Methodological considerations in surveys of older adults: technology matters

Abstract

Surveys of adults over the age of 55 have unique methodological considerations, which
typically concern the physiological and psychological factors associated with age-related
declines in cognitive functioning and health. Though significant, these comprise a limited
view of the way in which older adults differ from the general population. Lower rates of
adoption and use of newer information and communication technologies and concerns about
privacy in the online environment are other ways which this population is distinct, and should
be considered in survey research. As surveys grow more technologically advanced, older
adults too may regard data collection practices differently than younger adults, leading to
varying rates of participation and response. This paper reviews the literature on the
methodological issues inherent in surveying older adults, and analyzes data collected in a
large telephone survey to provide further evidence that technology adoption and use should
be considered as a cultural practice of this subpopulation, a view which holds additional
methodological implications for survey research.

Full text

International Journal of Emerging Technologies and Society
Vol. 8, No. 2, 2010, pp: 114 – 133
©International Journal of Emerging Technologies and Society 2010
ISSN 1835-8780
http://www.swin.edu.au/ijets
Methodological Considerations in Surveys of Older Adults:
Technology Matters
Kelly Quinn is a PhD candidate in the Department of Communication at the University of
Illinois at Chicago
Abstract
Surveys of adults over the age of 55 have unique methodological considerations, which
typically concern the physiological and psychological factors associated with age-related
declines in cognitive functioning and health. Though significant, these comprise a limited
view of the way in which older adults differ from the general population. Lower rates of
adoption and use of newer information and communication technologies and concerns about
privacy in the online environment are other ways which this population is distinct, and should
be considered in survey research. As surveys grow more technologically advanced, older
adults too may regard data collection practices differently than younger adults, leading to
varying rates of participation and response. This paper reviews the literature on the
methodological issues inherent in surveying older adults, and analyzes data collected in a
large telephone survey to provide further evidence that technology adoption and use should
be considered as a cultural practice of this subpopulation, a view which holds additional
methodological implications for survey research.
Keywords:
Methodological Issues – Surveys – Older Adults – Technology Use

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
115
http://www.swin.edu.au/ijets
Methodological Considerations in Surveys of Older Adults:
Technology Matters
Introduction
The dynamic demographic structures of many societies emphasize the need for continual
revaluation of subpopulations and subcultures in the effort to surface potential systemic
biases in survey research. Frequently, these examinations consider differences in
subgroups related to regional language and cultural practices, such as those found in
immigrant populations or minority ethnic groups. Often overlooked however are
considerations related to age and ageing, particularly with older adults over the age of 55.
As a sub-segment of society that is not only increasing as a percentage of the total
population, but also one that carries significant political and economic weight, this group has
markedly different characteristics than the general populace. Generally, the methodological
considerations made by survey researchers for this age group are confined to the
physiological and psychological factors associated with declines in cognitive functioning and
health. While these considerations are significant, they comprise a limited view of the ways
in which this group may differ from the general population. In recent years, one important
area in which this age group differs markedly in practices is in their adoption and use of
newer information and communication technologies, or ICTs (Pew Internet 2009).
Lower participation rates by older adults in newer technologies are often treated as a literacy
issue by many survey researchers, and are compensated for by an adoption of survey
administration to paper-and-pencil or face-to-face forms. But these approaches diminish the
differences as a proficiency matter, and overlook their meaning as cultural practices that
have emerged in light of progressively mediated communications. As survey methodologies
grow more technologically-dependent these variances become a potential source of bias and
an increasingly important component of survey data analysis. The purpose of this paper is to
highlight some of the more conventionally-understood methodological issues involved in
surveying older adults and, through an examination of survey data on Internet use in a US
population, raise additional considerations for researchers in light of increased use of
technologically sophisticated data collection mechanisms.
The Older Adult Population
The interest in the older adult population, defined as those individuals over the age of 55, is
fuelled by its sheer size and rapid growth as a percentage of the overall populace over the
past 30 years. In the United States (US), for example, adults over the age of 55 comprised
approximately 20.7% of the population in 1979 (US Census Bureau 1979) but had grown to
23.1% of the population by 2007 (US Census Bureau 2007a). Older adults bring
considerable financial muscle to the consumer economy and their voting rates are
significantly higher than those of the population on average – 62.5% in the US, as contrasted
with 36.3% of adults under the age of 55 (US Census Bureau 2008). In 2005, the US
Census Bureau estimated that the rate of growth in the resident population of those age 55
and older between the years 2000 and 2005 was 13%, four times that of those persons
younger than the age of 55 (US Census Bureau 2007b), and it anticipated a continuation in
this high rate as more Baby Boomers mature into this age segment. Because they comprise
such a rapidly expanding and potent social, political, and economic force, interest in this age
group has increased dramatically from a social sciences perspective.

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
116
In addition, due to advances in medical technologies, the health status of this age group has
undergone considerable transformation over the past 50 years, with a significantly increased
lifespan and marked improvements in the quality of life at advanced ages (Martin et al.
2009). These epidemiological shifts have sparked increased inquiry by health researchers,
due to the greater numbers of older persons generally and more widespread occurrences of
physical and psychological afflictions at older ages. This interest by health researchers
taken together with the social sciences focus reveals the considerable attention that the older
adult segment of the population commands from researchers from a variety of disciplines, as
they try to understand and predict behaviours and outcomes of this population segment.
Surveys are a common tool to gather epidemiological, behavioural and social data on all
populations, including older persons. And like other types of research, it may be prone to
specific weaknesses depending on the processes used to select samples, gather data and
interpret results. Survey research on older persons faces some unique challenges not found
with surveys of the general population due to the strong associations between age, cognition
and health status. But using age as a demographic benchmark to segment out these issues
has it limitations. Age can be a conflated variable because it represents three distinct time-
related dimensions: (a) the ageing effect, or physical and cognitive change associated with
maturation; (b) the period effect, or the consequence of influences that occur through time
and which tend to be uniform across cohorts; and (c) the cohort effect, or the effect that
results from the unique socio-historical time at which a group of individuals is born
(Settersten 2003).
Much of the research on surveys with older adults has concentrated on the consequences of
ageing, or the physiological and psychological changes associated with maturation, which
contribute to systemic difference in data collection and interpretation. Often overlooked,
however, are the differences attributable to this group’s unique placement in socio-historic
time, which has resulted in a distinctive frame of reference and patterned way of living that
distinguishes older adults from their younger counterparts. One defining criterion for
examining these group-related differences is the Internet’s rise to common adoption as a
communication medium, as it has resulted in unique and learned practices and meanings for
each group.
The Pew Internet & American Life Project (2009) reports that while Internet usage rates
generally trend downward as age increases, a sharper and more rapid decline is evidenced
in adults aged 55 years and older, as evidenced in Figure 1.
The abrupt arrival and rapid deployment of the Internet as a commonplace communication
medium ensured that the process to acquire the necessary skills for everyday use was
substantively different for older adults than what was experienced by younger adults. As a
result, older adults have developed cultural practices, such as lower technology adoption and
usage rates, and cultural values, such as a concern for the privacy of personal information
captured in the online environment, that are distinguishable from those of younger adults.
These differences hold implications for research strategies and data analysis in surveys that
use of technology for data collection with an older adult population, and raise additional
considerations for survey administration, sampling, and response in research in this
population apart from the traditional approaches that consider the physiological and
psychological effects of ageing.

