Too little and too much self-efficacy is bad for interest
November 28, 2013 at 1:25 am 11 comments
What an interesting paper! (Pun slightly intended.) In this paper from Paul Silvia, he found experimentally that self-efficacy and interest are related on a bell-shaped curve. Too little self-efficacy makes a task seem too daunting and uninteresting. Too much makes the task boring. This is important because we know that self-efficacy is among the most significant factors influencing non-majors success in learning to program. It’s clear that there’s a sweet spot that we’re aiming for.
Entry filed under: Uncategorized. Tags: computing for everyone, educational psychology, non-majors.
Computing Ed Research - Guzdial's Take 
1.
alanone1 | November 28, 2013 at 6:57 am
It is difficult to see what this adds to Mike Csikszentmihalyi’s work from many decades ago. The author references Mike’s 1970s book “Beyond Boredom and Anxiety”, but there seems to be little new content beyond this, and the subsequent “Flow” investigations.
2.
Ken Bauer | November 28, 2013 at 9:16 am
Perhaps so. I don’t have enough experience in this literature but I am at least interested to know that the results still hold these many years later. Society always talks about how much technology has changed the stage for teaching and learning. More research giving us data points like this are a good thing and should not necessarily be discouraged.
That said, the author is claiming some new findings “Why has past research not found a non-linear relation between self-efficacy and interest?” I leave it to those with more knowledge of the research body to judge whether this is true; for myself, I trust that the editorial process made sure this was true.
3.
Mark Guzdial | November 28, 2013 at 9:24 am
I don’t think self-efficacy is the same construct as Flow. It’s a sub-construct. In Susan Wiedenbeck’s work (linked above), male students with too high self-efficacy failed CS because they misjudged their abilities. That didn’t have anything to do with a state of Flow.
4.
alanone1 | November 28, 2013 at 10:26 am
Right, I didn’t think it was the same as Flow, but I also didn’t realize it was a technical term (I took it literally as “one’s ability to do something”, and this is closely allied with one of Mike’s two axes). Now I see what they are trying to label with the term.
Still, one of the influences on the GUI at PARC was to look at the two axes of Challenge vs Ability, and to note that the Flow state can be very narrow and both the anxiety and boredom states wide. In a UI you want to widen the situations where Flow can happen, and you can do that by making the challenge side more safe (e.g. with UNDO) and by making the ability side more interesting (e.g. to introduce esthetics, play, contemplation, helping to notice more, etc.)
5.
Mark Guzdial | November 29, 2013 at 1:04 pm
Hi Alan,
Completely agreed on the issues in the GUI at PARC. I think that work on self-efficacy (and I should have mentioned that it was a technical term, a particular psychological construct) really picked up after Carol Dweck’s work on fixed vs. growth mindset. Self-efficacy is related. A fixed mindset is a belief that self-efficacy is fixed — you are only effective at some tasks, at some level, and nothing can change that. Growth mindset is a belief that self-efficacy is malleable, that it can be expanded.
Cheers,
Mark
6.
shriramkrishnamurthi | November 29, 2013 at 1:23 pm
Interesting that you related this to Carol Dweck’s, but then also a little scary, given how hard it is to change those mindsets…
7.
alanone1 | November 29, 2013 at 11:36 pm
Hi Mark
Seems as though the older terms “ability” and “perceived ability” cover this well (though please correct me if I’m wrong). Then “fixed ability” and “improvable ability” would cover the other two jargon phrases.
I would certainly be interested in seeing the statistics on how many people don’t think that practice at something would improve their abilities. I’m guessing that the deeper problem is the worry that for some tasks practice would not improve ability enough, or that it would take too much time to improve enough.
I’d also like to see the statistics on the accuracy of these self-assessments.
There is probably something that is analogous to Vygotsky’s Zone of Proximal Development (a very opaque jargon term itself) for the extension of ability that can be achieved by a “reasonable” amount of practice/effort.
Just a guess, but I’ll bet that social acceptance and judgement is a much much bigger influence on motivation than any of these (e.g. amateur sports).
Cheers,
Alan
8.
Mark Guzdial | November 30, 2013 at 5:46 pm
Hi Alan,
The current interest in self-efficacy seems to be due to Albert Bandura’s work in social psychology. His definition of self-efficacy is “the belief in one’s capabilities to organize and execute the courses of action required to manage prospective situations.” Self-efficacy is close to “perceived ability,” but Bandura’s work emphasizes the social influences, e.g., the social pressure from statements like “women aren’t good at math” are an influence a female student’s self-efficacy, perhaps in spite of having high ability. Bandura explored world views that emphasize a lack of control by the individual in their future (e.g., a belief in destiny), where practice may not be relevant. Yes, I would agree that the social acceptance and judgment are a major factor on motivation and self-efficacy.
Cheers,
Mark
9.
shriramkrishnamurthi | November 28, 2013 at 9:22 am
The consequences seem to apply to programming in general, where (as you know) we HtDP folks have been arguing that we need to overcome the “blank page syndrome”. The new Captain Teach interface (which Kathi previewed at GA Tech) arguably shifts the focus not just from one large task to several smaller tasks, but from one where perceived self-efficacy may be low (“can you write this program?”) to one where it may be higher (“can you tests that describe the input-output behavior of this program even if you don’t know how to write the program itself?”).
The many components of the HtDP design recipe (data definitions, contract, headers, tests, etc.) are arguably reductions of not only size but also kind, and may hence help a student feel more adequate. Of course I’m speculating, but it may be worth measuring.
10.
Mark Guzdial | November 29, 2013 at 1:07 pm
It certainly would be worth exploring. You’re also reducing cognitive load. The question is whether students have interest in writing tests (and which students do, and which don’t). Self-efficacy influences interest, but isn’t the same thing. Have you explored epistemic frames? I think that might be a useful theory in your work — you speak to an epistemic frame of professional practice.
11.
shriramkrishnamurthi | November 29, 2013 at 1:27 pm
Thanks for the pointer. The term wasn’t familiar to me (except inasmuch as I can infer a meaning from the constituent words). Will go do some reading.
Note that testing is only one part of the design recipe; in principle one could skip it while still retaining the other parts (though in practice most people skip the other parts while keeping that one part!).