Adam Robbins Education

Praise is used in schools in a multitude of ways. The psychology of praise is fascinating and confounding. Often things we praise students for are poorly thought out and can have unintended consequences. I want to unpick a few of these ideas, challenge some and offer a rule of thumb I think might be useful. 

To make life easy I am going to focus on praise tokens, not verbal praise. Although for the record the psychology is the same, it’s just that verbal praise is more transient and therefore harder to build into policy. This year I have been trying to only write about things people can actually use and implement, so I am narrowing the remit here.

Praise tokens come in a variety of forms from stickers, sweets, merits, achievement points, postcards home, certificates etc..they are all essentially a concrete acknowledgement of a message and that message is ‘well done’. From here on I will refer to these as achievement points, but again it’s all basically the same.

Schools use achievement points in a variety of different ways but the overall goal of all of them is to motivate students and encourage certain behaviours. In that way schools can often be described as utilising a behaviourist approach to motivation. The basic idea is that students will modify their behaviour as a result of a potential positive stimulus. It’s pretty old and in some respects very useful. On the whole these systems do work. 

Another psychological approach to motivation looks at a cost:benefit calculation. It argues students will look at the potential rewards of a behaviour and make a choice to see if it is worth the effort they have to invest. This effort is not just the physical effort of the act but also the potential costs to social status, opportunity costs etc.. According to this approach a student who gets rewarded for their homework will complete the homework providing getting the reward/ avoiding the sanction is worth the loss of time and possible negative peer effects (if the group as a whole is anti-homework).

Both of these approaches have examples where they seem to allow us to make accurate predictions of how people will behave, but they also are far from universal. 

A third theory tries to bridge these gaps. Self Determination Theory (SDT) is a raft of mini theories that try to explain how intrinsic motivation arises. It is often boiled down to the broad psychological needs of autonomy, competence and relatedness. While this is broadly true it does lead to some over-generalisations of its implications. Today we will look at a couple of ideas within SDT and what they tell us about rewards. 

1. Cognitive Evaluation Theory (CET)

CET is a sub-theory of SDT that focuses specifically on how external events, like rewards and praise, influence intrinsic motivation. According to CET rewards and praise can have different effects depending on how they are perceived by the individual. People can perceive rewards as controlling or informational.

Controlling Aspect: If rewards or praise are perceived as controlling (e.g., “If you do X, you’ll get Y”), they can undermine autonomy and reduce intrinsic motivation. This is because the activity may start to be seen as a means to an end, rather than as inherently valuable. So if you say to a student “if you collect all the pens up I will give you an achievement point” then the student is doing it because of the reward. This can make them less likely to repeat the behaviour of their own volition in future. 

Informational Aspect: If rewards or praise are perceived as informational (e.g., “You did a great job because of the effort you put in”), they can enhance feelings of competence without undermining autonomy. This can actually boost intrinsic motivation if it reinforces a sense of mastery or self-improvement. So saying to a student “Thank you for picking up all the pens when I asked, here is an achievement point for respecting the classroom environment.”

2. Organismic Integration Theory (OIT)

OIT addresses the internalisation of extrinsic motivation, describing how external motivations can be assimilated into one’s sense of self. OIT helps explain how external factors can become intrinsically motivational to some people much more easily than others. 

OIT outlines a continuum of motivation from external regulation (where behaviour is driven purely by external rewards) to integrated regulation (where behaviour is fully assimilated into the self).

OIT suggests that the way rewards and praise are framed can influence where on this continuum a person’s motivation lies:

External Regulation: If rewards are used to control behaviour, motivation remains external and can lead to compliance without engagement. Think of this as a kind of behaviourist interpretation. For example, I do my homework to avoid a detention.

Introjected Regulation: Praise that induces guilt or pride can lead to a form of motivation where the person acts to avoid negative feelings, rather than out of true internalisation. For example a student might complete their homework to avoid feeling guilty when the teacher confronts them about incomplete work. Completing your homework to make your mother proud would also be an example.

Identified and Integrated Regulation: When rewards and praise are aligned with the person’s values and self-concept, they can promote internalisation, leading to more self-determined forms of motivation. In this case we could think of this as a student doing their homework because they want to do well in their studies.

NB: If you are thinking there is a developmental aspect to OIT as you would expect yr7 to be externally regulated but yr11 to be integrated then you are probably right. There is no scale put on these as everyone is different and even an individual’s motivation will vary according to the task. However the idea that we can learn these skills is key to why it is worth trying to develop them in the first place.

It is worth making it clear that SDT focuses on intrinsic motivation. That is what will our students be motivated to do on their own in the future. Behaviourist systems are not inherently wrong in schools because they have the systems in place to ensure the conditions of consequence are always in place. But what if we could avoid some of the undermining of intrinsic motivation and enhance our systems to grow intrinsic motivation over time? Would this improve school attendance, or discretionary effort, for example? 

