Printer Friendly

Testing a moderator-type research model on the use of high speed Internet.

INTRODUCTION AND BACKGROUND FOR THE STUDY

The vast technological possibilities of the Internet are at the basis of the fast progress of the information society (Al-Omoush & Shaqrah, 2010). It has become one of the most important means of new forms of cooperation and competition in the various subsystems of society (Al-Omoush & Shaqrah, 2010). Anderson (2008) argues that Internet has a great influence on people's connections to friends, families, and their communities, on the social system of formal and informal support, and on the working of groups and teams. It is also the valuable instrument of scientific, social, marketing researches, and business development (Al-Omoush & Shaqrah, 2010). In addition, the Internet, as an information and entertainment technology, affects education, government, publishing, retail industry, banking, broadcast services, health care delivery, and so on (Al-Omoush & Shaqrah, 2010). So, the scope of the Internet is now worldwide and in all sectors of the society, and then forces to deliver this essential resource to people in households. In order to provide the reader with a good overall view of the actual Internet population, we have regrouped in Table 1 the different generations (as conventionalized by Strauss and Howe's book: Generations: The History of America's Future, 1584 to 2069) and the percentages of adult both total and online.

As shown in Table 1, "Contrary to the image of Generation Y as the 'Net Generation,' Internet users in their twenties do not dominate every aspect of online life. Generation X is the most likely group to bank, shop, and look for health information online. Boomers are just as likely as Generation Y to make travel reservations online. And even Silent Generation Internet users are competitive when it comes to e-mail (although teens might point out that this is proof that e-mail is for old people). The Web continues to be populated largely by younger generations, as more than half of the adult Internet population is between 18 and 44 years old. But larger percentages of older generations are online now than in the past and they are doing more activities online, according to the Pew Research Center's Internet & American Life Project surveys taken from 2006-2008." (Jones & Fox, 2009, p. 1)

As the Internet population is continually growing since its infancy, the need for always faster and performing telecommunications networks allowing people to communicate and to perform their daily activities at a satisfying pace has become an everyday concern for all the countries, hence the apparition of high speed Internet in the middle of the 1990s. For example, for more than a decade, the Government of Canada has been developing strategies to enable Canadians to become participants in the information society (Government of Canada, 1999; Government Online Advisory Panel, 2003; quoted in Middleton & Ellison, 2006, p. 1). As part of this strategy, it was recommended that broadband Internet access be made available to all Canadian households (National Broadband Task Force, 2001; quoted in Middleton & Ellison, 2006, pp. 1-2). There are still many not served and underserved areas in the country (CRACIN, 2005), and the Telecommunications Policy Review Panel has urged the federal government to "reaffirm its commitment to maintaining Canada's global broadband leadership and to ensuring that broadband access is available everywhere in the country" (Telecommunications Policy Review Panel, 2006, p. 8-5; quoted in Middleton & Ellison, 2006, p. 2). Similarly, as a public issue, broadband has taken on a higher profile in recent months because of President Obama's decision to include funding for broadband in the American Recovery and Reinvestment Act (ARRA); also ARRA included $7.2 billion for broadband with the goal of accelerating the deployment of broadband Internet access in the United States (Horrigan, 2009). These are just two examples here of how much all of the countries are putting a strong emphasis on strategies capable of promoting the acceleration of the deployment of high speed Internet everywhere in the world. As a result, the number of households with a broadband Internet connection continues to grow with one in five households worldwide expected to have a fixed broadband connection by the end of 2009, according to Gartner Research (Digital Home, 2009).

In 2009, Canada ranked fifth in broadband penetration. In fact, leading the high speed Internet race at the beginning of 2009 was South Korea at 86%, followed by the Netherlands (80%), Denmark (75%), and Hong Kong (72%), followed by Canada and Switzerland (69%). Rounding out the top ten were Norway (67%), New Zealand (65%), France, Singapore and the UK (62%). The United States ranked 14th at 60%. Overall, in 2009, approximately 21 countries had high speed Internet connections in at least 50 percent of homes. (Digital Home, 2009)

But, what is high speed Internet?

According to Gill (2010), ever wonder what a company means when it says its Internet service is "high speed"? Then check out Table 2 that documents the plethora of technologies that the Federal Communications Commission (FCC) counts as "broadband"--be warned, speeds can vary by as much as 2,000 percent (for example, regarding Canada, Dunn (2010) argues that high speed Internet access is expensive and slow)! In short, "broadband" is defined by the FCC as anything other than "dial up"--and "high speed" has no commonly-agreed-to definition (Gill, 2010). So, in this paper, when we talk about "high speed" Internet or "broadband", we then consider the two as the same telecommunications technology which is faster than "dial up".

What are the anticipated benefits of high speed Internet?

Here we provide the reader with only a brief overview of the anticipated benefits of high speed Internet given that the studies presented in Table 3 are further putting in evidence some benefits. Thus, we describe in the next paragraphs three relevant examples of benefits that high speed Internet can bring: one at the environmental level, one at the social level, and one at the economic level.

First, a study realized by Fuhr and Pociask released in October 2007 by the American Consumer Institute praises the benefic impact that a widely spread use of broadband could have on the environment by cutting greenhouse gas emissions. The study focuses on the different behaviors that the use of high speed Internet allows, such as buying online, telecommuting, e-materializing, teleconferencing, videoconferencing, as well as distance learning, and converts their benefits into saving of greenhouse gas emissions by mainly cutting on energy. According to Fuhr and Pociask (2007; quoted in Labriet-Gross, 2007), using high speed Internet mainly influences the amount of travel, space and material needed when you buy, work or lean based on rather simple findings. "Indeed, instead of going to five or 6 stores to find who has a product or who has the best price, you can just search on the Internet and buy online, so it cuts back on the pollution linked to the commute", explain these authors. As for the supply chain, it decreases the inventory, which means less storage facilities, so less need for heat, air conditioning, and lighting. For instance, Dell has increased its sales by 36 times, while its facilities space has been reduced by 4. Overall, e-commerce already cuts 37.5 million tons of CO2 emissions (and only in business to business (B2B) and business to consumers (B2C) transactions), and could save 206.3 million tons in 2017, argue Fuhr and Pociask (2007; quoted in Labriet-Gross, 2007).

