Testing a moderator-type research model on the use of mobile phone.
Fillion, Gerard ; Ekionea, Jean-Pierre Booto
INTRODUCTION
Since numerous years, mobile phone is used for different
professional purposes, particularly by senior managers in the workplace.
And this technology is more and more used in the workplace since mobile
applications have been integrated to actual enterprise business
strategies. Individual adoption of technology has been studied
extensively in the workplace (Brown & Venkatesh, 2005). Far less
attention has been paid to adoption of technology in the household
(Brown & Venkatesh, 2005). Obviously, mobile phone is now integrated
into our daily life. According to the more recent forecast of Gartner
Research, 1.22 billion of mobile phones have been sold throughout the
world in 2008, a 6 percent increase over 2007 sales (Gartner Newsroom,
2008). And, as the tendency is showing up, mobile phone use will be
still increasing in the future. The purpose of this study is then to
investigate who really are the users of a mobile phone and what are the
determining factors who make such that they are using a mobile phone?
Few studies have been conducted until now which investigate the
intention to adopt a mobile phone by people in household (in the case of
those who do not yet own a mobile phone) or the use of mobile phone in
the everyday life of people in household (in the case of those who own a
mobile phone). Yet we can easily see that mobile phone is actually
completely transforming the ways of communication of people around the
world. It is therefore crucial to more deeply examine the determining
factors in the use of mobile phone by people in household. This is the
aim of the present study. The related literature on the actual research
area of mobile phone is summarized in Table 1.
In addition to the summary of literature on the actual research
area of mobile phone presented in Table 1, other researchers have
identified some factors which may increase the use of mobile phone by
people in household. For example, in a large study conducted in 43
countries of the world, Kauffman and Techatassanasoontorn (2005) noted a
faster increase in the use of mobile phone in countries having a more
developed telecommunications infrastructure, being more competitive on
the wireless market, and having lower wireless network access costs and
less standards regarding the wireless technology. And a study involving
208 users by Wei (2008) showed that different motivations predict
diverse uses of mobile phone. According to the Wei's findings,
mobile phone establishes a bridge between interpersonal communication
and mass communication.
As we can see in the summary of literature related to mobile phone
presented above, few studies until now examined the determining factors
in the use of mobile phone by people in household. Thus, the present
study brings an important contribution to fill this gap as it allows a
better understanding of the impacts of mobile phone usage into
people's daily life. It focuses on the following two research
questions: (1) Who are the buyers of mobile phone for household use? (2)
What are the determining factors in the use of mobile phone by people in
household?
The paper builds on a framework suggested by Fillion (2004) in the
conduct of hypothetico-deductive scientific research in organizational
sciences, and it is structured as follows: first, the theoretical
approach which guides the study is developed; second, the methodology
followed to conduct the study is described; finally, the results of the
study are reported and discussed.
THEORETICAL APPROACH
This study is based on the theoretical foundations developed by
Venkatesh and Brown (2001) to investigate the factors driving personal
computer adoption in American homes as well as those developed by Brown
and Venkatesh (2005) to verify the determining factors in intention to
adopt a personal computer in household by American people. In fact,
Brown and Venkatesh (2005) performed the first quantitative test of the
recently developed model of adoption of technology in households (MATH)
and they proposed and tested a theoretical extension of MATH integrating
some demographic characteristics varying across different life cycle
stages as moderating variables. With the exception of behavioural
intention (we included user satisfaction instead given people
investigated in this study own a mobile phone), all the variables
proposed and tested by Brown and Venkatesh (2005) are used in this
study. And we added two new variables in order to verify whether people
are using mobile phone for security and mobility. The resulting
theoretical research model is depicted in Figure 1.
[FIGURE 1 OMITTED]
Figure 1 shows that Brown and Venkatesh (2005) integrated MATH and
Household Life Cycle in the following way. MATH presents five
attitudinal beliefs grouped into three sets of outcomes: utilitarian,
hedonic, and social. Utilitarian beliefs are most consistent with those
found in the workplace and can be divided into beliefs related to
personal use, children, and work (we added beliefs related to security
and mobility). The extension of MATH suggested and tested by Brown and
Venkatesh (2005) presents three normative beliefs: influence of friends
and family, secondary sources, and workplace referents. As for control
beliefs, they are represented in MATH by five factors: fear of
technological advances, declining cost, cost, perceived ease of use, and
self-efficacy. And, according to Brown and Venkatesh (2005), integrating
MATH with a life cycle view, including income, age, child's age,
and marital status, allows to provide a richer explanation of household
personal computer adoption (household mobile phone usage in this study)
than those provided by MATH alone. Finally, as shown in Figure 1, the
dependant variable of the theoretical research model developed is
related to user satisfaction (satisfaction in the use of mobile phone by
people in household). All of the variables integrated in the theoretical
research model depicted in Figure 1 are defined in Table 2.
We can see in Table 2 that the definitions of MATH variables
integrated in the theoretical research model proposed in Figure 1 are,
in the whole, adapted from the theoretical foundations developed by
Venkatesh and Brown (2001) to investigate the factors driving personal
computer adoption in American homes. As for the definitions of the
variables related to the household life cycle, they were taken from
Danko and Schaninger (1990) as well as Wagner and Hanna (1983),
respectively. And the definitions of the two new independent variables
that we added to the model are from our own. In fact, we defined these
variables in accordance with which we wanted to measure regarding
security and mobility before to develop and validate items measuring
them on the basis of the definitions formulated.
In the reminder of the section, we develop eight research
hypotheses (H1-H8) related to the model suggested in Figure 1.
H1: Marital status and age will moderate the relationship between
applications for personal use and satisfaction of using a mobile phone
at home.
H2: Child's age will moderate the relationship between utility
for children and satisfaction of using a mobile phone at home.
H3: Age will moderate the relationship between utility for
work-related use and satisfaction of using a mobile phone at home.
H4: Age will moderate the relationship between applications for fun
and satisfaction of using a mobile phone at home.
H5: Age will moderate the relationship between status gains and
satisfaction of using a mobile phone at home.
H6: Age, marital status, and income will moderate the relationship
between the normative beliefs ((a) friends and family influences; (b)
secondary sources' influences; and (c) workplace referents'
influences) and satisfaction of using a mobile phone at home.
H7: Age and income will moderate the relationship between the
external control beliefs ((a) fear of technological advances; (b)
declining cost; and (c) cost) and satisfaction of using a mobile phone
at home.
H8: Age will moderate the relationship between the internal control
beliefs ((a) perceived ease of use; and (b) self-efficacy) and
satisfaction of using a mobile phone at home.
In the next section of the paper, we describe the methodology
followed to conduct the study.
METHODOLOGY
The study was designed to gather information concerning mobile
phone adoption decisions in Atlantic Canadian households. Indeed, the
focus of the study is on individuals who own a mobile phone. We
conducted a telephone survey research among individuals of a large area
in Atlantic Canada. In this section, we describe the instrument
development and validation, the sample and data collection, as well as
the data analysis process.