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
117
http://www.swin.edu.au/ijets
Figure 1 - Percentage of Americans online by age in 2009
Source: Pew Internet and American Life Project, Generations Online in 2009
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
12-17
18-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-75
76+
Age (years)
Percentage
Survey Design Issues
The ageing process is strongly and positively correlated to certain physiological issues,
including the incidence of chronic illness (Fried & Wallace 1992) and a decline in memory
function (Craik 1999). Moreover, older populations experience greater heterogeneity in
health status than any other (Zimmer et al. 1985), thus presenting the potential for a wider
variety of confounding variables (Carter et al. 1991; Fried & Wallace 1992). Chronic illness
raises hospital admission rates in this age demographic to levels disproportionate to the
population at large (Fox, Sidani & Streiner 2007), and exposes individuals to higher levels of
ethical drug use and the development of disability (Fried & Wallace 1992). Though it is
difficult to accurately assess the impact of chronic illness, the use of medications, and
institutional residency individually, each has the potential to impact data validity and reliability
in ways not typical of the general population (Fried & Wallace 1992). It is therefore critical for
researchers to recognize and consider their presence as possible modifying variables in the
research analysis. Capturing health status at a cursory level (in much the same way as other
demographic information such as age, education and occupational status), regardless of
whether it may be one of the research variables under consideration, allows researchers to
control for the incidence of illness and physiological disability in the data analysis, and
determine whether older survey participants are more or less healthy than older adults in the
general population (Schaie 1984).
A second physiological issue strongly associated with age is cognitive function. Cognitive
function affects not only reflected in the understanding of survey questions and instructions,
but also the respondent’s ability to provide accurate information through memory retrieval
processes (Herzog & Rodgers 1992). Cognitive function can be differentiated between
cognitive pragmatics which is based on cultural influences such as knowledge and
experience and cognitive mechanics or those functions which are based on the
neurophysiological architecture of the mind (Baltes, Lindenberger & Staudinger 1998). While

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
118
research on cognitive pragmatics demonstrates that capacities such as wisdom remain
stable throughout adulthood (Baltes 1993), research on cognitive mechanics reveals age-
related declines in processing and sensory abilities (Park 1999; Salthouse 1996). Racial
disparities in cognitive function have also been detected at older ages, and are not solely
attributable to social factors such as income and educational attainment, biological factors
such as risk of stroke, or functional limitations, but perhaps reflect the effects of accumulated
disadvantage experienced over the life course (Zsembik & Peek 2001).
Declines in certain forms of memory function are associated with increased age, and occur
with both distant and recent events and cued and free recall (Craik 1999). Age-related
declines in the working memory of older adults lead to potential problems in situations where
they must hold, manipulate and integrate moderate amounts of data over short time spans,
such as with complex sentences in telephone surveys or when information needs to be
retained in memory to evaluate later alternative answers (Craik 1999; Salthouse & Babcock
1991). Further, due to age-related declines in episodic memory, older adults encounter
greater difficulty remembering where or when they experienced an event or learned a fact
(Craik 1999). Restating the general context for a set of questions can be particularly helpful
to older adults (for example, asking “Where did you take your vacation last summer?” instead
of “What was the name of the hotel on your last vacation?”). In constructing survey items,
therefore, researchers must consider question complexity and evaluate the significance of
requiring the specifics of time and place as these may be areas for potential sources of error.
In addition, older adults have been found to be more vulnerable than younger persons to the
“truth effect,” or the tendency to find repeated statements as more valid and believable than
a statement presented only once (Law, Hawkins & Craik 1998). Difficulties in the ability to
recollect the source of information can also result in greater vulnerability to confusing
perceived and imagined experiences, or memory distortions, including false recall and
recognition (Schacter, Koutstaal & Norman 1997). And while data collected from older adults
are less likely to incur bias from question order, there is evidence that increases in response
order effects may be related to age (Knäuper 1999). Older adults may be less willing to draw
comparisons between themselves and others because admitting so would be considered
presumptuous (Rodgers & Herzog 1987). These factors should be considered by
researchers when designing individual survey items.
To sum up, researchers might consider capturing cursory health status information as a
survey item to enable its examination as a potential confounding factor. Moreover,
throughout the survey design process researchers should give considered attention to
question complexity and wording, the respondent’s need to recall specific information from
memory, and the potential for priming effects resulting from whether the respondent has
dealt recently with information in question. Care taken in the design phase of the research
process to incorporate the changes that older adults experience in health and cognitive
functioning will enable researchers to collect data of higher quality and reliability.
Sampling and Response Issues
Defining sample frames for older adult populations becomes a complex process due to
mortality and morbidity. After about age 65, mortality rates between men and women differ
such that women tend to outnumber men by 1.7 to 1 (Freeman et al. 1992). This requires
researchers evaluating gender differences at advanced ages to make adjustments in the
analytical stages of their research through weighting. In longitudinal studies, attrition of older
respondents attributable to the death or incapacitation of the respondent may also be an
issue. A variety of strategies have been devised to compensate for the bias that is

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
119
http://www.swin.edu.au/ijets
introduced by this phenomenon, including respondent replacement (Riedel-Heller, Busse &
Angermeyer 2000) and the use of multigenerational data (Feng et al. 2006). The use of
proxy respondents is an additional means to compensate for sample attrition due to
incapacitation; however this method introduces the potential for a systematic under- or over-
reporting of responses based on the question criteria and the nature of the relationship
between the research subject and the proxy if the proxy is not used at from the outset of the
study (Magaziner 1992).
The research question may require stratification strategies to minimize confounding factors of
age and disease. Christensen et al. (1992) found that medical conditions that might impair
cognition affected 35% of the randomly selected subjects for a study, and that these
individuals were older and more frequently male than their counterparts. Their findings
suggest that while random recruitment is feasible for older adult populations, stratification
may be necessary to obtain a representative sample of a target population.
Traditional household sample frames may underrepresent older adults. Herzog and Rodgers
(1992) note that surveys employing a household sampling frame underrepresent older adults
because a significant number of older persons live in institutionalized settings such as
retirement residences or hospitals. A further complication rests with adults living in these
institutionalized settings, as Camp, West and Poon (1989) point out: institutional settings
can affect cognitive performance, mood and morale because of more limited opportunities for
interaction and stimulation, making data obtained from individuals living in these
circumstances even less representative of the general population. But Riedel-Heller et al.
(2000) argue the importance of including institutionalized individuals in the sample frame
when performing community studies, as non-inclusion of these individuals can lead to
sample bias due to underrepresentation of those who are cognitively-impaired; researchers
should consider the importance of inclusion of this subpopulation with respect to the
individual research questions when determining the sample frame.
The choice of survey modality is also a consideration with older adult populations and ties
into rates of technology adoption. Surveys based on landline random digit dialing (RDD)
methods may underrepresent older adult populations due to institutionalized population’s
exclusion from RDD methodologies as non-households (Herzog & Kulka 1989). However,
the more recent trend of using both landline and cellphone sample frames introduces greater
complexity when considering the inclusion of older adult populations. A recent study found
that 54.6% of adults over the age of 55 exclusively use landline phones as compared to only
38.6% the general population, and only 1.8% of adults over the age of 55 use a cellphone
exclusively as compared to a mean of 5.4% (Tucker, Brick & Meekins 2007). These large
deviations from the means of the general population impose a need to evaluate each stage
of the frame to ensure adequate and appropriate representation of older adults in the overall
sample.
Refusal to participate is an area of particular interest to researchers because it is amenable
to change (Jacomb et al. 2002) and because those who do not participate are frequently
different from those who do participate (Herzog & Rodgers, 1988). A meta-analysis of studies
on cognitive ageing research found that non participation rates in random sample studies
were generally high for older adults, moreover those who did agree to participate were
younger, more highly educated, and more likely to be male than the nonparticipants
(Christensen et al. 1992). Grotzinger, Stuart and Ahern (1994) demonstrated that non
response increases with age, but also that blacks are less likely to respond than non-blacks,
and minority nonresponders tend to be in poorer health than responders. Jackson (1989)
points out that some of these non response issues are exacerbated in minority adult