Varon (2010) relates the experience of Case Western Reserve University who launched the University Circle Innovation Zone, a large project deploying gigabit fiber optic connections to residents of Cleveland's poorest neighborhoods. Beginning with 104 homes, local institutions, including hospitals, schools, electric utilities, and public safety agencies, will use the network to deliver cutting edge services to residents. Over the next 18 months, Case Western researchers will study changes in residents' health and other indicators of their standard of living. One project will use videoconferencing technology provided by LifeSize to enable residents with chronic conditions such as diabetes to consult with healthcare providers over high definition video. Patients will also be given devices that automatically monitor their health and transmit data to medical professionals. Broadband access could enable residents to take better care of them when they cannot visit their doctors. Other projects would provide science and math ... materials to students, deliver video feeds to police, and collect data to help residents manage energy usage. Case Western's technology and service provider partners are financing this social initiative.

And, a study conducted by Orazem (2005) measuring the impact of high speed Internet access on local economic growth suggests that this one increases growth in earnings per worker, aggregate earnings and the number of firms, but it lowers the rate of growth of employment. All of these are consistent with the presumption that high speed Internet access can lower firm costs, improve information flows with suppliers and consumers, and, at the same time, lower the need for employees specializing in sales or procurement. And all of these effects were larger in less density populated areas, then suggesting that rural areas do benefit disproportionately from high speed Internet access. A more recent study made by Majumdar (2008) is consistent with Orazem's (2005) findings concerning earnings per worker and employment, that is, he found that broadband diffusion within and between the firms over time has a positive and significant impact on wage levels, but its impact on employment is negative.

On the other hand, what are the main barriers to high speed Internet adoption?

According to the Pew Internet & American Life Project (see Horrigan, 2009), the factors positively correlated with home broadband adoption (in order of importance) are: income (household income greater than $75,000 annually); having college degree or more; parent of minor child in household); married or living with partner; and full-time employee. As for the factors negatively correlated with home broadband adoption (in order of importance), they are: having less than a high school degree; senior individual (aged of 65 or over); living in rural America; having a high school degree; and being African American (non-Hispanic).

Telecommunications industry is continually in a shift of change, alimented by technological innovation and consumers' demand for always better and faster communication tools. High speed Internet is now an integral part of everyday life of more than a billion people. And, as the tendency is showing up, its use will be still increasing in the future. Thus, this technology has and will continue to have major social and economic impacts. Individual adoption of technology has been studied extensively in the workplace, but far less attention has been paid to adoption of technology in household (Brown & Venkatesh, 2005). So, few studies have been conducted until now to verify satisfaction of household people using high speed Internet. It is therefore crucial to more deeply examine the determining factors in satisfaction of using high speed Internet by people in household. This is the aim of the present study. The related literature on the actual research area of high speed Internet is summarized in Table 3.

As we can see in the summary of literature related to high speed Internet presented in Table 3, very few studies until now examined the determining factors in satisfaction of using high speed Internet by people in household. Thus, the present study brings an important contribution to fill this gap given it allows a better understanding of the impacts of high speed Internet usage in people's everyday life. It focuses on the following research question: What are the determining factors in satisfaction of using high speed Internet 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 development of the study is presented; second, the methodology followed to conduct the study is described; finally, the results of the study are reported and discussed.

THEORETICAL DEVELOPMENT

This study is based on the theoretical foundations developed by Venkatesh and Brown (2001) to investigate the factors driving personal computer (PC) adoption in American homes as well as those developed by Brown and Venkatesh (2005) in order to verify the determining factors in intention to adopt a PC 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. And Brown et al. (2006) tested the same integrated model in the context of PC use. As pointed out by Brown et al. (2006), even though the technology of interest in MATH is PC, the model is expected to generalize to other information technology (IT) products and systems in the household context. Also, with the exception of behavioral intention (we included user satisfaction instead of behavioral intention given people investigated in this study already have high speed Internet access), all the variables proposed and tested by Brown and Venkatesh (2005) are used in this study. And we added a new variable, mobility, in order to verify whether or not it is a factor of satisfaction of household people using high speed Internet. The resulting theoretical research model is depicted in Figure 1.

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 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 PC adoption (household high speed Internet 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 high speed Internet by people in household). All of the variables integrated in the theoretical research model depicted in Figure 1 are defined in Table 4.

We can see in Table 4 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 PC 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 definition of the new independent variable that we added to the model is from our own. In fact, we defined this variable in accordance with which we wanted to measure regarding mobility before developing and validating items that measure the variable on the basis of the definition formulated.

[FIGURE 1 OMITTED]

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 high speed Internet at home.

H2: Child's age will moderate the relationship between utility for children and satisfaction of using high speed Internet at home.

H3: Age will moderate the relationship between utility for work-related use and satisfaction of using high speed Internet at home.

H4: Age will moderate the relationship between applications for fun and satisfaction of using high speed Internet at home.

H5: Age will moderate the relationship between status gains and satisfaction of using high speed Internet 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 high speed Internet 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 high speed Internet 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 high speed Internet at home.

In the next section of the paper, the methodology followed to conduct the study is described.

METHODOLOGY

The study was designed to gather information concerning high speed Internet satisfaction in Atlantic Canadian households. Indeed, the focus of the study is on individuals who have high speed Internet access at home. 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 empirically validated by Brown and Venkatesh (2005) to which we added two new scales, the first one measuring another dimension in satisfaction of using high speed Internet by people in household, that is, 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, few 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 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 variable 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 from a large area in Atlantic Canada, which have high speed Internet access. To do that, 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, to get the more diversified sample as possible (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 that end, the phone number of each respondent was put in a box for the drawing. Following this data collection process, 322 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 (322 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. First, 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 analyses to test the research hypotheses.

Participants

The participants in this study were relatively aged, with a mean of 40 years and a standard deviation of 13.7 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 (62.7%). Near from 80% of the participants were married (52.1%) or single (26.8%). The gross yearly income of the respondents in the study was in the range of $0 to $60,000. Indeed, 80.9% of the respondents were winning between $0 and $60,000, and, from this percentage, 47.3% were winning between $30,000 and $60,000. And 5.3% of the respondents were winning $100,000 or over. Concerning the level of education, 20.3% of the participants in the study got a high-school diploma, 28.3% got a college degree, 37.3% completed a baccalaureate, and 10.3% got a master. Only 2.9% of the participants completed a doctorate, which is relatively consistent with the whole population in general. Finally, the respondents in our study were mainly full-time employees (46.8%), part-time employees (14.6%), retired people (13.3%), self employed (8.9%), unemployed (7.6%), and students (5.4%).

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 and 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 (322 Atlantic Canadian people). The results are presented in Table 5.

As shown in Table 5, 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 5 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 0.74 to 0.99.