When rewards go wrong.

Harvard University ran a famous study on attendance rewards in the state of California. It looked at the idea of broadcasting rewards. The hypothesis was “If we advertise that students with good attendance will receive a reward at the end of the semester then attendance will improve.” This is how the study was set up

The outcomes were surprising. The prospective group had no improvement compared to the control group. The reward was not significant enough to change students’ behaviours and after the study was finished the students attendance was also unaffected. The intervention was completely pointless. The attendance of the retrospective group was also unchanged (as expected).The surprise was when tracking the attendance of the retrospective group after the letters were sent. The retrospective reward group actually missed more days of school. Absence increased by 8.3% compared to the control. This is a great example of two things, firstly that rewards communicate information, and secondly that in this case the information they communicated that the child was above average on their attendance. This sat in the parents’ brains and the idea that ‘they were above expectation’ caused more students to be allowed to miss school for various reasons as they somehow had a surplus of education. 

What makes an effective reward?

When looking at the ideas of SDT it can be hard to figure out what the most optimal reward system might be if we want to foster intrinsic motivation for certain behaviours moving forward. For example we could misinterpret the need for autonomy and think rewards need to involve choices, this is not the case. Autonomy is important but autonomy in SDT is not limited to being able to make your own choices. In OIT it tells us that if people have an integrated view of the external controls of the school systems then they maintain autonomy. To put another way, if a student has bought into the concept of education or believes in the values of the school then they are willing to ‘sacrifice’ autonomy for the feelings of competence and relatedness that they get in return. 

There is also a way of defusing the controlling aspect of rewards by when they are offered. Rewards that are awarded after the event are much less controlling that those that are advertised as a reason for completing the event in the first place. You have seen this in the example used in CET above.

How frequent should rewards be?

The frequency of rewards is not something that any of our theories has much to say on. Reward consistency is really important in all the models we have looked at. Often I think that the cost/benefit approach to rewards is mistakenly interpreted as the need for high value rewards. If you only give a couple of reward points a lesson then they are more appreciated than if you give loads. But rewards do not follow the same kind of supply and demand economics that consumer goods do. What is important is its comparative value. If a student gets a reward point for attending every day in a week, then the currency of that reward is very low. We are talking Lira-levels on the rewards exchange rate. That means that if a student also gets a point for producing 3 paragraphs of work independently and getting on to the extension we are accidentally communicating those two things are equivalent. The in-class reward is devalued by the tariff for attendance. The best way to avoid this is to ensure rewards are very low tariff. Rewards should be cheap as chips and just as plentiful. It is much better for students to receive lots of small rewards throughout the day than a big one once and while. This also prevents point discrepancies. Often in reward systems teacher expectations play a role, so SEN students find it easier to gain rewards than the more able, and disruptive students get rewarded for just sitting still. Having a low tariff reward system can sometimes lead to this problem of rewarding basic expectations, but there are ways to avoid this.

To summarise the key features of rewards:

1. They must be awarded for things that are above the basic level of expectations or they may lower expectations.
2. Rewards should not be high tariff. They should be attainable by all students all the time.
3. They must communicate competence or relatedness to the recipient.
4. They must be given after the event has happened and not pre-advertised as a way of motivating.
5. They do not need to facilitate autonomy in the truest sense of the word.
6. They can help foster greater buy-in to the school’s culture and as they do they will become less controlling and more intrinsically motivational. 

My rule of thumb for rewards.

How do we try to do all these things? Even more importantly, how do we try and do these things in a simple and sustainable way that can be implemented by all staff? 

My rule of thumb is you reward things that require the students to take risks.

By taking a risk I mean investing some social capital, mental effort or time. Essentially we reward participation that requires effort. If a student holds the door open, it’s a reward point. If a student is the first to get started in a class, it’s a reward point. Ask a really good question? Point. Answer a question clearly and correctly? Another point. If an EAL student attempts to pronounce ‘photosynthesis’ gets it wrong but then corrects it with your guidance, another point. Hopefully you get the idea.. 

This way any student can access the points and all students get these positive affirmations through the day. 

If you want to see this done fantastically go and visit Ark Soane Academy. Without a doubt they have the best reward culture I have seen. 

But our school doesn’t use rewards?

If your school does not use rewards but you have an effective culture then my guess would be that you have these systems in place, but just use verbal praise. Maybe go watch a few lessons and listen for how often students are thanked and praised. I would bet it’s loads and for similar things. I am not saying you must use physical reward tokens, the features of the system are what matters. 

Do you agree?

I am interested in your thoughts on this area. Does your school reward in this way? Are there any problems you find? Do you completely disagree with my rationale? Let me know either in the comments or on BlueSky/twitter/linkedIn/Threads