INSTRUMENT DEVELOPMENT AND VALIDATION
To conduct the study, we used the survey instrument developed and
validated by Brown and Venkatesh (2005) to which we added three new
scales, the first two measuring other dimensions in satisfaction in the
use of mobile phone by people in household, that is, utility for
security and mobility, and the last one measuring user satisfaction as
such. The survey instrument was then translated in French (a large part
of the population in Atlantic Canada is speaking French) and both the
French and English versions were evaluated by peers. This review
assessed face and content validity (see Straub, 1989). As a result,
changes were made to reword items and, in some cases, to drop items that
were possibly ambiguous, consistent with Moore and Benbasat's
(1991) as well as DeVellis's (2003) recommendations for scale
development. Subsequent to this, we distributed the survey instrument to
a group of 25 MBA students for evaluation. Once again, minor wording
changes were made. Finally, we performed some adjustments to the format
and appearance of the instrument, as suggested by both peers and MBA
students, though these minor changes had not a great importance here
given the survey was administered using the telephone. As the instrument
was already validated by Brown and Venkatesh (2005) and showed to be of
a great reliability, that we used the scale developed by Hobbs and
Osburn (1989) and validated in their study as well as in several other
studies to measure user satisfaction, and that we added only few items
to measure the new variables utility for security and mobility, then we
have not performed a pilot-test with a small sample. The evaluations by
both peers and MBA students were giving us some confidence that we could
proceed with a large-scale data collection.
SAMPLE AND DATA COLLECTION
First, in this study, we chose to survey people in household over
18 years taken from a large area in Atlantic Canada who own a mobile
phone. To do this, undergraduate and graduate students studying at our
faculty were hired to collect data using the telephone. A telephone was
then installed in an office of the faculty, and students, one at a time
over a 3 to 4-hour period, were asking people over the telephone to
answer our survey. And in order to get a diversified sample (e.g.,
students, retired people, people not working, people working at home,
and people working in enterprises), data were collected from 9 a.m. to 9
p.m. Monday through Friday over a 5-week period. Using the telephone
directory of the large area in Atlantic Canada chosen for the study,
students were randomly selecting people and asking them over the
telephone to answer our survey. The sample in the present study is
therefore a randomized sample, which is largely valued in the scientific
world given the high level of generalization of the results got from
such a sample. Once an individual had the necessary characteristics to
answer the survey and was accepting to answer it, the student was there
to guide him/her to rate each item of the survey on a seven points
Likert-type scale (1: strongly disagree ... 7: strongly agree). In
addition, the respondent was asked to answer some demographic questions.
Finally, to further increase the response rate of the study, each
respondent completing the survey had the possibility to win one of the
30 Tim Hortons $10 gift certificates which were drawn at the end of the
data collection. To this end, the phone number of each respondent was
put in a box for the drawing. Following this data collection process,
327 people in household answered our survey over a 5-week period.
DATA ANALYSIS PROCESS
The data analysis of the study was performed using a structural
equation modeling software, that is, Partial Least Squares (PLS-Graph
3.0). Using PLS, data have no need to follow a normal distribution and
it can easily deal with small samples. In addition, PLS is appropriate
when the objective is a causal predictive test instead of the test of a
whole theory (Barclay et al., 1995; Chin, 1998) as it is the case in
this study. To ensure the stability of the model developed to test the
research hypotheses, we used the PLS bootstrap resampling procedure (the
interested reader is referred to a more detailed exposition of
bootstrapping (see Chin, 1998; Efron & Tibshirani, 1993)) with an
iteration of 100 sub-sample extracted from the initial sample (327
Atlantic Canadian people). Some analyses were also performed using the
Statistical Package for the Social Sciences software (SPSS 13.5). The
results follow.
RESULTS
In this section of the paper, the results of the study are
reported. We begin to present some characteristics of the participants.
Then we validate the PLS model developed to test the research
hypotheses. Finally, we describe the results got from PLS analysis to
test the research hypotheses.
PARTICIPANTS
The participants in this study were either relatively aged or
relatively young, with a mean of 39.8 years and a large standard
deviation of 14.5 years. These statistics on the age of the participants
are, in fact, consistent with the growing old population phenomenon.
Near from two third of the participants were female (62%). Near from 80%
of the participants were married (50.9%) or single (28.4%). The gross
yearly income of the respondents in the study was in the range of $0 to
$50,000. Indeed, 72.4% of the respondents were winning between $0 and
$50,000, and, from this percentage, 35.5% were winning between $30,000
and $50,000. And 5.5% of the respondents were winning $100,000 or over.
Concerning the level of education, 25.5% of the participants in the
study got a high-school diploma, 26.4% got a college degree, and 39.6%
completed a baccalaureate. Only 2.1% of the participants got a
doctorate, which is relatively consistent with the whole population in
general. Finally, the respondents in our study were mainly full-time
employees (52.5%), retired people (12%), students (11.7%), self employed
(9%), part-time employees (7.4%), and unemployed (4.6%). These
statistics on the respondents' occupation help to explain the large
standard deviation on their age reported above. Indeed, 11.7% of the
respondents were young students, while 12% were retired people. So the
difference in age between these two groups is very large.
VALIDATION OF THE PLS MODEL TO TEST HYPOTHESES
First, to ensure the reliability of a construct or a variable using
PLS, one must verify the three following properties: individual item
reliability, internal consistency, and discriminant validity (for more
details, see Yoo & Alavi, 2001).
To verify individual item reliability, a confirmatory factor
analysis (CFA) was performed on independent and dependent variables of
the theoretical research model. A single iteration of the CFA was
necessary given all loadings of the variables were superior to 0.50 and
then none item was withdrawn nor transferred in another variable in
which the loading would have been higher. Indeed, in the whole, items
had high loadings, which suppose a high level of internal consistency of
their corresponding variables. In addition, loadings of each variable
were superior to cross-loadings with other variables of the model. Hence
the first criterion of discriminant validity was satisfied.
And to get composite reliability indexes and average variance
extracted (AVE) in order to satisfy the second criterion of discriminant
validity and to verify internal consistency of the variables, we used
PLS bootstrap resampling procedure with an iteration of 100 sub-sample
extracted from the initial sample (327 Atlantic Canadian people). The
results are presented in Table 3.
As shown in Table 3, PLS analysis indicates that all square roots
of AVE (boldfaced elements on the diagonal of the correlation matrix)
are higher than the correlations with other variables of the model. In
other words, each variable shares more variance with its measures than
it shares with other variables of the model. Consequently, discriminant
validity is verified. Finally, as supposed previously, we can see in
Table 3 that PLS analysis showed high composite reliability indexes for
all variables of the theoretical research model. The variables have
therefore a high internal consistency, with composite reliability
indexes ranging from 0.82 to 0.98.
HYPOTHESIS TESTING
First, to get the significant variables in the study and the
percentage of variance explained ([R.sup.2] coefficient) by all the
variables of the research model, we developed a PLS model similar to
those of Fillion (2005), Limayem and DeSanctis (2000), Limayem et al.
