A New Approach to Understanding the Digital Divide : Why two levels are not enough.

Todd Marshall
School of Information Studies
Syracuse University
245 Hinds HallSyracuse, NY 13244

Contextual Background
Over the past few decades, new information and communication technologies (ICT) have made inroads into business, education, and personal lives. The present generation has seen computing technology transition from the mainframe to the mobile phone. The technical capabilities of these systems that were originally only available to highly trained professionals are now in the hands of children. Technology is affecting how we communicate, store, retrieve, and process the information in virtually every aspect of our lives. During these decades of transition, technological possibilities have been increasing faster than our ability to understand their impact on the people using them. For various reasons, certain individuals and segments of society have not kept pace with these developments. Gaps have developed between the “haves” and the “have nots” as well as between the “users” and “non users.” Why have these gaps developed? How can we understand them? Are they predictable?
The purpose of this paper is to propose a new framework, the Behavioral Adoption Framework, to further our understanding of these gaps and the differentiations in adoption that have caused them. I am using the term “framework” as opposed to “model” because the factors I discuss are generic in nature and not specifically defined. Before this framework can be examined, it is necessary to explain the historical context and development of previous frameworks, models, and theories. Basic attempts to answer these questions about adoption have developed in two streams of academia. These streams are found in Informations Systems and education. It is my assumption that these two streams of investigation are not mutually exclusive, but are actually converging. My goal is to combine the strengths of each approach and synthesize a new framework that is broad enough to address technology adoption in any context without loosing its explanatory power. Why is a new framework necessary? One might argue that current theory is sufficient to explain this phenomenon. Therefore, discussion will begin with a survey of current theories which originated in the discipline of Information Systems (IS) during the 1970s and 1980s. The discussion will include an examination of previous models and theories and then a description of the proposed framework. Because the digital divide perspective is newer, less developed, and lacking the theorietical frameworks of the IS models, it will receive only enough attention to orient the reader to the fundamentals of that perspective.

Adoption from the TAM Perspective
The first stream of academic discussion is on the concept of adoption as usage. This has developed within the field of Information Systems and the business community. The goal here is the successful usage of technology. It generally assumes that the user already has access and concentrates on factors which affect intentions to use the technology already in hand. This began with the TAM, Technology Acceptance Model (Davis, 1989), and the TRA, Theory of Reasoned Action (Ajzen and Fishbein, 1973). This theory and its accompanying models have an older lineage than the digital divide discussion and also a more limited context for usage. Over time, scholars have added additional factors to increase accuracy of these theories, but they still have significant weaknesses (Venkatesh et al., 2003). Later discussions will examine this approach in detail.

Current Theories
TRA/TAM/TRB
The Technology Acceptance Model (TAM) is one of the most prominent theories in Information Systems (IS). The basic premise is that “perceived ease of use” and “perceived usefulness” combine to influence “behavioral intention” which in turn affects actual system “usage” (Davis, 1989). TAM is based on the Theory of Reasoned Action (TRA) (Ajzen and Fishbein, 1973). Another variant of TAM built on TRA is the Theory of Planned Behavior (TPB) (Azjen, 1985; Azjen, 1991). The fundamental assumption of both TRA and TPB is that attitude toward an act or behavior and perceived behavior control are the main factors affecting behavioral intention and ultimately behavior itself. The focus is on the user’s beliefs and attitudes. TBP also adds the factor of a subjective norm. TPB and TRA both focus on the factors which lead to intentional behavior. These theories are not identical, but use the same logic and same general approach predicting behavior. Figure 1 by Wixom and Todd (2005) represents this graphically.

Figure 1

It shows that how a person thinks about a behavior leads to intention to use. For the sake of brevity, these collected theories will be referred to generically as TAM/2.
TAM, as indicated by the name, focuses on the acceptance of new technology which leads to actual usage. TAM has been tested extensively and demonstrated significant results but has been subject to numerous revisions including TAM2 (Davis and Venkatesh, 2000). In this case, the originator of the TAM model realized that there were other significant factors external to the user and the user’s perceptions of usefulness or ease of use which did affect usage. TAM2 adds “social influence processes (subjective norm, voluntariness, and image) and cognitive instrumental processes (job relevance, output quality, result demonstrability, and perceived ease of use)” (Venkatesh and Davis, 2000, p. 187). This is a step in the right direction and attempts to provide a richer picture of factors affecting acceptance. However it is still not complete. For example, someone could perceive a system as not useful, but use it anyway because subjective norms require usage. In fact, Venkatesh and Davis found this very disconnect in their study. “Subjective norm” had a negative correlation (-0.047, p<.001) with “perceived usefulness” but a positive correlation (0.44, p<.001) with “intention to use” (2000, p. 197). Additionally, issues of access, ability, and policy which might hinder participation are not addressed by TAM/2. TAM/2 is designed for measuring usage in the workplace. In such a context, there are assumptions which would not apply outside a workplace context. Employees are assumed to be trained, given adequate equipment, and can be required or even forced to use a system. Many, if not most, of the most widely used technological systems are voluntary and outside the work environment. Personal usage may require upgrading one’s own hardware, self-training, and a strictly personal choice. Finally, TAM and TAM2 (together referred to as TAM/2) focus on prediction of usage and do not address barriers to usage. It is as if it assumes that usage will happen. TAM/2 is like a one-way street headed toward usage. The theory necessary to explain today’s context of usage should be a two-way street that explains factors which both enable and hinder usage. For these reasons, the TAM is not a sufficient theory to explain general technology usage.

