No Comments on reality, part 4

Before moving on, I wanted to add a little more to the story of the neuroscience of free will, including another experiment which I find particularly interesting.

Despite his findings, which I discussed in the previous post, Libet did not believe that his study disproved free will. Instead, he believed that even though the genesis of an action might be unconscious, the conscious mind still plays a role – namely, to approve or veto each action. Thus, free will (Or ‘free won’t’, as it is sometimes called in this model) plays the role of a gatekeeper for actions, rather than an initiator of them.

Unfortunately, ‘free won’t’ might not be a conscious process either. A 2008 experiment by Kühn and Brass asked subjects to press a button as quickly as possible when a monitor displayed a ‘go signal’ (such as a green light). In some cases a ‘decide signal’ (such as a yellow light) would be displayed some interval after the green light appeared, in which case subjects were told to either stop and not press the button or continue and press it. Because of the varying intervals before the decide signal was displayed, in some cases it would be displayed quickly and the subject would have plenty of time to make a decision, and in other cases it was displayed too late after the go signal for it to be even possible for the subject to decide (i.e. in those cases the subject could not react in time to stop the action of pressing the button).

This was reflected in the reaction times of the different outcomes: In ‘control’ trials where a decide signal was not shown, the average reaction time was 600 milliseconds. In trials where a decide signal was shown and the subject (theoretically) had time to decide to press the button, the average reaction time was a longer 1400 ms (Attributable to the time to make a decision). In trials where a decide signal was shown too late, the average reaction time was 600 ms – the same as the control, again because it was simply impossible for the subjects to react and stop the action. In trials where a decide signal was shown and the subject had time to decide to not press the button, no reaction time was measured (because no action was taken).

Here’s where it gets interesting. After each trial, subjects were asked whether they had truly made a decision (i.e. saw the go signal, began the action to press the button, saw the yellow light, and decided whether to press the button or not) or if the decide signal had occurred too late for them to be able to make a decision. It seems as if the identification would be easy – after all, we know when we’re making a decision and when we don’t have time to, right?

Evidently not, because a significant fraction of the 600 ms reaction time decide trials (That is, the trials where the subject theoretically could not react fast enough to stop themselves from pressing the button) were labeled true decisions rather than ‘too late to make a decision’. Now, is it possible that some of them were actually decisions – that the 600 ms reaction time didn’t necessarily imply an inability to make a decision? Maybe. But more likely in my opinion is the conclusion that Kühn and Brass came to – that in those trials, the action of pressing the button was completely unconscious impulse and only after the subjects became aware of the action did they believe they had made the decision.

It’s a frightening concept, that our own brains can trick us into believing we had autonomy over a decision that actually occurred subconsciously. And it’s tempting to extrapolate further, to conclude that all choice is illusion, a biological fail-safe to preserve the feeling of control and keep us from going insane. Indeed, if all possible realities exist, why does it particularly matter which one I experience?

Next: Why indeed?

Education (Part 2: Intrinsic and extrinsic motivation)

No Comments on Education (Part 2: Intrinsic and extrinsic motivation)

Part one

In working on this next part of my thoughts on education and learning I found myself trying to go in too many directions at once, developing ideas from the first part while also bringing up new things. As such I decided to limit myself to one topic at once. This part will be dedicated to the idea of intrinsic versus extrinsic motivation.

The topic of why we learn new things, which I brought up in the first few paragraphs of part one, is very much a discussion of intrinsic vs. extrinsic motivation. Should we learn because we’re curious and find new topics exciting, as I argued strongly for? Or should we do so in pursuit of good grades, goals, rewards? The various reasons that we learn – or, more broadly, that we do anything – can be categorized as either intrinsic or extrinsic motivations.

Extrinsic motivation is essentially operant conditioning: An action is motivated by the promise of reinforcement or punishment. In the context of learning and education this would be a parent promising to pay their child $20 for good grades, or threatening to ground them for bad ones. This is the style of learning that I associated with the stereotypical premed student, working hard in school so that you get good grades and the things associated with them: Acceptance to prestigious schools, scholarships, so on and so forth.

Intrinsic motivation is a little more elusive. It’s the desire to do something not because of external reinforcement or punishment, but because one wants to do that thing. The reward isn’t tangible because it’s internal: Curiosity, passion for a subject, that sort of thing. Intrinsic motivation is the time I was sitting outside in winter and wondered how much snow has fallen in the history of the Earth (My estimate was about 3×1026 kg of it, or roughly 50 Earth-masses of snow. I have that saved in a file somewhere, though I’ll need to look up the work I did to reach that number).