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 theoretical research model, we developed a PLS model similar to those of Fillion (2005), Fillion and Le Dinh (2008), Fillion et al. (2010a), Fillion and Booto Ekionea (2010b), 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 (322 Atlantic Canadian people). The PLS model is depicted in Figure 2.

As we can see in Figure 2, all of the variables of our theoretical research model, used as independent variables, are explaining 49.2% of the variance on the dependant variable user satisfaction. And about half of these variables are significant, that is, they are determining factors in satisfaction of using high speed Internet by people in household. More specifically, the five more significant variables (in order of significance) are mobility (t = 5.177, beta = 0.238, p < 0.001), cost (t = 3.839, beta = -0.177, p < 0.001), applications for fun (t = 3.504, beta = 0.218, p < 0.001), age (t = 3.009, beta = 0.245, p < 0.001), and perceived ease of use (t = 2.800, beta = 0.288, p < 0.001). And two other variables are significant at the level of significance required in this study, that is, p [less than or equal to] 0.05. They are fear of technological advances (t = 1.908, beta = -0.101, p < 0.05) as well as self-efficacy (t = 1.646, beta = 0.163, p < 0.05). As shown in Figure 2, our new variable mobility is by far the more significant variable in the global PLS structural model. So, in this study, the fact that high speed Internet allows using only this technology to perform all personal and professional activities is by far the more satisfying factor for Atlantic Canadian people when they choose to get access to high speed Internet from Internet services providers (ISP).

[FIGURE 2 OMITTED]

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 through 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 in information systems (IS) journals, Carte and Russell (2003) found nine errors frequently committed by researchers when they estimate such an effect, and provided solutions (see their paper for more details). We tried to avoid these nine errors in applying their solutions to test hypotheses 1 through 8. Indeed, among others, in the verification of hypotheses 1 through 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 [R.sup.2] calculated before the addition of interaction effect and those calculated after the addition of interaction effect, that is, A[R.sup.2] (named delta [R.sup.2])) is greater than 0.

For a matter of space, as the test of hypotheses 1 through 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, for each PLS model developed, we used the PLS bootstrap resampling procedure with an iteration of 100 sub-sample extracted from the initial sample (322 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 not significant (t = 0.882, beta = -0.181), 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 and age have not an influence on the relationship between applications for personal use and satisfaction of using high speed Internet by people in household. Hypothesis 1 is therefore not supported. The scenario is similar 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 = 0.219, beta = 0.039) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Also, contrary to what we expected, the moderator variable child's age has not an influence on the relationship between utility for children and satisfaction of using high speed Internet 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 = 1.646, beta = 0.457, p < 0.05) and there is a huge change in [R.sup.2] (^[R.sup.2] = 0.015). Thus, as we expected, the moderator variable age has an influence on the relationship between utility for work-related use and satisfaction of using high speed Internet by people in household. Hypothesis 3 is therefore supported. The scenario is similar for 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 significant (t = 1.695, beta = -0.334, p < 0.05) and there is a huge change in [R.sup.2] (^[R.sup.2] = 0.014). Thus, as we expected, the moderator variable age has an influence on the relationship between applications for fun and satisfaction of using high speed Internet by people in household. As a result, hypothesis 4 is also supported. And the scenario is still similar regarding 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 significant (t = 1.712, beta = -0.339, p < 0.05) and there is a huge change in [R.sup.2] (^[R.sup.2] = 0.011). Thus, as we expected, the moderator variable age has an influence on the relationship between status gains and satisfaction of using high speed Internet by people in household. Consequently, hypothesis 5 is also supported.

In the case of hypothesis 6a 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 = 0.894, beta = -0.122), but there is a change in [R.sup.2] (^[R.sup.2] = 0.004). Also, 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 high speed Internet by people in household. Hypothesis 6a is then not supported. The scenario is similar for hypothesis 6b 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 not significant (t = 0.263, beta = 0.041) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Contrary to what we expected, the moderator variables marital status, age, and income have not an influence on the relationship between secondary sources' influences and satisfaction of using high speed Internet by people in household. As a result, hypothesis 6b is not supported. The scenario is also similar concerning hypothesis 6c 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 not significant (t = 0.327, beta = 0.042) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Contrary to our expectations, the moderator variables marital status, age, and income have not an influence on the relationship between workplace referents' influences and satisfaction of using high speed Internet by people in household. Hypothesis 6c is therefore not supported.

As for hypothesis 7a 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 = 0.888, beta = -0.109), but there is a change in [R.sup.2] (^[R.sup.2] = 0.003). 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 high speed Internet by people in household. Hypothesis 7a is therefore not supported. The scenario is similar for hypothesis 7b 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 = 0.434, beta = 0.068) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Also, contrary to what we expected, the moderator variables age and income have not an influence on the relationship between declining cost and satisfaction of using high speed Internet by people in household. Consequently, hypothesis 7b is not supported. And the scenario is also similar for hypothesis 7c 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 = 0.021, beta = 0.003) and there is no change in [R.sup.2] (^[R.sup.2] = 0.000). Thus, contrary to our expectations, the moderator variables age and income have not an influence on the relationship between cost and satisfaction of using high speed Internet by people in household. As a result, hypothesis 7c is not supported.

Finally, concerning hypothesis 8a 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 = 0.801, beta = -0.324), but there is a change in [R.sup.2] (^[R.sup.2] = 0.003). 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 high speed Internet by people in household. As a result, hypothesis 8a is not supported. The scenario is different for hypothesis 8b 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 = 2.412, beta = -1.286, p < 0.01) and there is a huge change in [R.sup.2] (A[R.sup.2] = 0.033). So, as we expected, the moderator variable age has an influence on the relationship between self-efficacy and satisfaction of using high speed Internet by people in household. Consequently, hypothesis 8b is supported. In order to provide the reader with an overall view of the test of hypotheses, Table 6 presents a summary.

In summary, as shown in Table 6, four hypotheses have been supported in our study, that is, H3, H4, H5, and H8b. Thus, the moderator variable age had several moderating effects in this study. As for moderator variables marital status, income, and child's age, these ones had not a significant moderating effect on the relations between the independent and dependent variables involved. Hence hypotheses H1, H2, H6a, H6b, H6c, H7a, H7b, and H7c were not supported. And the moderator variable age had not a significant moderating effect on the relation between perceived ease of use and satisfaction of using high speed Internet at home. Hence hypothesis H8a was not supported.

In the next and last section of the paper, we discuss about the more important findings of the study, the theoretical and practical implications, the limitations, and the future directions.

DISCUSSION AND CONCLUSIONS

This last section is devoted to a discussion about the findings of the study and some conclusions. First, 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 7).

