PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 52nd ANNUAL MEETING—2008 139

OLDER ADULTS AND TECHNOLOGY ADOPTION

Sara J. Czaja, Ph.D.1, Chin Chin Lee, M.S., M.S.P.H.2, Sankaran

N. Nair, M.S.2, Joseph Sharit, Ph.D.3
1
Department of Psychiatry and Behavioral Sciences, University of Miami Miller School
of Medicine, 2Center on Aging, University of Miami Miller School of Medicine,
3
Industrial Engineer, University of Miami,

Access to computers and the Internet is a major public policy concern as technology
has become a significant aspect of economic, social and health equity. Recent data
suggest that although computer and Internet use is lower among older, as compared to
younger adults access is increasing among older people. This paper examines changes
in use of computers and the Internet over time (2000-2002 and 2006-2007) among two
samples (N=424) of older adults ranging in age from 50–85 yrs. Data are also reported
on changes in attitudes towards computers and how adoption is influenced by attitudes
and demographic characteristics. Technology adoption and attitudes towards computers
were assessed via questionnaire. Over time, although there was no difference in
percentage of participants who had computer experience, both breadth of computer use
and Internet use increased. Participants from the more recent time point also reported
more comfort with computers. The data also indicated that age, education, and comfort
with computers predicted breadth of computer and Internet use. Understanding factors
that influence access is important to the development of strategies to close the gap
between adopters and non-adopters.

INTRODUCTION Life Report, 2004, 2007). Although the use of

Copyright 2008 by Human Factors and Ergonomics Society, Inc. All rights reserved. 10.1518/107118108X351554

computers and the Internet is increasing among

Computers, the Internet and other older people there still remains a “digital divide.”
communication technologies are ubiquitous in most In 2005 about 26% of people age 65+ were Internet
aspects of our society. The Internet is rapidly users as compared to 67% of people age 50-64 and
becoming a major vehicle for communication and 80% of those 30-49 years old (Pew Internet and
information dissemination about health, American Life Project, 2005). Further, seniors who
community and government services. For example, use the Internet tend to be white, highly-educated
in 2006, 80% of Internet users or 113 million and living in households with higher incomes (Pew
Americans searched the Internet for information on Internet and American Life Project, 2004). In 2007
health issues such as illnesses, treatment options, or 32% of adults over the age of 65 reported using the
health insurance (Pew Internet and American Life Internet as compared to 65% of those age 50-64
Report, 2006). The U.S. government and yrs. and 83% of those aged 30-49 yrs. This is about
community agencies are also increasingly using the a 5% increase among those 65+ yrs.
Internet to provide information about, and access Lower rates of technology adoption among
to, programs and services. In addition, many older adults place them at a disadvantage in terms
businesses and consumer groups now rely on the of their ability to live and function independently.
Internet for information transmission. In fact, there is increasing concern that people who
In general, use of the computer and the are part of the digital divide will have less power as
Internet has increased substantially within the consumers, less access to health information, fewer
general population. In 2007, 71% of people in the options for dealing with government agencies and
United States used the Internet as compared to fewer opportunities for interaction. Development
about 22% in 1997(Pew Internet and American of strategies to help ensure that older people have
Downloaded from pro.sagepub.com at NORTHWESTERN UNIV LIBRARY on June 5, 2016
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 52nd ANNUAL MEETING—2008 140

