Wednesday, January 9, 2013

Part 4 - Psychology and Motivation

Now about psychology and, in particular, about motivation which are central to what makes gamification effective. Psychology needs to be discussed as people don't entirely know what makes them do something and motivation is complicated as people are complicated. We are not motivated by the same thing all the time but we are motivated by different things at different times. So we need to think about our players, about different ways to motivate them and about how to deploy motivation in a systematic way.

Prof. Kevin Werbach mentioned two major traditions in psychology: behaviorism and cognitivism. Behaviorism is about looking at behaviours, looking at what people do. Cognitivism is about mental states, what's internally going on in people's heads. Both of them are very relevant for gamification.

Now behaviorism says that when we look at a person we know that there are feelings , thoughts and emotions inside but we cannot focus on those internal mental states as they are not scientifically testable. So the behaviorist program is to restrict ourselves to what is external which is called the black box. The idea is what we can test is what goes in from outside and what comes out. Although behaviorism has some limitations when we focus on gamification, it is still instructive for some aspects of gamification design.

The basic notion of behaviorism is two things. One is stimulus. Something which happens externally creates a challenge, opportunity or reaction and that reaction is called response. This is when we can see or observe a certain type of behaviour and that is why this is called behaviorism in response to stimulus. This was first tested by Pavlov in what's called classical conditioning where he rang a bell and got the dogs to salivate. The idea in Pavlovian conditioning is that the stimulus is instinctively related to the response. So when the bell rings, it automatically produces this response of salivating whether there is food coming or not.

A more important kind of conditioning is associated with B.F.Skinner and is called operant conditioning which is about feedback loop. There is a stimulus and there is a response and based on this pattern there is learning. Subjects learn; they see the consequences of their actions and the consequences actually matter.

So these experiments have the notion that in certain cases people will respond to stimuli and will learn to make associations between two things especially if there are rewards. Observation is important here. We should observe what people actually do. If people respond to to stimulus in a certain way, we should learn something from that. Moreover, feedback loop is important. When the person involved sees some feedback,  it produces some response and a process of action. Feedback tends to motivate behavior which is very important for gamification.

Now let's look at FarmVille which managed to create what is called an appointment mechanic. The idea is that people know that they have to go back to their farm at certain times to water their crops otherwise the crops will wither. So FarmVille got people to learn, as a matter of habit, learning to regularly check their farms This is a behavioral kind of approach which worked well for FarmVille.

Although behaviorism talks conditioning and providing rewards, some benefits that seem valuable even though they are not tangible or worth any money, rewards are still only one piece of the game mechanics. Rewards are powerful and significant in gamification but they should not be the only thing to use in a gamified system.

But why are those rewards so powerful? This is related to brain chemistry, something called dopamine system - the structure of brain associated with pleasure and learning. Our brains release and reabsorb dopamine in response to certain activities (usually things that find rewarding or just surprising. It causes a learning process and makes people want to engage in the activity. Again this doesn't always work or doesn't always work for everyone, but when it does work it is the hit of dopamine.

If we talk about rewards themselves, firstly we should say that there are different ways to give rewards and there are many things that can be rewarded, so effective gamification will think about what can be rewarded, what kind of behaviour the designer want to reward and what the different options are. The goal is to give players/users a set of meaningful choices and a set of options to make the system feel more engaging. Secondly, there are different categories of rewards and one typology of different kinds of rewards is called Cognitive Evaluation Theory which comes from Richard Ryan and Ed Deci. What they developed is a typology of different kinds of rewards that can be used to motivate behaviour. One distinction that they point out is between tangible and intangible rewards. The second one is between expected and unexpected rewards. If the reward just happens, it is a surprise and our brain loves surprises which means that expected rewards are not so cool to our brain. The third set of distinctions is what rewards are dependent on: whether you get a reward without even doing anything, or you get one for simply engaging in an activity, or the reward depends on the completion of an activity, or rewards can be performance contingent, when the reward is given because the task has been done well.

In designing gamification the key is to think about different possibilities for rewards  and decide on the most effective ones that will motivate players/users and will ensure that there is a meaningful and rewarding experience for those players/users.

Let's move on to reward schedules which refer to when the reward is given as opposed to what it is or what it is based on. Behavioral studies suggest that the structuring of reward schedules has significant implications for the psychological reaction that the rewards produce. There are several possible reward schedules. One is continuous reward. It is given for every incidence of the action - it is automatic. The other kind of reward is a fixed ration reward when a reward is given when something happens a certain number of times. The third kind of reward is fixed interval reward (let's say after each second unit of a course). The fourth category of rewards is variable rewards which is on no fixed schedule. The other three kind of rewards, although they still have some psychological value, can become predictable, and thus dull, as the brain can pick up the pattern. But this last type is the most interesting out of all of them because as we said before our brain loves surprises. However, this type of reward should not be given for doing something that nobody can do, because then it would make it unachievable and players/users may just give up before they can get this type of reward.