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
120
populations due to the disproportionate socio-economic and educational disadvantages
present in this group.
A decrease in participation rates with increasing age has been documented for telephone
interviews (Massey, Barker & Hsiung 1981) as well as face-to-face interview surveys (Herzog
& Rodgers 1988), and older adults are less likely to participate in a telephone survey than
one conducted in person (Herzog, Rodgers & Kulka 1983). Moreover, those that do
participate in telephone surveys tend to be healthier and more educated than those that
participate in face-to-face surveys (Carter et al. 1991), another potential consideration for
surveys employing multiple modalities. Jacomb et al. (2002) linked refusal to participate in
health surveys to older adults to low verbal IQ scores and cognitive impairment, suggesting
that complexity of the survey process may also be a factor.
Finally, older adults are more resistant to reporting potentially embarrassing information, are
more inclined to overreport desirable information, and score higher on social desirability
scales (Gove & Geerken 1977; Rodgers & Herzog 1987). The incidence of item non
response and “don’t know” (DK) responses increases with increasing age (Colsher &
Wallace 1989) and older adults are more likely to decline to answer a question than younger
adults (Herzog & Rodgers 1988; Mercer & Butler 1967). Older adults may have an apparent
resistance with providing categorical responses to questions (Jobe & Mingay 1990), a
tendency which may arise either from the inability to retrieve information in a categorical
response format or from a lack of familiarity with the standardized testing procedures and
multiple choice test formats now commonly found in schools. Kohout (1992) also alluded to
older adults’ unfamiliarity with standardized testing as a reason for resistance to responding
to closed-ended questions, suggesting that this factor may play into the incidence of DK
responses and non response.
Taken together, this evidence suggests that survey researchers need to give attention to and
control for sources of bias within samples of older populations, paying particular attention to
non response and item non response. The use of sample stratification strategies may be
considered to control the refusal to participate and the underrepresentation inherent in
particular survey modalities with those in later life. Finally, Carter et al. (1991) emphasize the
importance of flexible data collection procedures due to the general decline in health status
for older persons. They suggest flexible follow up schedules, options for in-person interview
sessions for those respondents with hearing impairments or who object to telephone
interviews, and incorporating provisions for proxy respondents.
Data Quality Issues
There appears to be some evidence that validity of survey responses from older persons
appears to decrease with age, raising concerns about comparability of research variables
between birth cohorts. Andrews and Herzog (1986) demonstrate that as age increases
construct validity of measurement decreases, and two kinds of error—random and correlated
error—tend to increase, and they specifically found that these quality issues are not
explained by the lower educational attainment levels of older adults. In addition, the
challenges of institutional settings such as hospitals and retirement homes—a more frequent
occurrence with surveys of older adults—have also been found to affect the internal validity
of standardized tests with older adults, affecting comprehension, memory retrieval, and the
mapping of responses onto numeric response items (Fox et al. 2007). Consistent with this
research, more recently Holbrook, Cho and Johnson (2006) found age, and not gender, race
or education level, to be a significant predictor of mapping difficulties, or the ability of
respondents to map their judgment onto survey response formats.

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
121
http://www.swin.edu.au/ijets
Rodgers and Herzog (1987) suggest that threats to accuracy in survey data may theoretically
arise in older populations due to the increased loss of memory function associated with age,
yet results on studies concerning data reliability and age have been mixed, alternatively
demonstrating no evidence that reliability and age are correlated (Alwin 1999; Rodgers &
Herzog 1987), that item inconsistencies were higher at older ages with men but not women
(Colsher & Wallace 1989), and that reliability depends on the nature of the information being
ascertained (Kuczmarski, Kuczmarski & Najjar 2001) or the age-education correlate of the
respondent (Alwin 1999). These mixed results would suggest that researchers should be
cognizant of the potential for age-related error within their data, but that there is little support
that age alone contributes to greater measurement error in older population.
Cultural Factors of Education, Literacy and Technology Use
While the physiological and psychological factors associated with declines in cognitive
functioning and health have been well documented with respect to survey research with an
older adult population, one area that has received less attention is the cultural dimension of
being an older adult today. The cultural distinctions that differentiate older adults from their
younger counterparts result from a variety of influences, including the prolonged educational
experiences and exposure to standardized testing in younger persons, differing exposures to
mass media between older and younger populations, innovations in health care which has
changed life expectancies and expectations, and markedly differing rates of adoption and
use of ICTs.
A prominent culturally-related characteristic that differentiates younger and older persons in
the US lies in their educational attainment and, correspondingly, resultant literacy levels.
The US Census Bureau (Stoops 2004) reports that in 2003, the middle-aged population—45
to 49 year olds—had the highest levels of educational attainment of the US population as
measured by completion rates of both high school and a bachelor’s degree, but those over
the age of 75 had the lowest educational levels of any other age group (68% high school
completion rate, as compared to 89% of those between the ages of 45 and 49). High school
completion rates of adults over the age of 50 were comparable for both men and women.
The differences in educational experiences between the age cohorts are substantial, and
point toward resultant lower literacy rates for older persons (Brown et al. 1997).
But literacy is not a factor fully attributable to educational attainment levels. The 2003
National Assessment of Adult Literacy (Kutner et al. 2007) found that 23% of those aged 65
years and older performed at the lowest prose reading level, which translates into an inability
to perform basic reading tasks. Prose literacy is defined as those “skills needed to
understand and use information from texts that include editorials, news stories, poems, and
fiction; for example, finding a piece of information in a newspaper article, interpreting
instructions from a warranty, inferring a theme from a poem, or contrasting views expressed
in an editorial” (Brown et al., 1997, p. 3). Further, the authors found that 34% of those aged
65 and older had difficulty with quantitative reading tasks, or the ability to use arithmetic skills
on information that is embedded in a text, and 27% had difficulty with document related tasks
such as filling out forms, reading and following directions, and filling out schedules. They
noted these problems seem to increase in severity with advancing age.
Technology adoption and use by older adults is another culturally-related consideration that
holds implications for survey researchers. A May 2008 study by The Pew Internet &
American Life Project estimated that while 72% of adults in the US reported at least
occasional use of computers and the Internet, only 47% of adults over the age of 55 reported
these activities (Pew Internet 2008). Further, the cultural meaning of technology practices of