As shown in Table 7 (and Figure 2), the eighteen independent variables examined in the study explained 49.2 percent ([R.sup.2] = 0.492) of the variance in satisfaction of using high speed Internet at home. And we can also see in Table 7 that the seven variables who showed to be significant (see also the significant beta path coefficients in Figure 2), that is, mobility, cost, applications for fun, age, perceived ease of use, fear of technological advances, and self-efficacy explained alone 45.9 percent of the variance in satisfaction of using high speed Internet at home. Thus, these seven variables are assuredly very important factors to take into account in future studies on high speed Internet and on the part of high speed Internet providers, and more particularly self-efficacy and perceived ease of use which explained alone 33 percent of this variance (see Table 7). It is very interesting and surprising here to see that the new variable that we added to the Brown and Venkatesh's (2005) theoretical research model, that is mobility, showed to be the more significant (t = 5.177, beta = 0.238, p < 0.001; see Table 7) in satisfaction of using high speed Internet by people in household. Indeed, the present study showed that people are, to some extent, using high speed Internet for a matter of mobility (e.g., high speed Internet provides them with the possibility to use only this technology to perform all their personal and professional activities). So, here is a new variable that we can now assuredly include in the integrated research model of MATH and household life cycle characteristics suggested by Brown and Venkatesh (2005) as well as Brown et al. (2006) to test in future studies. In fact, we included this new variable mobility in the integrated model of MATH and household life cycle characteristics in several different studies (see Fillion & Berthelot, 2007; Fillion & Le Dinh, 2008; Fillion & Booto Ekionea, 2010b) and it always showed a very significant effect on the dependent variables involved. Of course, its inclusion in the integrated model will depend on its relevance to the technologies examined in the studies. For example, mobility can be included in studies on mobile phone, high speed Internet, or PC, but it cannot be integrated in studies on e-government services, e-learning, or course management software. On the practical point of view, this new variable mobility can be included in the sales marketing plan of high speed Internet 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 PC 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 a new independent variable to the model, that is, mobility. And we also used the household life cycle variables as moderating variables in our research model as did Brown and Venkatesh (2005). Finally, as we investigated the perceptions of people already using high speed Internet at home instead of those having the intention to adopt high speed Internet, as did Brown and Venkatesh (2005) for the PC, then we used the dependent variable user satisfaction instead of behavioral intention. And the model explained 49.2 percent of the variance in satisfaction of using high speed Internet by people in household (see Table 7 and Figure 2). Thus, in this study, using a different dependent variable than did Brown and Venkatesh (2005), that is user satisfaction instead of behavioral intention, our research model explained the same percentage of variance than those explained by MATH alone (e.g., without the household life cycle characteristics and using behavioral 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 independent variable mobility than we included in the present study in which we investigated satisfaction of using high speed Internet at home. And our model explained the same percentage of variance in intention to buy a mobile phone than in the present study in satisfaction of using high speed Internet, that is, 50 percent. According to this finding, we can then see that the variable user satisfaction is as much appropriate as dependent variable in the research model proposed by Brown and Venkatesh (2005) than is behavioral intention. And this finding is also consistent with what is argued by Brown et al. (2006), that is, the model is expected to generalize to other IT products and systems in the household context. However, when the dependent variable of the model is interchanged (e.g., user satisfaction instead of behavioral intention) as did Fillion and Le Dinh (2008) and Fillion and Booto Ekionea (2010b) in studies examining the determining factors in the use of mobile phone, it seems that high speed Internet (the technology involved in the present study) is a more appropriate technology to examine than is mobile phone, since the amount of variance explained by the model in the present study is largely superior, that is, 50 percent comparatively to 32 percent and 35 percent respectively for the two studies on the mobile phone quoted above. Besides, it is to be noted that, in the model we used in this study, less independent variables showed to be good predictors in satisfaction of using high speed Internet by people in household than in the two studies quoted above examining the predictors in satisfaction of using mobile phone by people in household. So, this study brings several interesting findings which contribute to the technology adoption and use literature by offering key insights regarding the differences between adoption, use, and satisfaction of using technology at home.

First, our main findings regarding user satisfaction of a certain technology are consistent with those got in Tao et al.'s (2009) study in the sense that several variables have a significant effect on user satisfaction, but other variables need an improvement on some elements, for example, consumer service, transmission line and connection stability need to be improved in Tao et al.'s (2009) study, while other's usage influences, utility for work-related use as well as applications for personal use need to be improved in our study. Second, we found seven very important variables that seem to be good predictors in satisfaction of using high speed Internet at home, and more particularly perceived ease of use, cost, applications for fun, age, and the new variable that we added to the Brown and Venkatesh's (2005) model, mobility (see Table 7). And the fact that the moderator variable age has been found a very significant predictor (taken as independent variable) in satisfaction of using high speed Internet and a very significant influencing factor (taken as moderator variable) in all hypotheses supported in the study provides additional evidence concerning the importance of integrating household life cycle stage in research examining household technology adoption and use. These seven variables are also important to take into account by high speed Internet providers in order to improve actual services, to offer new services still better adapted to people's needs, as well as to perform their sales marketing. Third, we found that people are, to some extent, using high speed Internet for a matter of mobility given our new variable mobility showed to be the more significant in the study (see Table 7). Fourth, we found that, depending on the technology studied, the dependent variables behavioral intention and user satisfaction might be interchanged in the model proposed by Brown and Venkatesh (2005) given the amount of variance explained by the models are quite varying across technologies and dependent variables observed. The dependent variable use behavior proposed by Thompson et al. (1991) and the dependent variable user satisfaction (examined in the present study) conceptualized in the work of Cyert and March (1963), and initially developed by Ives et al. (1983), may also be further tested in future studies. And, finally, we suggest the test of new independent variables which may explain a greater amount of variance in satisfaction of using high speed Internet by people in household in future studies. To that end, we recommend three new independent variables in the next paragraph.

Indeed, depending on the technology examined, it would be interesting in future studies to add a variable such as utility for security (in utilitarian outcomes) to the theoretical research model suggested by Brown and Venkatesh (2005) augmented with the new variable mobility that we tested in this study. This variable has been found very significant in the case of mobile phone technology in the studies conducted by Fillion and Berthelot (2007), Fillion and Le Dinh (2008), as well as Fillion and Booto Ekionea (2010b). Who knows, people might be also using high speed Internet for a matter of their own security and those of their family given this technology allows to rapidly communicate with helping people or organisms everywhere in the world. The variable social norm might be also added in social outcomes. Who knows, people might be using high speed Internet just to do as everybody. And the variable provider support might be added in external control beliefs. People might be according a great importance to the quality of support offered by the high speed Internet provider. It would be also interesting to test the actual model in other situations and with other populations. For example, with colleagues from Brasil (University of Lavras) and Cameroon (University of Yaounde I), we are now testing the actual model with people who are using a mobile phone at home. As in this study, we used the dependent variable user satisfaction since the respondents are already using a mobile phone. The results of these studies will follow in subsequent papers. It will be interesting to see whether the results remain the same as those got from people who are using high speed Internet in household.