equal opportunities to access computer and Internet working full or part-time than participants in 2000-
technologies depends on understanding factors that 2002.
influence technology adoption. Prior research
investigating this issue has suggested that adoption Materials
is influenced by socio-demographic factors (e.g., Attitudes towards computers were assessed
Pew Internet and American Life Report, 2005) and using the Attitudes Towards Computers
attitudinal issues (e.g., Czaja, Charness, Fisk, Questionnaire (ATCQ; Jay & Willis, 1992). The
Hertzog, Nair, Rogers and Sharit, 2006). ATCQ is a 35-item multidimensional scale
Given the explosive use of the Internet in assessing seven dimensions of attitudes toward
the past several years this paper examines if use of computers: comfort (feelings of comfort with the
computers and the Internet has changed among computer and its use); efficacy (feelings of
older adults and if factors other than age, such as competence with the computer); gender equality
education and attitudes influence technology (the belief that computers are important to both
adoption. We also examine change in use of men and women); control (the belief that people
various Internet applications. In addition, we control computers); interest (the extent to which
present data on change in computer attitudes over one is interested in learning about and using
time. It may be that as computers and the use of the computers); dehumanization (the belief that
Internet have become an integral part of routine computers are dehumanizing); and utility (the
activities that older adults are beginning to express belief that computers are useful). Each dimension
more comfort with this technology. is assessed by 5 or 6 items that are rated on a 5-
point Likert-type scale format. The Technology
METHOD and Computer Experience Questionnaire (Czaja et
al., 2006, 2007) assessed participant experience
Sample with computers and breadth of computer and
The sample included 424 adults (136 males Internet use. Breadth of computer experience was
and 288 females) ranging in age from 50-85 yrs. based on experience with various input devices,
(M= 66.76; SD = 9.47). Participants were recruited computer and window functions and computer
from two time periods: 2000-2002 (N= 225, M = applications. Breadth of Internet use was based on
65.28 yrs.; SD = 8.98); and 2006-2007, (N = 199, frequency of using the Internet and number of
M = 65.94 yrs.; SD = 9.35). The sample was fairly Internet-based activities.
well-educated; 78% had educations beyond high
school (trade school or some college) and 42% had Protocol
were college graduates. Eighty five percent of the All participants were recruited as part of
sample reported that they were in good to excellent on-going studies examining age differences in the
health and 27% were working either full or part performance of computer-based tasks (e.g., Czaja
time. Participants were community dwelling and et al., 2001). Recruitment occurred during the years
spoke English. All participants were screened for 2000-2002 and 2006-2007. Following telephone
cognitive impairment (a score of > 27) on the Mini- screening participants were scheduled for the
Mental Status Exam (Folstein, Folstein & laboratory testing session where they completed
McHugh, 1975). the questionnaires. During the testing session,
To determine if the samples were similar, participants were provided with an explanation of
the characteristics of the two groups were study activities and asked to read and sign a
compared across the two time periods. There were participant consent form. They completed the
no differences between the groups in age, demographic questionnaire, technology experience
education, health status, or gender composition (all questionnaire, and the ATCQ prior to participating
ps > .05). However there was a significant in any of the computer-based tasks.
difference between the groups in occupational
status (F (2,424) = 12.83, p < .05). A greater
number of participants from 2006-2007 were
Downloaded from pro.sagepub.com at NORTHWESTERN UNIV LIBRARY on June 5, 2016
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 52nd ANNUAL MEETING—2008 141

RESULTS Although there was no difference in

computer experience (p > .05), there were
Differences in categorical variables (e.g., significant main effects of time period for breadth
demographic characteristics, experience with of computer use (t (349) = - 5.29, p < .001) and
computers) were analyzed using Chi Square tests breadth of Internet use (t (289) = -2.96, p < .01)
with time as a grouping factor. Differences in (Figure 1). Overall, across the two time periods,
breadth of computer and Internet experience and both breadth of computer use and Internet use
the attitude subscales were analyzed using t-tests increased between the two samples of older adults.
with time period at two levels as the grouping The findings also indicated that the number of
factor. For this analysis, examination of change in hours spent using the Internet per week increased
attitudes towards computers was restricted to three over time (χ2 (3,401) = 26.3, p < .001) such that
subscales: interest, efficacy and comfort. Prior participants from time period 2 reported that they
research (Czaja et al., 2006) has shown that these spent more time using the Internet as compared to
dimensions of attitudes are important to technology participants from time period 1.
adoption. Differences in patterns of Internet use In terms of activities performed on the
were examined using descriptive statistics. Internet, the data indicate that overall the most
Correlations were also computed to determine common activities are communication, accessing
relationships among the variables. Finally, news and health information followed by
hierarchical regression analyses were conducted to entertainment and shopping. However, the results
examine factors that predict breadth of computer also show that use of the Internet for finding
and Internet use. Separate models were computed information about the community (χ2 (1, 291) =
for each outcome measure. In these analyses time 28.22, p < .001), government issues (χ2 (1, 291) =
period was entered in the first step as a control 24.52, p < .001), and shopping (χ2 (1, 291) = 16.03,
variable, demographic characteristics (age, p < .001) increased between the two time periods.
education, health status and occupational status) There was a slight trend indicating an increased use
were entered in the second step and the attitude of the Internet over time to find health related
subscale for comfort was entered in the third step. information (p = .07) (Figure 2).
The age by demographic characteristics and
attitude subscale interactions were also examined.
70
We did not enter the efficacy and interest subscales
in the equation as these subscales were highly 60
correlated with comfort. 50

40
Percent

8
7 30

6 20
5
4 10

3 0
2 Community Government  Shopping Health‐related
1 Issues

0
2000‐2002 2006‐2007
Breadth of PC Use Breadth of Internet Use

2000‐2002 2006‐2007 Figure 2. Internet Use by Time.

With respect to computer attitudes, the data

Figure 1. Breadth of PC Use and Internet Use by indicated that feelings of comfort with computers
Time Period. increased between the two time periods (t (422) = -
2.01, p < .05). There was no change in interest in
Downloaded from pro.sagepub.com at NORTHWESTERN UNIV LIBRARY on June 5, 2016
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 52nd ANNUAL MEETING—2008 142

computers or feelings of self-efficacy related to DISCUSSION

computers. The data indicated that computer self-
efficacy was strongly related to computer comfort Computers and the Internet are increasingly
(r (424) = .63, p < .001) and interest in computers becoming the focal point of many routine
was also related to comfort with computers (r (424) activities. In fact, access to technology is becoming
= .63, p < .001). People who reported higher self- an important public policy issue as the Internet is a
efficacy and greater interest in computers also significant component of economic, social and
reported more comfort with computers. health equity. Recent data indicates that despite
The results of the hierarchical regression increased use of computers and the Internet in the
analysis indicated that breadth of computer use was general population a digital divide still exists for
predicted by age, education, occupational status older adults. Unfortunately, as the use of these
and comfort with computers. In general, younger technologies is becoming important to success in
adults and those who were better educated and most aspects of life older adults who are “non-
were working had more breadth of computer users” of technology may become increasingly
experience. Also, a greater degree of comfort was disadvantaged with respect to their ability to
associated with greater breadth of use. Age, function in today’s digital world.
education and computer comfort were also
significant predictors of breadth of Internet Table 2. Final Hierarchical Regression Model for
experience. The relationships were in the same Breadth of PC and Internet Use
direction as for breadth of computer use (Tables 1
and 2). Independent Variables β F df p
Breadth of PC Use
Time 2 (2006-2007) .327 5.914 1 .000
Table 1. Hierarchical Regressions with R2 and Age -.012 -3.413 1 .001
Trade school/Some college± .249 2.966 3 .003
Increment of in R2 for Breadth of PC and Internet College graduate± .286 3.033 3 .003
Use Post graduate± .417 4.738 3 .000
Part-time⊥ -.173 -1.676 4 .095
Student⊥ -.098 -.525 4 .600
Models R2 Adj. R2 Δ R2 ΔF Retired⊥ -.242 -2.573 4 .011
Breadth of PC Use Other⊥ -.296 -2.885 4 .004
Model 1 .074 .072 .074 27.994*** General health status .025 .831 1 .407
Time Computer attitude - Comfort .066 10.347 1 .000
Model 2 .255 .233 .181 9.177***
Time, Age, Education, Breadth of Internet use
Occupation status, Time 2 (2006-2007) .949 3.182 1 .002
General health status Age -.055 -3.139 1 .002
***
Model 3 .434 .416 .179 107.055 Trade school/Some college± .192 .385 3 .701
Time period, Age, College graduate± 1.167 2.110 3 .036
Education, Occupation Post graduate± 1.505 2.931 3 .004
status, General health Part-time⊥ -1.073 -1.977 4 .049
status, Computer Student⊥ -.843 -.867 4 .387
attitude - Comfort
Retired⊥ -.678 -1.376 4 .170
Other⊥ -.772 -1.425 4 .155
Breadth of Internet use
General health status -.127 -.771 1 .441
Model 1 .029 .026 .029 8.747**
Computer attitude - Comfort .258 7.577 1 .000
Time ±
Model 2 .180 .151 .151 5.732*** For education, the reference group is high school education.
⊥
Time, Age, Education, For occupational status, the reference group is the full-time.
Occupation status,
General health status
Model 3 .320 .293 .140 57.413*** Overall, the data from this study indicate
Time, Age, Education, that although low in comparison to the general
Occupation status,
General health status, population, breadth of computer and Internet use
Computer attitude - increased in two comparable samples of older
Comfort
***
p < .001. **
p < .01 adults between the time periods 2000-2001 and
2006-2007. In addition, time spent using the
Internet per week increased over time. These
findings are consistent with recent data from the
Pew Internet and American Life Project and
Downloaded from pro.sagepub.com at NORTHWESTERN UNIV LIBRARY on June 5, 2016
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 52nd ANNUAL MEETING—2008 143