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
122
older generations differ markedly from those of youth, emphasizing functionality such as
information search and product purchases over socializing and entertainment (Jones & Fox
2009). These differences in technology practices between older and younger populations
have sometimes been interpreted by researchers to represent differing levels of
technological proficiency, but as Selwyn et al. (2003) point out, this simplifies the issue to a
dichotomy of abilities and ignores important factors such as access to computers and the
Internet and ICT relevance to everyday life. Previous research has demonstrated that
adoption and usage of more advanced ICTs such as the Internet has been linked to cultural
norms and values (Hermeking 2005; LaFerle, Edwards & Mizuno 2002; Rhee & Kim 2004),
suggesting that there is more to slower technology adoption rates than just a lack of
exposure and skills.
In the online environment specifically, concern for privacy is a norm that has shown these
cultural dimensions (DeBoni & Prigmore 2002; 2004; Harris, VanHoye & Lievens 2003), and
has been evidenced through reluctance to use and submit information over the Internet
(Harris, VanHoye & Lievens 2003). The widespread availability of all kinds of information in
the online environment has heightened concerns related to of privacy norms for older
American adults in particular. There is evidence of a growing age-related split in attitudes,
which hold consequent implications for technological practices between younger and older
adults. Such differences may have effects in the collection of survey data, particularly when
more technologically-oriented survey practices are employed.
Due to cost advantages and declining response rates, survey researchers may look to
introduce more sophisticated forms of technology into the research process, particularly in
the area of survey administration. Surveys administered through Internet-based and
electronic formats present new modalities for data capture which increase the efficiency of
data collection and analysis, and offer respondents flexibility in providing information.
However, differing cultural perceptions regarding such a survey modality may present
challenges to survey research through sample bias and non response. In society such as
found in the US, the implication of these newer strategies, for a populace already
complicated by a multi-ethnic and multi-cultural constituency, is that samples become even
more fractured as older and younger persons develop cultural values and practices that are
increasingly distinct.
As previously stated, the view of lower technology adoption rates solely as a function of skills
or ability gives researchers a limited view of how older adults differ from their younger
counterparts, as it dichotomizes the issue to one of proficiency. A more nuanced
understanding incorporates these differences as reflecting cultural practices and values, a
view that considers not only whether a respondent has proficiency in performing certain tasks
but also gives an indication of the comfort in navigating the Internet environment by
considering the range and frequency of a variety of Internet-related activities, such as
reading email, gathering political information and performing basic financial tasks in the
online environment. The technological sophistication of the individual respondent provides
one such means of viewing technology use as a cultural practice. A second means of
viewing the adoption and use of technology as a cultural practice lies in examining related
norms and values associated with such activities, such as the level of concern an individual
may have regarding the use of information stored by online service providers. Examining
technological sophistication and values regarding the online environment can provide
evidence of how different age groups may be more culturally distinct.

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
123
http://www.swin.edu.au/ijets
Method
To more closely examine levels of technology adoption and use as a cultural practice in the
US, data obtained through a May 2008 survey conducted by the Pew Internet & American
Life Project (Pew Internet 2008) on American adults’ use of the Internet was analyzed to
surface patterns of technology use and to measure the value of privacy of information in the
online environment. To emphasize the adoption and use of technology as a cultural practice,
the frequency and use of various Internet activities were examined as a measure of
technological sophistication. A series of questions concerning the treatment of personal
information captured by online service providers were also analyzed as an indication of a
basic cultural norms of privacy.
Sample
The survey was conducted Princeton Survey Research Associates International via
telephone interviews during the period of April 8, 2008 to May 11, 2008 from a sample of
2,251 adults aged 18 and older; of those sampled, 1,553 adults were Internet users (n =
1,553). A random digit sample of telephone numbers was employed. The response rate was
calculated to be a product of the contact rate (85.7%), the cooperation rate (33.3%), and the
completion rate (88.5%), for a total of 25.2%.
Respondents ranged in age from 18 to 94 years (mean = 47.1 years, sd = 18.0 years), with
slightly more than half (51.9%) female. The respondent group was racially mixed, with 77.7%
reporting as White, 12.1% Black or African American, and less than 3% each of Pacific
Islander, Native American and other. Nearly one-third of respondents (31.9%) reported living
with children under the age of 18, and 57.5% reported being married or living with a partner.
Sample demographic weighting, based on the US Census Bureau’s March 2007 Annual
Social and Economic Supplement, was used in the analysis of the data to compensate for
selection and non response bias.
Measures
The primary variables interest for this analysis were respondent demographic and socio-
economic status information; a measure of respondent’s technological sophistication and
Internet use, and a measure of the respondent’s attitudes toward the privacy of personal
information captured and stored online. These latter two variables were measured through
the construction of indices from a battery of questions relating to the frequency of use of
various Internet-related technologies and attitudes toward the use of personal information
supplied to online service providers as detailed below.
Statistical Analysis
The indices representing technological sophistication and privacy concerns were constructed
using the number of affirmative responses to a series of questions on each topic; index
reliability was measured using Cronbach’s alpha. One way analysis of variance (ANOVA)
was performed to determine mean differences of dependent variables in various age
subgroups. Hierarchical ordinary least squares regression (OLS) was employed to examine
whether technological sophistication and concern about privacy exhibited a systematic
relationship with demographic and socio-economic variables and broadband access to the
Internet.

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
124
Technological Sophistication
A series of questions involving the frequency of use of specific online activities were
considered, including obtaining news and political information, getting weather information,
sending instant messages, looking for job opportunities, using a search engine, visiting
governmental websites, making charitable donations online, creating online journals and
blogs, reading online journals or blogs created by others, using social networking sites,
downloading podcasts, watching clips on a video-sharing website and downloading files from
peer-to-peer networks. The use and frequency of use of these various applications can give
an indication of the level of sophistication of Internet use and functionality, i.e., more variety
in application use and more frequent use of the applications indicates a greater degree of
technological sophistication. The responses to this series of questions were used to
construct a Technological Sophistication Index (Cronbach’s  = .82), with resultant values
ranging 0 to 25 for this sample set (mean = 8.80, sd = 5.0).
Privacy Concerns
Respondents were also asked four questions relating to concerns they might have with the
use of personal information captured and stored in the online environment. On a four-point
Likert-type scale, the questions registered the respondent’s degree of concern with
information practices that an online provider might undertake, such as maintaining copies of
information after the respondent tried to delete it, using stored information in marketing
campaigns, analyzing stored information for use in targeted advertising, and selling stored
information to third parties. Responses to these questions were used to construct the
Privacy Concern Index (Cronbach’s = .82), with resultant values ranging from 0 to 12 (mean
= 10.4, sd = 2.59).
Results
Technological Sophistication
First, and perhaps unsurprisingly, there are distinct differences between adults over and
under the age of 55 in levels of technological sophistication (see Table 1).
Table 1. Age Comparison of Technological Sophistication Index
Mean
Item
55 and Over†
(n = 523)
¤
Under 55†
(n = 1828)
¤
t-value**
Technological Sophistication Index (Cronbach’s  = .82) 6.47 (4.29) 9.53 (5.01) 13.8*
Get news online .97 ( .80) 1.18 ( .80) 7.5*
Check weather reports and forecasts online .98 ( .73) 1.14 ( .69) 6.5*
Seek political information online .69 ( .79) .82 ( .80) 4.7*
Use online short/instant messaging .32 ( .57) .61 ( .74) 13.5*
Seek information about a job .28 ( .52) .61 ( .61) 17.5*
Visit government websites .70 ( .63) .83 ( .65) 5.9*
Donate to charities online .19 ( .42) .22 ( .44) 1.5
Use search engines 1.19 ( .69) 1.45 ( .65) 11.1*
Create an online journal or blog .09 ( .34) .20 ( .52) 7.9*
Read online journals or blogs .27 ( .57) .51 ( .72)
7.9*
Use social networking sites .09 ( .33) .53 ( .77) 27.3*
Download podcasts .12 ( .36) .26 ( .52) 10.4*
Watch online videos .32 ( .59) .79 ( .74) 21.7*
Download or share files using peer-to-peer networks .05 ( .25) .22 ( .50) 15.5*
†Standard deviations are in parentheses
¤
Sample sizes reflect weighted values *p < .01**Confidence interval of 95% (1 – )