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 of using high speed Internet 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. So, rest assured that 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

Al-Omoush, K.S. & A.A. Shaqrah (2010). An empirical study of household Internet continuance adoption among Jordanian users. International Journal of Computer Science and Network Security, 10(1), 32-44.

Anderson, B. (2008). The social impact of broadband household Internet access. Information, Communication, & Society, 11(1), 5-24.

Barclay, D., C. Higgins & R. Thompson (1995). The partial least squares (PLS) approach to causal modeling, personal computer adoption and use as an illustration. Technology Studies, 2(2), 285-309.

Brown, S.A., V. Venkatesh & H. Bala (2006). Household technology use: Integrating household life cycle and the model of adoption of technology in households. The Information Society, 22, 205-218.

Brown, S.A. & V. Venkatesh (2005). Model of adoption of technology in households: A baseline model test and extension incorporating household life cycle. MIS Quarterly, 29(3), 399-426.

Cambini, C. & Y. Jiang (2009). Broadband investment and regulation: A literature review. Telecommunications Policy, 33, 559-574.

Carte, T.A. & C.J. Russell (2003). In pursuit of moderation: Nine common errors and their solutions. MIS Quarterly, 27(3), 479-501.

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., B.L. Marcolin & P.R. Newsted (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.

Compeau, D.R. & C.A. Higgins (1995a). Application of social cognitive theory to training for computer skills. Information Systems Research, 6(2), 118-143.

Compeau, D.R. & C.A. Higgins (1995b). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19(2), 189-211.

CRACIN (2005). Written Submission to Telecommunications Policy Review Panel. Toronto: Canadian Research Alliance for Community Innovation & Networking.

Cyert, R.M. & J.G. March (1963). A Behavioral Theory of the Firm. Englewood Cliffs, NJ: Prentice-Hall.

Danko, W.D. & C.M. Schaninger (1990). An empirical evaluation of the Gilly-Enis updated household life cycle model. Journal of Business Research, 21, 39-57.

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.

Digital Home (2009). Canada ranked fifth in broadband penetration. Retrieved August 6, 2010, from http://www.digitalhome.ca/ 2009/09/canada-ranked-fifth-in-broadband- penetration/.

Dumitru, R.C., T. Burkle, S. Potapov, B. Lausen, B. Wiese & H.-U. Prokosch (2007). Use and perception of Internet for health related purposes in Germany: Results of a national survey. International Journal of Public Health, 52, 275-285.

Dunn, J. (2010). Digital home thoughts: High-speed Internet access in Canada: It's expensive and slow. Digital Home Thoughts, http://www.digitalhomethoughts.com/news/show/97187/high- speed-internet-access-incanada. Retrieved August 9, 2010.

Efron, B. & R.J. Tibshirani (1993). An Introduction to the Bootstrap. New York: Chapman and Hall.

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. (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., M. Limayem, T. Laferriere & R. Mantha (2010a). Onsite and online students' and professors' perceptions of ICT use in higher education. In N. Karacapilidis (Ed.), Novel Developments in Web-Based Learning Technologies: Tools for Modern Teaching (Chapter 6), Hershey, PA: IGI Global Publishing.

Fillion, G. & J.-P. Booto Ekionea (2010b). Testing a moderator-type research model on the use of mobile phone. Forthcoming in Academy of Information and Management Sciences Journal, 13.

Fillion, G. & T. Le Dinh (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.

Fillion, G. & S. Berthelot (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.

Fuhr, J.P. & S.B. Pociask (2007). Broadband services: Economic and environmental benefits. Retrieved August 9, 2010, from, http://www.theamericanconsumer.org/2007/10/31/broadband-services- economic-andenvironmental-benefits/.

Gill, K.E. (2010). What does 'high speed Internet' mean exactly?. Retrieved August 6, 2010, from http://wiredpen.com/2010/ 03/26/what-does-high-speed-internet-mean-exactly/.

Government of Canada (1999). Speech from the throne to open the second session of the 36th parliament of Canada. http://www.pco-bcp.gc.ca/ default.asp?Language=E&Page=InformationResources&sub=sftddt&doc= sftddt1999_e.htm. Retrieved August 6, 2010, from

Government On-Line Advisory Panel (2003). Connecting with Canadians: Pursuing Service Transformation. Ottawa: Treasury Board of Canada.

Helsper, E.(2010). Gendered Internet use across generations and life stages. Communication Research, 37(3), 352-374.

Hobbs, V.M. & D.D. Osburn (1989). Distance Learning Evaluation Study Report II: A Study of North Dakota and Missouri Schools Implementing German I by Satellite. ERIC ED 317 195.

Horrigan, J. (2009). Home Broadband Adoption 2009. Report for the Pew Internet & American Life Project, Retrieved August 6, 2010, from http://pewinternet.org/Reports/2009/10-Home-Broadband-Adoption2009.aspx. [Accessed 6 August 2010]

Howard, P.N. & N. Mahazeri (2009). Telecommunications reform, Internet use and mobile phone adoption in the developing world. World Development, 37(7), 1159-1169.

Ida, T. & K. Sakahira (2008). Broadband migration and lock-in effects: Mixed logit model analysis of Japan's high speed Internet access services. Telecommunications Policy, 32, 615-625.

Ives, B., M.H. Olson & J.J. Baroudi (1983). The measurement of user information satisfaction. Communications of the ACM, 26(10), 785-793.

Jones, S. & S. Fox (2009). Generations Online in 2009. Report for the Pew Internet & American Life Project, January 28, 2009.

Kwon, H.S. & L. Chidambaram (2000). A test of the technology acceptance model: The case of cellular telephone adoption. Proceedings of HICSS-34, Hawaii, January 3-6.

Labriet-Gross, H. (2007). High speed Internet: Can broadband save the planet?. L'Atelier North America, Retrieved August 9, 2010, from http://www.atelier-us.com/internet- usage/article/high-speed-internet-can-broadbandsave-the-planet.