(2002), and Yoo and Alavi (2001). And to ensure the stability of the
model, we used the PLS bootstrap resampling procedure with an iteration
of 100 sub-sample extracted from the initial sample (327 Atlantic
Canadian people). The PLS model is depicted in Figure 2.
As shown in Figure 2, all the variables of our theoretical research
model, used as independent variables, are explaining 34.7% of the
variance around the dependant variable user satisfaction. And half of
these variables are significant, that is, they are determining factors
in satisfaction of using a mobile phone by people in household. More
specifically, the two more significant variables are perceived ease of
use (t = 5.18, beta = 0.36, p < 0.001) and utility for security (t =
4.38, beta = 0.21, p < 0.001). Three other variables are a few less
significant than these first two, but they are also very significant.
These variables are child's age (t = 2.32, beta = 0.20, p <
0.01), utility for work-related use (t = 2.26, beta = -0.12, p <
0.01), as well as declining cost (t = 2.14, beta = 0.11, p < 0.01).
And four other variables are significant at the level of significance
requested in this study, that is, p < 0.05. These variables are
marital status (t = 1.89, beta = -0.11, p < 0.05), cost (t = 1.79,
beta = 0.13, p < 0.05), applications for fun (t = 1.76, beta = 0.09,
p < 0.05), and mobility (t = 1.70, beta = 0.08, p < 0.05).
Finally, to measure interaction effect of moderator variables (the
life cycle stage characteristics: income (I), marital status (MS), age
(A), and child's age (CA)) in order to verify hypotheses 1 to 8, we
used the PLS procedure proposed by Chin et al. (2003) (see the paper for
more details). On the other hand, in a review of 26 papers assessing
interaction effect of moderator variables published between 1991 and
2000 into information systems (IS) journals, Carte and Russell (2003)
found nine errors frequently committed by researchers while estimating
such an effect, and provided solutions (see the paper for more details).
So we tried to avoid these nine errors in applying their solutions to
test hypotheses 1 to 8. Indeed, among others, in the verification of
hypotheses 1 to 8 that follows, interaction effect of a moderator
variable is significant if, and only if, the path between the latent
variable (the multiplication of items of independent and moderator
variables forming interaction effect) and the dependent variable is
significant, as well as if the change in [R.sup.2] coefficient (the
difference between the R2 calculated before the addition of interaction
effect and those calculated after the addition of interaction effect,
that is, ^[R.sup.2)] is greater than 0.
[FIGURE 2 OMITTED]
For a matter of space, given that the test of hypotheses 1 to 8
required the development of several PLS structural equation models (two
models per hypothesis, that is, 16 models), we summarize PLS analyses to
test each hypothesis. And, as for the PLS model developed to get the
significant variables in the study and the percentage of variance
explained by all the variables of the theoretical research model
previously (see Figure 2), for each PLS model developed, we used the PLS
bootstrap resampling procedure with an iteration of 100 sub-sample
extracted from the initial sample (327 Atlantic Canadian people) to
ensure the stability of the model.
Concerning hypothesis 1 related to the independent variable
applications for personal use (APU), the path from the latent variable
APU*MS*A to the dependent variable user satisfaction is significant (t =
1.698, beta = -0.154, p < 0.05) and there is a change in [R.sup.2]
(^[R.sup.2] = 0.011). Thus, as we expected, the moderator variables
marital status and age have an influence on the relationship between
applications for personal use and satisfaction of using a mobile phone
by people in household. Also hypothesis 1 is supported. The scenario is
different for hypothesis 2 related to the independent variable utility
for children (UC). The path from the latent variable UC*CA to the
dependent variable user satisfaction is not significant (t = 0.188, beta
= 0.034) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). So,
contrary to our expectations, the moderator variable child's age
has not an influence on the relationship between utility for children
and satisfaction of using a mobile phone by people in household. As a
result, hypothesis 2 is not supported. For hypothesis 3 related to the
independent variable utility for work-related use (UWRU), the path from
the latent variable UWRU*A to the dependent variable user satisfaction
is significant (t = 1.743, beta = -0.267, p < 0.05) and there is a
change in [R.sup.2] (^[R.sup.2] = 0.005). Thus, as we expected, the
moderator variable age has an influence on the relationship between
utility for work-related use and satisfaction of using a mobile phone by
people in household. Hypothesis 3 is therefore supported. Regarding
hypothesis 4 related to the independent variable applications for fun
(AF), the path from the latent variable AF*A to the dependent variable
user satisfaction is not significant (t = 0.450, beta = -0.068) and
there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Contrary to our
expectations, the moderator variable age has not an influence on the
relationship between applications for fun and satisfaction of using a
mobile phone by people in household. As a result, hypothesis 4 is not
supported. And the scenario is similar for hypothesis 5 related to the
independent variable status gains (SG), the path from the latent
variable SG*A to the dependent variable user satisfaction is not
significant (t = 0.466, beta = 0.093), but there is a small change in
[R.sup.2] (^[R.sup.2] = 0.002). So, contrary to our expectations, the
moderator variable age has not an influence on the relationship between
status gains and satisfaction of using a mobile phone by people in
household. Consequently, hypothesis 5 is not supported.
In the case of hypothesis 6 (a) related to the independent variable
friends and family influences (FFI), the path from the latent variable
FFI*MS*A*I to the dependent variable user satisfaction is not
significant (t = 0.477, beta = -0.068), but there is a substantial
change in [R.sup.2] (^[R.sup.2] = 0.006). So, contrary to our
expectations, the moderator variables marital status, age, and income
have not an influence on the relationship between friends and family
influences and satisfaction of using a mobile phone by people in
household. As a result, hypothesis 6 (a) is not supported. Concerning
hypothesis 6 (b) related to the independent variable secondary
sources' influences (SSI), the path from the latent variable
SSI*MS*A*I to the dependent variable user satisfaction is significant (t
= 1.666, beta = -0.169, p < 0.05) and there is a change in [R.sup.2]
(^[R.sup.2] = 0.002). Thus, as we expected, the moderator variables
marital status, age, and income have an influence on the relationship
between secondary sources' influences and satisfaction of using a
mobile phone by people in household. And hypothesis 6 (b) is supported.
The scenario is similar for hypothesis 6 (c) related to the independent
variable workplace referents' influences (WRI), the path from the
latent variable WRI*MS*A*I to the dependent variable user satisfaction
is significant (t = 1.778, beta = -0.195, p < 0.05) and there is a
change in [R.sup.2] (^[R.sup.2] = 0.001). Thus, as we expected, the
moderator variables marital status, age, and income have an influence on
the relationship between workplace referents' influences and
satisfaction of using a mobile phone by people in household.
Consequently, hypothesis 6 (c) is supported.
