteristics. Below, we elaborate on these findings and discuss their policy implications.

The differences in Internet use patterns across countries may be partially explained by different levels of media literacy among Internet users as well as various research design issues, including different types of samples and different measures of Internet use patterns. Consistent with previous studies (Brandtzæg et al., 2011; Chiu, 2019), Leisure Users and Advanced Users were also identified in our study among older adults in China. The novel class, Socializers, is the most prevalent user class (47.67%) identified in our study. It was mainly due to the prevalence of WeChat (a communication application like WhatsApp) available in China since 2012, and older users usually use it for sending voice messages and making video calls. However, we did not identify instrumental users as a group among Chinese older adults, which was shown in European samples. Due to poor media literacy and accessibility, the current cohort of Chinese older Internet users is more likely to seek help from their family and friends compared with search for some kinds of information online, especially in rural areas with lower Internet accessibility. Meanwhile, e-healthcare services for older adults were underdeveloped in China in 2018. Further, although DiDi (a rideshare application like Uber) and online shopping applications were prevalent in China at that time, many older adults did not use these options because the application design was not age-friendly. These may partially explain the absence of instrumental users in our sample.

Earlier studies reported that Internet use was associated with higher levels of social support, life satisfaction, self-rated health, social participation, and lower levels of depression (Erickson & Johnson, 2011; Heo et al., 2015; Jin & Zhao, 2019; Lu & Li, 2022). Information and communication technologies intended to expand and sustain social contact and alleviate social isolation among older adults were demonstrated to be useful (Khosravi et al., 2016). Our study shed insight for this field and found that social networks of older online users may have benefitted differently from new technologies based on the underlying Internet use patterns reported here. Interestingly, Socializers have larger and more supportive social networks than Advanced Users. A recent study also reported that older adults using the Internet once a week or once a month were less likely to be socially isolated than everyday users because everyday users might either be online too frequently and/or for long duration, which might displace time spent in offline social activities and lead to greater social isolation (Stockwell et al., 2021). The relatively high frequency of Internet use across the activities identified in our study suggests that Advanced Users spend more time online, which may result in decreasing social relationships. In addition, Leisure Users may also enlarge their social networks through online leisure activities. However, the leisure activities captured in our study, such as watching news, listening to music/radio, or watching videos, lack a social interaction component. Therefore, Socializers manifested broader social networks than Advanced and Leisure Users.

Digital divide exists across rural–urban areas, racial/ethnic groups, and socioeconomic status. Rural older adults, ethnic minorities, and low SES are less likely to use the Internet (Choi et al., 2022; Yoon et al., 2020, 2021). Existing research further examined digital divide through the lens of intersectionality and reported that the negative impacts of rural residence on the Internet use were more pronounced in older minorities (Choi et al., 2022). Ethnic minority status combined with low SES substantially reduced the Internet use (Yoon et al., 2020). Our study added to the digital divide literature by examining the interaction effect of rural–urban residence and Internet use on social well-being. The positive effects of online social interactions on social well-being are more salient in rural older adults. China has witnessed a massive migration of young labors from rural to urban areas in search of employment, which leads to rural empty-nest older adults who are living in the originating place of residence after children migrate. When the geographical proximity between the older parent and adult child increases, in-person contact frequency, emotional ties, and family support will be reduced (Wang et al., 2021). However, intergenerational relationships can be maintained through online communication, even when the large geographic distance exits (Cotten et al., 2013). Older parents living in rural areas can use WeChat to make video calls or send voice messages to their children living in different cities or provinces to maintain social connectedness.

Our study has implications for gerontological research and practice. Most prior studies on Internet use patterns have been conducted in Western countries, with few exceptions (Chiu, 2019). This study is among the first to examine Internet use patterns among older adults in Mainland China. Understanding Internet use among Chinese older adults, as well as the ways they use have evolved over time, may clarify how best to support digital media use within this population, and thus Chinese older adults can truly benefit from the digital society. The findings reveal the heterogeneity in Internet use patterns among Chinese older adults. Internet use shapes a person’s social landscape, but little is known about which Internet use pattern contributes to social well-being. Our findings showed that different Internet use patterns provided distinctive benefits for social networks of older online users, whereby Socializers had larger and more supportive social networks than other user groups, especially for those living in rural areas. This research could set the basis for future studies to understand the relationship between Internet use patterns and various health outcomes in Chinese ageing populations.