UTAUT
UTAUT stands for Unified Theory of Acceptance and Use of Technology. UTAUT is built on the same basic model, but expands the number of factors leading to behavioral intention and use behavior (Venkatesh et al., 2003). The purpose for addressing this model separately is that it significantly expands the factors which lead to use. It includes: performance expectancy, effort expectancy, social influence, facilitating conditions, gender, age, experience, and voluntariness of use (Figure 2) (Venkatesh et al., 2003). While these factors continue to be a step in the right direction, it is this author’s opinion that they do not address enough factors when attempting to explain IS systems outside the workplace. The main focus for everything becomes behavioral intention.

Figure 2

There has proven to be a high correlation between intention and use, but one may ask “How is this really useful?” There is a definite tendency for people to do what they intended to do. The growing complexity of “everything before intention” seems to indicate a fundamental weakness of this basic focus on intention based model of usage. As the saying goes, “To the man who has a hammer, the whole world becomes a nail.” In this case, “To theories based on TRA, everything affects intention.” It would seem prudent to take a step back and look at those factors which affect intention. UTAUT purports to account for “as much as 70% of the variance in user intention,” however “future research should focus on identifying constructs that can add to the predication of intention and behavior over and above what is already known and understood” (Venkatesh et al., 2003). However, the solution may require going beyond intention all together.

Other Models
Several other recent models have attempted to modify TAM/2 to account for the newer, more complex user and usage environment. There are dozens of TAM/2 variants of various qualities which seek to incorporate various factors and constructs. Two were chosen specifically for their attempt to adapt TAM/2 to the complexities of more real-world environments.
The first example (Figure 3) crosses over from usage to participation (Yoo et al., 2002).

Figure 3

However, the similarity to TAM/2 is obvious. This model simply replaces one item with another: “perceived usefulness” with “managing strategy,” “perceived ease of use” with “IS quality,” “intention” with “sense of community,” “attitude” with “visit,” and “usage behavior” with “participation.” Participation is certainly a broader conceptthan behavioral usage and community aspects are vital, but this model still follows the original TAM/2 model.
The next model by Dholakia et al. (Figure 4) also tries to expand the TAM/2 model to include a richer representation of value perceptions and social influence variables (2004). This model shows progress in that it also steps outside of a work/job environment. Elements such as self-discovery, social enhancement, and entertainment value do begin to broaden the scope of usage to account for voluntary systems. However, there is a certain artificiality and seeming randomness to the directions and connections which represent causality.

Figure 4

Classification of Factors
As a first step toward proposing a new model, it would be appropriate to evaluate the current models and see which factors they include to begin to create a classification scheme or taxonomy of concepts. The following classifications are general in nature. The task of classifying factors is complicated by authors’ use of identical terms with separate meanings. This list of articles comes from an exhaustive article by Venkatesh et al. (2005) which surveys the major TAM/2 modifications from 1989 to 2003 and proposes the UTAUT model. The list of theories is given by Venkatesh et al., but the categorization of factors is original (Figure 5). The purpose is to see which factors have been addressed by TAM/2 developments and which factors may still be lacking. The four categories are from the Behavioral Adoption Model proposed in this paper. It is interesting to note that these models focus on personal factors because the focus is on the personal intention. However, as the diagrams and historical progression have demonstrated, this has not been satisfactory so subsequent research has had to chronologically backtrack. In reverse chronological order there is behavior, intention, attitude, and “other factors” (personal, contextual, etc.).

Figure 5

The next chart (Figure 6) includes some of the other factors which have been added to the “classic” theories and models given above. As one can see, there is greater diversity in the utilization of both the vocabulary and the categories.