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
125
http://www.swin.edu.au/ijets
A t-test examination of the Technological Sophistication Index reveals a significant mean
difference between the Technological Sophistication Index for those aged 55 years and over
(mean = 6.47, s = 4.29) and those under 55 years (mean = 9.53, s = 5.01), t(967) = 13.8, p <
.001,  = .05. This means that the older adult group has significantly less experience in the
variety and use of a wide range of Internet activities than younger adults. As shown in Table
1, this distinction in experience is not only generalized in nature, but appears across each of
the individual online activities surveyed, with the exception of online charitable giving.
These distinctions are reinforced in a more granular breakout of the Technological
Sophistication Index means by age groups. Breaking down age groups to reflect younger
adults (aged 18 to 34 years), mature adults (aged 35 to 54 years), younger older adults
(aged 55 to 64 years), post retirement adults (aged 65 to 74 years), and those in old age
(aged 75+ years), a one way ANOVA indicates statistically significant differences between
the groups for Technological Sophistication (F(4, 2346) = 7.39, p < .001).
As shown in Table 2, post hoc Games-Howell tests demonstrated significant differences
between adults of all age groups except those between age 55 and 64 years (mean = 7.15,
sd = 4.22) and age 65 and 74 years (mean = 5.97, sd = 3.95). Significant differences were
found between these age groups and younger adults aged 18 to 34 years (mean = 10.84, sd
= 5.13), mature adults aged 35 to 54 years (mean = 8.48, sd = 4.66), and older in age adults
aged 75+ years (mean = 3.7, sd = 4.07).
Table 2. Technological Sophistication Index Mean Difference Comparison by Age Group**
Age Group (in years)
Age Group
18 – 34
(n = 816)
35 – 54
(n = 1013)
55 – 64
(n = 348)
65 – 74
(n = 111)
75 +
(n = 64)
18 – 34 years - -2.36* -3.69* -4.88* -7.14*
35 – 54 years 2.36* - -1.33* -2.51* -4.78*
55 – 64 years 3.69* 1.33* - -1.18 -3.44*
65 – 74 years 4.88* 2.51* 1.18 - -2.26*
75 + years 7.14* 4.78* 3.44* 2.26* -
* Mean difference is significant at the .05 level **Confidence interval of 95% (1 – )
These data indicate that Technological Sophistication does vary by age and the gradations
between age groups are significant. Predictably, the youngest adults have the greatest
experience in a wide variety and use of Internet activities, significantly more than adults aged
35 to 54 years and than adults over age 55 years. Importantly, statistically significant
differences appear between younger adults age 18 to 34 years and those in the mature adult
group aged 35 to 54 year, indicating that the disparities in technology use may not just “age
out” of the population in a few years. The lack of a statistically meaningful difference in
Technological Sophistication between adults aged 55 to 64 years and adults aged 65 to 74
years is somewhat surprising, as the younger of the two groups is conceivably still in the
workforce and would have alternative points of access and opportunities for Internet use.
This may point to the significance of other relevant factors for Internet use at these age
levels.
A closer examination of the relationship between the Technological Sophistication Index and
demographic factors, socio-economic indicators and broadband Internet access was
performed using ordinary least squares (OLS) regression in a hierarchical manner to
highlight the cross-level interactions between the variables. Specifically, three models were
developed: the first looked only at the demographic background information of age, gender,
whether the respondent was married or living with a partner, and whether children under the

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
126
age of 18 lived in the home (Model 1); the second model added socio-economic indicators of
income and education (Model 2); and a third incorporated a measure of the respondent’s
access to the Internet via a broadband connection (Model 3). Age was treated as an interval
level variable.
Table 3. OLS Regression Analysis of Technological Sophistication
Technological Sophistication
Factor Model 1 Model 2 Model 3
B† 
§
B† 
§
B† 
§
Age -.13 (.01)*** -.37*** -.14 (.01)*** -.42*** -.14 (.01)*** -.40***
Gender .78 (.22)*** .08*** .57 (.21)** .06** .55 (.20)** .06**
Married, living with partner -.25 (.25) -.02 -1.02 (.24)*** -.10*** -1.02 (.24)*** -.10***
Children under 18 living in home -1.34 (.24)*** -.13*** -1.59 (.23)*** -.16*** -1.54 (.22)*** -.15***
Income - - .49 (.06)*** .20*** .41 (.06)*** .17***
Education - - .59 (.07)*** .19*** .50 (.07)*** .16***
Quality of internet access - - - - 2.39 (.27)*** .19***
Intercept 15.10 (.39) 10.91 (.46) 9.46 (.48)
Adj. R
2
.14 .23 .27
N 1801 1801 1801
† Unstandardized coefficients, standard errors are in parentheses
§
Standardized coefficients
* < .05, ** < .01, *** < .001
As summarized in Table 3, hierarchical regression modelling demonstrates the systematic
relationship between technological sophistication and certain respondent characteristics. In
Model 1, the baseline model which includes age, gender, whether respondent is
married/lives with a partner and whether children under the age of 18 live in the home, the
respondent’s age has a significant and strongly negative relationship ( = -.37,  < .001) with
Technological Sophistication; it is the most significant predictor among the demographic
variables. Model 2 adds in components related to socio-economic status, and reveals that
income ( =.20,  < .001) and educational attainment ( =.19,  < .001) also feature into
technological sophistication, yet neither of these is as significant a factor as age ( = -.42,  <
.001). Model 3 adds in whether the quality of Internet access available to the respondent
makes a difference to the model, and exposes that while broadband access is an important
factor ( =.19,  < .001), age continues to be the most significant predictor of technological
sophistication ( = -.40,  < .001). These models indicate that age is a strong predictor of
Technological Sophistication, followed by characteristics such as income, educational
attainment and broadband access to the Internet.
Privacy Concerns
With respect to concern for the privacy of information stored online, there are again distinct
differences between adults over and under the age of 55 years. Older adults display a higher
level of concern about the privacy of personal information stored online. Distinctions are
again found between the 55 and older age group and their younger counterparts in the
Privacy Concerns Index: a t-test reveals a mean difference between those aged 55 and over
(mean = 10.81, sd = 2.53) and those under 55 years (mean = 10.31, sd = 2.60), t(779) = 4.2,
p < .001,  = .05. This means that older adults are more concerned than their younger
counterparts about how their personal information may be retained, stored and used by
online services. As evidenced in Table 4, this heightened concern is reflected across several
areas: the retention of information by the provider after user deletion; the use of information
by the provider in their marketing activities; and the analysis of information by the provider for