This research increases the awareness of gerontological practitioners to the Internet use pattern of older adults in China, which is shown to enhance their social well-being in a developing country with the largest older adult population globally. It is important not only to promote access to Internet but also to provide adequate training to overcome technology barriers and encourage beneficial online activities. The Internet use construct identified in this study could be a basis for risk assessment and a means for determining the efficacy of interventions on maintaining or extending social networks. If an older adult changed from a Socializer to a Leisure User, this might be a warning signal that the older adult is at a higher risk of reducing social networks, potentially being related to social isolation and loneliness. More research is needed for this possibility. In contrast, if someone became more of an Internet Socializer, he or she is more likely to have increased social network size and support and thus achieve active ageing. This research could also inform interventions to increase social interaction of older adults on the Internet. Specifically, given our findings showing that Leisure and Advanced Users had smaller and less supportive social networks than Socializers, especially in rural areas where the rural–urban migration of adult children results in older parents living alone, online communication can play the same important role as a face-to-face interaction in preventing older adults from social isolation. Therefore, a special focus should be put on strengthening Internet-based social interactions for Leisure and Advanced Users, especially in the context of decreasing size of family households, the long distance of residence between older parents and adult children, and the outbreak of COVID-19, which have led to the decline in social networks among older adults.

We note several limitations of this study. First, although CLASS began to gather information on the Internet use among Chinese older adults since 2016, the items used to assess the Internet use in 2018 were changed. As such, we were unable to measure the typology of Internet users with the same indicators across two waves and examine the causal relationship between Internet use patterns and social networks over time. Future longitudinal studies could strengthen the casual relationship between Internet use pattern and social networks. Second, the findings might not be used to understand social network characteristics for people who do not use the Internet, as this study examined the Internet use patterns among Chinese older adults who use the Internet. Third, the four-class Internet users identified in this study are limited to the current cohort of older adults in China. Considering the increase in the number of older online users with more complex Internet use patterns in China, the findings might not be generalized to future ageing cohorts in China. Fourth, this study didn’t use clinical diagnosis to exclude participants with cognitive impairment. There might be potential recall bias for older adults with worse cognitive function.

Conclusion

As individuals use the Internet for diverse purposes, it is challenging to develop a deeper understanding of the sophisticated use patterns if we examine a single item of use. This is the first study to use a large population sample to understand the typology of Internet users according to diverse use behaviors and its social network implications among older adults in China. Our finding of the typology of Internet users reveals the spread and diversity of Internet use behavior among this population and also raises concerns about a new digital divide in terms of “a user typology divide.” In addition, our findings inform policy efforts to maintain and extend social networks for older adults by promoting their online social interaction use. New social media platforms (e.g., TikTok) increase rapidly, consisting of short-form lip-sync videos and enabling users to express oneself creatively and be connected with others. Some older adults become content creators and share experiences in later life, racking up millions of followers, dispelling the stereotype that older adults are passive users of social media. Future studies could further examine how new social media platforms reshape the Internet use behaviors of older adults.

Footnotes

Acknowledgments

The authors would like to thank Dr. Jeffrey Burr (Department of Gerontology, University of Massachusetts Boston) for providing feedback on our study. The baseline survey of CLASS was approved by the Ethics Review Committee of Renmin University of China. Date of the approval is May 15, 2014. The authors would like to thank the Population Development Studies Center and Institute of Gerontology at Renmin University of China for making the data available to the public.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Fund of China [21CRK003].

Data Availability Statement

The datasets are available from the principal investigator of CLASS on a reasonable request.

ORCID iDs

Yang Cao

Mengting Li

References

Akaike

(1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), 716–723. https://doi.org/10.1109/TAC.1974.1100705

Antonucci

T. C.

Ajrouch

K. J.

Manalel

J. A.

(2017). Social relations and technology: Continuity, context, and change. Innovation in Aging, 1(3), igx029. https://doi.org/10.1093/geroni/igx029

Boulton-Lewis

G. M.

Buys

Lovie-Kitchin

Barnett

David

L. N.

(2007). Ageing, learning, and computer technology in Australia. Educational Gerontology, 33(3), 253–270. https://doi.org/10.1080/03601270601161249

Brandtzæg

P. B.

Heim

Karahasanović

(2011). Understanding the new digital divide—a typology of Internet users in Europe. International Journal of Human-Computer Studies, 69(3), 123–138. https://doi.org/10.1016/j.ijhcs.2010.11.004

Carstensen

L. L.

Isaacowitz

D. M.

Charles

S. T.

(1999). Taking time seriously: A theory of socioemotional selectivity. The American Psychologist, 54(3), 165–181. https://doi.org/10.1037/0003-066X.54.3.165

Chen

Y.-R. R.

