Page 38 - IJPS-6-2

Internet use in older African Americans

Table 3. Ordinal logistic regression on internet use by gender.
Model 1 Model 2 Model 3 Model 4
Men Women Men Women Men Women Men Women
Demographic
Old-old (ref=young-old) 0.33*** 0.20*** 0.58 0.28*** 0.54 0.28*** 0.69 0.35***
Living alone (ref= yes) 0.57* 0.87 1.21 0.99 1.18 0.97 0.38 1.15
Socioeconomic
Years of education 1.32*** 1.42*** 1.32*** 1.41*** 1.18** 1.28***
Married/partnered (ref= yes) 1.52 0.99 1.61 0.99 0.76 1.32
Retired (ref= yes) 0.42*** 0.39*** 0.44** 0.50*** 0.39** 0.39***
Living in poverty (ref= yes) 0.26*** 0.39*** 0.27*** 0.42*** 0.12*** 0.57 +
Health-related
Self-rated health 1.14 1.19 1.12 1.15
Number of diseases 1.07 0.98 1.05 1.07
Difficulties in activities of daily 1.19 0.75* 0.86 0.72*
living
Difficulties in instrumental 0.62 0.93 0.97 1.18
activities of daily living
Mental health-related
Depression 0.77* 0.91
Discrimination 1.13 1.18
Cognitive functioning 1.17*** 1.12***
R square 0.022 0.034 0.115 0.141 0.125 0.151 0.212 0.168
-Log Likelihood 575.95*** 1038.79*** 500.78*** 892.61*** 472.57*** 841.84*** 238.65*** 476.83***
Effect sizes stand for odds ratio. p<0.10; *p<0.05; **p<0.01; ***p<0.001.
+
3.4. Sensitivity Analysis

Sensitivity analyses were conducted and presented in Appendix Tables A1-A3. First, the outcome variable, the internet
use, was analyzed as a continuous variable, and multilinear regressions were conducted (Appendix Table A1). Second,
the ordinal internet use variable (1-7) was recoded into two categories: internet nonuser (1) and user (2-7), and binary
logistic regressions were performed and reported in Appendix Table A2. Third, the ordinal internet use variable (1-7) was
categorized into a binary variable: active internet user (5-7) versus non-active user (1-4), and its binary logistic regressions
were conducted and summarized in Appendix Table A3. Results from all the three analyses remained largely the same.

4. Discussion
The purpose of this study was to examine internet use among African American older adults and investigate gender and
age differences on correlates of internet use. Internet users among older African American adults in this study were <70%,
and the active users, those who used the Internet at least once a week, was only about 55%. Gender and age differences
in internet use were identified in this study: older women and young-old adults had higher percentages of active internet
users and lower percentages of non-users than older men and old-old adults, respectively. Gender and age differences on
the correlates of internet use among older African Americans were revealed: being old-old and difficulties in ADL were
significant factors only for older women, whereas depression was predictive only for older men. Education and cognition
were the only two significant predictors for old-old adults. By contrast, for young-old adults, besides education and
cognition, being retired, living in poverty, and depression all affected their internet use.
Findings in this study indicated that membership within the old-old category was associated with decreased odds
of more frequent use of the Internet only among older African American women but not men. This difference may
be explained by gender differences on internalized ageist stereotypes. Compared to older men, older women are more
susceptible to internalized ageist stereotypes and are more likely to feel helpless, dependent, and weak and have reduced
self-esteem and self-efficacy because of their older age (Chrisler, Barney, and Palatino, 2016; Choi, Kim, Chipalo, et al.,
2020). Due to the negative effects of internalized ageist stereotypes, older women may have lower self-confidence and
believe that they are not capable of using the Internet when they become older, especially when they have functional

32 International Journal of Population Studies | 2020, Volume 6, Issue 2