Longitudinal Analysis of the Relationship Between Purposes of Internet Use and Well-being Among Older Adults

Longitudinal Analysis of the Relationship Between Purposes of Internet Use and Well-being Among... Abstract Purpose There is support for the role of Internet use in promoting well-being among older people. However, there are also contradictory findings which may be attributed to methodological issues. First, research has focused on frequency of online activity rather than how engagement in different types of online activities may influence well-being. Secondly, previous studies have used either cross-sectional designs, which cannot elucidate causality or intervention designs with uncontrolled extraneous variables. In this longitudinal observational study, we test the indirect impact of online engagement for social, informational, and instrumental purposes on older adults’ well-being via reducing loneliness and supporting social engagement. Design and Method A population sample of 1,165 adults aged 60–77 (M = 68.22, SD = 4.42; 52.4% female) was surveyed over 3 waves. Using longitudinal mediation analysis with demographic controls, the indirect effects of types of Internet use on well-being through loneliness and social engagement were estimated. Results Participants engaged online for 3 purposes: social (e.g., connecting with friends/family), instrumental (e.g., banking), and informational (e.g., reading health-related information). Social use indirectly impacted well-being via decreased loneliness and increased social engagement. Informational and instrumental uses indirectly impacted well-being through engagement in a wider range of activities; however, were unrelated to loneliness. Implications Findings highlight that Internet use can support older adults’ well-being; however, not every form of engagement impacts well-being the same way. These findings will inform the focus of interventions which aim to promote well-being. CASP, Internet use, Loneliness, Longitudinal mediation, New Zealand Health, Work & Retirement Study, Volunteering, Well-being As for younger generations, engagement with online platforms has become an important part of the lives of many older adults (aged 60 years or older). According to a recent Pew survey (Pew Research Center, 2017), 67% of American seniors use the Internet, and around 50% have a broadband connection at home. Similar trends have been observed internationally. In New Zealand, Internet access grew from 21% to 75% in the 65–74 age band between 2001 and 2013. Among those aged 75+, the number rose from 10% to almost 50% (Ministry of Social Development, 2016). The popularity of Internet use is in part attributable to its positive impacts on everyday life: It provides convenience to allow the completion of tasks from home, broadening opportunities for entertainment, and enabling to connect with friends and family in an increasingly global era. In response, gerontological research has paid greater attention to understanding the role of the Internet in maintaining and promoting well-being for older people. Internet Use and Well-being Evidence regarding the health- and well-being promoting aspects of Internet use in older age is mixed. Some studies link Internet use to better well-being and mental health in older populations. Using data from Americans aged 50 years and older in the Health and Retirement Study (HRS), Cotten, Ford, Ford, and Hale (2012, 2014) found that the likelihood of clinically relevant depression was substantially lower (by 20%–28%) among Internet users compared with nonusers. This finding was replicated longitudinally, indicating a 33% reduction in the probability of future depression. In a study with community-dwelling older adults (aged 60+) in Ohio, Internet users reported higher levels of personal growth and purpose in life and better self-reported health than nonusers (Chen & Persson, 2002). In contrast, other investigations reported no direct association between Internet use and various health outcomes. For example, analyzing the first wave of the National Health and Aging Trends Study (NHATS) of persons aged 65 and older, Elliot, Mooney, Douthit, and Lynch (2014) found that Internet use was not related to either depression symptoms or well-being. A meta-analysis by Huang (2010) reported a small negative impact of Internet use on psychological well-being. Although the analysis included studies from the general population, age was examined as a moderator. When considering the evidence such studies provide regarding the health-promoting qualities of Internet use, a limitation is their simplistic conceptualization of Internet use. Many compare the health outcomes of users and nonusers, who likely differ in other significant ways aside from Internet use. Similar issues arise with studies focusing on frequency and duration of Internet use. Analysis of persons aged 65 and older from the European Social Survey revealed a positive relationship of frequent Internet use with life satisfaction and happiness (Lelkes, 2013). In contrast, a study of Chinese older adults found no significant relationship between mental health and frequency of Internet use, experience with using the Internet or attitudes toward Internet use, although perceived ease of Internet use was associated with better psychological health (Wong, Yeung, Ho, Tse, & Lam, 2014). These results point to capabilities with technology as a factor associated with psychological health. Current gerontological research has shown that older adults engage with the digital world for different purposes, including personal administration, work, connecting with others, seeking information, and entertainment (Zheng, Spears, Luptak, & Wilby, 2015). This variation in the nature of Internet use was illustrated by van Boekel, Peek, and Luijkx (2017), who conducted a latent class analysis on the Internet use practices of older adults (aged 65+) in the Netherlands. They identified four groups: practical users (engaging in functional activities), minimizers (low-frequency use for email and information search), maximizers (high-frequency use for a range of purposes), and social users (communicating with friends and family). Practical users and maximizers reported the highest level of psychological well-being. Their findings suggest that how and for what purposes older people engage with the online environment might influence whether they experience benefits for well-being. Indeed, a handful of studies have shown that online engagement can impact well-being outcomes differently depending on the purpose of use. Erickson and Johnson (2011) surveyed community-dwelling Canadians aged 60 and older and measured their Internet use practices in terms of frequency, duration, and purpose (communication, information seeking, and entertainment). Greater Internet use for communication and information seeking was positively correlated with life satisfaction, self-efficacy, and social support, and negatively correlated with depression. However, there were no significant relationships between use for entertainment and psychological outcomes. A study focusing on the oldest old (Americans aged 80 years and older) found that using the Internet for social purposes was associated with higher levels of life satisfaction and greater goal attainment. Using the Internet to fulfill informational goals was associated with better physical and subjective health but not life satisfaction (Sims, Reed, & Carr, 2017). Thus, although there is increasing evidence linking Internet use for communication and information seeking purposes to better health and well-being, it is relatively unclear how such use exerts positive effects for older people. In the following sections, we discuss two potential mechanisms—loneliness and social engagement—through which online engagement might promote health and well-being in older adults. Internet Use and Loneliness One pathway through which Internet use might facilitate well-being in older adults is reduced feelings of loneliness. Qualitative research indicates that one of the reasons older people engage online is to seek and maintain social relationships (Nimrod, 2010; Pfeil, 2007). This is further supported by survey research highlighting the negative relationship of Internet use with loneliness and social isolation in older adults (Cotten, Anderson, & McCullough, 2013). Sum, Mathews, Hughes, and Campbell (2008) assessed Australians aged 55 and older about their well-being, loneliness, and Internet use practices through an online survey. They differentiated among five types of Internet use: finding new people, entertainment, commerce, communication, and seeking information. In general, spending more time on the Internet was associated with higher levels of social loneliness. However, time spent online to communicate with others was linked to reduced levels of loneliness. Similar findings were reported with older samples in Canada and the United States (Erickson & Johnson, 2011; Sims et al., 2017). Although seeking information and fulfilling instrumental goals were also primary practices, they were not associated with loneliness. Such findings highlight that overall frequency of use is not necessarily a good indicator of the psychological impact of online activity and that purpose of use may be a better predictor of outcomes. In a separate analysis of this data, Sum, Mathews, Pourghasem, and Hughes (2009) investigated the relationship between types of online engagement and participants’ sense of belonging to online and offline communities. Internet use for communication purposes was associated with a stronger sense of online community, and seeking information on the Internet was correlated with a stronger sense of offline community. In turn, sense of belonging positively predicted well-being. As above instrumental use was not significantly related to the sense of belonging or well-being. Recent studies have more formally assessed loneliness as a potential mediator between Internet use and well-being. Using data from the HRS of adults aged 65 and older, Heo, Chun, Lee, Lee, and Kim (2015) found that Internet use predicted higher social support, leading onto lower levels of loneliness and higher levels of psychological well-being and life satisfaction. In another study with Korean adults aged 50 and older Internet use was related to increased satisfaction with social relationships, which, in turn, was associated with decreased levels of depression symptoms (Jun & Kim, 2017). In sum, numerous studies have established a negative association between Internet use and loneliness in older adults. Furthermore, research highlights that online engagement for social purposes is particularly helpful to reduce loneliness (Erickson & Johnson, 2011; Sims et al., 2017; Sum et al., 2008). Mediation models tested by Heo and colleagues (2015) and Jun and Kim (2017) also provide support for loneliness as an underlying mechanism. However, the cross-sectional designs on which these analyses were conducted mean that these results should be interpreted with caution. Internet Use and Social Engagement In addition to reducing loneliness, it has been suggested that Internet use can contribute to creating social capital. Population studies have shown that frequent Internet use increases social and community engagement (Penard & Poussing, 2010; Wellman, Quan-Haase, Witte, & Hampton, 2001). Although there has been less research undertaken with