Regarding hypothesis 7 (a) related to the independent variable fear
of technological advances (FTA), the path from the latent variable
FTA*A*I to the dependent variable user satisfaction is not significant
(t = 0.493, beta = 0.092), but there is a small change in [R.sup.2]
(^[R.sup.2] = 0.001). Thus, contrary to our expectations, the moderator
variables age and income have not an influence on the relationship
between fear of technological advances and satisfaction of using a
mobile phone by people in household. Hypothesis 7 (a) is then not
supported. The scenario is similar for hypothesis 7 (b) related to the
independent variable declining cost (DC), the path from the latent
variable DC*A*I to the dependent variable user satisfaction is not
significant (t = 0.653, beta = -0.139), but there is a change in
[R.sup.2] (^[R.sup.2] = 0.003). So, contrary to our expectations, the
moderator variables age and income have not an influence on the
relationship between declining cost and satisfaction of using a mobile
phone by people in household. Consequently, hypothesis 7 (b) is not
supported. And the scenario is also similar for hypothesis 7 (c) related
to the independent variable cost (C), the path from the latent variable
C*A*I to the dependent variable user satisfaction is not significant (t
= 0.498, beta = -0.081), but there is a change in [R.sup.2] (^[R.sup.2]
= 0.004). Thus, contrary to our expectations, the moderator variables
age and income have not an influence on the relationship between cost
and satisfaction of using a mobile phone by people in household. As a
result, hypothesis 7 (c) is not supported.
Finally, concerning hypothesis 8 (a) related to the independent
variable perceived ease of use (PEU), the path from the latent variable
PEU*A to the dependent variable user satisfaction is not significant (t
= 0.816, beta = -0.334), but there is a substantial change in [R.sup.2]
(^[R.sup.2] = 0.005). Thus, contrary to our expectations, the moderator
variable age has not an influence on the relationship between perceived
ease of use and satisfaction of using a mobile phone by people in
household. As a result, hypothesis 8 (a) is not supported. The scenario
is different regarding hypothesis 8 (b) related to the independent
variable self-efficacy (SE), the path from the latent variable SE*A to
the dependent variable user satisfaction is significant (t = 1.726, beta
= -0.512, p < 0.05) and there is a substantial change in [R.sup.2]
(^[R.sup.2] = 0.006). So, as we expected, the moderator variable age has
an influence on the relationship between self-efficacy and satisfaction
of using a mobile phone by people in household. Consequently, hypothesis
8 (b) is supported.
In the next and last section of the paper, we discuss about some
implications of the more important findings of the study.
DISCUSSION AND CONCLUSIONS
This last section is devoted to a discussion about the results of
the study and some conclusions. And, to support our discussion and
conclusions, we provide the reader with a more detailed view of the PLS
structural equation model developed to get the significant variables in
the study, including the percentages of variance explained of variables
(see Table 4).
As shown in Table 4 (and Figure 2), the nineteen independent
variables examined in the study explained 34.7 percent ([R.sup.2] =
0.347) of the variance in satisfaction in the use of mobile phone by
people in household. And we can also see in Table 4 that the nine
variables who showed to be significant (see also the significant beta
path coefficients in Figure 2), that is, utility for work-related use,
utility for security, mobility, applications for fun, declining cost,
cost, perceived ease of use, marital status and child's age,
explained alone 24.6 percent of the variance in satisfaction of using a
mobile phone by people in household. Thus, these nine variables are
assuredly very important factors to take into account in future studies
on the mobile phone and on the part of mobile phone providers, and more
particularly perceived ease of use and utility for security which
explained alone 18.7 percent of this variance (see Table 4). It is very
interesting to see here that the two new variables that we added to the
Brown and Venkatesh's (2005) research model, that is utility for
security and mobility, showed to be very significant (p < 0.001 and p
< 0.05, respectively; see Table 4) in satisfaction of using a mobile
phone by people in household. Indeed, the present study showed that
people are, to some extent, using a mobile phone for a matter of
security (the mobile phone is useful for their own security and those of
their families) and mobility (the mobile phone provides them with the
possibility to use only this telephone to perform all their personal and
professional activities). So here are two new variables which we can add
to the integrated research model of MATH and household life cycle
characteristics suggested by Brown and Venkatesh (2005) to test in
future studies. In addition, these two new variables may be included in
the sales marketing plan of mobile phone providers.
In the large-scale study in which Brown and Venkatesh (2005)
integrated MATH and some household life cycle characteristics (as
moderating variables), the integrated model explained 74 percent of the
variance in intention to adopt a personal computer for home use, a
substantial increase of 24 percent over baseline MATH that explained 50
percent of the variance. In the present study, we used the integrated
model proposed by Brown and Venkatesh (2005). We also added two new
independent variables to the model, that is, utility for security and
mobility. And we used the household life cycle variables as moderating
variables in the research model as did Brown and Venkatesh (2005).
Finally, given that we investigated the perceptions of people already
using a mobile phone instead of those having the intention to adopt a
mobile phone, as did Brown and Venkatesh (2005) for the personal
computer, we used the dependent variable user satisfaction instead of
behavioural intention. And the model explained 34.7 percent of the
variance in satisfaction of using a mobile phone by people in household
(see Table 4 and Figure 2). Thus, in this study, our theoretical
research model explained a smaller percentage of variance than those
explained by MATH alone (without the household life cycle
characteristics and using behavioural intention as dependent variable).
Further, in a previous study in which we investigated the intention
to buy a mobile phone by people in household (see Fillion &
Berthelot, 2007), we also used the theoretical research model suggested
by Brown and Venkatesh (2005) to which we added the same two independent
variables utility for security and mobility than we included in the
present study in which we investigated satisfaction in the use of mobile
phone by people in household. And our model explained 50 percent of the
variance in intention to buy a mobile phone, while in the present study
our model explained 34.7 percent of the variance in satisfaction of
using a mobile phone. Of course, the dependent variable was different in
the two studies. Indeed, we used behavioural intention in the previous
study and user satisfaction in the present study. Hence we can see that
the variable behavioural intention is probably more appropriate as
dependent variable in the research model proposed by Brown and Venkatesh
(2005) than is user satisfaction, even when the model is augmented of
some new independent variables. Further, with the addition of the life
cycle stage variables income, marital status, age and child's age
as moderating variables to the model, as did Brown and Venkatesh (2005),
to test our research hypotheses, we have just observed a 3.1 percent
increase in variance explained, that is, 37.8 percent. However, it is to
be noted that, in the model we used in this study, more independent
variables showed to be good predictors in satisfaction of using a mobile
phone by people in household than did independent variables in the model
we used in the previous study in intention to adopt a mobile phone for
household use. So, although the result of our test seems, at first, not
to be very conclusive, in this study, we found several interesting
things to advance knowledge in this new and exciting field of adoption
and use of technology in households.
