See this in PDF format; revised 28 October 2008.
As a professional working in Information Technology, I often encounter hesitation and trepidation on the part of functional1 users to engage my employer’s information system, an ERP system known as Banner. The engagement of technology—especially for older generations of users, in which was not inculcated the idea of information systems (i.e. the Internet) as pleasurable or entertaining—has been a focus of information systems and organizational behavior research for some time. In much the same way as MacGregor’s landmark work (Montana & Charnov, 2000, pp. 251-52) split the concept of innate employee behavior into two extremes—essentially wicked and lazy on one end, earnest and self-actualizing on the other (“Abrahamic,” to borrow Herzberg’s phrase)—so traditional IT adoption research has rather myopically divided all impetuses for system use into either endogenous or exogenous antecedents of behavior.
Malhotra, Galletta, and Kirsch’s “How Endogenous Motivations Influence User Intentions” expands upon and empirically tests Deci and Ryan’s organismic integration theory (1985, as cited in Malhotra et al, 2008, p. 269). The proposed model ostensibly offers a better prediction of actual system use than the traditional “Motivational Model,” which accounts in large part for the existing dichotomy of external factors, such as reward for work, and internal factors such as the inherent joy of performing the task (1992, p. 111-1132, as cited in Malhotra et al., 2008, 270). Malhotra et al. rightly question the continued validity of such a model, citing the inherent variability of the effect of extrinsic motivators on different groups of people as well as the possible interplay of extrinsic and intrinsic factors of motivation (2008, p. 270). While perhaps useful as a rule of thumb, or popular as a crass generalization, this existing dichotomy “seems to have limited our understanding of user motivation” (2008, p. 270).
Admittedly, some of this is a difficult sell: Malhotra et al. strain to find examples of such an “endogenous notion of volitional extrinsic motivation,” finally offering such internal thoughts as “good employees do not play computer games as work” (2008, p. 271), an idea which inhabits several strange belief schema at once. With very little prestidigitation, such an example an internalized external motivator can be externalized once more by tying it deftly to fear of reprisal or loss of other purely extrinsic motivators. Because it is extraordinarily difficult to map such internal representations and—as the authors themselves note—to consistently calculate the effect such external motivators, internalized or not, this is a very precarious concept. In the case of their rather poor example, the idea of a good employee is either entirely an intrinsic motivation, or it is an external demand: that the ultimate damnation possibly comes in the form of self-reprisal rather than external punishment (to wit: one’s own goals for baseline accomplishment meet or exceed institutional goals but fail to meet internalized representation of said accomplishment) in no way necessitates that this motivator be extrinsic in nature and internal in representation.
Venkatesh, Brown, Maruping, and Bala (2008) speak to this in a limited fashion by reinforcing the theory of “behavioral expectation,” which takes into account not only an employee’s internal schema of beliefs (called “behavior intention”) but also the extrinsic factors which are subject both to change and to amorphous degrees of influence on intrinsic motivations (p. 486). Their findings indicate a weak relationship between the internalized belief systems of behavioral intent and the frequency and intensity of system use but a strong relationship with intensity alone (p. 497).
What strikes me as fundamentally odd about Malhotra et al.’s study is the semantic ambiguity. The authors’ eventual semantic summation of intrinsic and extrinsic stimuli (or more importantly, the dynamic between the two) is “perceived locus of causality,” or PLOC (2008, p. 271), usually a dialectic reached by some internal confluence of motivations. This perceived locus of causality is by necessity an internal representation; as such, the authors find it grounds to argue that unequivocally extrinsic motivators are therefore functions of this concept, and no longer purely extrinsic in nature. This, therefore, is the difference between “volition” and “compulsion” (2008, p. 272); since compulsion may exist anywhere along a wide swath, the authors admit that attempting to identify or measure extrinsic motivation (and therefore qualify the PLOC of which it is constituent) is difficult (2008, p. 274). There exists a relatively small slice of intrinsic motivation, and my experience in the field of information systems would indicate that it—that is, a desire to use the system out of an inward desire to learn and explore it—is almost never to be found in functional (to wit: the majority) of users. Even our so-called “power-users,” who may explore the system and test its capabilities beyond the scope of their stated duties generally do so in the context of an internal goal which is externally motivated—i.e. a job function.
I find this to some degree a largely irrelevant when discussing information system use, since such system use is usually mandated by employment but entirely volitional insofar as the employee has chosen to fulfill said employment and in the nature prescribed by the information system. However, the flexibility of PLOC allows for it to account for and describe such flux states as an “introjected PLOC,” wherein a user’s desired action does not match the system’s mandated action (2008, pp 272-273). At the University of St. Francis, Banner is usually the cause of such states, since it represents the absolute nadir of human interaction design. While this makes the use of Banner unpleasant, and results in an elevated level of required support and training, it has not in our experience actually lowered use of Banner by those functional users whose job requires it2. This is precisely what Venkatesh et al. refer to when they introduce “behavior expectation” into their proposed nomological framework. Behavior expectation is “an individual’s self-reported subjective probability of his or her performing a specified behavior, based on her or her cognitive appraisal of volitional and non-volitional behavioral determinants” (Warshaw and Davis, 1984, p. 111, as cited in Venkatesh et al., 2008, p. 484). Behavioral expectation, in accounting for the user’s internalized representation of extrinsic, non-volitional factors (to wit: the often confusing or counterintuitive nature of a given information system), has influenced a new conceptualization of system use. Their research indicates that when mediated by measures of behavior expectation, there arises no discernible relationship between external factors (“facilitating conditions” in their parlance) and actual system use (2008, p 494-95), meaning that regardless of the qualities of an employee’s perceived locus of control, their behavior with respect to mandatory use of an information system can be accurately predicted insofar as the employee understands the relationship of the system to his or her job.