Majumdar, S.K. (2008). Broadband adoption, jobs and wages in the US telecommunications industry. Telecommunications Policy, 32, 587-599.

Matthews, D. & L. Schrum (2003). High-speed Internet use and academic gratifications in the college residence. The Internet and Higher Education, 6(2), 125-144.

Middleton, C.A. & J. Ellison (2006). All Broadband Households Are Not the Same: Why Scope and Intensity of Use Matter. Report for Ryerson University, Toronto.

Moore, G.C. & I. Benbasat (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, 2(3), 192-222.

National Broadband Task Force (2001). The New National Dream: Networking the Nation for Broadband Access. Ottawa: Industry Canada.

Orazem, P.F. (2005). The Impact of High-Speed Internet Access on Local Economic Growth. Research Report Prepared for University of Kansas School of Business, Topeka, Kansas.

Perry, T.T., L.A. Perry & K. Hosack-Curlin (1998). Internet use by university students: An interdisciplinary study on three campuses. Internet Research, 8(2), 136-141.

Platt, R.G., W.B. Carper & M. McCool (2010). Outsourcing a high speed Internet access project: An information technology class case study in three parts. Journal of Information Systems Education, 21(1), 15-25.

Rains, S.A. (2008). Health at high speed: Broadband Internet access, health communication, and the digital divide. Communication Research, 35(3), 283-297.

Rosston, G., S.J. Savage & D.M. Waldman (2010). Household Demand for Broadband Internet Service. Final Report for the Broadband.gov Task Force, Federal Communications Commission (FCC).

Selouani, S.. & H. Hamam (2007). Social impact of broadband Internet: A case study in the Shippagan area, a rural zone in Atlantic Canada. Journal of Information, Information Technology, and Organizations, 2, 79-94.

Straub, D.W. (1989). Validating instruments in MIS research. MIS Quarterly, 13(2), 147-169.

Tao, C.J., S.C. Chen & L. Chang (2009). Apply 6-Sigma methodology in measuring the competition quality of satisfaction performance--An example of ISP industry. Quality & Quantity, 43, 677-694.

Taylor, S. & P.A. Todd (1995). Understanding information technology usage: A test of competing models. Information Systems Research, 6(2), 144-176.

Telecommunications Policy Review Panel (2006). Telecommunications Policy Review Panel--Final Report 2006. Ottawa: Industry Canada.

Thompson, R.L., C.A. Higgins & J.M. Howell (1991). Personal computing: Toward a conceptual model of utilization. MIS Quarterly, 15(1), 124-143.

Varon, E. (2010). Testing the benefits of high-speed Internet access. Appeared on NetworkWorld, Retrieved August 9, 2010, from http://www.netorkworld.com/news/2010/051710- testing-the-benefits-of-high-speed.html.

Venkatesh, V. & S.A. Brown (2001). A longitudinal investigation of personal computers in homes: Adoption determinants and emerging challenges. MIS Quarterly, 25(1), 71-102.

Wagner, J. & S. Hanna (1983). The effectiveness of family life cycle variables in consumer expenditure research. Journal of Consumer Research, 10, 281-291.

Webster, J. & J.J. Martocchio (1993). Turning work into play: Implications for microcomputer software training. Journal of Management, 19(1), 127-146.

Webster, J. & J.J. Martocchio (1992). Microcomputer playfulness: Development of a measure with work place implications. MIS Quarterly, 16(2), 201-226.

Windhausen Jr., J. (2008). A Blueprint for Big Broadband. Report for EDUCAUSE, Washington D.C., Retrieved August 9, 2010, from http://www.educause.edu.

Yoo, Y. & M. Alavi (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: Generations Explained
(adapted from Jones & Fox, 2009, p. 1)

Generation       Birth Years,       % of Total        % of
Names            Ages in 2009         Adult      Internet-Using
                                    Population     Population

Gen Y            Born 1977-1990,        26             30
(millennials)    Ages 18-32

Gen X            Born 1965-1976,        20             23
                 Ages 33-44

Younger          Born 1955-1964,        20             22
Boomers          Ages 45-54

Older Boomers    Born 1946-1954,        13             13
                 Ages 55-63

Silent           Born 1937-1945,        9               7
Generation       Ages 64-72

G.I.             Born -1936,            9               4
Generation       Age 73+

Source: Pew Research Center's Internet & American
Life Project December 2008 survey. N = 2,253 total
adults, and margin of error is [+ or -] 2%. N =
1,650 total Internet users, and margin of error is
[+ or -] 3%.

Table 2: Main High Speed or Broadband Technologies
(adapted from Gill, 2010, pp. 3-4)

Technologies    Speeds (1)

Cable           Basic: 4 Mbps to 6 Mbps
                High End: 12 Mbps to 16 Mbps and faster

DSL             Basic: 768 Kbps to 1.5 Mbps
                High End: 3 Mbps to 7 Mbps

Fiber Optic     15 Mbps to 25 Mbps
Cable

Mobile--EDGE    Up to 58 Kbps, average 22 Kbps

Mobile--3G      AT&T: Download: 700 Kbps to 1.7 Mbps;
                  Upload: 500 Kbps to 1.2 Mbps
                Sprint: Download: 600 Kbps to 1.4 Mbps
                Verizon: 600 Kbps to 1.4 Mbps

Mobile--4G      Download: 3 to 6 Mbps

Satellite       10 to 20 Kbps

WiMax           Download: 3 to 6 Mbps
(like Clear)

South Korea     1 Gbps (2012)

Japan           Average: 93.6 Mbps (2007)

France          Average: 44.1 Mbps (2007)

(1) One kilobit per second (Kbps) is 1,000 bits
per second (bps). One megabit per second (Mbps) is
1,000 Kbps or 1,000,000 bps. One gigabit per
second (Gbps) is 1,000 Mbps or 1,000,000 Kbps or
1,000,000,000 bps.

Table 3: Related Literature Survey

Research Area                              References

High speed Internet social impact and      Orazem (2005)
economic growth.                           Selouani & Hamam (2007)
                                           Anderson (2008)

High speed Internet and wages and          Majumdar (2008)
employment.

High speed Internet and health.            Dumitru et al. (2007)
                                           Rains (2008)

High speed Internet and regulation.        Cambini & Jiang (2009)
                                           Howard & Mahazeri (2009)

High speed Internet migration,             Ida & Sakahira (2008)
implementation, and support.               Platt et al. (2010)

High speed Internet adoption and use.      Perry et al. (1998)
                                           Matthews & Schrum (2003)
                                           Middleton & Ellison (2006)
                                           Dumitru et al. (2007)
                                           Windhausen Jr. (2008)
                                           Horrigan (2009)
                                           Howard & Mahazeri (2009)
                                           Al-Omoush & Shaqrah (2010)
                                           Helsper (2010)
                                           Rosston et al. (2010)

High speed Internet (e.g., ISP) user       Tao et al. (2009)
satisfaction.