Figure 6

While there is still a definite orientation to a work environment, especially in terms of technology, there is a significant increase in personal competence and cultural values. One can also see from figures 5 and 6 is that most models and theories do not address all for sets of factors: technology, competency, cultural values, and personal values. They have emphasized peronsal values and cultural values with only relatively recent consideration of technology and competency. What is needed at this point is a framework which explains factors that influence motivation and ability and move beyond the relationship between intention and action.

Adoption from the Digital Divide Perspective
The second stream of academic research on adoptiong focuses on access and has developed into discussions around the “digital divide” (Warschauer, 2002). The digital divide perspective on adoption began with discussions about access to ICT. Much of the attention was on access to broadband as well as socioeconomic and political issues. Investigation then shifted to usage skills which are referred to as information, digital, or technical literacy. These are the first and second levels of the digital divide.
The original concept was that the primary barrier to adoption was access. Initial digital divide discussions were about whether schools had sufficient access to ICT. It was couched in terms of “haves” and “have nots.” This led to the push for increased access to the internet in public schools, universities, and libraries (Hargitai, 2002; Warschauer, 2002). The assumption was that if people just had access, then they would take advantage of the opportunity and adopt these new technologies. However, it has been demonstrated that having access does not guarantee usage (Crump and McIlroy, 2003). Bridging the access gap still did not create adoption and integreation (Crump, 2003). Possession of ICT did not guarantee adoption of ICT.
Following the “bridging” of the first level digital divide, attention shifted to information literacy skills and the educational component of adoption. Access alone was deemed insufficient (Hargittai, 2002; Waschauer, 2002). This conclusion has led to an emphasis on technical competencies and educational initiatives (Dewan and Riggins, 2005). Because of progress in briding the first level divide over the past decade, ICT access is commonplace in schools, public libraries, and many home settings. Most of the academic work at this time is investigating competencies and education so that students will know how to take full advantage of this technology.
This is basically where the conversation stands at this point. Most studies have been weak on theory and demonstrated little explanatory power (Becker, 1999). I propose that is due to issues related to policy, culture, values, and beliefs that needed to be addressed. These factors have not gone unnoticed, but on the other hand they have not been integrated into a broader framework, model, or theory (Becker, 2007; Warschauer, 2002; Mardis et al., 2008). Some have pointed out the importance of the political or environmental context (Marcovitz, 2006), but this is not yet widely accepted or identified as a third level (Korupp and Szydlik, 2005). Personal values, what I would call a fourth level, have been identified as significant to adoption (Garthwait, 2005; Warschauer, 2002). Other than an earlier paper using the framework proposed in this paper (Mardis et al., 2008), the concept of a fourth level digital divide has not yet emerged. The Behavioral Adoption Framework seeks to integrate first and second level digitial divide issues with what it calls third and fourth level divides. The third level consists of cultural values including the school, government, and societal factors that affect adoption. The fourth level is the individual teacher or student’s choice to adopt based on individual, personal values.

Proposed Framework
The theories and models discussed earlier have advanced the study of usage and adoption. However, it is time to reframe the conversation and look at a new framework which accounts for a greater number of factors, but has the flexibility necessary for an increasingly more complex contextual environment. In terms of the digital divide issues, this framework would represent a four level digital divide: technology, competency, cultural values, and personal values. In terms of classical Information Systems discussion, this is a further evolution of TAM/2 type models which seek to explain dynamics of usage. Over time, researchers have attempted to enhance the original TAM model and account for more environmental factors. However, these models have made fundamental assumptions concerning access and knowledge/skills in particular. In particular, they have assumed that the technology and competency are not contributing factors. In a business environment, this might be the case. However, in the broader world of usage and users, these are vital concepts. The proposed framework does not assume that potential users have access or that they have the necessary knowledge, skills, or competence necessary for adoption. In terms of the digital divide, the factors correspond to the four levels of the digital divide. It may seem obvious that access to technology is required, but it is also obvious that intention to use must precede use, as in TAM/2. The framework which is proposed here is represented by four factors, each of which is representative of two types of factors. This is the Behavioral Adoption Framework (BAF). Figure 7 depicts this framework in terms of relationships.

Figure 7

This framework may also be represented as a four quadrant matrix that classifies the factors (Figure 8).

