Source: The Naughty Seat

Source: The Naughty Seat

Often in today’s society when a parent is faced with a child who is disobedient, not using the toilet as the parent would like, not doing homework, not cleaning up, or any number of things that parents want their children to do, they are told to start up a reward chart.  You know, the kind where kids earn stickers or stars or even a checkmark for every time they do something “good” and after a certain number, they get a tangible reward.  It could be a trip to the park, a dessert, a new toy, whatever.  You get the picture.

[Notably, reward charts are different from tracking charts which are the same or similar but without the reward at the end.  These can be incredibly helpful for children.  So if your child is tracking behaviour  without any (tangible) reward or punishment (verbal or otherwise), then it’s okay to use something to help them remember things.  This may also help them develop their own sense of autonomy regarding their behaviour.  Rewards are critical to reward charts.  Hence the name.  I would also like to take this moment to say this post refers to neurotypical children as there are children with special needs for whom reward charts have been found to be immensely beneficial.  I do not intend this to refer to these cases at all.  They are the reason I added ‘almost’ in the title.]

The basis of reward charts stems from early behaviourist theory (yes, the same theory that told us holding children too much will spoil them or that you need to ‘condition’ your child to not cry at night by not responding).  The idea is that children will receive a reward which will reinforce the behaviour you desire.  And not surprisingly it works.  When there is a reward, children will work towards that goal, sometimes even harder than they might if there were no reward.

So what’s the problem?  Isn’t this great?

Well, I suppose it is if you always have a reward and you can guarantee that your children will always be rewarded for what they do.  And not just feel-good rewards, but extrinsic, tangible rewards of things they want.

But what if there isn’t this reward?  Most parents don’t want to have to keep giving rewards for cleaning up or doing homework.  They want their child to internalize the acts so they do it on their own and you can eventually get rid of that nasty chart and the rewards that go with it.  This is where the concept of intrinsic motivation comes into play.

For those unaware, the term “intrinsic motivation” refers to the idea that some activities provide their own inherent reward or enjoyment, and thus motivation to do these activities is not based on external rewards or control[1][2].  If I take EP, as an example, when I started writing, I had no external rewards.  No one was really coming to the site and it certainly didn’t make money (though that hasn’t changed too much).  But I loved writing the articles and doing the research and just expanding my own mind in the process – I was intrinsically motivated.  As a child I was the same with puzzles or logic problems or math problems.  I could sit and spend hours doing them simply because I enjoyed them.  Didn’t need any type of external reward for them.  Generally people have at least a few things that they feel this way about and many other things they don’t feel this way about (and probably many of them are things we were rewarded for as children).  This is what most of us report wanting for our kids.  We want them to find some level of internal satisfaction in what they do, whether it’s homework, cleaning, playing, etc.  When they learn new skills, we want them to learn to enjoy them for what they are so they’ll continue to try and learn and master new things as they get older.

Now imagine if you were working on something new or learning a new skill and every time you engaged in it, you were really working towards a external reward.  Then the reward was taken away.  Would you continue to do that work?  Well, what research has shown us is that you might not, at least not to the same degree as if you had never been given that reward in the first place (e.g., [1]).  The intrinsic motivation is reduced, even if the level of enjoyment reported for the task remains the same.

This is not just something parents have noticed (though yes they have noticed it over the years), it has been tested experimentally in many, many studies.  One of the more recent meta-analyses includes 128 studies on the topic[3] with results I’ll cover in a moment.  But first I want to mention that there are different types of rewards that can be given and may not be equal.  In one of the first studies on the topic (i.e., [1]), an external reward of money decreased intrinsic motivation in college students, however, verbal rewards in the form of praise for the work done seemed to increase intrinsic motivation so if you use a tracking chart (as mentioned above) and do a verbal celebration or encouragement, that’s awesome.  One of the important areas, then, to consider is the type of reward given, something addressed in this meta-analysis.

So… what does the research find?

Type of Reward

Effect on Intrinsic Motivation (IM)

All rewards together

Free Choice Behaviour (as measure of IM): Small but significant effect lowering IM.
Self-Reported Interest: No effect on IM.

Tangible Rewards

Free Choice Behaviour: Small-moderate effect decreasing IM, but this was tempered by differences in unexpected and expected tangible rewards (see below).
Self-Reported Interest: No effect on IM.

Unexpected Tangible Rewards

Free Choice Behaviour: No effect on IM.
Self-Reported Interest: No effect on IM.

Expected Tangible Rewards

Free Choice Behaviour: Moderate-small effect decreasing IM.
Self-Reported Interest: No effect on IM.

Verbal Rewards

Free Choice Behaviour: Small-moderate effect increasing IM; however, results differed by age.  Children showed no effect while college students showed a moderate effect.
Self-Reported Interest: Small-moderate effect increasing IM regardless of age group.

The authors also compared the type of contingency with rewards, but the results are quite similar to what you see here so I’ll leave it at that.  One thing I would like to point out is that “small” effects are not negligible effects.  The authors, and thus I, use effect size terminology which explains effect sizes in practical terms.  This means a “small” effect is practical and you will notice a small difference in life (on average, of course, as stats are based on large populations).  This is different from statistical significance (although all the effects here are statistically significant as well) which can be influenced greatly by sample size.  So what we’re seeing are real effects.  Take a group of people and the differences will be noticeable in their behaviour, although it will be small.