Table 4: Variables and Definitions

Beliefs and           Variables               Definitions
Characteristics

Dependent         User Satisfaction    According to Cyert and March
Variable                               (1963, p. 126), an information
                                       system or information
                                       technology which meets the
                                       needs of its user will
                                       reinforce satisfaction with
                                       that system or technology. If
                                       the system or technology does
                                       not provide the needed
                                       information or service, the
                                       user will become dissatisfied
                                       and look elsewhere.

Attitudinal       Applications for     The extent to which using high
Beliefs           Personal Use         speed Internet enhances the
(independent                           effectiveness of household
variables)                             activities (Venkatesh & Brown,
                                       2001).

                  Utility for          The extent to which using high
                  Children             speed Internet enhances the
                                       children's effectiveness in
                                       their activities (Venkatesh &
                                       Brown, 2001).

                  Utility for          The extent to which using high
                  Work-Related Use     speed Internet enhances the
                                       effectiveness of performing
                                       work-related activities
                                       (Venkatesh & Brown, 2001).

                  Mobility             The extent to which high speed
                                       Internet allows using only
                                       this technology to perform all
                                       personal and professional
                                       activities.

                  Applications for     The pleasure derived from high
                  Fun                  speed Internet use (Venkatesh
                                       & Brown, 2001). These are
                                       specific to high speed
                                       Internet usage, rather than
                                       general traits (Brown &
                                       Venkatesh, 2005; see Webster &
                                       Martocchio, 1992, 1993).

                  Status Gains         The increase in prestige that
                                       coincides with the purchase of
                                       high speed Internet access for
                                       home use (Venkatesh & Brown,
                                       2001).

Normative         Friends and Family   "The extent to which the
Beliefs           Influences           members of a social network
(independent                           influence one another's
variables)                             behavior" (Venkatesh & Brown,
                  Secondary Sources'   2001, p. 82). In this case,
                  Influences           the members are friends and
                                       family (Brown & Venkatesh,
                                       2005). The extent to which
                                       information from TV,
                                       newspaper, and other secondary
                                       sources influences behavior
                                       (Venkatesh & Brown, 2001).

                  Workplace            The extent to which coworkers
                  Referents'           influence behavior (Brown &
                  Influences           Venkatesh, 2005; see Taylor &
                                       Todd, 1995).

Control Beliefs   Fear of              The extent to which rapidly
(independent      Technological        changing technology is
variables)        Advances             associated with fear of
                                       obsolescence or apprehension
                                       regarding high speed Internet
                                       access purchase (Venkatesh &
                                       Brown, 2001).

                  Declining Cost       The extent to which the cost
                                       of high speed Internet access
                                       is decreasing in such a way
                                       that it inhibits adoption
                                       (Venkatesh & Brown, 2001).

                  Cost                 The extent to which the
                                       current cost of high speed
                                       Internet access is too high
                                       (Venkatesh & Brown, 2001).

                  Perceived Ease       The degree to which using high
                  of Use               speed Internet is free from
                                       effort (Davis, 1989; see also
                                       Venkatesh & Brown, 2001).

                  Self-Efficacy        The individual's belief that
                  (or Requisite        he-she has the knowledge
                  Knowledge)           necessary to use high speed
                                       Internet. This is closely tied
                                       to computer self-efficacy
                                       (Compeau & Higgins, 1995a,
                                       1995b; see also Venkatesh &
                                       Brown, 2001).

Life Cycle        Income               The individual's year gross
Characteristics                        income (see Wagner & Hanna,
(moderator                             1983).
variables)
                  Marital Status       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 5: Means, Standard Deviations, Composite
Reliability Indexes, Correlations, and Average
Variance Extracted of Variables

Variables             M      SD     Reliability      Correlations
                                       Index            and
                                                       Average
                                                      Variance
                                                     Extracted (e)

                                                     1       2

1.  Applications    5.17    1.87        0.82       0.78
    for Personal
    Use

2.  Utility fo      3.07    3.00        0.99        .13    0.98
    Children

3.  Utility for
    Work-Related    4.47    2.63        0.89        .21     .12
    Use

4.  Mobility        4.70    2.63        0.89        .11     .17

5.  Applications    4.83    1.95        0.86        .29     .09
    For Fun

6.  Status Gains    2.88    2.08        0.94        .08     .06

7.  Friends and
    Family          3.75    2.48        0.93        .06     .01
    Influences

8.  Secondary
    Sources'        3.73    2.20        0.85        .12     .09
    Influences

9.  Workplace
    Referents'      3.15    3.00        0.91        .15     .07
    Influences

10. Fear of
    Technological   2.87    2.07        0.87        .03     .11
    Advances

11. Declining       3.93    2.07        0.86        .16     .08
    Cost

12. Cost            4.77    1.93        0.74       -.22    -.07

13. Perceived
    Ease            5.73    1.45        0.83        .30    -.04
    of Use

14. Self-Efficacy   6.37    1.05        0.93        .29     .01

15. Income (a)       NA      NA          NA         .10     .25

16. Marital          NA      NA          NA        -.07    -.29
    Status (b)

17. Age (c)         40.00   13.70        NA        -.05     .08

18. Child's         14.91   8.89         NA        -.06    -.20
    Age (d)

19. User            5.72    1.38        0.90        .28     .06
    Satisfaction

Variables                            Correlations
                                         and
                                       Average
                                      Variance
                                     Extracted (e)

                      3       4       5       6       7       8

1.  Applications
    for Personal
    Use

2.  Utility fo
    Children

3.  Utility for
    Work-Related    0.85
    Use

4.  Mobility         .28    0.85

5.  Applications    -.02     .30    0.78
    For Fun

6.  Status Gains     .11     .17     .29    0.92

7.  Friends and
    Family           .10     .20     .27     .36    0.87
    Influences

8.  Secondary
    Sources'         .05     .20     .37     .26     .39    0.81
    Influences

9.  Workplace
    Referents'       .29     .19     .11     .21     .42     .27
    Influences

10. Fear of
    Technological    .09    -.04     .01     .13     .16     .11
    Advances

11. Declining        .20     .22     .13     .16     .14     .10
    Cost

12. Cost            -.19    -.11    -.11    -.12    -.15    -.12

13. Perceived
    Ease             .12     .25     .26     .10     .07     .17
    of Use

14. Self-Efficacy    .11     .24     .20     .07     .04     .11

15. Income (a)       .13     .01    -.20    -.11    -.10    -.09

16. Marital         -.01     .16     .13     .05     .09     .11
    Status (b)

17. Age (c)         -.07    -.28    -.22    -.05    -.08    -.18

18. Child's         -.12    -.20    -.07    -.01     .01    -.08
    Age (d)

19. User             .13     .38     .39     .18     .15     .17
    Satisfaction

Variables                            Correlations
                                         and
                                       Average
                                      Variance
                                     Extracted (e)