Figure 8

It may seem obvious that access to technology is required, but it is absent from the TAM/2 models. However, usage in terms of digital divide has focused on the lack of technology or access. In terms of the digital divide, the quadrants represent the four levels of the digital divide: access, education/training, culture/policy, and personal desire. Some authors, such as Korupp and Szydlik (2005) have tentatively identified this third level. However, the concept of a fourth level digital divide is a new idea.
The purpose of these quadrants is not to identify every possible factor, but to propose a taxonomy which could be both explanatory and predictive. In this sense, it can serve as a taxonomy of factors, aid in discovery, and serve as a basis for further discussions. Specifics in each can and will change based on the context, but the basic issues should remain the same. The framework also has philosophical symmetry, addressing issues which are technical and social as well as issues which are internal and external. This may provide face validity, but the framework still must be tested to be proven accurate. The following is a proposed formula for BAF if one were to be able to create a quantifiable scale for each factor.
Technology * Competency = Ability
Cultural Values * Personal Values * Competency = Motivation
Motivation * Values = Usage
This is similar to Vroom’s Expectancy Theory: Expectancy * Instrumentality * Valence = Motivation (Vroom, 1964). Effects of weak or null values demonstrate that all factors must be necessary for usage. If one of the four factors is weak, usage will be significantly affected.
Finally, there are several assumptions which are implied in this framework. First, genuine adoption requires both ability and motivation. Second, all four factors must cooperate at some level for usage to occur. Technical factors are required for the ability to use and social factors in combination with competency are required for motivation. Third, each the contents of each factor can change from context to context. This provides flexability for broad application regardless of the technology or the environment.

Four Types of Factors
Technical
Technical factors involve technological artifacts and a person’s ability to use them. The technology includes the hardware, software and infrastructure as well as the systems which are greater than the sum of their parts. It is the combination of energy, information, and matter which accomplishes work. These are the technical artifacts. The relationship of these artifacts to the individual is the second example of a technical factor.

Social
Social factors involve the content and method of human opinion. These are the thoughts of others and the thoughts of the individual. This is culture, law, taboo, policy, values beliefs, aesthetics, virtue, truth and associated ideas at the national, societal, historical, organizational, and familial level. It is what “others” think and what “self” thinks.

External
External factors are anything outside of the individual. It is the world of technology and the world of other people. It is the objective artifacts and the people who use them. It is that which is “other.”

Internal
Internal factors are those which are unique to the individual user. It involves everything a person thinks, knows, and can do. It is turning on a switch, reading a screen, installing a drive, or typing a sentence on a keyboard. It is the mental and physical qualities of each and every person. It is what the individual values as oppose to those around him.

The Four Factors
Technology
This factor lies at intersection of the technical and the external. This is the objective technological artifact and its environment. It is the connectivity, the hardware, software, and the physical setting of usage. This is where the typical issues related to the level one digital divide reside. The issues which affect usage include access to the necessary technology including internet access and suitability of hardware and software for the participatory environment. It is technical support and the reliability of systems. As an example, a person may desire to participate in a virtual community such as Second Life, however if they do not have broadband and a sufficiently fast computer, their technological environment prevents their usage. However, if they do have broadband and their computer has insufficient memory and continually crashes during participation, then this will affect their usage as well. Technology consists of the artifacts and resources necessary for adopting, including the ability to purchase that technology.

Competency
The second factor lies at the intersection of the technology and the user. Even if a user has broadband and a sufficiently fast computer, they still must have the knowledge and skills necessary to operate that technology. The focus here is on factors which require the user to interact with the technology. This will vary from user to user even in a single technical environment. For example, in a single department of one company, there will be varying levels of aptitude, knowledge, and experience with a given system. Everyone may have the same technology and same connectivity, but some people are more competent than others. Since this ability varies from person to person, these factors are technical but also very personal. They may vary based on physical ability or disability, education, training, or any other HCI factor. Does the user have the competency to use the technology?
As seen in Figure 7, competency affects usability and motivation. This is based on the concept that competence and the ability to freely engage in an activity are directly related to intrinsic motivation and the self-determination model of motivation (Deci, 1975; Ryan and Deci, 2000; Deci and Ryan, 2000; Vallerand et al., 1997). Individuals who feel competent to perform a certain action (adopt technology) will derive greater pleasure, a full sense of enjoyment, and a feeling of autonomy when they perceive themselves to be competent. This increases motivation. High perceptions of competence are also indicative of higher actual competence (Deci, 1975), thereby increasing ability. As such, the factor of competency incorporates both actual competency and perceived competency.

