One of my areas of expertise is in the field of technology adoption; why do people use technology? To be honest, this is a somewhat difficult subject, riddled with puzzles wrapped inside of enigmas…oops, sorry about that.
My dissertation topic revolves around technology acceptance and prediction; specifically, I’m going to apply a model of adoption to a group of users to see how well it can explain their acceptance of email. So, like, what are some of these models?
The most widely used model of adoption was formulated back in 1986 by a guy named Fred Davis. Davis managed to revolutionize how we were able to measure technology adoption, and as a corollary, how we can predict this adoption. Predicting usage is very important, especially back in 1986 when even simple desktop computers could cost tens of thousands of dollars. Who the hell wants to spend a cool million bucks on software that no one wants to use?
So Davis used two models from the social sciences to formulate his model, the Theory of Reasoned Action and the Theory of Planned Behavior; for those keeping score, information systems is not a computer science but a social science, so this was perfectly acceptable. Now, I won”t bore you with the details of the TRA and TPB, but suffice to say, those are the roots of Davis’ model.
Davis formulated a very simple approach to adoption, a pattern if you will which has since been validated by the likes of Venkatesh, Brown, Rodgers, among others. Perception of technology leads towards the intention to use or not use that technology. Thus, the intention to use or not use technology leads towards the actual behavior of using or not using the technology. In other words, if you intend to use technology, you’re more likely to actually do so than if you don’t intend to use it. Seems pretty simple, right? So, what is it that leads one to intend to use technology?
This is where Davis hit the gold mine: he posited that there are really only two things that lead you to want to use technology: is it useful and is it easy to use? Yup, that’s it! Just those two little constructs were proposed to affect intention directly. Does it get more parsimonious?
Specifically, Davis called these Perceived Usefulness and Perceived Ease of Use, or PU and PEOU for short. What he put forward was a positive correlation between these two constructs and intention; as PU and PEOU increased, so did the intention to use technology.
And in 1989, Davis validated his model by…wait for it…TESTING IT! Yes, that’s right, he put it to the test by measuring these two constructs and then gauging acceptance. And what he found was that by examining these two constructs alone, he obtained an explanation of variability of between 42 and 44 percent; this means that we can predict whether or not someone will use technology 42 to 44 percent of the time based only on these two constructs. Not too shabby. So of course it was given an awesome name, the Technology Acceptance Model, or TAM.
The problem is, of course, that here we are 22 years later and the IS literature is littered with study after study after study using Davis’ TAM, but each time modifying it to account for specific technologies or to include demographic information. But the TAM itself has never changed. Not in 22 years. Problem.
So my study focuses on a new model proposed by Venkatesh, Morris, Davis and Davis (yes, THAT Davis) called the Unified Theory of Acceptance and Use of Technology, or UTAUT. Why? Well, because that 42 to 44 percent explanatory power of the TAM simply falls before the 70+ percent explanatory power of the UTAUT. Oh yeah, baby.
The UTAUT replaced PU and PEOU with Performance Expectancy (PE) and Effort Expectancy (EE), roughly the same thing, but not quite (not getting into it here, but notice the word “perception” was not used). Plus, it added two new constructs, Social Influence (SI) and Facilitating Conditions (FC). SI refers to how others view your usage of the technology (thing group norms from social theory) and FC refers to a support infrastructure. So as SI is a positive view and as FC is more trusted, we should see greater intention to use, right? Well, sort’ve.
There are actually four variables we look at as well: gender, age, experience with technology, and voluntariness of use. These variables actually impact the constructs, which in turn impact intention. SI actually doesn’t matter if technology is voluntary; if technology is mandatory, then SI has an impact, otherwise it doesn’t really matter. And FC? Well, it’s only important to older people with less experience with technology (not significant, statistically, otherwise).
So you can see, we’ve gone from a highly parsimonious model to one that is a bit more complicated, but obviously is much more effective at predicting usage. But of course, reality is not black and white, so we get some discrepancies.
For instance, Schwager et al did a study at a university of business to see what affected the adoption of tablet PCs amongst the faculty. What did they find? The only thing that matters was PE. Yup, that’s right, as long as the faculty found the tablet useful, they were willing to put forth the effort to learn how to use it.
So we have a lot to learn about what drives people to use technology, but we’ve got some good models to use. My study will see if the unified theory (UTAUT) can accurately explain the adoption of email. I’m excited to see the results, but more excited to actually get to chapter four of the dissertation 🙂
Until next time.