What do we take from this?

Well, first, rewarding children’s behaviour with tangible rewards will at best not increase their intrinsic motivation and at worst will decrease it.  All this in spite of still verbally maintaining the same level of enjoyment of an activity.  This is not too surprising to many developmental psychologists as they know that from birth onward, children are naturally curious and inquisitive and eager to learn; in short, they have an abundance of intrinsic motivation[4].  It’s up to us to maintain that and while there are other factors (such as the tendency for an individual to attribute success to an internal locus of control versus external[5]), reward charts are one way we can interfere with this innate sense of intrinsic motivation.  Interestingly, some research shows that individuals with an external locus of control are particularly susceptible to the effects of reward charts (while those with autonomous control may not show any ill-effects)[6]. (This raises another question of for whom reward charts may be particularly negative or have no real effect, a question that deserves further research.)

[Side note: A quick example of external locus of control versus autonomous control would be comparing two children who dislikes homework but do it anyway.  The child with an external locus of control would do it because his parents make him/her – s/he is under their control.  The child with autonomous control does it because, although s/he dislikes it, s/he sees the value to her/his future plans and thus derives the motivation to do it for this external reason.  Neither of these children are showing intrinsic motivation and this is where the locus of control aspect enters the discussion as we all have things we do – or at least start doing – for external reasons.  Why we continue them, however, is the question.  And clearly the reward chart is a very clear example of not only external reward, but an external locus of control as well as it is set up and rewarded by someone outside of the individual.]

As parents or teachers or caregivers, I would hope we are cognizant of the effects we have on our children.  Fostering an external locus of control and the use of reward charts are ways in which we inadvertently reduce our children’s drive to be curious, to learn, to explore, and to master the skills they are interested in or just simply need to master.  These types of teaching do nothing to help a child grow up to feel responsible for his or her own actions and more importantly to learn to enjoy and find inherent pleasure in many of the more mundane acts.

So ditch the reward.  It may take longer to achieve the types of behaviours expected of a child, but in the end it will be completely worth it.

Update: A reader pointed out that the meta-analysis cited herein was updated by other researchers (it’s a bit of an ongoing battle or so it seems) with different results.  This meta-analysis (which discusses the Deci et al. meta-analysis and the previous one by Cameron et al.) found that the effects of rewards were more varied than Deci et al. found.  Notably, there were negative effects of rewards when the task was initially high-interest, when the reward was tangible, and when the reward was expected.  Furthermore, when breaking it down further, the strongest negative effects were when a reward was given either for “doing well”, “doing a task”, “for each unit solved”, or for “finishing a task”.  The effects of reward were non-existent (not positive, not negative) when they were task non-contingent (just showing up; not relevant to the issue herein) or for surpassing a certain score on a task.  The effects were positive for “exceeding others” (competition based).  For low-interest tasks, the reward increased intrinsic motivation.  Similar to what I reported above, verbal rewards were associated with increased motivation.  Notably, “high” and “low” interest are relative as most of the tasks given were not super-exciting, but the students were happy to try them.
What can we take from the new research?  In my opinion, it comes down to how much natural curiosity and interest we give to our kids.  If we’re starting with older kids who don’t have this (or have lost it) and we can’t find a way to get them interested in a task on their own (i.e., make it a high-interest task), perhaps the findings from the low-interest tasks will help.  However, the majority of times that we use tasks are for things like simply finishing a task or doing it or doing well, the reward is tangible and expected, and children are often at the start curious and interested even if they aren’t succeeding as much as we’d like (i.e., they are not low-interest tasks).  This means even the results of the new meta-analysis suggest negative consequences for the majority of times when reward charts are used.
My final suggestion is such: Start with a tracking chart, see how it goes.  If your child is struggling, try to understand why and work with them on it.  Adding in rewards should be done only when absolutely necessary and you are certain the task is so low-interest to your child that they need it.  Unfortunately, most of the tasks we give our children really shouldn’t be “low-interest” or without other “rewards” that they can see themselves.  Homework has the reward of better grades, going to the bathroom has the reward of cleanliness, and so on.  Our job as parents is to foster the initial intrinsic motivation and then work to help them see the natural rewards that exist.

[1] Deci EL.  Effects of externally mediated rewards on intrinsic motivation.  Journal of Personality and Social Psychology 1971; 18: 105-15.

[2] Ryan RM, Deci EL.  Intrinsic and extrinsic motivations: classic definitions and new directions.  Contemporary Educational Psychology 2000; 25: 54-67.

[3] Deci EL, Koestner R, Ryan RM.  A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation.  Psychological Bulletin 1999; 125: 627-68.

[4] Harter S.  Effectance motivation reconsidered: toward a developmental model.  Human Development 1978; 1: 661-9.

[5] deCharms R.  Personal Causation.  New York: Academic Press (1968).

[6] Hagger MS, Chatzisarantis NLD.  Causality orientations moderate the undermining effect of rewards on intrinsic motivation.  Journal of Experimental Social Psychology 2011; 47: 485-9.