                      9      10      11      12      13      14

1.  Applications
    for Personal
    Use

2.  Utility fo
    Children

3.  Utility for
    Work-Related
    Use

4.  Mobility

5.  Applications
    For Fun

6.  Status Gains

7.  Friends and
    Family
    Influences

8.  Secondary
    Sources'
    Influences

9.  Workplace
    Referents'      0.92
    Influences

10. Fear of
    Technological    .06    0.83
    Advances

11. Declining        .15    -.01    0.83
    Cost

12. Cost            -.16     .18    -.29    0.70

13. Perceived
    Ease             .14    -.26     .13    -.23    0.76
    of Use

14. Self-Efficacy    .08    -.25     .13    -.18     .47    0.90

15. Income (a)       .04    -.12     .03    -.06     .04     .05

16. Marital          .14    -.13    -.12    -.05     .12     .06
    Status (b)

17. Age (c)         -.30     .19    -.06     .15    -.22    -.26

18. Child's         -.20     .07    -.08     .11    -.13    -.20
    Age (d)

19. User             .11    -.20     .20    -.32     .56     .49
    Satisfaction

Variables                        Correlations
                                     and
                                  Average
                                  Variance
                                Extracted (e)

                     15      16      17      18      19

1.  Applications
    for Personal
    Use

2.  Utility fo
    Children

3.  Utility for
    Work-Related
    Use

4.  Mobility

5.  Applications
    For Fun

6.  Status Gains

7.  Friends and
    Family
    Influences

8.  Secondary
    Sources'
    Influences

9.  Workplace
    Referents'
    Influences

10. Fear of
    Technological
    Advances

11. Declining
    Cost

12. Cost

13. Perceived
    Ease
    of Use

14. Self-Efficacy

15. Income (a)       NA

16. Marital         -.26     NA
    Status (b)

17. Age (c)          .28    -.27     NA

18. Child's         -.02    -.05     .34     NA
    Age (d)

19. User             .03     .02    -.08    -.07    0.77
    Satisfaction

(a) This variable was coded as an ordinal
variable. It was measured in terms of non
quantified distinct ordered categories.

(b) This variable was coded as a nominal variable.
It was measured in terms of non quantified
distinct categories.

(c) This variable was coded as a continuous
variable. It was measured using the respondents'
birth date.

(d) This variable was coded as a numeral. It was
measured using the age of the respondents'
youngest child.

(e) 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.

NA: Not applicable.

Table 6: Summary of the Test of Hypotheses

Hypotheses                       Results             Software
                                                     (beta sig.)

H1-Marital status and age will   Not supported       PLS (-0.181)
moderate the relationship
between applications for
personal use and satisfaction
of using high speed Internet
at home.

H2-Child's age will moderate     Not supported       PLS (0.039)
the relationship between
utility for children and
satisfaction of using high
speed Internet at home.

H3-Age will moderate the         Supported           PLS (0.457*)
relationship between utility
for work-related use and
satisfaction of using high
speed Internet at home.

H4-Age will moderate the         Supported           PLS (-0.334*)
relationship between
applications for fun and
satisfaction of using high
speed Internet at home.

H5-Age will moderate the         Supported           PLS (-0.339*)
relationship between status
gains and satisfaction of
using high speed Internet at
home.

H6-Age, marital status, and      a- Not supported    PLS (-0.122)
income will moderate the         b- Not supported    PLS (0.041)
relationship between the         c- Not supported    PLS (0.042)
normative beliefs  ((a)
friends and family influences;
(b)  secondary  sources'
influences;  and (c) workplace
referents' influences) and
satisfaction of using high
speed Internet at home.

H7-Age and income will           a- Not supported    PLS (-0.109)
moderate the relationship        b- Not supported    PLS (0.068)
between the external control     c- Not supported    PLS (0.003)
beliefs ((a) fear of
technological advances; (b)
declining cost; and (c) cost)
and satisfaction of using high
speed Internet at home.

H8-Age will moderate the         a- Not supported    PLS (-0.324)
relationship between the         b- Supported        PLS (-1.286**)
internal control beliefs ((a)
perceived ease of use; and (b)
self-efficacy) and
satisfaction of using high
speed Internet at home.

* p < 0.05; ** p < 0.01.

Table 7: Beta Path Coefficients, T-Values, and
Percentages of Variance Explained of Variables

Variables                       Beta         t-values    [R.sup.2]
                            Coefficients    (one-tail)

Applications for Personal       0.032         0.566        0.001
Use

Utility for Children           -0.031         0.729        0.001

Utility for Work-Related       -0.016         0.317        0.000
Use

Mobility                     0 238 ****       5.177        0.037

Applications for Fun         0.218 ****       3.504        0.047

Status Gains                    0.036         0.823        0.004

Friends and Family              0.005         0.117        0.004
Influences

Secondary Sources'             -0.017         0.380        0.005
Influences

Workplace Referents'            0.028         0.588        0.002
Influences

Fear of Technological          -0.101*        1.908        0.003
Advances

Declining Cost                  0.012         0.258        0.005

Cost                         -0 177 ****      3.839        0.038

Perceived Ease of Use        0.288 ****       2.800        0.081

Self-Efficacy                  0.163 *        1.646        0.249

Income                         -0.041         0.819        0.006

Marital Status                 -0.016         0.300        0.000

Age                          0 245 ****       3.009        0.004

Child's Age                    -0.066         0.903        0.005

* p < 0.05; **** p < 0.001.
COPYRIGHT 2012 The DreamCatchers Group, LLC
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fillion, Gerard; Ekionea, Jean-Pierre Booto
Publication:Academy of Information and Management Sciences Journal
Article Type:Report
Geographic Code:1CANA
Date:Jan 1, 2012
Words:12039
Previous Article:Predicting information technology adoption in small businesses: an extension of the technology acceptance model.
Next Article:Multi-factor user interface components layout problem with genetic algorithm.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters