Here is a response to Roy Bohlin's paper. I am going to argue that the ideas and work it discusses are fundamentally misconceived and misplaced. Is it because it's a bad paper? no, I shall keep this paper and probably refer to it, not least because it is a thoughtful application of the ARCS model. Is it because I've never done work in this area? no, this is much more systematic (i.e., better) than some rather casual use of attitude surveys about using computers in teaching that I have done. Is it because the topic is unimportant? not at all: in the UK the government has recently instituted a policy that in effect will make IT use compulsory by all teachers. What psychologists call anxiety, but in common language is fear and loathing, can be expected as a widespread response. The question is what does this mean, and what should be the response.
No, my opposition to the paper is because this year at CAL'99 I had a bit of a conversion experience, and it is with the fervor of a convert that this is written. I tell you this as you may wish to take it with a pinch -- or a shovel full -- of salt. But the problem is not any lack of sincerity, but perhaps a surfeit of it.
[quoting Bohlin's paper] Assuming that good instruction in teacher preparation programs will always change anxiety, confidence, and attitudes is naive. These are very important variables, because attitudes and confidence are strong predictors of actual voluntary behaviors (such as decisions regarding the integration of technology into their curriculum). If computer anxiety is combined with low confidence, low motivation, or negative attitudes, individuals will strive to avoid interactions with computers. ....
It would seem that the best approach to addressing severe computer avoidance would be to use a multi-pronged approach, which attempts to impact all of these factors.
And most bluntly of all...
To properly integrate computers into the classroom curriculum, teachers need the proper disposition.
Despite some guarded language, the paper in fact makes the basic error of inferring causality from a correlation. Many school teachers (called "individuals") avoid computers: true. Many also show anxiety and negative attitudes: true. Roy Bohlin however infers, or rather assumes, that the negative attitudes cause computer avoidance, and furthermore that changing the attitudes will change the behavior, and has and is carrying out a program of action based on this fallacious inference.
My view is that anxiety is a message like pain. If you just take medication "for pain", serious damage may follow, even though the patient will have felt better for a while. To put it in a more homely manner, a heart attack is God's way of telling you to slow down, and anxiety was his way of whispering an early warning. To take anxiety as a symptom to be overcome is a) patronizing: it is saying teachers' feelings are meaningless and technophiles know best, b) as noted, it is logically an elementary fallacy, c) from the viewpoint of trying to understand education, IT use, organizational change, etc. it is a failure to seek an understanding of the factors that are actually large, active, and having a big effect in the real situation.
In terms of the psychological literature, I think this is a case of "the fundamental attribution error": the widespread tendency of people to attribute their own actions to outside circumstances but other people's to in built traits or attitudes. If we survey people about their attitudes to computers (as I have) we are setting ourselves up to make Roy Bohlin's assumption that these attitudes determine their behavior. We should consider that other external factors might be important in affecting computer "avoidance" (what a value-neutral term!).
Is my argument sensible? Roy's paper does a good job in conjuring up an image of an anxious computer user. But not all the arguments actually support the overall position. Firstly, it is easy to believe that anxiety disables learning. If you try to teach a class and then tip out a sack full of rattlesnakes round their ankles, no doubt cortisol measurements would peak and most learning about computers would stop. It doesn't follow that the computers caused the anxiety. Secondly, in fact, a lot of the anxiety observed will be about the learning situation, not the computers per se. You could probably reduce my learning by making me do it in class, as a job requirement, with a bunch of my peers, and with an instructor who has never experienced "computer anxiety" and is thus actually wholly without insight or understanding of his (yes, probably his) students. It may be relevant to use ARCS for this, but it doesn't add anything to support the original assumption that anxiety pre-dated the instructional situation and is the main bottleneck to be addressed.
It should be noted that some years back there was a lot of verbiage in the IT literature about how middle managers were "failing" to use IT themselves, and then as now this was attributed to their bad attitudes, fear of seeming incompetent in front of their subordinates, etc. etc. Then we didn't hear any more about this. This was not because Roy Bohlin's predecessors had designed anxiety-dissolving instruction. It seems instead to have been because the technology developed, so that instead of just being fun for nerds, it actually started to do things that were useful to those middle managers in their real jobs even after allowing for the real learning costs. If this is a precedent, then it suggests that computer avoidance by teachers in fact stems from the fact that computers don't, on balance, help teachers do their jobs. Avoidance is the rational, even if not rationally articulated, response to a "tool" whose benefits fail to outweigh their costs. Anxiety is the rational, if visceral, response to being pressured to do something experience has taught you will make your job or life worse. A more rational response by those wishing to increase uptake, would be to make computers more useful to teachers in the real situation of use, not to treat the symptoms.
Here is a guess about alternative kinds of important factor determining computer use by way of illustrating what might be discovered by research that hadn't fallen into the trap of presupposing that teachers are the problem. (Actually, not just a guess, but inspired by Lyn Dawes' data.)
For computers to be used regularly, then in preparing a lesson computers have to be an option as convenient to use as the alternatives. Where are lessons prepared? I understand that some are prepared at school, and many are prepared at home. So obviously there needs to be a computer always available for each teacher in school, and another one in every teacher's home. Actually, as someone recently and wisely remarked on ITForum, there need to be two in the teacher's home: one so that they don't have to negotiate to use it, and another for other family members firstly to avoid that negotiation/competition and secondly to have a competent user community there to provide help and support. So every plan for school computer use should clearly give each teacher three personal machines besides those for the children.
Have you done a survey asking about this, and correlating availability of personal machines with "anxiety" and "avoidance"?
Do you think this unreasonable? Only those members of ITForum who do not have access to a personal machine are qualified to answer. Those who do must ask themselves how many of their computer activities they would still bother with if they had to book a machine at constrained times in order to do so. (This message wouldn't be written and sent, for example.)
Although the main issue here is an enormous trap and failing in IT/educational research (due no doubt to bad attitudes and traits in researchers), it is worth mentioning that it is also an example of the problem with IT use in general, at least according to Landauer. [Landauer, T.K. (1995) The trouble with computers: Usefulness, usability, and productivity (MIT press; Cambridge, MA)]. (I've only just read this, but will now make it the main text on courses I teach). He argues persuasively that most IT investment has failed dismally to be worthwhile, and that this is because there is almost always a failure to investigate the situation of use and to modify both the design, development AND deployment to fit in with that situation.
My conversion experience was caused by hearing Lyn Dawes' paper "First connections: factors influencing the acquisition of network literacy by teachers during the introduction of the National Grid for Learning". (This has been submitted to, but as far as I know has not yet emerged from, the journal Computers and Education. You could try asking her for a copy "ldawes@dmu.ac.uk").
She has absorbed some of the situativity literature, and applied it to teacher uptake of IT. In her perspective it isn't a question of whether teachers will take it up (it's now more or less a job requirement), but of where they are in the apprenticeship / community model: out beyond the firelight, in the outer circle, the inner circle, etc. (Her actual categories of user are: potential, participant, involved, adept, integral.) We can then ask about all sorts of factors, and correlate any possible factor with degree of use. What I like about this so much is that it is quite possible to ask about attitudes AND about "situation" variables such as availability and a supportive community of other knowledgeable users, along with asking about use (i.e., actual behavior, current practice). Then we can test what is only assumed in the kind of attitude survey I have done in the past, and which Roy Bohlin's work is founded on.
Although Lyn was a school teacher, and has a private agenda of combating teacher-bashing attitudes in both government policies and research literature, this is well concealed in her work (as my views are not in this rant). As I said, what is so good about the approach she suggests is that it can be neutral and treat attitude and situation factors equally, in a way that attitude surveys alone can not.
As for me, I hope in the next year to collect some data that may allow me to convict a lot of past work as guilty of the fundamental attribution error. But so far, that's just a hope. I haven't done the work either, yet.
IT training for teachers may be important; it probably in fact will turn out to be a significant situational factor in IT uptake. But this should be demonstrated, not assumed. And the same research that could demonstrate it should also show how big a factor it is: what in fact the bottlenecks are, and where training ranks among them.
Roy Bohlin's paper, like much other literature, assumes that the teacher is the bottleneck. This attribution is not based on evidence, and itself prevents looking for evidence. It is predictable from the psychological literature, and says more about the unexamined attitudes of those making the attribution than about teachers. If it turns out to be unwarranted, efforts based on it will turn out to have been directed at ineffective aspects of the situation, and will not change IT uptake.