Schulz

P. J.

(2016). The effect of information communication technology interventions on reducing social isolation in the elderly: A systematic review. Journal of medical Internet research, 18(1), Article e4596. https://doi.org/10.2196/jmir.4596

China Internet Network Information Center. (2022). The 49th statistical report on China’s internet development. Retrieved February 25 from https://www.cnnic.net.cn/hlwfzyj/hlwxzbg/hlwtjbg/202202/P020220407403488048001.pdf

Chiu

C.-J.

(2019). Relationship between internet behaviors and social engagement in middle-aged and older adults in Taiwan. International Journal of Environmental Research and Public Health, 16(3), 416. https://doi.org/10.3390/ijerph16030416

Choi

E. Y.

Kanthawala

Kim

Y. S.

Lee

H. Y.

(2022). Urban/rural digital divide exists in older adults: Does it vary by racial/ethnic groups? Journal of Applied Gerontology, 41(5), 1348–1356. https://doi.org/10.1177/07334648211073605

10.

Choi

N. G.

Dinitto

D. M.

(2013). Internet use among older adults: Association with health needs, psychological capital, and social capital. Journal of medical Internet research, 15(5), Article e97–e97. https://doi.org/10.2196/jmir.2333

11.

Cotten

S. R.

Anderson

W. A.

McCullough

B. M.

(2013). Impact of internet use on loneliness and contact with others among older adults: Cross-sectional analysis. Journal of medical Internet research, 15(2), Article e2306. https://doi.org/10.2196/jmir.2306

12.

Erickson

Johnson

G. M.

(2011). Internet use and psychological wellness during late adulthood. Canadian journal on aging, 30(2), 197–209. https://doi.org/10.1017/S0714980811000109

13.

Fox

Purcell

(2010). Chronic disease and the internet. Pew Research Center. Retrieved August 9 from https://www.pewresearch.org/internet/2010/03/24/chronic-disease-and-the-internet/

14.

Gabbiadini

Baldissarri

Durante

Valtorta

R. R.

De Rosa

Gallucci

(2020). Together apart: The mitigating role of digital communication technologies on negative affect during the COVID-19 outbreak in Italy. Frontiers in Psychology, 12, 2763. https://doi.org/10.3389/fpsyg.2021.701130

15.

Guo

Stensland

Dong

(2020). Parent-adult child relations of chinese older immigrants in the United States: Is there an optimal type? The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 75(4), 889–898. https://doi.org/10.1093/geronb/gbz021

16.

Heo

Chun

Lee

K. H.

Kim

(2015). Internet use and well-being in older adults. Cyberpsychology, Behavior and Social Networking, 18(5), 268–272. https://doi.org/10.1089/cyber.2014.0549

17.

Hülür

Macdonald

(2020). Rethinking social relationships in old age: Digitalization and the social lives of older adults. The American Psychologist, 75(4), 554–566. https://doi.org/10.1037/amp0000604

18.

Iliffe

Kharicha

Harari

Swift

Gillmann

Stuck

A. E.

(2007). Health risk appraisal in older people 2: The implications for clinicians and commissioners of social isolation risk in older people. British journal of general practice, 57(537), 277–282.

19.

Jin

Zhao

(2019). Internet use and the elderly’s active aging in China: A study based on 2016 China longitudinal aging social survey. Population Journal, 41(238), 44–55. https://doi.org/10.16405/j.cnki.1004-129X.2019.06.004

20.

Khosravi

Rezvani

Wiewiora

(2016). The impact of technology on older adults’ social isolation. Computers in Human Behavior, 63, 594–603. https://doi.org/10.1016/j.chb.2016.05.092

21.

(2022). The impact of internet use on the life satisfaction of the elderly in China: Based on CLASS 2018 data. Northwest Population Journal, 43(5), 1–12.

22.