However useful behaviorist approach is, there are some serious limitations to it. One of the problems with it that it leaves out what people think or feel or what is going on in their heads or what really motivates them when they act in a certain way. Now game thinking, we said, is about putting the player in the centre, but if we take a purely behaviorist approach, we might move away from the notion that the player is a human being, so this focus on rewards tends to have some problems as well.

One of the problems with this kind of system is that it may be designed to manipulate people, to make them do things they may not necessarily want to do. It may become like an addiction to them, which is already not a good thing. If we can design a gamified system that may addict people on this rewards system, then that does not mean that we should. Prof. Kevin Werbach compares this kind of behaviorist approach to that of casino owners.

The second problem with this is what is called Hedonic Treadmill. The idea here is once we start focusing on giving people rewards in order to give them pleasure, we will have to keep doing it. Because if people learn to respond to rewards, they are only going to respond to them. And as certain rewards get familiar or boring, we will have to come up with new rewards: make them more interesting, more challenging. So this can put a significant burden on the designer. There are some studies which look into what actually happens when dopamine system is activated. The studies showed that the dopamine system is not really about rewards, it is about anticipation of rewards.

If a gamified system focuses too heavily on rewards, it tends to miss some other kinds of benefits that can be delivered through a gamified system. PBLs are about status and not everyone is interested in or moved by that. We do things for many other reasons: social reasons, altruistic reasons, tangible reasons; so these need to be exploited as well.

The alternative to behaviorist approach is Cognitivism which focuses on opening up the black box to find out what really motivates people to behave in certain ways. In order to understand this, we need to distinguish between different kinds of motivation and rewards. In particular, we need to think about two broad categories called Intrinsic and Extrinsic Motivation.

Intrinsic motivation is about doing something for its own sake, not because there is any external stimuli, but because we find it rewarding, engaging, fun, or motivating. Here we are not focused on the consequences or any other kinds of benefit that we might get.

Extrinsic motivation is about doing something for some other reason rather than the activity itself. The reasons may be different: it could be money, fame or fortune; or it might be because somebody asked you to do it and you value the person so you do it for them. Gabe Zichermann talks about four kinds of extrinsic reward motivators which he calls SAPS: Status, Access, Power and Stuff. Status as a motivator is what makes us think that it is cool, this is done with leaderboards in particular. Access is about getting access to something that other people don't have: content unlocking could create the feeling of gaining access to something. Power enables users to do certain things that was a result of their activity: for example, edit other people's posts or submit posts bypassing moderation. Stuff is about tangible rewards, real things that the user can get in response to their actions. However, motivating these are, we still need to keep in mind the problems discussed above.

The biggest danger of rewards system is that it can actually demotivate by crowding out intrinsic motivation that was already there as it acts like an extrinsic motivator. This is sometimes called over-justification effect and it's a danger in any kind of system that uses rewards: a substitution effect where the intrinsic motivation goes away and is replaced by a less effective and problematic extrinsic motivation of rewards. The studies discussed in the link for 'over-justification effect' demonstrate this quite well.

In the 1970s Ed Deci and Richard Ryan developed the basis for Self-Determination Theory - a comprehensive theory of human motivation which, through many studies, has shown that people are not necessarily always motivated by rewards and, in fact, intrinsic motivation is a more powerful and more effective way to encourage people to act in certain ways.

The screenshot of Prof. Kevin Werbach's video lecture explains the spectrum of motivational types.

Amotivation is when the users have no motivation whatsoever, they are totally indifferent to the activity.
The broader category of extrinsic motivators is in the middle.
External Regulation is when users don't really want to do something or maybe they are indifferent to it, but they do it because someone tells them to do it.
Introjection is when we take external motivators and make them our own, mainly because other people will think that I am cool.
Identification is when we take external motivators and make then our own but not because it is important what other people will think, but because we  can see some value in it. This is when it is somehow aligned with our own personal goals.
Intrinsic Motivation is when the users do the activity for the pure reason of loving it., because it is rewarding in itself. This is the strongest motivation because it takes nothing external to get them to do the activity.

Now all of these may come useful in gamification, but what we should think about is how to utilize these different kinds of motivations appropriately and how to push towards approaches that are more dependent on intrinsic motivation. Under self-determination theory, there are three characteristics of intrinsic motivation: three factors that when they are present suggest that the activity is worthwhile. One of these factors is Competence which is about a person's sense of ability, the sense that they are accomplishing something, solving problems, overcoming obstacles. The sense that they are achieving something within the activity. The second one is Autonomy which is about a person's sense of being in control. Users should feel that it's them making the choices. And the third one is Relatedness, the sense that the activity one is doing is somehow connected to something beyond themselves that could have some meaning, sense or purpose. Fitocracy demonstrates this quite well.

We were offered to read two books if we wanted to learn more about Self-Determination Theory.
1. Drive by Daniel Pink
2. Glued to Games by Richard Ryan and Scott Rigby

Further reading:
David Freedman, Alfie Kohn, Sebastian Deterding

No comments:

Post a Comment