Vaske and Grantham report that measurable intelligence leads to varying results in perceived locii of control; specifically, intelligent individuals tended to report higher levels of control (1990, p. 58). Vaske and Grantham posit a particular conceit on the part of intelligent people, in that they believe there are fewer variables outside of their control; in addition, internal causality becomes the default attribution for successful experiences, while external causality remains the scapegoat for failure (1990, p. 58), reinforcing that there is a consistent dichotomy persisted by individuals between their intrinsic and extrinsic forces, even if there is a nebulous area wherein those two mingle, outside the scope of the individual’s self-analysis. Malhotra et al. use this as their first hypothesis (2008, p. 274).
Vekantesh et al. (2008) warn against drawing any conclusions based on this sort of experimental data, citing the “productivity paradox,” the theory that greater system use is not equatable to greater productivity (p. 499). Indeed, all this talk of motivations has drifted far away from what is perhaps the more pressing question to managers, namely “What will maximize productivity?” Is work done out of fear of retribution, termination of employee, ridicule of peers, or the motivation to more respect and better salary substantially different from that done out of technically curiosity or some other elusive intrinsic motivation? To their credit, Malhotra et al. (2008) do consider this possibility, measure breaking down initial adoption and continued use of a given information system (in their case, Blackboard, a web-based course management system) based on the composite reliabilities of various PLOC categories and other factors (p. 281-82). This, of course, is an experiment with immediate interest to the University of St. Francis as a whole, since its Information Technology department will be moving to Blackboard in 2009. Given a population who is not required by employment, but rather by education requirements, the confluence of personal and educational (in this case substituting for intrinsic and extrinsic, respectively) factors is of great importance, since we may to some degree predict the success of our online learning program on the relationship of motivation to continued use. In fairness, I cannot think of any theoretical student who would continue taking classes online only because he or she enjoyed the look of Blackboard and so fulfilled a peculiar internal drive, but for those students who generate introjected PLOC (perceived locii of control) from their interaction with the system, the difference may be enough to potentially reduce enrollment. For a system which, though mandatory, is part of a larger structure which is volitional, the quality of the information system, and the University’s approach to managing and promoting it, is potentially a thousand- or million-dollar issue.
Malhotra et al. (2008) ultimately conclude that “motivation may influence behaviors in more complex ways… [than] this simple dichotomy [between intrinsic and extrinsic factors]” (p. 290). Like Venkatesh et al. (2008, pp. 497-98), they suggest that while behavioral expectation is a reliable predictor of usage than quantifiable extrinsic factors, which seems not only correct, but obvious to anyone who has worked either in management or in an capacity relating to information systems. However, Malhotra et al. (2008) appear to suggest that extrinsic motivators almost never exist in isolation, but almost always as internal representations interpreted by a particular employee. I take issue with such a characterization, as I believe it relies too heavily on a particular sampling of a particular kind of employee in a particular kind of organization. Perhaps most the beneficial conclusion, though surely not the most novel, is the emphasis of an introjected PLOC’s negative effect on a user’s attitude and future intentions with respect to using to system (p. 291).
- Malhotra, Y., Galletta, D. F., & Kirsch, L. J. (2008). How Endogenous Motivations Influence User Intentions: Beyond the Dichotomy of Extrinsic and Intrinsic User Motivations. Journal of Management Information System, 25(1), 267-299. Retrieved October 27, 2008, from Business Source Elite (33245054).
- Montana, P. J., & Charnov, B. H. (2000). Management. Hauppauge, NY: Barron’s.
- Vaske, J. J., & Grantham, C. E. (1990). Socializing the Human-Computer Environment. Bristol: Intellect Books.
- Venkatesh, V., Brown, S. A., Maruping, L. M., & Bala, H. (2008). Predicting Different Conceptualizations of System Use: The Competing Roles of Behavioral Intention, Facilitating Conditions, and Behavioral Expectation. MIS Quarterly, 32(3), 483-502. Retrieved October 27, 2008, from Business Source Elite (33436540).
- In this case, a “functional user” is a person who uses an Information System to accomplish a particular function; this is distinct from a technical user, who administrates, builds, or extends said system.[↩]
- That the Banner team has been slowly replacing certain often-used screens in Banner with more user-friendly web interfaces may also play a part.[↩]