First, we found nine very important variables that seem to be good
predictors in satisfaction of using a mobile phone by people in
household, and more particularly perceived ease of use, utility for
work-related use, declining cost as well as the two new variables that
we added to the Brown and Venkatesh's (2005) model, utility for
security and mobility (see Table 4). These nine variables are also very
important to take into account by mobile phone providers to design new
mobile phones still better adapted to people's needs and to perform
their sales marketing. Second, we found that people are, to some extent,
using a mobile phone for a matter of security and mobility, given our
two new variables utility for security and mobility showed to be very
significant (see Table 4). Third, we found that it is probably much more
appropriate to use the dependent variable behavioural intention instead
of user satisfaction in the model proposed by Brown and Venkatesh
(2005), even augmented of our two new independent variables utility for
security and mobility, given the percentage of variance explained in
intention to adopt a mobile phone for household use in our previous
study is relatively higher. But, according to us, it is also appropriate
to include user satisfaction as dependent variable in the model given we
found more good predictors in satisfaction in the use of mobile phone in
the present study than in the previous one in which we used behavioural
intention as dependent variable. The dependent variable use behaviour
proposed by Thompson et al. (1991) may also be tested in future studies.
Also, we suggest the test of new independent variables which may explain
a greater percentage of variance in satisfaction of using a mobile phone
by people in household in future studies. To this end, we recommend
three new independent variables in the next paragraph. Finally, the
results of this study provided the evidence that it is far better to use
the household life cycle variables as moderating variables in the model,
as did Brown and Venkatesh (2005), given the percentage of variance
explained in intention to adopt a new technology in household by the
model tested by these authors was significantly higher. Indeed, we used
the household life cycle variables as moderating variables in the
theoretical research model of this study instead of independent
variables, as we have made in the previous study, and the percentage of
variance explained by the model in satisfaction of using a mobile phone
by people in household has been higher.
It would be interesting in future studies to add some other new
variables to the actual theoretical research model (those suggested by
Brown and Venkatesh (2005) augmented with the two new variables that we
tested in the present study, depending on the technology examined
naturally) in order to try to explain still more variance in
satisfaction of using a new technology in household. For example, the
variable attention might be added in social outcomes (a lot of people,
particularly young and old people, are feeling to be alone in the actual
stressing world, in which both men and women are working and get very
busy, so the mobile phone might be a good way to communicate with other
people every time and everywhere to get the feeling to be less alone),
the variable social norm might also be added in social outcomes (who
knows, people might be using a mobile phone just to do as everybody!),
and the variable control might be added in utilitarian outcomes (some
people might be using a mobile phone to control other people in their
family or others; maybe another kind of Big Brother!). It would be also
interesting to test the actual model in other situations and with other
populations. For example, in a subsequent study, we tested the actual
model with Atlantic Canadian people who are using high speed Internet at
home. As in this study, we used the dependent variable user satisfaction
given the respondents were already using high speed Internet. The
results of the study will follow in a subsequent paper. It will be
interesting to see whether the results remain the same as those got from
people who are using a mobile phone at home.
Regarding the limitations of this study, as pointed out by Brown
and Venkatesh (2005), the primary limitation is the reliance on a single
informant. It is possible that other members of the household would have
provided different responses concerning the motivations to use a mobile
phone at home. Future research in household use of technology should
incorporate responses from multiple members of the household to truly
assess the nature of household use. A second limitation of the study is
that it was conducted in only one area in Atlantic Canada. If the study
would have been carried out in the whole Atlantic Canada, its results
would be of a higher level of generalization. But the fact that the
sample of the study was a randomized sample allows a high level of
generalization of its results. Another limitation of the study is the
administration of the survey instrument over the telephone. Some
respondents might have not very well understood some items of the survey
instrument over the telephone and then provided more or less precise
ratings on these items, introducing the possibility of some response
bias. But the method we privileged in this study to administer the
survey instrument is not an exception to the rule: each method has its
own limitations!
To conclude, much more research will be needed on the use of
technology in households in order to better understand its impacts on
people's daily life. The research will allow, among others, at
least to minimize, if not to remove, some negative impacts of technology
in people's daily life in the future and to develop new
technologies still better adapted to people's needs. We will
continue to inquire into this new and exciting field.
ACKNOWLEDGMENTS
The authors would sincerely like to thank Professor Wynne W. Chin
(University of Houston at Texas) who kindly offered to us a license of
the last version of his structural equation modeling software PLS to
perform the data analysis of this study. We are also grateful to the
Faculte des Etudes Superieures et de la Recherche (FESR) at the
University of Moncton for its financial contribution to this study.
REFERENCES
Andonova, V. (2006). Mobile phones, the Internet and the
institutional environment. Telecommunications Policy, 30, 29-45.
Aoki, K. & Downes, E.J. (2003). An analysis of young
people's use of and attitudes toward cell phones. Telematics and
Informatics, 20(4), 349-364.
Balik, H.H., Turgut-Balik, D., Balikci, K. & Ozcan, I.C.
(2005). Some ocular symptoms and sensations experienced by long term
users of mobile phones. Pathologie Biologie, 53, 88-91.
Balikci, K., Ozcan, I.C., Turgut-Balik, D. & Balik, H.H.
(2005). A survey study on some neurological symptoms and sensations
experienced by long term users of mobile phones. Pathologie Biologie,
53, 30-34.
Barclay, D., Higgins, C. & Thompson, R. (1995). The partial
least squares (PLS) approach to causal modeling, personal computer
adoption and use as an illustration. Technology Studies, 2(2), 285-309.
Botelho, A. & Costa Pinto, L. (2004). The diffusion of cellular
phones in Portugal. Telecommunications Policy, 28, 427-437.
Brown, S.A. & Venkatesh, V. (2005). Model of adoption of
technology in households: A baseline model test and extension
incorporating household life cycle. MIS Quarterly, 29(3), 399-426.
Butt, S. & Phillips, J.G. (2008). Personality and self reported
mobile phone use. Computers in Human Behavior, 24, 346-360.
Campbell, S.W. & Russo, T.C. (2003). The social construction of
mobile telephony: An application of the social influence model to
perceptions and uses of mobile phones within personal communication
networks. Communication Monographs, 70(4), 317-334.
Carte, T.A. & Russell, C.J. (2003). In pursuit of moderation:
Nine common errors and their solutions. MIS Quarterly, 27(3), 479-501.
Centrone, F., Goia, A. & Salinelli, E. (2007). Demographic
processes in a model of innovation diffusion with a dynamic market.
Technological Forecasting & Social Change, 74, 247-266.
Chae, M. & Kim, J. (2004). Do size and structure matter to
mobile users? An empirical study of the effects of screen size,
information structure, and task complexity on user activities with
standard web phones. Behaviour & Information Technology, 23(3),
165-181.
Chen, C.H., Khoo, L.P. & Wan, W. (2003). Evaluation of
multicultural factors from elicited customer requirements for new
product development. Research in Engineering Design, 14, 119-130.
Chin, W.W. (1998). The partial least squares approach to structural
equation modeling. In G.A. Marcoulides (Ed.), Modern Methods for
Business Research (pp. 295-336), Mahwah, NJ: Lawrence Erlbaum
Associates.
Chin, W.W., Marcolin, B.L. & Newsted, P.R. (2003). A partial
least squares latent variable modeling approach for measuring
interaction effects: Results from a Monte Carlo simulation study and an
electronic-mail emotion/adoption study. Information Systems Research,
14(2), 189-217.