Lubben

Blozik

Gillmann

Iliffe

von Renteln Kruse

Beck

J. C.

Stuck

A. E.

(2006). Performance of an abbreviated version of the lubben social network scale among three european community-dwelling older adult populations. The Gerontologist, 46(4), 503–513. https://doi.org/10.1093/geront/46.4.503

23.

Medlock

Eslami

Askari

Arts

D. L.

Sent

de Rooij

S. E.

Abu-Hanna

(2015). Health information-seeking behavior of seniors who use the internet: A survey. Journal of medical Internet research, 17(1), Article e10–e10. https://doi.org/10.2196/jmir.3749

24.

C. K. L.

Karsh

B.-T.

(2009). A systematic review of patient acceptance of consumer health information technology. Journal of the American Medical Informatics Association: JAMIA, 16(4), 550–560. https://doi.org/10.1197/jamia.M2888

25.

Park

N. S.

Jang

Lee

B. S.

Chiriboga

D. A.

Chang

Kim

S. Y.

(2018). Associations of a social network typology with physical and mental health risks among older adults in South Korea. Aging & Mental Health, 22(5), 631–638. https://doi.org/10.1080/13607863.2017.1286456

26.

Pew Research Center. (2021). Internet/Broadband fact sheet. Retrieved July 11 from https://www.pewresearch.org/internet/fact-sheet/internet-broadband/

27.

Raftery

A. E.

(1986). Choosing models for cross-classifications. American Sociological Review, 51(1), 145–146. https://doi.org/10.2307/2095483

28.

Shakya

H. B.

Christakis

N. A.

(2017). Association of facebook use with compromised well-being: A longitudinal study. American Journal of Epidemiology, 185(3), 203–211. https://doi.org/10.1093/aje/kww189

29.

Silverstein

Zhen Cong Shuzhuo

Li.

(2006). Intergenerational transfers and living arrangements of older people in rural China: Consequences for psychological well-being. Journals of Gerontology Series B: Psychological Sciences & Social Sciences, 61(5), S256–S266. https://doi.org/10.1093/geronb/61.5.S256

30.

Sims

Reed

A. E.

Carr

D. C.

(2017). Information and communication technology use is related to higher well-being among the oldest-old. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 72(5), 761–770. https://doi.org/10.1093/geronb/gbw130

31.

Stockwell

Stubbs

Jackson

S. E.

Fisher

Yang

Smith

(2021). Internet use, social isolation and loneliness in older adults. Ageing & Society, 41(12), 2723–2746. https://doi.org/10.1017/S0144686X20000550

32.

Szabo

Allen

Stephens

Alpass

(2019). Longitudinal analysis of the relationship between purposes of internet use and well-being among older adults. The Gerontologist, 59(1), 58–68. https://doi.org/10.1093/geront/gny036

33.

Tang

Lin

Chen

(2020). Moving beyond living arrangements: The role of family and friendship ties in promoting mental health for urban and rural older adults in China. Aging & Mental Health, 24(9), 1523–1532. https://doi.org/10.1080/13607863.2019.1602589

34.

Wang

Han

S. H.

Mutchler

J. E.

Burr

J. A.

(2021). Family, friendship, and loneliness among older Chinese adults: Urban-rural comparisons. In Silverstein

M. D.

(Ed.), Aging families in Chinese society. Routledge.

35.

Yoon

Jang

Kim

Speasmaker

Nam

(2021). Trends in internet use among older adults in the United States, 2011–2016. Journal of Applied Gerontology, 40(5), 466–470. https://doi.org/10.1177/0733464820908427

36.

Yoon

Jang

Vaughan

P. W.

Garcia

(2020). Older adults’ internet use for health information: Digital divide by race/ethnicity and socioeconomic status. Journal of Applied Gerontology, 39(1), 105–110. https://doi.org/10.1177/0733464818770772

37.

Liu

Chi

(2021). Harnessing mobile technology to support type 2 diabetes self-management among Chinese and hispanic immigrants: A mixed-methods acceptability study. Journal of Ethnic & Cultural Diversity in Social Work, 1–14. https://doi.org/10.1080/15313204.2021.1949775

38.

Zhang

Kim

Silverstein

N. M.

Song

Burr

J. A.

(2021). Social media communication and loneliness among older adults: The mediating roles of social support and social contact. The Gerontologist, 61(6), 888–896. https://doi.org/10.1093/geront/gnaa197