Cultural Values
This factor, level three, is at the intersection of external and philosophical factors. This is the atmosphere surrounding the user. It involves factors related to policy, values, beliefs, culture, voluntariness, and social influences. One can think of it as the culture in which the user must function. It could be the culture of a school, a business, a home or any other context which affects the user. The context tells the user what behavior is legal or illegal, required or voluntary, encouraged or discouraged, and so on. As well, a user may function in several different atmospheres. An activity that is against school and library policy may be permitted at home. Each environment has its own cultural values. There is also the aspect of the physical atmosphere and the virtual atmosphere. Just as the physical environment has its own set of beliefs, pressures, and social interactions which may promote or inhibit participation, the same can be said of the virtual atmosphere as well. Whether it is a forum, e-mail, or virtual reality, there are also social norms, policies, and peer pressures which affect usage. The user may or may not choose to follow their cultural values but they are a significant factor which does affect the user and his or her motivation.

Personal Values
Factor four concerns issues which are both personal and social. This is the user’s attitude toward the behavior. Here is intention, desire, pleasure, enjoyment, fulfillment, anxiety and all the other factors which distinguish one user from another and the user from their environment. In short, this is “what’s going on” in the user’s head when they form opinions and ultimately decide whether they will or will not use a given technology. Classical IS theories of usage and participation have focused on this quadrant because this is the location of the factors, including intention, which are seen to lead to the final decision regarding behavior.
Some may view this quadrant as most significant because they perceive ultimate decisions to adopt as happening here. However, there are situations when other factors outweigh personal values. For example, full intention to use and high motivation are always limited by the technology, the context, and competency. Usage may face geographic, infrastructure, or economic barriers that prohibit usage. In this situation, technology is a stronger factor than personal values or intention. They have a greater affect on usage. Just because one wants to adopt does not mean that one can or is able. This concept is not addressed in the typical TAM/2, TRA, and TRB models of usage. They generally assume that there are no barriers between intention and behavior.
Factors in quadrant number four may lead the individual to change their technology, learn new skills, or change their environment. However, such a dynamic is not directly leading to adoption but to changing oneself or one’s environment. So, this issue is not directly addressed by this framework.
Finally, personal values should not be confused with personal characteristics such as demographics. Studies continually compare usage to age, income, race, education, geography, and similar factors (Lenhart, 2002). This framework does not ignore those factors, but instead understands them to influence each of the four quadrants. Race and education will certainly affect one’s cultural and personal values and the technology at one’s disposal. However, such factors do not cause usage. Demographics alone are not deterministic of usage. Rather, they shape an individual’s values and abilities. From the standpoint of this framework such factors are secondary.

Motivation
Motivation can have many forms. While some would bisect motivation into internal and external motivation, there are also those who see motivation as a spectrum of ideas (Deci, 1975; Deci and Ryan, 2000; Ryan and Deci, 2000) (Figure 9). The advantage of this approach is that it encompasses three loci of causality (impersonal factors, external factors, and internal factors) and defines three types of motivation (amotivation, extrinsic motivation, and intrinsic motivation) (Ryan and Deci, 2000, p. 72).

Figure 9

The acknowledgment of amotivation which is impersonal and competency related is especially significant because it helps to account for the non-motivated user, accidental user, or the non-user. TAM/2 models do not address these types of users
In the proposed framework, motivation is affected by three factors: cultural values, personal values, and competency. This fits with the Self-Determination theory and provides a richer representation of motivation. Self-determination theory loads across all three of the four factors: cultural values, personal values, and competency. Technology is not considered as a contributor to motivation because it lies in the quadrant which is both external and impersonal. It is not part of the individual’s relationship with self or with others. Motivation is ultimately a personal as opposed to an object oriented dynamic. A person may see a shiny new technology, but it does not in and of itself motivate anyone. Values are what motivate the people to use, acquire, or participate. If two people see a new SUV, one may be motivated to drive it while the other person may be motivated to condemn it. Motivation is based on personalm, competency and cultural values, not on artifacts themselves.
In the framework proposed by this paper, motivation functions much like intention in the TAM/2 related models. However, motivation is a richer concept with a fuller theoretical history and a well developed psychological concept. Some may even see intention as a weak concept which is not necessarily that helpful. For example, people generally do what they intend to do. Similarly, people generally do what they plan to do, unless there are intervening circumstances. Someone may intend to go to lunch or plan a lunch break. In these cases it is legitimate to ask if such a construct is truly predictive or explanatory. In contrast motivation has a broad spectrum of content which provides a richer construct encompassing amotivation, extrinsic motivation, and intrinsic motivation. Intention can only be intrinsic. In keeping with the overall tenor of the proposed framework, motivation is more suited to explaining contextual factors which affect individual behavior and usage.