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
127
http://www.swin.edu.au/ijets
the display of advertising. Predictably however, all adults surveyed share a concern about
the sale of information to third parties.
Table 4. Age Comparison of Privacy Concerns Index
Mean
Item
55 and Over†
(n = 544)
¤
Under 55†
(n = 2821)
¤
t-value**
Privacy Concerns Index (Cronbach’s  = .82) 10.8 (2.5) 10.3 (2.6) 4.2*
Concerned or very concerned that online service would:
• Keep copies of files even after deletion 2.5 (.92) 2.4 (.94) 3.3*
• Use photos or other information for marketing 2.8 (.73) 2.6 (.82) 3.2*
• Analyze information held for display of ads 2.7 (.77) 2.5 (.90 5.4*
• Sell files to others 2.8 (.65) 2.8 (.60) .4
†Standard deviations are in parentheses
¤
Sample sizes reflect weighted values *p < .01**Confidence interval of 95% (1 – )
A more detailed breakout of the Privacy Concern Index means by age groups is somewhat
more revealing. Again breaking down age groups to reflect younger adults (aged 18 to 34
years), mature adults (aged 35 to 54 years), younger older adults (aged 55 to 64 years), post
retirement adults (aged 65 to 74 years), and those in old age (aged 75+ years), a one way
ANOVA showed significant differences between the various age groups (F(4, 3359) = 11.42,
p < .001). As detailed in Table 5, post hoc Games-Howell tests demonstrated significant
differences between young adults aged 18 to 34 years (mean = 10.07, sd = 2.84), mature
adults aged 35 to 54 years (mean = 10.57, sd = 2.28), and younger older adults aged 55 to
64 years (mean = 10.88, sd = 2.45). Somewhat surprisingly, differences in the older two age
groups did not reflect significant differences between any of the age groups. These results
may be interpreted to indicate that concerns about the retention, storage and use of personal
information by online service providers increase with age, but perhaps only to a point where
the lack of technological sophistication makes awareness of these concerns improbable.
Table 5. Privacy Concerns Index Mean Difference Comparison by Age Group**
Age Group (in years)
Age Group 18 – 34
(n = 816)
35 – 54
(n = 1013)
55 – 64
(n = 348)
65 – 74
(n = 111)
75 +
(n = 64)
18 – 34 years - .51* .82* .64 .55
35 – 54 years -.51* - .31 .13 .05
55 – 64 years -.82* -.31 - -.18 -.26
65 – 74 years -.64 -.13 .18 - -.08
75 + years -.55 -.05 .26 .08 -
* Mean difference is significant at the .05 level **Confidence interval of 95% (1 – )
Hierarchical OLS modeling was also employed to demonstrate the systematic relationship
between concerns about the privacy of online information and certain respondent
characteristics, and the results are summarized in Table 6. Again, three models were
developed: the first examined the effects of demographic information including age, gender,
whether the respondent was married or living with a partner, and whether children under the
age of 18 lived in the home (Model 1); the second model added socio-economic indicators of
income and education (Model 2); and the third incorporated a measure of the respondent’s
access to the Internet via a broadband connection (Model 3).
These results demonstrate that while age is an important and positive characteristic when
considering privacy concerns (Model 1:  = .09,  < .001; Model 2:  =.09,  < .001; Model 3:
 =.09,  < .001), demographic factors do not adequately explain the variance in concerns
about the privacy of personal information in the online environment (Adj R² = .02).

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
128
Table 6. OLS Regression Analysis of Privacy Concerns
Privacy Concerns
Factor Model 1 Model 2 Model 3
B† 
§
B† 
§
B† 
§
Age .02 (.00)*** .09*** .02 (.00)*** .09*** .02 (.00)*** .10***
Gender -.33 (.10)** -.07** -.34 (.10)
***
-.0
7
***
-.37 (.10)
***
-.07
***
Married, living with partner .08 (.11) .02 .06 (.11) .01 .07 (.11) .01
Children under 18 living in home -.17 (.11) -.03 -.17 (.11) -.03 -.17 (.11) -.03
Income - - .03 (.03) .02 .03 (.04) .01
Education - - -.05 (.03) -.03 -.07 (.03) -.04
Quality of internet access - - - - .64 (.14)*** .09***
Intercept 10.01 (.16) 10.08 (.21) 9.66 (.23)
Adj. R
2
.01 .01 .02
N 1283 1283 1283
† Unstandardized coefficients, standard errors are in parentheses
§
Standardized coefficients
* < .05, ** < .01, *** < .001
Technological Sophistication and Privacy Concerns
Interestingly, the correlation between Technological Sophistication and Privacy Concerns is
weakly negative, as indicated by a Spearman’s rho test (r = -.08, p < .001, n = 1623). This
indicates that as Technological Sophistication increases, concerns about the privacy of
personal information stored online decrease only slightly. A Fisher r to z transformation
indicates that there is no significant differences between the age groups of under 55 years
and over 55 years in this correlation (z = 1.31, p = .10).
Discussion
These results raise two important points, especially as research moves into a more
technologically-oriented environment to optimize research costs and time horizons. First,
because technological sophistication is so closely correlated with age and the mean
differences in the Technological Sophistication Index are sufficiently distinct in the under 55
years and 55 and over age groups, survey researchers who use technologically
sophisticated data collection measures must consider whether their respondent base of older
adults adequately represents the characteristics of older adults in the broader population:
older adults sufficiently comfortable with survey administration technology may not be
representative of older adults generally. On a related note, the adoption of new survey
technologies should be considered carefully in the research planning process, particularly if
the sample is to include older persons. Survey researchers should be sensitive to this
potential for sample bias, and make appropriate adjustment to the sample and/or analysis.
Second, for survey researchers, the age-related distinction in concern for privacy in the
online environment also becomes an important consideration. As dependence on web-
based data collection procedures become more prevalent, researchers should carefully
consider the importance of communicating privacy protection measures to potential
respondents, and particularly to older adults. The increased level of concern regarding the
privacy of information in the online environment may lead older adults to regard
technologically-oriented data collection practices differently than younger adults, potentially
leading to differing levels of participation and non response between age groups. Results
should be evaluated for the presence of these forms of bias. Further research on survey
modalities, non response, participation rates and age is warranted to understand this

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
129
http://www.swin.edu.au/ijets
dynamic; however researchers should be aware of the potential for differences when
analyzing survey data.
Several shortcomings of this analysis should be mentioned. The data analyzed herein were
collected through telephone interviews of adults selected through a random digit sample of
telephone numbers of exchanges in the continental US. Consequently, limitations to this
study include the liabilities inherent in random digit dialling sampling methods and also in
self-reported behaviour. Further, to compensate for potential non response bias, the sample
data were weighted in the analysis to reflect population parameters of the Census Bureau’s
March 2007 Annual Social and Economic Supplement; the resultant parameters reflect the
demographics of adults living in households with telephones, and may not fully reflect the US
adult population.
Conclusions
With the ageing of the Baby Boomer cohort, older adults are an increasing proportion of the
general population and the research focus on this age segment has been mounting across
all disciplines. The use of survey research explicitly faces special methodological
considerations when dealing with the older adult population that may require compensation
for the physiological and psychological factors associated with ageing, particularly in the
areas of sample selection, examination for response bias and evaluation of data reliability
and validity. And yet, often overlooked in this process are the distinctive cultural
considerations that distinguish this population from younger adults, such as literacy rates,
technology adoption and use, and values regarding the privacy of personal information
stored in the online environment.
Viewing technology adoption and use as a cultural practice provides strong impetus for
researchers to evaluate the use of technology in the research process, as it may provide
additional sources for potential bias in survey data collected from older adults. Researchers
should consider whether older adult respondents sufficiently sophisticated to participate in
technologically enabled data collection mechanisms are representative of their peers.
Communication of privacy protection measures may be an important component to
increasing participation rates and reducing non response rates for older adults. As survey
researchers introduce more technologically sophisticated tools into their research habits,
understanding of cultural distinctions becomes increasingly important because of their
potential significance to the quality of survey results.
Acknowledgements
An earlier version of this paper was accepted for presentation at the Midwest Association of
Public Opinion Researchers conference, Chicago, Illinois, Nov 20-21, 2009. Thanks go to the
anonymous reviewers for their insightful and helpful comments.