Chuang, M.C., Chang, C.C. & Hsu, S.H. (2001). Perceptual
factors underlying user preferences toward product form of mobile
phones. International Journal of Industrial Ergonomics, 27, 247-258.
Davie, R., Panting, C. & Charlton, T. (2004). Mobile phone
ownership and usage among pre-adolescents. Telematics and Informatics,
21(4), 359-373.
Davis, F.D. (1989). Perceived usefulness, perceived ease of use and
user acceptance of information technology. MIS Quarterly, 13(3),
319-340.
DeVellis, R.F. (2003). Scale Development: Theory and Applications
(2nd ed.). Thousand Oaks, CA: Sage Publications.
Eby, D.W., Vivoda, J.M. & St-Louis, R.M. (2006). Driver
hand-held cellular phone use: A four-year analysis. Journal of Safety
Research, 37, 261-265.
Efron, B. & Tibshirani, R.J. (1993). An Introduction to the
Bootstrap. New York: Chapman and Hall.
Ehlen, J. & Ehlen, P. (2007). Cellular-only substitution in the
United States as lifestyle adoption. Public Opinion Quarterly, 71(3),
717-733.
Fillion, G. (2004). Publishing in the organizational sciences: An
extended literature review on the dimensions and elements of an
hypothetico-deductive scientific research, and some guidelines on
"how" and "when" they should be integrated. Academy
of Information and Management Sciences Journal, 7(1), 81-114.
Fillion, G. (2005). L'integration des TIC dans la formation
universitaire : une etude des resultats educationnels des etudiants dans
les contextes de presence et de non presence en classe. Doctoral Thesis
(Ph.D.), Faculty of Administration, Laval University, Quebec.
Fillion, G. & Berthelot, S. (2007). An extended model of
adoption of technology in households: A model test on people's
intention to adopt a mobile phone. Management Review: An International
Journal, 2(2), 4-36.
Fillion, G. & Le Dinh, T. (2008). An extended model of adoption
of technology in households: A model test on people using a mobile
phone. Management Review: An International Journal, 3(1), 58-91.
Funk, J. (2005). The future of the mobile phone Internet: An
analysis of technological trajectories and lead users in the Japanese
market. Technology in Society, 27, 69-83.
Gartner Newsroom (2008). Gartner says worldwide mobile phone sales
grew 6 per cent in 2008, but sales declined 5 per cent in the fourth
quarter. Retrieved July 8, 2009, from
http://www.gartner.com/it/pagejsp?id=904729.
Han, S.H. & Wong, S.W. (2003). A systematic approach for
coupling user satisfaction with product design. Ergonomics, 46(13/14),
1441-1461.
Han, S.H., Kim, K.J., Yun, M.H., Hong, S.W. & Kim, J. (2004).
Identifying mobile phone design features critical to user satisfaction.
Human Factors and Ergonomics in Manufacturing, 14(1), 15-29.
Hobbs, V.M. & Osburn, D.D. (1989). Distance learning evaluation
study report II: A study of North Dakota and Missouri schools
implementing German I by satellite. ERIC ED 317 195.
Isiklar, G & Buyukozhan, G. (2007). Using multi-criteria
decision making approach to evaluate mobile phone alternatives. Computer
Standards & Interfaces, 29(2), 265-274.
Karjaluoto, H., Pakola, J., Pietila, M. & Svento, R. (2003). An
exploratory study of antecedents and consequences of mobile phone usage
in Finland. Proceedings of the AMA Conference (pp. 170-178), Chicago,
August 1518.
Kauffman, R.J. & Techatassanasoontorn, A.A. (2005). Is there a
global digital for digital wireless phone technologies? Journal of the
Association for Information Systems, 6(12), 338-382.
Kurniawan, S. (2008). Older people and mobile phones: A
multi-method investigation. International Journal of Human-Computer
Studies, 66(12), 889-901.
Kwon, H.S. & Chidambaram, L. (2000). A test of the technology
acceptance model: The case of cellular telephone adoption. Proceedings
of HICSS-34, Hawaii, January 3-6.
Lai, H.H., Lin, Y.C., Yeh, C.H. & Wei, C.H. (2006).
User-oriented design for the optimal combination on product design.
International Journal of Production Economics, 100(2), 253-267.
LaRose, R. (1989). Advances in telecommunications. In H. Sawhney
& G. Barnett (Eds.), Understanding Personal Telephone Behavior,
Progress in Communication Science, 15.
Lee, Y.S., Hong, S.W., Smith-Jackson, T.L., Nussbaum, M.A. &
Tomioka, K. (2006). Systematic evaluation methodology for cell phone
user interfaces. Interacting with Computers, 18, 304-325.
Liljander, V., Polsa, P. & Forsberg, K. (2007). Do mobile CRM
services appeal to loyalty program customers? International Journal of
E-Business Research, 3(2), 24-40.
Limayem, M. & DeSanctis, G. (2000). Providing decisional
guidance for multicriteria decision making in groups. Information
Systems Research, 11(4), 386-401.
Limayem, M., Banerjee, P. & Ma, L. (2002). Impact of GDSS:
Opening the black box. Proceedings of the 6th Pacific Asia Conference on
Information Systems (PACIS 2002), Tokyo, Japan, September 2-4.
Mazzoni, C., Castaldi, L. & Abdeo, F. (2007). Consumer behavior
in the Italian mobile telecommunication market. Telecommunications
Policy, 31, 632-647.
Moore, G.C. & Benbasat, I. (1991). Development of an instrument
to measure the perceptions of adopting an information technology
innovation. Information Systems Research, 2(3), 192-222.
Palen, L., Salzman, M. & Youngs, E. (2000). Going wireless:
Bevavior & practice of new mobile phone users. Proceedings of
ACMCSCW (pp. 201-210), Philadelphia, December 2-6.
Peters, C., Amato, C.H. & Hollenbeck, C.R. (2007). An
exploratory investigation of consumers' perceptions of wireless
advertising. Journal of Advertising, 36(4), 129-145.
Repacholi, M.H. (2001). Health risks from the use of mobile phones.
Toxycology Letters, 120, 323-331.
Rosenbloom, T. (2006). Driving performance while using cell phones:
An observational study. Journal of Safety Research, 37, 207-212.
Salvucci, D.D. & Macuga, K.L. (2002). Predicting the effects of
cellular phone dialling on driver performance. Cognitive Systems
Research, 3, 95-102.
Selwyn, N. (2003). Schooling the mobile generation: The future for
schools in the mobile-networked society. British Journal of Sociology
and Education, 24(2), 131-142.
Sohn, S. Y. & Kim, Y. (2008). Searching customer patterns of
mobile service using clustering and quantitative association rule.
Expert Systems with Applications, 34, 1070-1077.
Straub, D.W. (1989). Validating instruments in MIS research. MIS
Quarterly, 13(2), 147-169.
Sullman, M.J.M. & Baas, P.H. (2004). Mobile phone use amongst
New Zealand drivers. Transportation Research Part F, 7, 95-105.