Ability
The construct of Ability in the context of this framework combines technology and competency to enable usage. The technology must exist, be accessible, and functioning and the user must have the necessary knowledge and skills to operate a given technology. As demonstrated by Figure 7, ability is a combination of technology and competency. Without either one, ability is incomplete. The remaining two factors of cultural values and personal values do not absolutely affect ability in the negative or positive sense. First, in the positive sense, they do not create or enable ability. They do not have the power to create the technology or skills necessary for usage. Second, in the negative sense, they can not prevent ability. Cultural values and personal values may see usage as illegal or distasteful, but that does not change a person’s ability. A reformed criminal hacker may have conformed personal values to cultural values, but that does not necessarily destroy ability. Values affect motivation, not ability.
In the context of the digital divide, there was the idea that technology alone would create ability and lead to usage. First, it was just the technology. Giving the technology was viewed as a direct path to usage. This became know as the first level digital divide (Korupp and Szydlik, 2005). However, the presence of technology alone did not create ability nor did it address the issue of motivation (Korupp and Szydlik, 2005). Thus, the second level divide (Korupp and Szydlik, 2005). Technology requires training. A fantastic system with “perfect” technology still requires user competence. Most people still don’t know how to program their VCR or use half the features on their cell phone. Access to technology with training for competency can create the ability to use, but motivating the user is something else entirely.

Behavior
The concept of behavior must have a broad meaning. While usage may have originally been limited to computing in a business environment (Davis, 1989), it must now be expanded to include virtually any interaction with technology. People were accepting new technology long before the computer existed. Whether the technology is the telephone, a card punch machine, a laptop, the internet, or a neural interface; there will always be a need for a framework to evaluate usage. There must be a framework which allows for explanation and prediction. It must not be limited by the artifact or the individual. It must be flexible enough to deal with changing environments, changing technology, and changing people. Usage may be as simple as using a pen or as complex as a virtual team building the Airbus 380 in a collaborative online environment. Every time a dramatic technology is introduced, the same four divides will exist. During the last decade it was the internet. One hundred years ago it was the automobile.

Possible Research Methods
As mentioned earlier the goal of this paper is to propose a framework for assessing factors that promote and inhibit participation. This requires input not only from users, but also from non-users. For most technologies, it can be very difficult to understand or measure the potential community unless the technology is mandatory. For example, in a distance education course there may be a discussion forum, but usage would be at least somewhat mandatory. In a work situation, such systems might be voluntary (Ardichvili, 2003) or mandatory. Since the proposed framework has not yet been tested, it would be helpful to have a well controlled environment. This way, one or more of the factor can act as a control. For instance, using the same technology (Q I) for each individual would assist in more precisely measuring the other three factors (Q II-IV). This might be a single Learning Management System (LMS) across several campuses in different countries. The same could logic could be applied for controlling several variables. A study could target a sub-population with the same personal values (QIV) and the same competencies (QII), with the same cultural context (QIII), and see how they react to given technologies (QI). An example might be eight grade Hispanic children in a large metropolitan school. Such planned studies would reduce variance among the controlled variables and enhance variance in the particular factor being examined. The next section of this paper is proposed example of such a study.

Limitations and Future Research Related to BAF
The highest priority is to empirically test the validity of the BAF framework. It should also be tested against other TAM/2 models to determine if it is a better predictor of usage and if it accounts for a greater percentage of variance in usage. Testing the framework should also include a close examination of self-determination theory as a basis for measuring the context of technology. Self-determination theory has been used in a variety of disciplines, but has not yet been applied to technology usage. Another issue is the determination of the precise scope of each factor. Factors are not specific variables so the relationship of individual variables such a demographics to each factor needs to be explored. The primary limitation of the proposed framework is that it has not been empirically tested. It may have face validity and some theoretical connection to tested TAM/2 models, but there may be other factors or relationships which are not considered by the framework.
Second, this framework must be tested in a variety of contexts. This study needs to be repeated in similar and in different types of organizations with differing types of populations. Studies should also be completed in a number of different environments where usage is both mandatory and voluntary, where cultures vary significantly, and where there are significant differences in competency. Until then, one must question the relevance of this framework for other communities. Motivation to use reflects a limited breadth of personal values and even narrower set of cultural values. However, performing similar studies with the same methodology in different cultural setting would aid in the understanding of cultural values on motivation and usage.
If the Behavioral Adoption Framework is verified, it offers the advantage of a standardized, but extremely flexible framework. It could possibly lead to something like an e-readiness assessment for individual users. Such scales are usually used at the national level (Dada, 2006) and the organizational level as well (Huang et al., 2004). An e-readiness scale for an individual could possibly predict an individual’s likelihood to use a given technology.