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
130
References
Alwin DF 1999, ‘Aging and errors of measurement’. In: N Schwarz, D Park, B Knäuper & S
Sudman (eds), Cognition, Aging, and Self-Reports, pp. 365-385, Taylor & Francis,
Philadelphia, PA: .
Andrews FM & Herzog AR 1986, ‘The quality of survey data as related to age of respondent’,
Journal of the American Statistical Association, vol. 81, no. 394, pp. 403-410.
Baltes PB 1993, ‘The aging mind: potential and limits,’ Gerontologist, vol. 33, no. 5, pp. 580-
594.
Baltes PB, Lindenberger U & Staudinger UM 1998, ‘Life-span theory in developmental
psychology’. In: W Damon (Ed-in-Chief) & RM Lerner (Vol Ed), Handbook of Child
Psychology: Vol. 1. Theoretical Models of Human Development, pp. 1029–1041, Wiley,
New York.
Brown H, Prisuta R, Jacobs B & Campbell A 1997, Literacy of Older Adults in America, US
Department of Education, National Center for Education Statistics, Washington, DC.
Camp CJ, West RL & Poon LW 1989, ‘Recruitment practices for psychological research in
gerontology’. In: MP Lawton & AR Herzog (eds), Special Research Methods for
Gerontology, pp. 163-189, Baywood Publishing, Amityville, NY.
Carter WB, Elward K, Malmgren J, Martin ML & Larson E 1991, ‘Participation of older adults
in health programs and research: A critical review of the literature’, The Gerontologist,
vol. 31, no. 5, pp. 584-592.
Christensen KJ, Moye J, Armson RR & Kern TM 1992, ‘Health screening and random
recruitment for cognitive aging research’, Psychology and Aging, vol. 7, no. 2, pp. 204-
208.
Colsher PL & Wallace RB 1989, ‘Data quality and age: Health and psychobehavioral
correlates of item non response and inconsistent responses’, Journal of Gerontology:
Psychological Sciences, vol. 44, no. 2, pp. P45-P52.
Craik FIM 1999, ‘Memory, aging and survey measurement’. In: N Schwarz, D Park, B
Knäuper & S Sudman (eds), Cognition, Aging, and Self-Reports, pp. 95-115, Taylor &
Francis, Philadelphia, PA.
De Boni M & Prigmore M 2002, ‘Cultural aspects of Internet privacy’. In: B Howell & G
Orarge (eds), Information Systems Research, Teaching and Practice: Proceedings of
the seventh UK Academy of Information Systems Conference, Leeds Metropolitan
University, Leeds, UK.
De Boni M & Prigmore M 2004, ‘A Hegelian basis for privacy as an economic right’,
Contemporary Political Theory, vol. 3, no. 2, pp. 168-187.
Feng D, Silverstein M, Giarrusso R, McArdle JJ & Bengtson VL 2006, ‘Attrition of older adults
in longitudinal surveys: Detection and correction of sample selection bias using
multigenerational data’, Journal of Gerontology, vol. 61B, pp. S323-S328.
Fox MT, Sidani S & Streiner D 2007, ‘Using standardized survey items with older adults
hospitalized for chronic illness’, Research in Nursing & Health, vol. 30, pp. 468-481.
Freeman DH, Bruce ML, Leaf P & Berkman L 1992, ‘Sampling strategies for studying older
populations’. In: RB Wallace & RF Woolson (eds), The Epidemiologic Study of the
Elderly, pp. 46-59, Oxford University Press, Oxford, UK.

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
131
http://www.swin.edu.au/ijets
Fried LP & Wallace RB 1992, ‘The complexity of chronic illness in the elderly: From clinic to
community’. In: RB Wallace & RF Woolson (eds), The Epidemiologic Study of the
Elderly, pp. 10-19, Oxford University Press, Oxford, UK.
Gove WR & Geerken MR 1977, ‘Response bias in surveys of mental health’, American
Journal of Sociology, vol. 82, no. 6, pp. 1289-1317.
Grotzinger KM, Stuart BC & Ahern F 1994, ‘Assessment and control of non response bias in
a survey of medicine use by the elderly’. Medical Care, vol. 22, no. 10, pp. 989-1003.
Harris MM, VanHoye G & Lievens F 2003, Privacy and attitudes towards Internet-based
selection systems: A cross–cultural comparison’ International Journal of Selection and
Assessment, vol. 11, no. 2/3, pp. 230-236.
Hermeking M 2005, ‘Culture and Internet consumption: Contributions from cross-cultural
marketing and advertising research’, Journal of Computer-Mediated Communication,
vol. 11, no. 1, article 10. Retrieved 14 September 2009, from http://jcmc.indiana.edu/
vol11/issue1/hermeking.html.
Herzog AR & Kulka RA 1989, ‘Telephone and mail surveys with older populations: A
methodological overview’. In: MP Lawton & AR Herzog (eds), Special Research
Methods for Gerontology, pp. 63-90, Baywood Publishing, Amityville, NY.
Herzog AR & Rodgers WL 1988, ‘Age and response rates to interview survey samples’,
Journal of Gerontology, vol. 43, no. 6, pp. S200-S205.
Herzog AR & Rodgers WL 1992, ‘The use of survey methods in research on older
Americans’. In: RB Wallace & RF Woolson (eds), The Epidemiologic Study of the
Elderly, pp. 60-90, Oxford University Press, New York.
Herzog AR, Rodgers WL & Kulka RL 1983, ‘Interviewing older adults: A comparison of
telephone and face-to-face modalities’, Public Opinion Quarterly, vol. 47, no. 3, pp.
405-418.
Holbrook A, Cho YI & Johnson T 2006, ‘The impact of question and respondent
characteristics on comprehension and mapping difficulties’, Public Opinion Quarterly,
vol. 70, no. 4, pp. 565-595.
Jackson JS 1989, ‘Methodological issues in survey research on older minority adults’. In: MP
Lawton & AR Herzog (Eds), Special Research Methods for Gerontology, pp. 137-161,
Baywood Publishing, Amityville, NY.
Jacomb PA, Jorm AF, Korten AE, Christensen H & Henderson AS 2002, ‘Predictors of
refusal to participate: A longitudinal health survey of the elderly in Australia’ BMC
Public Health, vol. 2, pp. 4-9.
Jobe JB & Mingay DJ 1990, ‘Cognitive laboratory approach to designing questionnaires for
surveys of the elderly’ Public Health Reports, vol. 105, no. 5, pp. 518-524.
Jones S & Fox S 2009, Generations Online in 2009, Pew Internet & American Life Project,
Washington, DC. Retrieved 17 June 2009, from http://pewinternet.org/~/
media//Files/Reports/2009/PIP_Generations_2009.pdf
Knäuper B 1999, ‘Age differences in question and response order effects’. In: N Schwarz, D
Park, B Knäuper & S Sudman (eds), Cognition, Aging, and Self-Reports, pp. 341-363,
Taylor & Francis, Philadelphia, PA.