Tam, M.C.Y. & Tummala, V.M.R. (2001). An application of the AHP
in vendor selection of a telecommunications system. Omega, 29, 171-182.
Thompson, R.L., Higgins, C.A. & Howell, J.M. (1991). Personal
computing: Toward a conceptual model of utilization. MIS Quarterly,
15(1), 124-143.
Tornros, J. & Bolling, A. (2006). Mobile phone use--Effects of
conversation on mental workload and driving speed in rural and urban
environments. Transportation Research Part F, 9, 298-306.
Treffner, P.J. & Barrett, R. (2004). Hands-free mobile phone
speech while driving degrades coordination and control. Transportation
Research Part F, 7, 229-246.
Tucker, C., Brick, J.M. & Meekins, B. (2007). Household
telephone service and usage patterns in the United States in 2004:
Implications for telephone samples. Public Opinion Quarterly, 71(1),
3-22.
Venkatesh, V. & Brown, S.A. (2001). A longitudinal
investigation of personal computers in homes: Adoption determinants and
emerging challenges. MIS Quarterly, 25(1), 71-102.
Wang, S.E. & Sung, Y.C. (2003). Mobile phone consumption
behavior analysis. Proceedings of the Asia Pacific Management Conference
(pp. 1-11), China, April 23-25.
Wei, R. (2007). Motivations for using the mobile phone for mass
communications and entertainment. Telematics and Informatics, 25(1),
36-46.
Weinberger, Z. & Richter, E.D. (2002). Cellular phones and
effects on the brain: The head as an antenna and brain tissue as a radio
receiver. Medical Hypotheses, 59(6), 703-705.
Westerman, R. & Hocking, B. (2004). Diseases of modern living:
Neurological changes associated with mobile phones and radiofrequency
radiation in humans. Neuroscience Letters, 361, 13-16.
White, M.P., Eiser, J.R., Harris, P.R. & Pahl, S. (2007). Who
reaps the benefits, who bears the risks? Comparative optimism,
comparative utility, and regulatory preferences for mobile phone
technology. Risk Analysis, 27(3), 741-753.
Wilska, T. (2003). The role of m-services in the well-being of
young people. Journal of Consumer Policy, 26(4), 441-463.
Yoo, Y. & Alavi, M. (2001). Media and group cohesion: Relative
influences on social presence, task participation, and group consensus.
MIS Quarterly, 25(3), 371-390.
Gerard Fillion, University of Moncton
Jean-Pierre Booto Ekionea, University of Moncton
Table 1: Related Literature Survey (adapted from Isiklar & Buyukozhan,
2007, p. 267; and updated)
Research Area References
Mobile phone diffusion and its LaRose (1989)
impacts on people's daily life. Kwon & Chidambaram (2000)
Botelho & Costa Pinto (2004)
Funk (2005)
Andonova (2006)
Centrone et al. (2007)
Ehlen & Ehlen (2007)
Fillion & Berthelot (2007)
Nasar et al. (2007)
Fillion & Le Dinh (2008)
Kurniawan (2008)
Mobile phone ownership and usage. LaRose (1989)
Kwon & Chidambaram (2000)
Palen et al. (2000)
Aoki & Downes (2003)
Selwyn (2003)
Davie et al. (2004)
Mazzoni et al. (2007)
Peters et al. (2007)
Tucker et al. (2007)
Sohn & Kim (2008)
Mobile phone ownership and usage from Karjaluoto et al. (2003)
a behavioural and psychological Wilska (2003)
perspective. Davie et al. (2004)
Liljander et al. (2007)
White et al. (2007)
Butt & Phillips (2008)
Effects on human health and daily Repacholi (2001)
activities. Salvucci & Macuga (2002)
Weinberger & Richter (2002)
Sullman & Baas (2004)
Treffner & Barrett (2004)
Westerman & Hocking (2004)
Balik et al. (2005)
Balikci et al. (2005)
Eby et al. (2006)
Rosenbloom (2006)
Tornros & Bolling (2006)
Evaluation and design of mobile phone Chuang et al. (2001)
features for user interface and user Chen et al. (2003)
satisfaction. Han & Wong (2003)
Chae & Kim (2004)
Han et al. (2004)
Lee et al. (2006)
Analytical evaluations of mobile Tam & Tummala (2001)
phone-related observations. Campbell & Russo (2003)
Han & Wong (2003)
Wang & Sung (2003)
Lai et al. (2006)
Table 2: Variables and Definitions
Beliefs and
Characteristics Variables Definitions
Attitudinal Applications for The extent to which using
Beliefs Personal Use a mobile phone enhances
the effectiveness of
household activities
(adapted from Venkatesh &
Brown, 2001).
Utility for Children The extent to which using
a mobile phone enhances
the children's
effectiveness in their
activities (adapted from
Venkatesh & Brown, 2001).
Utility for Work-Related The extent to which using
Use a mobile phone enhances
the effectiveness of
performing work-related
activities (adapted from
Venkatesh & Brown, 2001).
Utility for Security The extent to which using
a mobile phone increases
the security of its user
and his/her family.
Mobility The extent to which a
mobile phone allows to
use only this telephone
to perform all personal
and professional
activities.
Applications for Fun The pleasure derived from
mobile phone use (adapted
from Venkatesh & Brown,
2001). These are specific
to mobile phone usage,
rather than general
traits (adapted from
Brown & Venkatesh, 2005;
see Webster & Martocchio,
1992, 1993).
Status Gains The increase in prestige
that coincides with the
purchase of a mobile
phone for home use
(adapted from Venkatesh &
Brown, 2001).
Normative Friends and Family "The extent to which the
Beliefs Influences members of a social
network influence one
another's behaviour"
(Venkatesh & Brown, 2001,
p. 82). In this case, the
members are friends and
family (Brown &
Venkatesh, 2005).
Secondary Sources' The extent to which
Influences information from TV,
newspaper, and other
secondary sources
influences behaviour
(Venkatesh & Brown,
2001).
Workplace Referents' The extent to which
Influences coworkers influence
behaviour (Brown &
Venkatesh, 2005; see
Taylor & Todd, 1995).
Control Fear of Technological The extent to which
Beliefs Advances rapidly changing
technology is associated
with fear of obsolescence
or apprehension regarding
a mobile phone purchase
(adapted from Venkatesh &
Brown, 2001).
Declining Cost The extent to which the
cost of a mobile phone is
decreasing in such a way
that it inhibits adoption
(adapted from Venkatesh &
Brown, 2001).
Cost The extent to which the
current cost of a mobile
phone is too high
(adapted from Venkatesh &
Brown, 2001).
Perceived Ease of Use The degree to which using
the mobile phone is free
from effort (Davis, 1989;
also adapted from
Venkatesh & Brown, 2001).
Self-Efficacy (or The individual's belief
Requisite Knowledge) that he-she has the
knowledge necessary to
use a mobile phone. This
is closely tied to
computer self-efficacy
(Compeau & Higgins,
1995a, 1995b; see also
Venkatesh & Brown, 2001).
Life Cycle Income The individual's year
Characteristics Marital Status gross income (see Wagner
& Hanna, 1983). The
individual's family
status (married, single,
divorced, widowed, etc.)
(see Danko & Schaninger,
1990).
Age The individual's age (see
Danko & Schaninger,
1990). In this case, age
is calculated from the
individual's birth date.
Child's Age The age of the
individual's youngest
child (see Danko &
Schaninger, 1990). In
this case, age is
represented by a numeral.
Table 3: Means, Standard Deviations, Composite Reliability Indexes,
Correlations, and Average Variance Extracted of Variables
Relia- Correlations
Variable M SD bility and Average
Index Variance
Extracted (d)
1 2
1. Applications for Personal
Use 3.84 1.92 0.82 0.77
2. Utility for Children 2.07 2.52 0.96 .27 0.94
3. Utility for Work-Related
Use 3.17 2.46 0.91 .39 .10
4. Utility for Security 5.62 1.68 0.89 .21 .16
5. Mobility 3.55 2.03 0.88 .30 .05
6. Applications for Fun 2.88 1.96 0.89 .35 .05
7. Status Gains 2.45 1.72 0.93 .18 .15
8. Friends and Family
Influences 3.66 2.27 0.93 .26 .05
9. Secondary Sources'
Influences 3.24 2.25 0.90 .17 .09
10. Workplace Referents'
Influences 3.12 2.41 0.98 .26 -.03
11. Fear of Technological
Advances 3.21 1.97 0.83 -.06 .10
12. Declining Cost 4.14 1.88 0.89 .17 .13
13. Cost 4.38 1.83 0.96 .07 .01
14. Perceived Ease of Use 5.69 1.45 0.88 .19 -.05
15. Self-Efficacy 6.39 1.02 0.93 .18 -.14
16. Income (a) NA NA NA .04 .11
17. Marital Status (a) NA NA NA -.04 -.03
18. Age (b) 39.80 14.50 NA .12 -.24
19. Child's Age (c) 16.29 9.09 NA .11 .09
20. User Satisfaction 5.46 1.41 0.86 .18 .04
Correlations and Average Variance
Extracted (d)
Variable
3 4 5 6 7 8
1. Applications for Personal
Use
2. Utility for Children
3. Utility for Work-Related
Use 0.88
4. Utility for Security -.04 0.85
5. Mobility .23 .09 0.84
6. Applications for Fun .23 .13 .25 0.82
7. Status Gains .19 .13 .31 .37 0.90
8. Friends and Family
Influences .16 .13 .19 .43 .40 0.88
9. Secondary Sources'
Influences .08 .10 .09 .25 .23 .36
10. Workplace Referents'
Influences .37 .04 .19 .31 .29 .53
11. Fear of Technological
Advances .04 .10 -.09 .04 .15 .13
12. Declining Cost .08 .12 .12 .06 .05 .04
13. Cost .04 .16 .13 .04 .22 .16
14. Perceived Ease of Use .09 .15 .27 .24 .18 .17
15. Self-Efficacy .04 .12 .18 .12 .03 .11
16. Income (a) .09 -.12 -.11 -.32 -.23 -.24
17. Marital Status (a) -.22 .27 -.02 .09 -.06 .06
18. Age (b) .20 .04 .21 .46 .22 .31
19. Child's Age (c) -.03 -.06 .02 .33 .12 .24
20. User Satisfaction -.09 .31 .20 .21 .11 .16
Correlations and Average Variance
Extracted (d)
Variable
9 10 11 12 13 14
1. Applications for Personal
Use
2. Utility for Children
3. Utility for Work-Related
Use
4. Utility for Security
5. Mobility
6. Applications for Fun
7. Status Gains
8. Friends and Family
Influences
9. Secondary Sources'
Influences 0.87
10. Workplace Referents'
Influences .33 0.98
11. Fear of Technological
Advances .15 .16 0.79
12. Declining Cost .13 .08 -.04 0.85
13. Cost .07 .10 .24 -.09 0.96
14. Perceived Ease of Use -.02 .20 -.11 .15 .00 0.80
15. Self-Efficacy -.08 .12 -.12 .15 -.00 .66
16. Income (a) -.05 -.04 -.07 .02 -.11 -.05
17. Marital Status (a) .00 -.04 .04 -.11 .03 -.03
18. Age (b) .16 .37 -.05 -.03 .10 .31
19. Child's Age (c) .13 -.09 -.06 -.07 .09 .16
20. User Satisfaction .06 .03 -.10 .21 -.06 .40
Correlations and Average Variance
Extracted (d)
Variable
15 16 17 18 19 20
1. Applications for Personal
Use
2. Utility for Children
3. Utility for Work-Related
Use
4. Utility for Security
5. Mobility
6. Applications for Fun
7. Status Gains
8. Friends and Family
Influences
9. Secondary Sources'
Influences
10. Workplace Referents'
Influences
11. Fear of Technological
Advances
12. Declining Cost
13. Cost
14. Perceived Ease of Use
15. Self-Efficacy 0.91
16. Income (a) -.00 NA
17. Marital Status (a) -.02 -.22 NA
18. Age (b) .18 -.41 .16 NA
19. Child's Age (c) .14 -.21 .07 -.08 NA
20. User Satisfaction .27 -.13 .10 .06 .07 0.71
(a) This variable was coded as a nominal variable. It was measured in
terms of non quantified distinct categories.
(b) This variable was coded as a continuous variable. It was measured
using the respondents' birth date.
(c) This variable was coded using the age of the respondents' youngest
child.
(d) Boldfaced elements on the diagonal of the correlation matrix
represent the square root of the average variance extracted (AVE).
For an adequate discriminant validity, the elements in each row and
column should be smaller than the boldfaced element in that row or
column.
Table 4: Beta Path Coefficients, T-Values, and Percentages of Variance
Explained of Variables
Variable Beta t [R.sup.2]
Applications for Personal Use 0.054 0.925 0.002
Utility for Children -0.039 0.726 0.001
Utility for Work-Related Use -0.119 ** 2.263 0.008
Utility for Security 0.210 **** 4.379 0.104
Mobility 0.079 * 1.704 0.004
Applications for Fun 0.089 * 1.764 0.004
Status Gains -0.029 0.511 0.000
Friends and Family Influences 0.022 0.357 0.000
Secondary Sources' Influences 0.015 0.248 0.001
Workplace Referents' Influences -0.020 0.314 0.000
Fear of Technological Advances -0.059 0.551 0.005
Declining Cost 0.111 ** 2.137 0.011
Cost 0.125 * 1.794 0.021
Perceived Ease of Use 0.355 **** 5.180 0.083
Self-Efficacy -0.019 0.271 0.069
Income -0.059 0.687 0.001
Marital Status -0.111 * 1.892 0.000
Age -0.076 0.640 0.014
Child's Age 0.201 ** 2.323 0.011
* p < 0.05; ** p < 0.01; **** p < 0.001.