Conclusion
At this point, it is necessary to summarize the Behavioral Adoption Framework and identity its unique features.
First, this framework is specific in its intent. It does not measure acceptance (TAM/2) or usage but behavior adoption. Instead, it seeks to explain usage. It is not the “what” or the “how,” but the “why.” It also does not measure success. Because the focus of the behavior is not strictly usage, this new framework can also address the failure of usage or what leads to the end of usage.
Second, this framework is relevant for the digital divide situation. Ultimately, the digital divide is about increasing usage, not about the presence or absence of technology. It concerns ability and motivation. It is a complex set of philosophical and technical issues involving personal and cultural values. BAF also explains the loci of barriers to usage. This an important aspect which is overlooked by the majority of frameworks.
Third, the BAF provides a richer description of usage and non-usage. It not only explains the individual’s perceptions and intentions, but takes into account the artifact, the social context, and the skill set of the user. It also explains why motivation (or intention) alone is an insufficient cause of usage. How helpful is it to distinguish between “behavioral beliefs,” “behavioral attitude,” and “intention” (Wixom and Todd, 2005) (Figure 1)? As mentioned earlier, TAM/2’s emphasis on intention fails to account for the accidental, casual, non-intentional, coerced, or non-user.
Fourth, this framework is scalable. It has the ability to explain simple and complex technologies as well as simple and complex types of usage behavior. It can account for virtual and physical culture. It is not bound to a work environment or to a the concept of productivity. It’s proposed explanation is not tied to one purpose. It may explain usage for entertainment or work. Many items in the TAM model relate to work, job, economy, quality of work, etc. In contrast a virtual gaming community may not have any of these benefits for the user. Instead, it may be based on pleasure, fun, sense of companionship, gossip, etc. In this sense the factors leading to usage are not work based, but pleasure based. This entails personal satisfaction as opposed to job satisfaction. For non-work situations, the typical TAM/2 model would predict failure to use.