suggest that as the baby boomers age the digital older adults generally report less comfort and
gap may decrease. However, the results from the competence with computers than younger people
regression analysis indicated that age was an (Nair, Lee & Czaja, 2005; Czaja et al., 2006).
important predictor of use and that use is lower Understanding use of technology and factors that
among the older people in our sample. Our data influence technology access is important to the
also show that usage is influenced by factors other design of intervention strategies to foster
than age such as education, and people with higher technology adoption.
levels of education report greater use. It is
important to point out that our sample was fairly ACKNOWLEDGEMENTS
well educated; a majority of people had beyond a
high school education. It may be that our findings This research is supported by National
regarding an increase in use would not be found for Institute on Aging Grants # AG11748-05 and
those who are less educated. These findings are AG17211-0252.
also consistent with the findings from the Pew REFERENCES
study and suggest access problems for large
subgroups of the older adult population. Czaja, S. J., Charness, N., Fisk, A. D., Hertzog, C.,
Interestingly, the data also indicate that Nair, S. N., Rogers, W., & Sharit, J. (2006).
among our participants the Internet is increasingly Factors Predicting the Use of Technology:
being used to find information about community Findings from the Center for Research and
programs and government services. This finding Education on Aging and Technology
may be due to necessity. In 2002 the United States Enhancement (CREATE). Psychology and
government passed the E-Government Act of 2002, Aging, 21, 333-352.
and the United States government has since been Czaja, S. J., Sharit, J. Ownby, R., Roth, D. L., &
increasingly using the Internet to exchange Nair, S. N. (2001). Examining age
information and services with citizens and differences in performance of a complex
businesses. People can now file taxes on-line and information search and retrieval task.
enroll in programs such as Social Security and Psychology & Aging, 16, 564-579.
Medicare on-line. The Internet is also a focal point Folstein, M. F., Folstein, S. A., & McHugh, P. R.
for many health care activities. There was also an (1975). Mini-Mental State: A practical
increased use of the Internet for shopping. Again, method for grading the cognitive state of
this may be due to necessity or to the fact that patients for the clinician. Journal of
comfort with computers increased across the two Psychiatric Research, 12, 189-198.
time periods. This finding lends support to the idea Jay, G. M. & Willis, S. L. (1992). Influence of
that use of the Internet may increase the direct computer experience on older adults’
independence of older adults, especially those with attitudes toward computers. Journal of
restricted mobility as access to this technology may Gerontology: Psychological Sciences, 47,
enable them to engage in activities such as 250-257.
shopping. Pew Internet & American Life (2004, March 25).
Our findings also suggest that comfort with Older Americans and the Internet.
computers increased across the two samples and, Retrieved May 1, 2004, from
consistent with other data (e.g., Czaja et al., 2006), http://www.pewinternet.org/pdfs/PIP_Senio
comfort was an important predictor of breadth of rs_Online_2004.pdf
computer and Internet use. Surprisingly, there was
no change in Interest in computers or computer
self-efficacy. This is an important finding as the
data indicated that computer self-efficacy is
strongly related to comfort with computers. Also, it
is important to note that although comfort with
computers increased across the two time periods,
Downloaded from pro.sagepub.com at NORTHWESTERN UNIV LIBRARY on June 5, 2016