iJETS
Vol. 8, No. 2, 2010, pp: 114 – 133
132
Kohout FJ 1992, ‘The pragmatics of survey field work among the elderly’. In: RB Wallace &
RF Woolson (eds), The Epidemiologic Study of the Elderly, pp. 91-119, Oxford
University Press, New York.
Kuczmarski MF, Kuczmarski RJ & Najjar M 2001, ‘Effects of age on validity of self-reported
height, weight and body mass index: Findings from the third National Health and
Nutrition Examination Survey, 1988-1994’, Journal of the American Dietetic
Association, vol. 101, no. 1, pp. 28-34.
Kutner M, Greenberg E, Jin Y, Boyle B, Hsu Y & Dunleavy E 2007, Literacy in Everyday Life:
Results from the 2003 National Assessment of Adult Literacy (NCES 2007–480), US
Department of Education, National Center for Education Statistics, Washington, DC.
LaFerle C, Edwards S & Mizuno Y 2002, ‘Internet diffusion in Japan: Cultural considerations’,
Journal of Advertising Research, vol. 42, no. 2, pp. 65-79.
Law S, Hawkins SA & Craik FIM 1998, ‘Repetition-induced belief in the elderly:
Rehabilitating age-related memory deficits’, Journal of Consumer Research, vol. 25, no.
2, pp. 91-107.
Magaziner J 1992, ‘The use of proxy respondents in health studies of the aged’. In: RB
Wallace & RF Woolson (Eds), The epidemiologic study of the elderly, pp. 120-129,
Oxford University Press, New York.
Martin LG, Freedman VA, Schoeni RF & Andreski PM 2009, ‘Health and functioning among
baby boomers approaching 60’, Journal of Gerontology: Social Sciences, vol. 64B, no.
3, pp. 369-377.
Massey JT, Barker PR & Hsiung S 1981, ‘An investigation of response in a telephone
survey’, In: Proceedings of the Survey Research Methods Section, American Statistical
Association, pp. 426-431. Retrieved 21 March 2009, from http://www.amstat.org/
sections/SRMS/Proceedings/
Mercer JR & Butler EW 1967, ‘Disengagement of the aged population and response
differentials in survey research’, Social Forces, vol. 46, no. 1, pp. 89-96.
Park DC 1999, ‘Cognitive aging, processing resources, and self report’. In: N Schwarz, D
Park, B Knäuper & S Sudman (eds), Cognition, Aging, and Self-Reports, pp. 45-69,
Taylor & Francis, Philadelphia, PA.
Pew Internet & American Life Project 2008, Cloud Computing, Politics and Adult Social
Networking [Data file], Washington, DC. Retrieved 13 March 2009, from
http://www.pewinternet.org/Shared-Content/Data-Sets/2008/May-2008--Cloud-
computing-politics-and-adult-social-networking.aspx
Pew Internet & American Life Project 2009, Usage Over Time [Data file], Washington, DC.
Retrieved 17 June 2009, from http://pewinternet.org/Trend-Data/Usage-Over-
Time.aspx
Rhee KY & Kim WB 2004, ‘The adoption and use of the Internet in South Korea’, Journal of
Computer Mediated Communication, vol. 9, no. 4. Retrieved 14 September 2009, from
http://jcmc.indiana.edu/vol9/issue4/rhee.html
Riedel-Heller SG, Busse A & Angermeyer MC 2000, ‘Are cognitively impaired individuals
adequately represented in community surveys? Recruitment challenges and strategies
to facilitate participation in community surveys of older adults. A review’, European
Journal of Epidemiology, vol. 16, no. 9, pp. 827-835.

Quinn: Surveys of Older Adults: Technology Matters
International Journal of Emerging Technologies and Society 2010
133
http://www.swin.edu.au/ijets
Rodgers WL & Herzog AR 1987, ‘Interviewing older adults: The accuracy of factual
information’, Journal of Gerontology, vol. 42, no. 4, pp. 387-394.
Salthouse TA 1996, ‘The processing-speed theory of adult age differences in cognition,’
Psychological Review, vol. 103, no. 3, pp. 403-428.
Salthouse TA & Babcock RL 1991, ‘Decomposing adult age differences in working memory,’
Developmental Psychology, vol. 27, no. 5, pp. 763-776.
Schacter DL, Koutstaal W & Norman KA 1997, ‘False memories and aging’, Trends in
Cognitive Sciences, vol. 1, no. 6, pp. 2292-36.
Schaie KW 1984, ‘Methodological issues in aging research: An introduction’. In: KW Schaie,
RT Campbell, W Meredith & SC Rawlings (eds), Methodological Issues in Aging
Research, pp. 1-11, Springer Publishing, New York.
Selwyn N, Gorard S, Furlong J & Madden L 2003, ‘Older adults’ use of information and
communication technology in everyday life’, Ageing & Society, vol. 23, pp. 561-582.
Settersten RA 2003 ‘Propositions and controversies in life-course scholarship’. In: RA
Settersten (ed), Invitation to the Life Course: Toward New Understandings of Later Life,
pp. 15-45, Baywood, Amityville, NY.
Stoops N 2004, ‘Educational attainment in the United States: 2003’, Current Population
Survey, (P20-550), US Census Bureau, Washington, DC.
Tucker C, Brick JM & Meekins B 2007, ‘Household telephone service and usage patterns in
the United States in 2004: Implications for telephone samples’, Public Opinion
Quarterly, vol. 71, no. 1, pp. 3-22.
US Census Bureau 1979, Population Estimates, National Estimates by Age, Sex, Race:
1900-1979. [Data file], Washington, DC. Retrieved 11 September 2009, from
http://www.census.gov/popest/archives/pre-1980/PE-11.html
US Census Bureau 2007a, 2005-2007 American Community Survey 3-Year Estimates –
Table S0101. Age and Sex [Data file], Washington, DC. Retrieved 24 March 2009,
from http://factfinder.census.gov/servlet/STTable?_bm=y&-geo_id=01000US&-gr-name
=ACS_2007_3YR_G00_S0101&-ds_name=ACS_2007_3YR_G00_
US Census Bureau 2007b, Population Profile of the United States: Dynamic Version, Part VI:
Special Populations Older Adults in 2005. US Government Printing Office, Washington,
DC, Retrieved 24 March 2009, from http://www.census.gov/population/www/pop-
profile/files/dynamic/OLDER.pdf.
US Census Bureau 2008, Voting and Registration in the Election of November 2006
(Current Population Report No. P20-557) US Government Printing Office, Washington,
DC. Retrieved 24 March 2009, from http://www.census.gov/prod/2008pubs/p20-
557.pdf.
Zimmer AW, Calkins E, Hadley E, Ostfeld AM, Kaye JM & Kaye D 1985, ‘Conducting clinical
research in geriatric populations’, Annals of Internal Medicine, vol. 103, pp. 276-283.
Zsembik BA & Peek MK, 2001, ‘Race differences in cognitive functioning among older
adults’, Journal of Gerontology, vol. 56B, no. 5, pp. S266-S274.