How do we apply these ideas? As it turns out, extrinsic and intrinsic motivation tend to be most effective in their own particular domains.

A 1987 study by Malone & Lepper suggests that intrinsic motivation is most common for activities involving challenge, curiosity, control, cooperation and competition, and recognition (Although I personally would characterize competition and recognition as extrinsic rewards – the desire to be superior to others and to be recognized as so seem more like external things to me). Challenge and curiosity, to me, can be characterized as tasks involving what I’ll refer to as higher order mental processes: Analysis, investigation, planning, invention, that sort of thing. Control and cooperation encompass tasks that offer a sense of meaning or purpose. So, simplistically, intrinsic motivation is best for tasks that make you think or feel.

Extrinsic motivation, on the other hand, is generally going to work best for simple, repetitive tasks that require little higher-order thinking. Memorizing the multiplication tables or learning the basic rules of English grammar are boring tasks that are unlikely to ever inspire any sort of internal passion. An interesting caveat here is the overjustification effect, which says that offering extrinsic rewards can actually reduce one’s intrinsic motivation for activities. Lepper, Greene and Nisbett (1973) found that children offered rewards for an activity that they could be expected to enjoy, drawing, showed less interest when the offer of reward was retracted than children who either received a reward without being notified before the activity or those who were neither offered nor given a reward. For this “expected reward” group, interest in the activity was only maintained through continued offer of reward. Although this seems strange at first blush, it actually makes a lot of sense. The offer of extrinsic reward for something you already enjoy doing essentially “cheapens” the experience, and when the reward is removed the activity simply seems less appealing. Put another way, we can imagine that there’s only a certain amount of motivation you can have to do something – we’ll call it 100% motivation. If you’re already at 100% motivation intrinsically, the addition of an extrinsic reward doesn’t push you to 150% motivation, but instead it makes you 50% intrinsically and 50% extrinsically motivated, for a total of 100%. Removal of that extrinsic reward now leaves you at only 50% motivation. On a more amusing, though less applicable note (and confirming the behavior of teenagers everywhere), the threat of extrinsic punishment can actually increase one’s intrinsic interest in something.

So what do we do with all this? It seems like intrinsic motivation should be the clear winner. After all, curiosity and critical thinking are the buzzwords of the day (or year, or whatever) in education. We need to teach kids to be active learners, foster curiosity, bla bla bla. But most elementary school children have no intrinsic interest whatsoever in math or science, challenge and curiosity be damned. Malone and Lepper suggest designing classroom instruction so it promotes their big themes: challenge, curiosity, control, cooperation and competition, and recognition. For the most part I agree here: we need to teach in a way that challenges students, that provides new experiences and promotes curiosity about the world, that makes students feel like they are in control and accomplishing something.

The problem is that we’ve either completely forgotten about extrinsic motivation or are angling to get rid of it altogether. Sure, it’s true that throwing gold stars and participation ribbons all over the place is terrible if we want kids to be self-directed and self-motivated. But remember those things I said extrinsic motivation is good for? Repetitive and simple tasks, like memorizing the multiplication tables or learning basic spelling and grammar? Yeah, it turns out that these things are the foundations for all our activities that are best motivated intrinsically. It doesn’t matter how curious a kid is if he doesn’t know basic arithmetic or the alphabet. These are things that you have to know to be able to do all that other stuff, and it turns out that in the vast majority of cases there is simply no way to get a kid excited about addition and multiplication – it’s something that just needs to be drilled in, and extrinsic motivation is going to be the best way to do that. It’s also going to be the best way to motivate kids to do things until they begin to develop intrinsic interest. Note that this doesn’t contradict the overjustification effect: Here we aren’t trying to exceed the 100% motivation cap. Instead we’re starting at 0% motivation and bringing it up a bit with extrinsic motivation until the student recognizes the intrinsic value of the task.

So both intrinsic and extrinsic motivation have a place in education. Rewards and punishments are necessary to motivate students to learn foundational knowledge that simply cannot be motivated intrinsically, and to create some “buffer motivation” while intrinsic interest develops. Once a student has recognized the value of a subject, however, extrinsic motivation needs to be scaled back so that they can properly reap the rewards of curiosity, challenge, bla bla etcetera. Something I touched on in part one and that will be developed later is grades: They can be categorized as extrinsic motivators, so based on what we’ve said here, should they be removed? For now, though, I think that’s plenty for one post.

Some ideas I want to develop in future parts: Testing and accountability, group vs. individual work. As always let me know if you’ve got other things you want me to consider.