Bibliography
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl, & J. Beckmann (Eds.), Springer series in social psychology (pp. 11-39). Berlin: Springer.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes 50(2), 179-211.
Ajzen, I., and Fishbein, M. (1973). Attitudinal and normative variables as predictors of specific behavior. Journal of Personality and Social Psychology 27(1): 41-57.
Ardichvili, A., Page, V., and Wentling. T. (2003). Motivation and barriers to participation in virtual knowledge-sharing communities of practice. Journal of Knowledge Management 7(1): 64-77.
Fishbein, M., and Ajzen, I. (1975). Believe, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addison-Wesley, Reading, MA.
Baber, B. (2004) Power and empowerment: A radical theory of participatory democracy. University of California Press.
Becker, J.D., (2007). Digital equity in education: A multilevel examination of difference in and relationships between computer access, computer usae and stat-level technology policies. Education Policy Analysis Archives 15(3): 1-36.
Becker, H.J., & Ravitz, J.L. (1999). The influence of computer and Internets use on teachers’ pedagogical practives and perceptions. Journal of Research on Computing in Education 31(4): 356-384.
Castells, M. The Rise of the Network Society. Cambridge, Mass: Blackwell Publishers, 1996.
Compeau, D. R., and Higgins, C. A. (1995a). Application of social cognitive theory to training for computer skills. Information Systems Research 6(2): 118-143.
Compeau, D. R., and Higgins, C. A. (1995b). Computer self-efficacy: development of a measure and initial test. MIS Quarterly 19(2): 189-211.
Conley, S. (1991). Review of research on teacher participation in school decision making. Review of Research in Education 17: 225-266.
Cornwall, A. and Jewkes, R. (1995) What is participatory research? Social Science and Medicine 41(12): 1667-1676.
Crump, B, and McIlroy, A. (2003). The digital divide: why the “don’t –want-tos” won’t compute: lessons from a New Zealand ICT project. First Monday 8(12). Retrieved April 17, 2007 from http://firstmonday.org/issues/issue7_7/warschauer/index.html.
Dada, D. (2006) E-readiness for developing countries: moving focus from the environment to the users. The Electronic Journal on Information Systems in Developing Countries 27(6): 1-14. Retrieved May 2, from http://www.ejisdc.org/ojs2/index.php/ejisdc/article/viewFile/219/184.
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 13(3): 319-339.
Davis, F. D., Bagozzi, R. P., and Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to user computers in the workplace. Journal of Applied Social Psychology 22(14): 1111-1132.
Davis, F.D. and Venkatesh, V. (2000). A theoretical extension of the technology acceptance model: four longitudinal field studies. Management Science 46(2): 186-2004.
Deci, E.L. (1975). Intrinsic Motivation. New York: Plenum.
Deci, E.L. and Ryan, R.M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry 11(4): 227-268.
deHan, J. (2004). A multifaceted dynamic model of the digital divide. IT & Society 1(7): 66-88.
Dewan, S., and Riggins, F.J. (2005). The digital divide: current research and future research directions. Journal of the Association for Information Systems 6(2): 298-337
Dholakia, U.M., Bagozzi, R.P., and Pearo, L.K. (2004). A social influence model of consumer participation in network- and small-group-based virtual communities. International Journal of Research in marketing 21: 241-263.
Epstein, R., Alper, B., and Quill, T. (2004). Communicating evidence for participatory decision making. JAMA 291(19): 2359-2366.
Garthwait, A., & Weller, H.G. (2005). A year in the life: Two seventh grade teachers implement one-to-one computing. Journal of Research on Technology in Education 37(4): 361-377.
Hargittai, E. (2002). Second-level digital divide: differences in people’s online skills. First Monday 7(4). Retrieved April 18, 2007 from http://www.firstmonday.org/issues/issue7_4/hargittai/.
Huang, J. H., Huang, W. W., Zhao, C. J., and Huang, H. (2004). An e-readiness assessment framework and two field studies. Communication of the Association for Information Systems 14: 364-386.
Lenhart, Amanda. (2002). Barriers to internet access: from the non-user and new user perspective. Pew Internet & American Life Project. Retrieved on May 5, 2007 from http://tprc.org/papers/2002/44/Lenhart_Barriers_TPRC.pdf.
Levin, M., Wheelan, S.A., Take, A., and Hammond, D. (1998). Participation action research. Systemic Practice and Action Research 11(2): 207-222.
Lin, H. The role of online and offline features in sustaining virtual communities: an empirical study. Internet Research 17(2): 119-138.
Macy, B.A., Peterson, M.F., and Norton L.W. (1989). A test of participation theory in a work re-design field setting: degree of participation and comparison site contrasts. Human Relations 42(12): 1095-1165.
Mardis, M., Hoffman, E, & Marshall, T.E. (2008). A new framework for understanding educational digitual library use: Re-examining digital divides in U.S. Schools. International Journal on Digital Libraries. (in press).
Marcovitz, D.M. (2006). Changing schools with technology: What every school should know about innovation. In: Tettegah, S.V., Hunter, R.C. (eds.). Technology and education: Issues in administration, policy, and applications in K12 schools. Elsevier, Amsterdam, Netherlands, pp. 3-15.
McKeen, J.D., Guimaraes, T, Witherbe, J.C. (1994). The relationship between user participation and user satisfaction: an investigation of four contingency factors. MIS Quarterly 18(4): 427-451.
McKeen, J.D., and Guimaraes, T. (1997). Successful strategies for user participation in systems development. Journal of Management Information Systems 4(2): 133-150.
Moore, G. C., and Benbasat, I. Development of an Instrument to Measure the perceptions of adopting an information technology innovation. Information Systems Research 2(3): 192-222.
Publications. Self-Determination Theory: an approach to human motivation & personality. Retrieved April 24, 2007 from http://www.psych.rochester.edu/SDT/publications/index.html.
Ryan, R.M. and Deci, E.L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist 55(1): 68-78.
Taylor, S., and Todd, P.A. Understanding information technology usage: a test of competing models. Information Systems Research 6(4): 144-176.
Thompson, R. L., Higgins, C.A., and Howell, J.M. (1991). Personal computing: toward a conceptual model of utilization. MIS Quarterly 15(1): 124-143.
Vallerand, R.J., Fortier, M.S., and Guay, F. (1997). Self-determination and persistence in real-life setting: toward a motivational model of high school dropouts. Journal of Personality and Social Psychology 72(5): 1161-1176.
Venkatesh,V., Morris, M. G., Davis,G. B., and Davis,F. D., “User acceptance of information technology: Toward a unified view.” MIS Quartery 2003, 27, 3, 425-478.
Vijayasarathy, L.R. (2004). Predicting consumer intentions to use on-line shopping: the case for an augmented technology acceptance model. Information and Management 41: 747-762.
Vroom, V. (1964). Work and motivation. John Wiley & Sons, New York.
Warschauer, M. (2002). Reconceptualizing the digital divide. First Monday 7(7). Retreived April 18, 2007 from http://firstmonday.org/issues/issue7_7/warschauer/index.html.
Wixom, B. and Todd, P. A. (2005). A theoretical integration of user satisfaction and technology acceptance. Information Systems Research 16(1): 85-102.
Yoo, W., Suh, K., and Lee, M. (2002). Exploring the factors enhancing member participation in virtual communities. Journal of Global Information Management 10(3): 55-71.