older adults, both qualitative and quantitative investigations have provided support for a positive relationship between Internet use and community engagement in this age group (Hogeboom, McDermott, Perrin, Osman, & Bell-Ellison, 2010; Kim, Lee, Christensen, & Merighi, 2017; Russell, Campbell, & Hughes, 2008). Choi and DiNitto (2013) analyzed data from the NHATS and found a positive association between Internet use for informational purposes (i.e., searching health-related information) and formal volunteering. In contrast, Internet use for instrumental tasks, such as shopping and banking, was negatively associated with community engagement, such as church attendance. Similar findings were reported by Ihm and Hsieh (2015), who sampled adults aged 60 and older from the Chicago area. They defined instrumental use as obtaining information, services, or other resources without direct social interaction. Social use was described as purposeful interactions with others using information technology. They found that instrumental use was positively related to social engagement in activities, such as volunteering. It is, however, impossible to tell whether this association was driven by Internet use for information seeking, completion of instrumental tasks, or a combination of the two. Social use was unrelated to community engagement. Methodological Limitations Although the findings of previous studies are illuminating, they mostly examined associations between online engagement and psychological constructs using cross-sectional data, often based on relatively small, self-selected samples. These investigations demonstrate the concurrent relationship of Internet use with well-being, loneliness, and social engagement, but they cannot demonstrate any causal effect of online engagement (for exception, see Cotten et al., 2014). Without providing evidence for the temporal nature of the effects, we cannot rule out the possibility that older adults who are healthy, have greater well-being and stronger social ties are also more likely to explore the digital environment. Numerous intervention studies have been implemented to address this methodological issue. However, their validity has often been compromised by uncontrolled extraneous variables. Dickinson and Gregor (2006) provided a critical review of Internet intervention research targeted to improve the well-being of older adults. They concluded that published studies rarely reported a significant change in well-being indicators after the intervention. When significant differences were found, they were confounded by training effects. Specifically, it is impossible to tell whether the positive effects are attributable to using the Internet or to more frequent social interactions as a result of engaging in a group activity. As a response to Dickinson and Gregor (2006), efforts have been made to disentangle training and intervention effects (Czaja, Boot, Charness, Rogers, & Sharit, 2017; Shapira, Barak, & Gal, 2007; Slegers, van Boxtel, & Jolles, 2008). However, no clear and systematic evidence has been found for the effect of intervention or training on well-being after short-term follow-ups. A small meta-analysis by Choi, Kong, and Jung (2012) has shown that Internet use interventions for older adults significantly reduced feelings of loneliness and social isolation but there was no direct impact on depression symptoms. This provides further support for loneliness as a driving mechanism in the relationship between Internet use and well-being outcomes. Although intervention studies could elucidate the potential positive effect of Internet use on well-being, they are often contaminated by extraneous variables which are hard to control. So far, it has been shown that interventions are effective at reducing social isolation, but evidence regarding their impact on well-being is inconclusive (Choi et al., 2012). This could partly be explained by short-term follow-up assessments (months after the intervention) and the focus on the impact of using the technology as opposed to understanding the impact of how older people engage with the online space. Another way to investigate how Internet use affects well-being through different pathways is to employ longitudinal mediation analysis. This technique provides insight into the temporal relations among the variables of interest without requiring experimental manipulation. Present Study Using data from three waves of the New Zealand Health, Work, and Retirement Study (NZHWR), we tested a longitudinal mediation model of Internet use (independent variable), loneliness, and social engagement assessed as volunteering (mediator variables) and well-being (dependent variable). We use a definition and measure of well-being emphasizing satisfaction of human needs. Four domains of well-being are assessed: control, autonomy, pleasure, and self-realization. Each represents capabilities to achieve outcomes valued by the individual and is separate from health, social relationships and material resources. We estimated how Internet use for three different purposes (social, informational, and instrumental) predicted change in loneliness and social engagement, and how these two mediators, in turn, predicted change in well-being. We hypothesized that: Online engagement for social purposes (e.g., communicating with family) would predict reduced levels of loneliness over time. Online engagement for informational use (e.g., reading news) would predict increased social engagement over time. Reductions in loneliness and improvements in social engagement would predict increased well-being over time. Regarding instrumental use, conclusive evidence upon which predictions could be formulated is not available; therefore, the longitudinal analysis of the impact of instrumental use (e.g., shopping) is exploratory. The analytic models controlled for self-rated health at baseline and demographic variables that have been shown to influence Internet use practices among seniors: socioeconomic status (SES), age, gender, marital status, working status, and education (Chang, McAllister, & McCaslin, 2015; Elliot et al., 2014; Erickson & Johnson, 2011; Gell, Rosenberg, Demiris, LaCroix, & Patel, 2015; Hogeboom et al., 2010; Kim et al., 2017). Method Design and Sample The data were drawn from the 2013, 2014, and 2016 waves of the NZHWR, a prospective cohort study of community-dwelling older adults. The NZHWR commenced in 2006 as a postal survey of persons aged 55–70, randomly selected from the New Zealand electoral roll. This original cohort included 6,662 participants, 46% of whom were re-approached for participation in subsequent waves. Data collection has been conducted biennially, with an off-wave survey focusing on Internet use administered to the original cohort in 2013. The 2013 survey was completed by 1,345 participants. Of the total sample, 1,165 (52.4% female, Mage = 68.22, SDage = 4.42) provided information about their Internet use practices and were included in the final analyses. Attrition due to death from 2013 to 2014 and from 2014 to 2016 was 0.3% (n = 3) and 0.4% (n = 4), respectively. Attrition due to noncompletion from 2013 to 2014 and from 2014 to 2016 was 6.9% (n = 80) and 6% (n = 65), respectively. Attrition from 2013 to 2014 was associated with Internet use for social and instrumental purposes, economic living standards, well-being, work status, and self-rated health. Those who completed the 2014 survey scored significantly higher on all scales, were more likely to be working and had better health in 2013 than the dropouts. Attrition from 2014 to 2016 was significantly associated with well-being, working status, self-rated health, and education such that survivors reported greater well-being and better health in 2014. They had a higher educational level and were more likely to be in active employment (statistics are reported in Supplementary Table 1). Measures Information regarding the number of items, response options, coding, and scale range is reported in Table 1. Reliability is reported in Table 2. Table 1. Information About Scale Properties: Number of Items, Response Options, Coding Procedure, and Maximum Range   No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36    No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36  View Large Table 1. Information About Scale Properties: Number of Items, Response Options, Coding Procedure, and Maximum Range   No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36    No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36  View Large Table 2. Descriptive Statistics and Bivariate Correlations   2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87    2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87  aSpearman’s rho was calculated to test the relationship between education and the other study variables. *p < .05; **p < .01. View Large Table 2. Descriptive Statistics and Bivariate Correlations   2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87    2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87  aSpearman’s rho was calculated to test the relationship between education and the other study variables. *p < .05; **p < .01. View Large Control Variables Sociodemographic variables included age, gender, marital status, work status, education, and SES. SES was assessed with the Economic Living Standards Index Short Form (ELSI-SF; Jensen, Spittal, Crichton, Sathiyandra, & Krishnan, 2002), specifically designed for the New Zealand context. Participants rate levels of consumption, possession of household items, availability of resources for social participation, economizing behaviors, global self-ratings of living standards and satisfaction with finances. Self-rated health was measured with a single item: “In general would you say your health is…?” Internet Use Frequency of Internet use for social, informational, and instrumental purposes was each assessed with four items. Social use included connecting with friends, connecting with family, making new connections/friends, and sharing photographs/data. Information seeking was assessed as reading news, seeking health/illness information, seeking other types of information, and searching for music/entertainment. Instrumental use comprised Internet use for work, business, banking, and shopping. Loneliness The short form version of the de Jong Gierveld Loneliness scale consists of six items assessing emotional and social loneliness (de Jong Gierveld & van Tilburg, 2006). Volunteering Participants were asked to indicate how often they give their time to 12 different types of organizations (e.g., sports’ clubs, community/service organizations, religious organization). Two indices were created: (a) diversity of volunteering based on the number of organizations engaged in; and (b) frequency of volunteering based on the average time spent volunteering in any number of organizations. Well-being The CASP-12 was developed by Wiggins, Netuveli, Hyde, Higgs, and Blane (2008) as an older adult specific measure of well-being/quality of life assessing “control,” “autonomy,” “self-realization,” and “pleasure.” Statistical Analysis Statistical analyses were performed using Mplus and SPSS. First, a confirmatory factor analysis was conducted to test the construct validity of the Internet use questionnaire. Next, descriptive statistics were calculated to examine relationships among the study variables and control variables. Finally, three focused longitudinal mediation analyses were employed for hypothesis testing (Jose, 2016). Mediators (loneliness and volunteering) and the dependent variable (well-being) were residualized to assess change over time. Acceptable model fit was determined based on the following fit criteria: χ2/df lower than 5, comparative fit index (CFI) higher than .95, root mean square error of approximation (RMSEA) lower than .06, and standardized root mean square residual (SRMR) lower than .08 (Hu & Bentler, 1999). Missing data (6.2% of the total data set) were treated with the full information maximum likelihood function in Mplus (Geiser, 2013). Results Confirmatory Factor Analysis The analysis indicated a good fit to the data and supported the construct validity of the measure. Further results and fit indices are reported in the Supplementary Material. Descriptive Statistics Bivariate correlations among the key variables are reported in Table 2. All three types of Internet use were significantly related to younger age, better economic living standards, higher levels of education, and better self-rated health. Types of Internet use also showed weak relationships with loneliness, diversity of volunteering, and well-being. Social use was weakly related to frequency of volunteering. There were significant gender differences in Internet use for information seeking and instrumental purposes; t(1,160) = −4.22, p < .001, Cohen’s d = 0.25 and t(1,152) = −5.29, p < .001, Cohen’s d = 0.31, respectively. Men reported more frequent Internet use for both informational (M = 11.7045, SD = 5.20) and instrumental (M = 10.67, SD = 5.87) purposes compared with women (informational: M = 10.47, SD = 4.79; instrumental: M = 8.98, SD = 5.03). Significant differences were found based on working status in all three purposes of use; social: t(1,098) = −2.25, p = .025, Cohen’s d = 0.14; informational: t(1,102) = −2.11, p = .035, Cohen’s d = 0.13; and instrumental: t(752.06) = −15.88, p < .001, Cohen’s d = 1.01. Workers engaged in social (M = 11.96, SD = 4.65), informational (M = 11.48, SD = 4.82), and instrumental uses (M = 12.85, SD = 5.65) to a greater extent than nonworkers (social: M = 11.30, SD = 4.70; informational: M = 10.83, SD = 5.11; instrumental: M = 7.77, SD = 4.34). With respect to marital status, significant differences were found in informational and instrumental uses; t(482.67) = −2.99, p = .003, Cohen’s d = 0. 21 and t(1,125) = −2.90, p = .004, Cohen’s d = 0.20, respectively. Those in a relationship used the Internet for informational (M = 11.39, SD = 4.86) and instrumental (M = 10.12, SD = 5.49) purposes more than their single peers (informational: M = 10.33, SD = 5.35; instrumental: M = 9.04, SD = 5.49). Demographic differences in other study variables are reported in Supplementary Material. Longitudinal Mediation Internet Use for Social Purposes Social use predicted a decrease in loneliness (β = −.073, p = .001) and an increase in the diversity of volunteering activities (β = .077, p = .001) but it was unrelated to the average time spent volunteering (β = −.010, n.s.). Model fit reported in Figure 1. Figure 1. View largeDownload slide Longitudinal mediation model with Internet use for social purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 220.324, df = 50, χ2/df = 4.406, CFI = .954, RMSEA [90% CI] = .054 [.047; .061]; SRMR = .026. Figure 1. View largeDownload slide Longitudinal mediation model with Internet use for social purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 220.324, df = 50, χ2/df = 4.406, CFI = .954, RMSEA [90% CI] = .054 [.047; .061]; SRMR = .026. Internet Use for Informational Purposes Informational use was associated with greater diversity of volunteering activities over time (β = .062, p = .008) but was unrelated to loneliness (β = −.022, n.s.) and average time spent volunteering (β = −.044, n.s.). Model fit reported in Figure 2. Figure 2. View largeDownload slide Longitudinal mediation model with Internet use for informational purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 222.800, df = 50, χ2/df = 4.416, CFI = .953, RMSEA [90% CI] = .054 [.047; .062]; SRMR = .026. Figure 2. View largeDownload slide Longitudinal mediation model with Internet use for informational purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 222.800, df = 50, χ2/df = 4.416, CFI = .953, RMSEA [90% CI] = .054 [.047; .062]; SRMR = .026. Internet Use for Instrumental Purposes Instrumental use was not significantly associated with loneliness (β = −.009, n.s.) or time spent volunteering (β = −.043, n.s.). However, it had a significant association with diversity of volunteering activities (β = .059, p = .011). Model fit reported in Figure 3. Figure 3. View largeDownload slide Longitudinal mediation model with Internet use for instrumental purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 215.526, df = 50, χ2/df = 4.311, CFI = .956, RMSEA [90% CI] = .053 [.046; .061]; SRMR = .026. Figure 3. View largeDownload slide Longitudinal mediation model with Internet use for instrumental purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 215.526, df = 50, χ2/df = 4.311, CFI = .956, RMSEA [90% CI] = .053 [.046; .061]; SRMR = .026. Predictors of Quality of Life Diversity of volunteering activities (β = .048, p = .032) was associated with increments in well-being, while loneliness (β = −.100, p < .001) predicted a reduction in well-being over time. Average time spent volunteering did not significantly predict well-being (β = −.004, n.s.; Figures 1–3). Discussion The article aimed to gain a better understanding of the well-being promoting aspects of online activity for older adults by clarifying the temporal relationships among Internet use, loneliness and social engagement, and well-being. We tested a theoretical model, in which online activity promotes well-being through two pathways; by reducing loneliness and by facilitating social engagement. Instead of focusing on overall time spent online, we explored how frequency of engagement in three types of online activities influenced well-being: (a) social use to interact with others; (b) informational use to search for information and services; and (c) instrumental use to manage personal administration and daily activities. Frequency of engagement in these online activities was moderately correlated, suggesting that engagement in each domain reflects more than an overall pattern of use for older adults. Social use predicted greater well-being over time by reducing loneliness and increasing the breadth of community engagement. In addition to confirming previous cross-sectional research linking socially focused Internet use to lower levels of loneliness (Erickson & Johnson, 2011; Sims et al., 2017; Sum et al., 2008), this finding provides support for the role of online social activity in creating offline social capital. Being able to connect with family and friends helps older adults maintain their sense of belonging and receive social support, which can be especially useful in an era when families are often geographically disconnected. Furthermore, our data suggest that these online interactions may also enable older adults to partake in more diverse volunteer activities. Informational use also predicted increases in the number of volunteer activities participants engaged in, thus promoting well-being. Previous studies converge to link online information seeking to community engagement and volunteering (Choi & DiNitto, 2013), and our results confirm that the effect holds longitudinally. Using the Internet to obtain information about things of their interest can encourage older people to participate in a wider range of activities in the community. For example, by searching information about their chronic health conditions they might find existing support groups in their neighborhood. Similarly, they can gather information online about activities they might like to pursue that they would be unable to obtain otherwise. Using the Internet for instrumental purposes was not predictive of loneliness, but it was associated with an increase in the diversity of volunteer activities. While skills such as banking and shopping online displayed a cross-sectional association with lower levels of loneliness and greater well-being, there was no evidence for long-term benefits over the current follow-up period. Findings suggest that using the Internet for any of the three purposes can encourage older adults to diversify their interests and to explore a wider range of community activities. However, none of the activities had any impact on the frequency of community engagement. The time people can spend on volunteering activities is likely to be influenced by other factors not measured in this study. Results also indicate that being involved in many different organizations, even if with low frequency, is more beneficial for older adults’ well-being than volunteering actively for a few. Social use was the only predictor of loneliness over time, which highlights the significant role online activity for social purposes can play in promoting greater well-being for older adults. In addition, demographic differences associated with Internet use were observed. Age was negatively and weakly related to the social and informational use and had a moderately strong negative association with instrumental use. This is a common finding, indicating that younger older adults use the Internet more frequently (Chang et al., 2015; Choi & DiNitto, 2013; Hogeboom et al., 2010). Men and those in a relationship reported more online engagement for informational and instrumental purposes than women and single participants. Workers were more likely to engage in all three types of activities. Similar differences have been shown in previous research (Chang et al., 2015; Cotten et al., 2014; Jun & Kim, 2017; Kim et al., 2017). Higher SES and education, and better self-reported health were positively related to all three types of online activity, which supports previous findings (Chang et al., 2015; Choi & DiNitto, 2013; Cotten et al., 2014; Ihm & Hsieh, 2015). A shortcoming of the current analytic model is that Internet use was assessed only in 2013. This restricted us to testing a specific temporal order of variables (Jose, 2016). The results show that online engagement for social, informational, and instrumental purposes can promote well-being over time because it is associated with decreased loneliness and/or increased social engagement. However, it does not tell us whether these effects are unidirectional or reciprocal. A feedback loop might exist among any of the variables involved. For example, by reducing loneliness over time, socially focused Internet use can be reinforced as an effective behavior, which might also lead to increased frequency of use over time. To fully understand the direction of the effects and the causality of the relationships, a full longitudinal design is necessary, in which every variable is assessed at each time point, and cross-lagged paths are estimated simultaneously. Potential cohort effects should also be modeled in future investigations. Although we controlled for age, the sample included young-old adults with a relatively limited age range (60–77 years). It is unclear whether the effects would generalize for the oldest old. Future studies should also examine different outcome variables. We assessed well-being with a measure that focuses on older adults’ capacity for control, autonomy, self-realization, and pleasure. Although it is an older-adult specific measure with an emphasis on capabilities (Wiggins et al., 2008), gaining insights into the impact of Internet use on other well-being outcomes or cognitive functioning (Zhang, Grenhart, McLaughlin, & Allaire, 2017) would be informative for intervention purposes. In conclusion, by using longitudinal data from a large sample, we were able to test temporal relationships of Internet use with loneliness, social engagement, and well-being, thus providing evidence for the well-being promoting effects of online engagement among seniors. The results confirm cross-sectional findings that Internet use promotes social engagement and reduces loneliness, and provide empirical support for the direction of the effects. Furthermore, our research highlights that being online is not necessarily sufficient to benefit older people’s well-being. It matters how seniors engage with the digital space. The importance of social use provides a useful focus for interventions which introduce older people to the Internet to enhance their well-being. Supplementary Material Supplementary data are available at The Gerontologist online. Funding This work was funded by the New Zealand Ministry of Science and Innovation (MAUX1205), the New Zealand Ministry of Business, Innovation and Employment (MAUX1403), and the New Zealand Earthquake Commission (2014). Conflict of Interest None reported. References Chang, J., McAllister, C., & McCaslin, R. ( 2015). Correlates of, and barriers to, Internet use among older adults. Journal of Gerontological Social Work , 58, 66– 85. doi: 10.1080/01634372.2014.913754 Google Scholar CrossRef Search ADS   Chen, Y., & Persson, A. ( 2002). Internet use among young and older adults: Relation to psychological well-being. Educational Gerontology , 28, 731– 744. doi: 10.1080/03601270290099921 Google Scholar CrossRef Search ADS   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, e97. doi: 10.2196/jmir.2333 Google Scholar CrossRef Search ADS   Choi, M., Kong, S., & Jung, D. ( 2012). Computer and internet interventions for loneliness and depression in older adults: a meta-analysis. Healthcare Informatics Research , 18, 191– 198. doi:10.4258/hir.2012.18.3.191 Google Scholar CrossRef Search ADS   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, e39. doi: 10.2196/jmir.2306 Google Scholar CrossRef Search ADS   Cotten, S. R., Ford, G., Ford, S., & Hale, T. M. ( 2012). Internet use and depression among older adults. Computers in Human Behavior , 28, 496– 499. doi: 10.1016/j.chb.2011.10.021 Google Scholar CrossRef Search ADS   Cotten, S. R., Ford, G., Ford, S., & Hale, T. M. ( 2014). Internet use and depression among retired older adults in the United States: A longitudinal analysis. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 763– 771. doi: 10.1093/geronb/gbu018 Google Scholar CrossRef Search ADS   Czaja, S. J., Boot, W. R., Charness, N., Rogers, W. A., & Sharit, J. ( 2017). Improving social support for older adults through technology: Findings from the PRISM randomized controlled trial. The Gerontologist . doi: 10.1093/geront/gnw249 de Jong Gierveld, J., & van Tilburh, T. ( 2006). A 6-item scale for overall, emotional, and social loneliness: Confirmatory tests on survey data. Research on Ageing , 28, 582– 598. doi: 10.1177/0164027506289723 Google Scholar CrossRef Search ADS   Dickinson, A., & Gregor, P. ( 2006). Computer use has no demonstrated impact on the well-being of older adults. International Journal of Human-Computer Studies , 64, 744– 753. doi: 10.1016/j.ijhcs.2006.03.001 Google Scholar CrossRef Search ADS   Elliot, A. J., Mooney, C. J., Douthit, K. Z., & Lynch, M. F. ( 2014). Predictors of older adults’ technology use and its relationship to depressive symptoms and well-being. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 667– 677. doi: 10.1093/geronb/gbt109 Google Scholar CrossRef Search ADS   Erickson, J., & Johnson, G. M. ( 2011). Internet use and psychological wellness during late adulthood. Canadian Journal on Aging = La revue Canadienne du Vieillissement , 30, 197– 209. doi: 10.1017/S0714980811000109 Google Scholar CrossRef Search ADS   Geiser, C. ( 2013). Data analysis with Mplus . New York: Guilford. Gell, N. M., Rosenberg, D. E., Demiris, G., LaCroix, A. Z., & Patel, K. V. ( 2015). Patterns of technology use among older adults with and without disabilities. The Gerontologist , 55, 412– 421. doi: 10.1093/geront/gnt166 Google Scholar CrossRef Search ADS   Heo, J., Chun, S., Lee, S., Lee, K. H., & Kim, J. ( 2015). Internet use and well-being in older adults. Cyberpsychology, Behavior and Social Networking , 18, 268– 272. doi: 10.1089/cyber.2014.0549 Google Scholar CrossRef Search ADS   Hogeboom, D. L., McDermott, R. J., Perrin, K. M., Osman, H., & Bell-Ellison, B. A. ( 2010). Internet use and social networking among middle aged and older adults. Educational Gerontology , 36, 93– 111. doi: 10.1080/03601270903058507 Google Scholar CrossRef Search ADS   Hu, L. T., & Bentler, P. M. ( 1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling—A Multidisciplinary Journal , 6, 1– 55. doi: 10.1080/10705519909540118 Google Scholar CrossRef Search ADS   Huang, C. ( 2010). Internet use and psychological well-being: A meta-analysis. Cyberpsychology, Behavior and Social Networking , 13, 241– 249. doi: 10.1089=cyber.2009.0217 Google Scholar CrossRef Search ADS   Ihm, J., & Hsieh, Y. P. ( 2015). The implications of information and communication technology use for the social wellbeing of older adults. Information, Communication & Society , 18, 1123– 1138. doi: 10.1080/1369118X.2015.1019912 Google Scholar CrossRef Search ADS   Jensen, J., Spittal, M., Crichton, S., Sathiyandra, S., & Krishnan, V. ( 2002). Direct measurement of living standards: The New Zealand ELSI scale . Ministry of Social Development: Wellington, New Zealand. Retrieved November 15, 2017 from https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/monitoring/living-standards/living-standards-elsi.html Jose, P. E. ( 2016). The merits of using longitudinal mediation. Educational Psychologist , 51, 331– 341. doi: 10.1080/00461520.2016.1207175 Google Scholar CrossRef Search ADS   Jun, H. J., & Kim, M. Y. ( 2017). What accounts for the relationship between Internet use and suicidal ideation of Korean older adults? A mediation analysis. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 846– 855. doi: 10.1093/geronb/gbw163 Kim, J., Lee, H. Y., Christensen, M. C., & Merighi, J. R. ( 2017). Technology access and use, and their associations with social engagement among older adults: Do women and men differ? The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 836– 845. doi: 10.1093/geronb/gbw123 Lelkes, O. ( 2013). Happier and less isolated: Internet use in old age. Journal of Poverty and Social Justice , 21, 33– 46. doi: 10.1332/175982713X664047 Google Scholar CrossRef Search ADS   Ministry of Social Development. ( 2016). The social report 2016 . Ministry of Social Development: Wellington, New Zealand. Retrieved November 11, 2017 from http://socialreport.msd.govt.nz/documents/2016/msd-the-social-report-2016.pdf Nimrod, G. ( 2010). Seniors’ online communities: A quantitative content analysis. The Gerontologist , 50, 382– 392. doi: 10.1093/geront/gnp141 Google Scholar CrossRef Search ADS   Penard, T., & Poussing, N. ( 2010). Internet use and social capital: The strength of virtual ties. Journal of Economic Issues , 44, 569– 595. doi: 10.2753/JEI0021-3624440301 Google Scholar CrossRef Search ADS   Pew Research Center. ( 2017). Tech adoption climbs among older adults. Retrieved November 15, 2017 from http://assets.pewresearch.org/wp-content/uploads/sites/14/2017/05/16170850/PI_2017.05.17_Older-Americans-Tech_FINAL.pdf Pfeil, U. ( 2007). Online social support for older people. ACM SIGACCESS Accessibility and Computing , 88, 3– 8. doi: 10.1145/1278234.1278235 Google Scholar CrossRef Search ADS   Russell, C., Campbell, A., & Hughes, I. ( 2008). Ageing, social capital and the Internet: Findings from an exploratory study of Australian ‘silver surfers’. Australasian Journal on Ageing , 27, 78– 82. doi: 10.1111/j.1741-6612.2008.00284.x Google Scholar CrossRef Search ADS   Shapira, N., Barak, A., & Gal, I. ( 2007). Promoting older adults’ well-being through Internet training and use. Aging & Mental Health , 11, 477– 484. doi: 10.1080/13607860601086546 Google Scholar CrossRef Search ADS   Sims, T., Reed, A. E., & Carr, D. C. ( 2017). Information and communication technology use is related to higher well-being among the oldest-old. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 761– 770. doi: 10.1093/geronb/gbw130 Slegers, K., van Boxtel, M. P., & Jolles, J. ( 2008). Effects of computer training and Internet usage on the well-being and quality of life of older adults: A randomized, controlled study. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 63, 176– 184. doi: 10.1093/geronb/63.3.P176 Google Scholar CrossRef Search ADS   Sum, S., Mathews, R. M., Hughes, I., & Campbell, A. ( 2008). Internet use and loneliness in older adults. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society , 11, 208– 211. doi: 10.1089/cpb.2007.0010 Sum, S., Mathews, R. M., Pourghasem, M., & Hughes, I. ( 2009). Internet use as a predictor of sense of community in older people. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society , 12, 235– 239. doi: 10.1089/cpb.2008.0150 van Boekel, L. C., Peek, S. T., & Luijkx, K. G. ( 2017). Diversity in older adults’ use of the Internet: Identifying subgroups through latent class analysis. Journal of Medical Internet Research , 19, e180. doi: 10.2196/jmir.6853 Google Scholar CrossRef Search ADS   Wellman, B., Quan-Haase, A., Witte, J., & Hampton, K. ( 2001). Does the Internet increase, decrease, or supplement social capital? Social networks, participation, and community commitment. American Behavioral Scientist , 45, 436– 455. doi: 10.1177/00027640121957286 Google Scholar CrossRef Search ADS   Wiggins, R. D., Netuveli, G., Hyde, M., Higgs, P., & Blane, D. ( 2008). The evaluation of a self-enumerated scale of quality of life (CASP-19) in the context of research on ageing: A combination of exploratory and confirmatory approaches. Social Indicators Research , 89, 61– 77. doi: 10.1007/s11205-007-9220-5 Google Scholar CrossRef Search ADS   Wong, C. K., Yeung, D. Y., Ho, H. C., Tse, K. P., & Lam, C. Y. ( 2014). Chinese older adults’ Internet use for health information. Journal of Applied Gerontology: The Official Journal of the Southern Gerontological Society , 33, 316– 335. doi: 10.1177/0733464812463430 Zhang, S. H., Grenhart, W. C. M., McLaughlin, A. C., & Allaire, J. C. ( 2017). Predicting computer proficiency in older adults. Computers in Human Behavior , 67, 106– 112. doi: 10.1016/j.chb.2016.11.006 Google Scholar CrossRef Search ADS   Zheng, R., Spears, J., Luptak, M., & Wilby, F. ( 2015). Understanding older adults’ perceptions of Internet use: An exploratory factor analysis. Educational Gerontology , 41, 504– 518. doi: 10.1080/03601277.2014.1003495 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Gerontologist Oxford University Press

Longitudinal Analysis of the Relationship Between Purposes of Internet Use and Well-being Among Older Adults

Loading next page...
 
/lp/ou_press/longitudinal-analysis-of-the-relationship-between-purposes-of-internet-7IAmyinfU0
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0016-9013
eISSN
1758-5341
D.O.I.
10.1093/geront/gny036
Publisher site
See Article on Publisher Site

Abstract

Abstract Purpose There is support for the role of Internet use in promoting well-being among older people. However, there are also contradictory findings which may be attributed to methodological issues. First, research has focused on frequency of online activity rather than how engagement in different types of online activities may influence well-being. Secondly, previous studies have used either cross-sectional designs, which cannot elucidate causality or intervention designs with uncontrolled extraneous variables. In this longitudinal observational study, we test the indirect impact of online engagement for social, informational, and instrumental purposes on older adults’ well-being via reducing loneliness and supporting social engagement. Design and Method A population sample of 1,165 adults aged 60–77 (M = 68.22, SD = 4.42; 52.4% female) was surveyed over 3 waves. Using longitudinal mediation analysis with demographic controls, the indirect effects of types of Internet use on well-being through loneliness and social engagement were estimated. Results Participants engaged online for 3 purposes: social (e.g., connecting with friends/family), instrumental (e.g., banking), and informational (e.g., reading health-related information). Social use indirectly impacted well-being via decreased loneliness and increased social engagement. Informational and instrumental uses indirectly impacted well-being through engagement in a wider range of activities; however, were unrelated to loneliness. Implications Findings highlight that Internet use can support older adults’ well-being; however, not every form of engagement impacts well-being the same way. These findings will inform the focus of interventions which aim to promote well-being. CASP, Internet use, Loneliness, Longitudinal mediation, New Zealand Health, Work & Retirement Study, Volunteering, Well-being As for younger generations, engagement with online platforms has become an important part of the lives of many older adults (aged 60 years or older). According to a recent Pew survey (Pew Research Center, 2017), 67% of American seniors use the Internet, and around 50% have a broadband connection at home. Similar trends have been observed internationally. In New Zealand, Internet access grew from 21% to 75% in the 65–74 age band between 2001 and 2013. Among those aged 75+, the number rose from 10% to almost 50% (Ministry of Social Development, 2016). The popularity of Internet use is in part attributable to its positive impacts on everyday life: It provides convenience to allow the completion of tasks from home, broadening opportunities for entertainment, and enabling to connect with friends and family in an increasingly global era. In response, gerontological research has paid greater attention to understanding the role of the Internet in maintaining and promoting well-being for older people. Internet Use and Well-being Evidence regarding the health- and well-being promoting aspects of Internet use in older age is mixed. Some studies link Internet use to better well-being and mental health in older populations. Using data from Americans aged 50 years and older in the Health and Retirement Study (HRS), Cotten, Ford, Ford, and Hale (2012, 2014) found that the likelihood of clinically relevant depression was substantially lower (by 20%–28%) among Internet users compared with nonusers. This finding was replicated longitudinally, indicating a 33% reduction in the probability of future depression. In a study with community-dwelling older adults (aged 60+) in Ohio, Internet users reported higher levels of personal growth and purpose in life and better self-reported health than nonusers (Chen & Persson, 2002). In contrast, other investigations reported no direct association between Internet use and various health outcomes. For example, analyzing the first wave of the National Health and Aging Trends Study (NHATS) of persons aged 65 and older, Elliot, Mooney, Douthit, and Lynch (2014) found that Internet use was not related to either depression symptoms or well-being. A meta-analysis by Huang (2010) reported a small negative impact of Internet use on psychological well-being. Although the analysis included studies from the general population, age was examined as a moderator. When considering the evidence such studies provide regarding the health-promoting qualities of Internet use, a limitation is their simplistic conceptualization of Internet use. Many compare the health outcomes of users and nonusers, who likely differ in other significant ways aside from Internet use. Similar issues arise with studies focusing on frequency and duration of Internet use. Analysis of persons aged 65 and older from the European Social Survey revealed a positive relationship of frequent Internet use with life satisfaction and happiness (Lelkes, 2013). In contrast, a study of Chinese older adults found no significant relationship between mental health and frequency of Internet use, experience with using the Internet or attitudes toward Internet use, although perceived ease of Internet use was associated with better psychological health (Wong, Yeung, Ho, Tse, & Lam, 2014). These results point to capabilities with technology as a factor associated with psychological health. Current gerontological research has shown that older adults engage with the digital world for different purposes, including personal administration, work, connecting with others, seeking information, and entertainment (Zheng, Spears, Luptak, & Wilby, 2015). This variation in the nature of Internet use was illustrated by van Boekel, Peek, and Luijkx (2017), who conducted a latent class analysis on the Internet use practices of older adults (aged 65+) in the Netherlands. They identified four groups: practical users (engaging in functional activities), minimizers (low-frequency use for email and information search), maximizers (high-frequency use for a range of purposes), and social users (communicating with friends and family). Practical users and maximizers reported the highest level of psychological well-being. Their findings suggest that how and for what purposes older people engage with the online environment might influence whether they experience benefits for well-being. Indeed, a handful of studies have shown that online engagement can impact well-being outcomes differently depending on the purpose of use. Erickson and Johnson (2011) surveyed community-dwelling Canadians aged 60 and older and measured their Internet use practices in terms of frequency, duration, and purpose (communication, information seeking, and entertainment). Greater Internet use for communication and information seeking was positively correlated with life satisfaction, self-efficacy, and social support, and negatively correlated with depression. However, there were no significant relationships between use for entertainment and psychological outcomes. A study focusing on the oldest old (Americans aged 80 years and older) found that using the Internet for social purposes was associated with higher levels of life satisfaction and greater goal attainment. Using the Internet to fulfill informational goals was associated with better physical and subjective health but not life satisfaction (Sims, Reed, & Carr, 2017). Thus, although there is increasing evidence linking Internet use for communication and information seeking purposes to better health and well-being, it is relatively unclear how such use exerts positive effects for older people. In the following sections, we discuss two potential mechanisms—loneliness and social engagement—through which online engagement might promote health and well-being in older adults. Internet Use and Loneliness One pathway through which Internet use might facilitate well-being in older adults is reduced feelings of loneliness. Qualitative research indicates that one of the reasons older people engage online is to seek and maintain social relationships (Nimrod, 2010; Pfeil, 2007). This is further supported by survey research highlighting the negative relationship of Internet use with loneliness and social isolation in older adults (Cotten, Anderson, & McCullough, 2013). Sum, Mathews, Hughes, and Campbell (2008) assessed Australians aged 55 and older about their well-being, loneliness, and Internet use practices through an online survey. They differentiated among five types of Internet use: finding new people, entertainment, commerce, communication, and seeking information. In general, spending more time on the Internet was associated with higher levels of social loneliness. However, time spent online to communicate with others was linked to reduced levels of loneliness. Similar findings were reported with older samples in Canada and the United States (Erickson & Johnson, 2011; Sims et al., 2017). Although seeking information and fulfilling instrumental goals were also primary practices, they were not associated with loneliness. Such findings highlight that overall frequency of use is not necessarily a good indicator of the psychological impact of online activity and that purpose of use may be a better predictor of outcomes. In a separate analysis of this data, Sum, Mathews, Pourghasem, and Hughes (2009) investigated the relationship between types of online engagement and participants’ sense of belonging to online and offline communities. Internet use for communication purposes was associated with a stronger sense of online community, and seeking information on the Internet was correlated with a stronger sense of offline community. In turn, sense of belonging positively predicted well-being. As above instrumental use was not significantly related to the sense of belonging or well-being. Recent studies have more formally assessed loneliness as a potential mediator between Internet use and well-being. Using data from the HRS of adults aged 65 and older, Heo, Chun, Lee, Lee, and Kim (2015) found that Internet use predicted higher social support, leading onto lower levels of loneliness and higher levels of psychological well-being and life satisfaction. In another study with Korean adults aged 50 and older Internet use was related to increased satisfaction with social relationships, which, in turn, was associated with decreased levels of depression symptoms (Jun & Kim, 2017). In sum, numerous studies have established a negative association between Internet use and loneliness in older adults. Furthermore, research highlights that online engagement for social purposes is particularly helpful to reduce loneliness (Erickson & Johnson, 2011; Sims et al., 2017; Sum et al., 2008). Mediation models tested by Heo and colleagues (2015) and Jun and Kim (2017) also provide support for loneliness as an underlying mechanism. However, the cross-sectional designs on which these analyses were conducted mean that these results should be interpreted with caution. Internet Use and Social Engagement In addition to reducing loneliness, it has been suggested that Internet use can contribute to creating social capital. Population studies have shown that frequent Internet use increases social and community engagement (Penard & Poussing, 2010; Wellman, Quan-Haase, Witte, & Hampton, 2001). Although there has been less research undertaken with older adults, both qualitative and quantitative investigations have provided support for a positive relationship between Internet use and community engagement in this age group (Hogeboom, McDermott, Perrin, Osman, & Bell-Ellison, 2010; Kim, Lee, Christensen, & Merighi, 2017; Russell, Campbell, & Hughes, 2008). Choi and DiNitto (2013) analyzed data from the NHATS and found a positive association between Internet use for informational purposes (i.e., searching health-related information) and formal volunteering. In contrast, Internet use for instrumental tasks, such as shopping and banking, was negatively associated with community engagement, such as church attendance. Similar findings were reported by Ihm and Hsieh (2015), who sampled adults aged 60 and older from the Chicago area. They defined instrumental use as obtaining information, services, or other resources without direct social interaction. Social use was described as purposeful interactions with others using information technology. They found that instrumental use was positively related to social engagement in activities, such as volunteering. It is, however, impossible to tell whether this association was driven by Internet use for information seeking, completion of instrumental tasks, or a combination of the two. Social use was unrelated to community engagement. Methodological Limitations Although the findings of previous studies are illuminating, they mostly examined associations between online engagement and psychological constructs using cross-sectional data, often based on relatively small, self-selected samples. These investigations demonstrate the concurrent relationship of Internet use with well-being, loneliness, and social engagement, but they cannot demonstrate any causal effect of online engagement (for exception, see Cotten et al., 2014). Without providing evidence for the temporal nature of the effects, we cannot rule out the possibility that older adults who are healthy, have greater well-being and stronger social ties are also more likely to explore the digital environment. Numerous intervention studies have been implemented to address this methodological issue. However, their validity has often been compromised by uncontrolled extraneous variables. Dickinson and Gregor (2006) provided a critical review of Internet intervention research targeted to improve the well-being of older adults. They concluded that published studies rarely reported a significant change in well-being indicators after the intervention. When significant differences were found, they were confounded by training effects. Specifically, it is impossible to tell whether the positive effects are attributable to using the Internet or to more frequent social interactions as a result of engaging in a group activity. As a response to Dickinson and Gregor (2006), efforts have been made to disentangle training and intervention effects (Czaja, Boot, Charness, Rogers, & Sharit, 2017; Shapira, Barak, & Gal, 2007; Slegers, van Boxtel, & Jolles, 2008). However, no clear and systematic evidence has been found for the effect of intervention or training on well-being after short-term follow-ups. A small meta-analysis by Choi, Kong, and Jung (2012) has shown that Internet use interventions for older adults significantly reduced feelings of loneliness and social isolation but there was no direct impact on depression symptoms. This provides further support for loneliness as a driving mechanism in the relationship between Internet use and well-being outcomes. Although intervention studies could elucidate the potential positive effect of Internet use on well-being, they are often contaminated by extraneous variables which are hard to control. So far, it has been shown that interventions are effective at reducing social isolation, but evidence regarding their impact on well-being is inconclusive (Choi et al., 2012). This could partly be explained by short-term follow-up assessments (months after the intervention) and the focus on the impact of using the technology as opposed to understanding the impact of how older people engage with the online space. Another way to investigate how Internet use affects well-being through different pathways is to employ longitudinal mediation analysis. This technique provides insight into the temporal relations among the variables of interest without requiring experimental manipulation. Present Study Using data from three waves of the New Zealand Health, Work, and Retirement Study (NZHWR), we tested a longitudinal mediation model of Internet use (independent variable), loneliness, and social engagement assessed as volunteering (mediator variables) and well-being (dependent variable). We use a definition and measure of well-being emphasizing satisfaction of human needs. Four domains of well-being are assessed: control, autonomy, pleasure, and self-realization. Each represents capabilities to achieve outcomes valued by the individual and is separate from health, social relationships and material resources. We estimated how Internet use for three different purposes (social, informational, and instrumental) predicted change in loneliness and social engagement, and how these two mediators, in turn, predicted change in well-being. We hypothesized that: Online engagement for social purposes (e.g., communicating with family) would predict reduced levels of loneliness over time. Online engagement for informational use (e.g., reading news) would predict increased social engagement over time. Reductions in loneliness and improvements in social engagement would predict increased well-being over time. Regarding instrumental use, conclusive evidence upon which predictions could be formulated is not available; therefore, the longitudinal analysis of the impact of instrumental use (e.g., shopping) is exploratory. The analytic models controlled for self-rated health at baseline and demographic variables that have been shown to influence Internet use practices among seniors: socioeconomic status (SES), age, gender, marital status, working status, and education (Chang, McAllister, & McCaslin, 2015; Elliot et al., 2014; Erickson & Johnson, 2011; Gell, Rosenberg, Demiris, LaCroix, & Patel, 2015; Hogeboom et al., 2010; Kim et al., 2017). Method Design and Sample The data were drawn from the 2013, 2014, and 2016 waves of the NZHWR, a prospective cohort study of community-dwelling older adults. The NZHWR commenced in 2006 as a postal survey of persons aged 55–70, randomly selected from the New Zealand electoral roll. This original cohort included 6,662 participants, 46% of whom were re-approached for participation in subsequent waves. Data collection has been conducted biennially, with an off-wave survey focusing on Internet use administered to the original cohort in 2013. The 2013 survey was completed by 1,345 participants. Of the total sample, 1,165 (52.4% female, Mage = 68.22, SDage = 4.42) provided information about their Internet use practices and were included in the final analyses. Attrition due to death from 2013 to 2014 and from 2014 to 2016 was 0.3% (n = 3) and 0.4% (n = 4), respectively. Attrition due to noncompletion from 2013 to 2014 and from 2014 to 2016 was 6.9% (n = 80) and 6% (n = 65), respectively. Attrition from 2013 to 2014 was associated with Internet use for social and instrumental purposes, economic living standards, well-being, work status, and self-rated health. Those who completed the 2014 survey scored significantly higher on all scales, were more likely to be working and had better health in 2013 than the dropouts. Attrition from 2014 to 2016 was significantly associated with well-being, working status, self-rated health, and education such that survivors reported greater well-being and better health in 2014. They had a higher educational level and were more likely to be in active employment (statistics are reported in Supplementary Table 1). Measures Information regarding the number of items, response options, coding, and scale range is reported in Table 1. Reliability is reported in Table 2. Table 1. Information About Scale Properties: Number of Items, Response Options, Coding Procedure, and Maximum Range   No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36    No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36  View Large Table 1. Information About Scale Properties: Number of Items, Response Options, Coding Procedure, and Maximum Range   No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36    No. of items  Response options  Coding  Range  Living standards index  25  A combination of 3-, 4-, and 5-point Likert scales  Algorithm developed by Jensen et al. (2005)  0–31  Education  1  (1) No qualification; (2) secondary school; (3) postsecondary certificate; (4) university degree  —  1–4  Self-rated health  1  (1) Excellent; (2) very good; (3) good; (4) fair; (5) poor  —  1–5  Internet use   Social  4  (1) Never; (2) once every few months; (3) about once a month; (4) several times a month; (5) several times a week; (6) daily  Sum all items  1–24   Informational  4  Sum all items  1–24   Instrumental  4  Sum all items  1–24  Loneliness  6  (0) “No”; (1) “more or less” or “yes”  Sum all items  0–6  Volunteering   Diversity  12  (0) Never; (1) once a year; (2) twice a year; (3) four times a year; (4) monthly; (5) weekly; (6) daily  Count number of organizations  0–12   Frequency  12  Average time across all items  0–6  Well-being  12  (0) Never; (1) not often; (2) sometimes; (3) often  Sum all items  0–36  View Large Table 2. Descriptive Statistics and Bivariate Correlations   2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87    2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87  aSpearman’s rho was calculated to test the relationship between education and the other study variables. *p < .05; **p < .01. View Large Table 2. Descriptive Statistics and Bivariate Correlations   2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87    2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  M  SD  α  1. Age  −.011  −.087**  .066*  −.116**  −.148**  −.350**  −.004  −.017  .029  .020  .090**  .087**  −.015  −.029  −.114**  68.22  4.42  —  2. Living standards    .149**  −.298**  .094**  .129**  .170**  −.328**  −.352**  −.053  −.022  −.026  −0.045  .562**  .556**  .492**  24.56  5.69  —  3. Educationa      −.061*  .115**  .221**  .287**  −.010  −.032  .145**  .151**  .121**  .117**  .110**  .125**  .062  —  —  —  4. Self-assessed health        −.126**  −.065*  −.157**  .219**  .195**  −.037  .006  −.026  .012  −.454**  −.425**  −.425**  2.43  0.87  —  5. Social use          .578**  .403**  −.137**  −.169**  .133**  .166**  .106**  .041  .129**  .109**  .090**  11.53  4.75  .81  6. Informational use            .484**  −.024  −.067*  .065*  .111**  .039  −.022  .099**  .098**  .047  11.06  5.02  .77  7. Instrumental use              −.061*  −.083**  .102**  .105**  .030  −.040  .176**  .160**  .157**  9.79  5.51  .76  8. Loneliness ‘13                .650**  −0.031  −.076*  −.010  −.021  −.488**  −.482**  −.470**  1.65  1.73  .76  9. Loneliness ‘14                  −0.052  −.065*  −.041  −.058  −.433**  −.508**  −.428**  1.73  1.63  .69  10. No. of organizations ‘13                    .660**  .799**  .325**  .095**  .036  .076*  3.47  2.70  —  11. No. of organizations ‘14                      .631**  .301**  .080*  .064*  .083*  3.05  2.45  —  12. Volunteering freq. ‘13                        .415**  .087**  .048  .067*  1.06  0.92  —  13. Volunteering freq. ‘14                          .056  .042  .036  1.47  1.43  —  14. CASP ‘13                            .696**  .702**  28.92  5.18  .81  15. CASP ‘14                              .715**  28.87  5.45  .86  16. CASP ‘16                                28.84  5.35  .87  aSpearman’s rho was calculated to test the relationship between education and the other study variables. *p < .05; **p < .01. View Large Control Variables Sociodemographic variables included age, gender, marital status, work status, education, and SES. SES was assessed with the Economic Living Standards Index Short Form (ELSI-SF; Jensen, Spittal, Crichton, Sathiyandra, & Krishnan, 2002), specifically designed for the New Zealand context. Participants rate levels of consumption, possession of household items, availability of resources for social participation, economizing behaviors, global self-ratings of living standards and satisfaction with finances. Self-rated health was measured with a single item: “In general would you say your health is…?” Internet Use Frequency of Internet use for social, informational, and instrumental purposes was each assessed with four items. Social use included connecting with friends, connecting with family, making new connections/friends, and sharing photographs/data. Information seeking was assessed as reading news, seeking health/illness information, seeking other types of information, and searching for music/entertainment. Instrumental use comprised Internet use for work, business, banking, and shopping. Loneliness The short form version of the de Jong Gierveld Loneliness scale consists of six items assessing emotional and social loneliness (de Jong Gierveld & van Tilburg, 2006). Volunteering Participants were asked to indicate how often they give their time to 12 different types of organizations (e.g., sports’ clubs, community/service organizations, religious organization). Two indices were created: (a) diversity of volunteering based on the number of organizations engaged in; and (b) frequency of volunteering based on the average time spent volunteering in any number of organizations. Well-being The CASP-12 was developed by Wiggins, Netuveli, Hyde, Higgs, and Blane (2008) as an older adult specific measure of well-being/quality of life assessing “control,” “autonomy,” “self-realization,” and “pleasure.” Statistical Analysis Statistical analyses were performed using Mplus and SPSS. First, a confirmatory factor analysis was conducted to test the construct validity of the Internet use questionnaire. Next, descriptive statistics were calculated to examine relationships among the study variables and control variables. Finally, three focused longitudinal mediation analyses were employed for hypothesis testing (Jose, 2016). Mediators (loneliness and volunteering) and the dependent variable (well-being) were residualized to assess change over time. Acceptable model fit was determined based on the following fit criteria: χ2/df lower than 5, comparative fit index (CFI) higher than .95, root mean square error of approximation (RMSEA) lower than .06, and standardized root mean square residual (SRMR) lower than .08 (Hu & Bentler, 1999). Missing data (6.2% of the total data set) were treated with the full information maximum likelihood function in Mplus (Geiser, 2013). Results Confirmatory Factor Analysis The analysis indicated a good fit to the data and supported the construct validity of the measure. Further results and fit indices are reported in the Supplementary Material. Descriptive Statistics Bivariate correlations among the key variables are reported in Table 2. All three types of Internet use were significantly related to younger age, better economic living standards, higher levels of education, and better self-rated health. Types of Internet use also showed weak relationships with loneliness, diversity of volunteering, and well-being. Social use was weakly related to frequency of volunteering. There were significant gender differences in Internet use for information seeking and instrumental purposes; t(1,160) = −4.22, p < .001, Cohen’s d = 0.25 and t(1,152) = −5.29, p < .001, Cohen’s d = 0.31, respectively. Men reported more frequent Internet use for both informational (M = 11.7045, SD = 5.20) and instrumental (M = 10.67, SD = 5.87) purposes compared with women (informational: M = 10.47, SD = 4.79; instrumental: M = 8.98, SD = 5.03). Significant differences were found based on working status in all three purposes of use; social: t(1,098) = −2.25, p = .025, Cohen’s d = 0.14; informational: t(1,102) = −2.11, p = .035, Cohen’s d = 0.13; and instrumental: t(752.06) = −15.88, p < .001, Cohen’s d = 1.01. Workers engaged in social (M = 11.96, SD = 4.65), informational (M = 11.48, SD = 4.82), and instrumental uses (M = 12.85, SD = 5.65) to a greater extent than nonworkers (social: M = 11.30, SD = 4.70; informational: M = 10.83, SD = 5.11; instrumental: M = 7.77, SD = 4.34). With respect to marital status, significant differences were found in informational and instrumental uses; t(482.67) = −2.99, p = .003, Cohen’s d = 0. 21 and t(1,125) = −2.90, p = .004, Cohen’s d = 0.20, respectively. Those in a relationship used the Internet for informational (M = 11.39, SD = 4.86) and instrumental (M = 10.12, SD = 5.49) purposes more than their single peers (informational: M = 10.33, SD = 5.35; instrumental: M = 9.04, SD = 5.49). Demographic differences in other study variables are reported in Supplementary Material. Longitudinal Mediation Internet Use for Social Purposes Social use predicted a decrease in loneliness (β = −.073, p = .001) and an increase in the diversity of volunteering activities (β = .077, p = .001) but it was unrelated to the average time spent volunteering (β = −.010, n.s.). Model fit reported in Figure 1. Figure 1. View largeDownload slide Longitudinal mediation model with Internet use for social purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 220.324, df = 50, χ2/df = 4.406, CFI = .954, RMSEA [90% CI] = .054 [.047; .061]; SRMR = .026. Figure 1. View largeDownload slide Longitudinal mediation model with Internet use for social purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 220.324, df = 50, χ2/df = 4.406, CFI = .954, RMSEA [90% CI] = .054 [.047; .061]; SRMR = .026. Internet Use for Informational Purposes Informational use was associated with greater diversity of volunteering activities over time (β = .062, p = .008) but was unrelated to loneliness (β = −.022, n.s.) and average time spent volunteering (β = −.044, n.s.). Model fit reported in Figure 2. Figure 2. View largeDownload slide Longitudinal mediation model with Internet use for informational purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 222.800, df = 50, χ2/df = 4.416, CFI = .953, RMSEA [90% CI] = .054 [.047; .062]; SRMR = .026. Figure 2. View largeDownload slide Longitudinal mediation model with Internet use for informational purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 222.800, df = 50, χ2/df = 4.416, CFI = .953, RMSEA [90% CI] = .054 [.047; .062]; SRMR = .026. Internet Use for Instrumental Purposes Instrumental use was not significantly associated with loneliness (β = −.009, n.s.) or time spent volunteering (β = −.043, n.s.). However, it had a significant association with diversity of volunteering activities (β = .059, p = .011). Model fit reported in Figure 3. Figure 3. View largeDownload slide Longitudinal mediation model with Internet use for instrumental purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 215.526, df = 50, χ2/df = 4.311, CFI = .956, RMSEA [90% CI] = .053 [.046; .061]; SRMR = .026. Figure 3. View largeDownload slide Longitudinal mediation model with Internet use for instrumental purposes presenting standardized regression coefficients (beta weights). Covariances were estimated but omitted from the figure to enhance readability. Only significant paths are reported. Model fit: χ2 = 215.526, df = 50, χ2/df = 4.311, CFI = .956, RMSEA [90% CI] = .053 [.046; .061]; SRMR = .026. Predictors of Quality of Life Diversity of volunteering activities (β = .048, p = .032) was associated with increments in well-being, while loneliness (β = −.100, p < .001) predicted a reduction in well-being over time. Average time spent volunteering did not significantly predict well-being (β = −.004, n.s.; Figures 1–3). Discussion The article aimed to gain a better understanding of the well-being promoting aspects of online activity for older adults by clarifying the temporal relationships among Internet use, loneliness and social engagement, and well-being. We tested a theoretical model, in which online activity promotes well-being through two pathways; by reducing loneliness and by facilitating social engagement. Instead of focusing on overall time spent online, we explored how frequency of engagement in three types of online activities influenced well-being: (a) social use to interact with others; (b) informational use to search for information and services; and (c) instrumental use to manage personal administration and daily activities. Frequency of engagement in these online activities was moderately correlated, suggesting that engagement in each domain reflects more than an overall pattern of use for older adults. Social use predicted greater well-being over time by reducing loneliness and increasing the breadth of community engagement. In addition to confirming previous cross-sectional research linking socially focused Internet use to lower levels of loneliness (Erickson & Johnson, 2011; Sims et al., 2017; Sum et al., 2008), this finding provides support for the role of online social activity in creating offline social capital. Being able to connect with family and friends helps older adults maintain their sense of belonging and receive social support, which can be especially useful in an era when families are often geographically disconnected. Furthermore, our data suggest that these online interactions may also enable older adults to partake in more diverse volunteer activities. Informational use also predicted increases in the number of volunteer activities participants engaged in, thus promoting well-being. Previous studies converge to link online information seeking to community engagement and volunteering (Choi & DiNitto, 2013), and our results confirm that the effect holds longitudinally. Using the Internet to obtain information about things of their interest can encourage older people to participate in a wider range of activities in the community. For example, by searching information about their chronic health conditions they might find existing support groups in their neighborhood. Similarly, they can gather information online about activities they might like to pursue that they would be unable to obtain otherwise. Using the Internet for instrumental purposes was not predictive of loneliness, but it was associated with an increase in the diversity of volunteer activities. While skills such as banking and shopping online displayed a cross-sectional association with lower levels of loneliness and greater well-being, there was no evidence for long-term benefits over the current follow-up period. Findings suggest that using the Internet for any of the three purposes can encourage older adults to diversify their interests and to explore a wider range of community activities. However, none of the activities had any impact on the frequency of community engagement. The time people can spend on volunteering activities is likely to be influenced by other factors not measured in this study. Results also indicate that being involved in many different organizations, even if with low frequency, is more beneficial for older adults’ well-being than volunteering actively for a few. Social use was the only predictor of loneliness over time, which highlights the significant role online activity for social purposes can play in promoting greater well-being for older adults. In addition, demographic differences associated with Internet use were observed. Age was negatively and weakly related to the social and informational use and had a moderately strong negative association with instrumental use. This is a common finding, indicating that younger older adults use the Internet more frequently (Chang et al., 2015; Choi & DiNitto, 2013; Hogeboom et al., 2010). Men and those in a relationship reported more online engagement for informational and instrumental purposes than women and single participants. Workers were more likely to engage in all three types of activities. Similar differences have been shown in previous research (Chang et al., 2015; Cotten et al., 2014; Jun & Kim, 2017; Kim et al., 2017). Higher SES and education, and better self-reported health were positively related to all three types of online activity, which supports previous findings (Chang et al., 2015; Choi & DiNitto, 2013; Cotten et al., 2014; Ihm & Hsieh, 2015). A shortcoming of the current analytic model is that Internet use was assessed only in 2013. This restricted us to testing a specific temporal order of variables (Jose, 2016). The results show that online engagement for social, informational, and instrumental purposes can promote well-being over time because it is associated with decreased loneliness and/or increased social engagement. However, it does not tell us whether these effects are unidirectional or reciprocal. A feedback loop might exist among any of the variables involved. For example, by reducing loneliness over time, socially focused Internet use can be reinforced as an effective behavior, which might also lead to increased frequency of use over time. To fully understand the direction of the effects and the causality of the relationships, a full longitudinal design is necessary, in which every variable is assessed at each time point, and cross-lagged paths are estimated simultaneously. Potential cohort effects should also be modeled in future investigations. Although we controlled for age, the sample included young-old adults with a relatively limited age range (60–77 years). It is unclear whether the effects would generalize for the oldest old. Future studies should also examine different outcome variables. We assessed well-being with a measure that focuses on older adults’ capacity for control, autonomy, self-realization, and pleasure. Although it is an older-adult specific measure with an emphasis on capabilities (Wiggins et al., 2008), gaining insights into the impact of Internet use on other well-being outcomes or cognitive functioning (Zhang, Grenhart, McLaughlin, & Allaire, 2017) would be informative for intervention purposes. In conclusion, by using longitudinal data from a large sample, we were able to test temporal relationships of Internet use with loneliness, social engagement, and well-being, thus providing evidence for the well-being promoting effects of online engagement among seniors. The results confirm cross-sectional findings that Internet use promotes social engagement and reduces loneliness, and provide empirical support for the direction of the effects. Furthermore, our research highlights that being online is not necessarily sufficient to benefit older people’s well-being. It matters how seniors engage with the digital space. The importance of social use provides a useful focus for interventions which introduce older people to the Internet to enhance their well-being. Supplementary Material Supplementary data are available at The Gerontologist online. Funding This work was funded by the New Zealand Ministry of Science and Innovation (MAUX1205), the New Zealand Ministry of Business, Innovation and Employment (MAUX1403), and the New Zealand Earthquake Commission (2014). Conflict of Interest None reported. References Chang, J., McAllister, C., & McCaslin, R. ( 2015). Correlates of, and barriers to, Internet use among older adults. Journal of Gerontological Social Work , 58, 66– 85. doi: 10.1080/01634372.2014.913754 Google Scholar CrossRef Search ADS   Chen, Y., & Persson, A. ( 2002). Internet use among young and older adults: Relation to psychological well-being. Educational Gerontology , 28, 731– 744. doi: 10.1080/03601270290099921 Google Scholar CrossRef Search ADS   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, e97. doi: 10.2196/jmir.2333 Google Scholar CrossRef Search ADS   Choi, M., Kong, S., & Jung, D. ( 2012). Computer and internet interventions for loneliness and depression in older adults: a meta-analysis. Healthcare Informatics Research , 18, 191– 198. doi:10.4258/hir.2012.18.3.191 Google Scholar CrossRef Search ADS   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, e39. doi: 10.2196/jmir.2306 Google Scholar CrossRef Search ADS   Cotten, S. R., Ford, G., Ford, S., & Hale, T. M. ( 2012). Internet use and depression among older adults. Computers in Human Behavior , 28, 496– 499. doi: 10.1016/j.chb.2011.10.021 Google Scholar CrossRef Search ADS   Cotten, S. R., Ford, G., Ford, S., & Hale, T. M. ( 2014). Internet use and depression among retired older adults in the United States: A longitudinal analysis. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 763– 771. doi: 10.1093/geronb/gbu018 Google Scholar CrossRef Search ADS   Czaja, S. J., Boot, W. R., Charness, N., Rogers, W. A., & Sharit, J. ( 2017). Improving social support for older adults through technology: Findings from the PRISM randomized controlled trial. The Gerontologist . doi: 10.1093/geront/gnw249 de Jong Gierveld, J., & van Tilburh, T. ( 2006). A 6-item scale for overall, emotional, and social loneliness: Confirmatory tests on survey data. Research on Ageing , 28, 582– 598. doi: 10.1177/0164027506289723 Google Scholar CrossRef Search ADS   Dickinson, A., & Gregor, P. ( 2006). Computer use has no demonstrated impact on the well-being of older adults. International Journal of Human-Computer Studies , 64, 744– 753. doi: 10.1016/j.ijhcs.2006.03.001 Google Scholar CrossRef Search ADS   Elliot, A. J., Mooney, C. J., Douthit, K. Z., & Lynch, M. F. ( 2014). Predictors of older adults’ technology use and its relationship to depressive symptoms and well-being. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 667– 677. doi: 10.1093/geronb/gbt109 Google Scholar CrossRef Search ADS   Erickson, J., & Johnson, G. M. ( 2011). Internet use and psychological wellness during late adulthood. Canadian Journal on Aging = La revue Canadienne du Vieillissement , 30, 197– 209. doi: 10.1017/S0714980811000109 Google Scholar CrossRef Search ADS   Geiser, C. ( 2013). Data analysis with Mplus . New York: Guilford. Gell, N. M., Rosenberg, D. E., Demiris, G., LaCroix, A. Z., & Patel, K. V. ( 2015). Patterns of technology use among older adults with and without disabilities. The Gerontologist , 55, 412– 421. doi: 10.1093/geront/gnt166 Google Scholar CrossRef Search ADS   Heo, J., Chun, S., Lee, S., Lee, K. H., & Kim, J. ( 2015). Internet use and well-being in older adults. Cyberpsychology, Behavior and Social Networking , 18, 268– 272. doi: 10.1089/cyber.2014.0549 Google Scholar CrossRef Search ADS   Hogeboom, D. L., McDermott, R. J., Perrin, K. M., Osman, H., & Bell-Ellison, B. A. ( 2010). Internet use and social networking among middle aged and older adults. Educational Gerontology , 36, 93– 111. doi: 10.1080/03601270903058507 Google Scholar CrossRef Search ADS   Hu, L. T., & Bentler, P. M. ( 1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling—A Multidisciplinary Journal , 6, 1– 55. doi: 10.1080/10705519909540118 Google Scholar CrossRef Search ADS   Huang, C. ( 2010). Internet use and psychological well-being: A meta-analysis. Cyberpsychology, Behavior and Social Networking , 13, 241– 249. doi: 10.1089=cyber.2009.0217 Google Scholar CrossRef Search ADS   Ihm, J., & Hsieh, Y. P. ( 2015). The implications of information and communication technology use for the social wellbeing of older adults. Information, Communication & Society , 18, 1123– 1138. doi: 10.1080/1369118X.2015.1019912 Google Scholar CrossRef Search ADS   Jensen, J., Spittal, M., Crichton, S., Sathiyandra, S., & Krishnan, V. ( 2002). Direct measurement of living standards: The New Zealand ELSI scale . Ministry of Social Development: Wellington, New Zealand. Retrieved November 15, 2017 from https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/monitoring/living-standards/living-standards-elsi.html Jose, P. E. ( 2016). The merits of using longitudinal mediation. Educational Psychologist , 51, 331– 341. doi: 10.1080/00461520.2016.1207175 Google Scholar CrossRef Search ADS   Jun, H. J., & Kim, M. Y. ( 2017). What accounts for the relationship between Internet use and suicidal ideation of Korean older adults? A mediation analysis. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 846– 855. doi: 10.1093/geronb/gbw163 Kim, J., Lee, H. Y., Christensen, M. C., & Merighi, J. R. ( 2017). Technology access and use, and their associations with social engagement among older adults: Do women and men differ? The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 836– 845. doi: 10.1093/geronb/gbw123 Lelkes, O. ( 2013). Happier and less isolated: Internet use in old age. Journal of Poverty and Social Justice , 21, 33– 46. doi: 10.1332/175982713X664047 Google Scholar CrossRef Search ADS   Ministry of Social Development. ( 2016). The social report 2016 . Ministry of Social Development: Wellington, New Zealand. Retrieved November 11, 2017 from http://socialreport.msd.govt.nz/documents/2016/msd-the-social-report-2016.pdf Nimrod, G. ( 2010). Seniors’ online communities: A quantitative content analysis. The Gerontologist , 50, 382– 392. doi: 10.1093/geront/gnp141 Google Scholar CrossRef Search ADS   Penard, T., & Poussing, N. ( 2010). Internet use and social capital: The strength of virtual ties. Journal of Economic Issues , 44, 569– 595. doi: 10.2753/JEI0021-3624440301 Google Scholar CrossRef Search ADS   Pew Research Center. ( 2017). Tech adoption climbs among older adults. Retrieved November 15, 2017 from http://assets.pewresearch.org/wp-content/uploads/sites/14/2017/05/16170850/PI_2017.05.17_Older-Americans-Tech_FINAL.pdf Pfeil, U. ( 2007). Online social support for older people. ACM SIGACCESS Accessibility and Computing , 88, 3– 8. doi: 10.1145/1278234.1278235 Google Scholar CrossRef Search ADS   Russell, C., Campbell, A., & Hughes, I. ( 2008). Ageing, social capital and the Internet: Findings from an exploratory study of Australian ‘silver surfers’. Australasian Journal on Ageing , 27, 78– 82. doi: 10.1111/j.1741-6612.2008.00284.x Google Scholar CrossRef Search ADS   Shapira, N., Barak, A., & Gal, I. ( 2007). Promoting older adults’ well-being through Internet training and use. Aging & Mental Health , 11, 477– 484. doi: 10.1080/13607860601086546 Google Scholar CrossRef Search ADS   Sims, T., Reed, A. E., & Carr, D. C. ( 2017). Information and communication technology use is related to higher well-being among the oldest-old. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 72, 761– 770. doi: 10.1093/geronb/gbw130 Slegers, K., van Boxtel, M. P., & Jolles, J. ( 2008). Effects of computer training and Internet usage on the well-being and quality of life of older adults: A randomized, controlled study. The Journal of Gerontology, Series B: Psychological Sciences and Social Sciences , 63, 176– 184. doi: 10.1093/geronb/63.3.P176 Google Scholar CrossRef Search ADS   Sum, S., Mathews, R. M., Hughes, I., & Campbell, A. ( 2008). Internet use and loneliness in older adults. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society , 11, 208– 211. doi: 10.1089/cpb.2007.0010 Sum, S., Mathews, R. M., Pourghasem, M., & Hughes, I. ( 2009). Internet use as a predictor of sense of community in older people. Cyberpsychology & Behavior: The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society , 12, 235– 239. doi: 10.1089/cpb.2008.0150 van Boekel, L. C., Peek, S. T., & Luijkx, K. G. ( 2017). Diversity in older adults’ use of the Internet: Identifying subgroups through latent class analysis. Journal of Medical Internet Research , 19, e180. doi: 10.2196/jmir.6853 Google Scholar CrossRef Search ADS   Wellman, B., Quan-Haase, A., Witte, J., & Hampton, K. ( 2001). Does the Internet increase, decrease, or supplement social capital? Social networks, participation, and community commitment. American Behavioral Scientist , 45, 436– 455. doi: 10.1177/00027640121957286 Google Scholar CrossRef Search ADS   Wiggins, R. D., Netuveli, G., Hyde, M., Higgs, P., & Blane, D. ( 2008). The evaluation of a self-enumerated scale of quality of life (CASP-19) in the context of research on ageing: A combination of exploratory and confirmatory approaches. Social Indicators Research , 89, 61– 77. doi: 10.1007/s11205-007-9220-5 Google Scholar CrossRef Search ADS   Wong, C. K., Yeung, D. Y., Ho, H. C., Tse, K. P., & Lam, C. Y. ( 2014). Chinese older adults’ Internet use for health information. Journal of Applied Gerontology: The Official Journal of the Southern Gerontological Society , 33, 316– 335. doi: 10.1177/0733464812463430 Zhang, S. H., Grenhart, W. C. M., McLaughlin, A. C., & Allaire, J. C. ( 2017). Predicting computer proficiency in older adults. Computers in Human Behavior , 67, 106– 112. doi: 10.1016/j.chb.2016.11.006 Google Scholar CrossRef Search ADS   Zheng, R., Spears, J., Luptak, M., & Wilby, F. ( 2015). Understanding older adults’ perceptions of Internet use: An exploratory factor analysis. Educational Gerontology , 41, 504– 518. doi: 10.1080/03601277.2014.1003495 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The GerontologistOxford University Press

Published: Apr 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off