A few weeks ago when listening to a gaming podcast, I heard the hosts describe a particular game as “giving them their shots of dopamine” in terms of the pleasure they had experienced with the game, and their desire to keep on playing (dopamine being a neurotransmitter i.e. a chemical used in the brain). The comment was made off-hand but reflects a common view – that having dopamine released is related to pleasure and reward, and therefore is relevant to gameplay. But is this view correct?
Well, if we go back around 30+ years, the view that dopamine is the chemical related to pleasure and reward was being presented by researchers. One classic experiment that led to this view of dopamine being related to pleasure comes from even further back and involved rats that had electrodes in their brain stimulating brain areas that (it turns out later) can be responsible the production of dopamine . These rats could press a lever to get this part of the brain activated. In response to this self-administered brain stimulation the rats would push this lever at the expense of basically anything else. For example, they would rather press the lever than eat, be social, sleep, and so on. This, and other later evidence, led to this area of the brain being labelled as the ‘pleasure centre’ and seemed pretty convincing.
So, the conclusion at the time is that dopamine was the most important chemical that made us enjoy rewards and at the same time could motivate us to seek them out . As such, the idea that dopamine is a reward and pleasure chemical, spread and is now mentioned off hand by podcasters (a very scientific metric of cultural spread, I know).
Unfortunately (well, actually in the long run, fortunately) the brain is not that simple. Science has moved on and things have changed. Indeed, a leading researcher in the area has, jokingly, suggested that that the best answer to the question “what does dopamine do?” is “confuse neuroscientists” .
That answer, while amusing (I laughed at least), doesn’t really help those working in games to understand what their games may or may not be doing to the brains of those playing them. As such, the aim of this blog is for me to as clearly as possible explain what Science currently says about the role of dopamine and rewards. More than that though, I will also try and provide some comment in terms of what all this neurotransmitter stuff actually means for people who make games.
Please note, in the following blog I will be limiting myself to talking about what is known about the effect of dopamine (the neurotransmitter) itself and not discussing what is known about the brain areas that are related to its production or suppression. I have also limited myself to only discussing the most relevant and interesting (in my opinion) examples and experiments as this is not the place for a full academic review of the subject (If that is what you want, check out the academic reviews listed in the reference section at the end of this blog [4-8]).
Liking, learning, or wanting rewards in gameplay
One useful way to approach the question of dopamine, and what it does, is to break down people’s reaction to rewards in terms of liking, learning, or wanting. Which is to say, that when it comes to how you react to a reward (such as achieving something in a game) then whether you like the reward (i.e. is it fun), have learned the appropriate way to get the reward (i.e. can you actually play the game), and if you want to work to get the reward (i.e. is it motivating to play), can be completely different, and independent, things.
Liking to play
It seems to make sense that if you do something for a reward then you must like that reward. This is indeed part of the reasoning that led early researchers looking at dopamine to assume that dopamine was to do with pleasure (‘liking’) . Basically, they thought, why else would the rats have pushed that lever for so long if they weren’t enjoying themselves?
The idea that the self-stimulating rats were enjoying themselves may have been an understandable assumption. However, it turns out that while rewards are often liked they do not have to be liked in order to be effective in changing behaviour and, furthermore, that dopamine itself does not appear to be directly involved in ‘liking’ and pleasure.
Specifically, as research has moved on, it has become clear that animals (usually mice and rats) that have had their ability to produce dopamine stopped or restricted (either via drugs, surgery, or genetic alteration) can still be seen to demonstrate that they ‘enjoy’ and ‘like’ things. One demonstration of this would be that mice that have been genetically altered to not produce dopamine (a type of mutant “knock out” mouse, or an X-Mouse if you prefer) still show a preference for sugary water and other foods . In that, these mice like and seem to show signs of enjoying the taste of sugary water and when given the choice will pick to drink it over plain, non-sugary, water. Furthermore, you can also genetically engineer mutant mice to have excess dopamine and these animals do not show any additional/enhanced signs of ‘liking’ different foods despite all the dopamine floating around in their brains .
So, that is mice but what about humans? Well, researchers aren’t really allowed to make mutant humans and it is not easy to get permission to give drugs or do brain surgery on people. But, we can look at patients with Parkinson’s disease, which is characterised by problems with dopamine production. These patients, much like the rats and mice, also do not appear to show any decreases in the liking of rewards, such as sweet tastes .
Given the above findings, and many others (e.g. see the reviews of [4-8]), it became hard for researchers to continue with the idea that dopamine is a pleasure or ‘liking’ chemical. In fact, by 1990, Roy Wise, a leading researcher and initial proponent of the idea that dopamine is related to pleasure stated:
“I no longer believe that the amount of pleasure felt is proportional to the amount of dopamine ﬂoating around in the brain” ( - p. 35).
Indeed, rather than dopamine it seems that other neurotransmitters, such as opioids (Endorphins! Remember when they were trendy to talk about?) and cannabinoids are actually more often involved in ‘liking’ a reward [4, 10, 13-15]). Although, it should be noted at this point that opioid release in the brain can, indirectly, also lead to reactions in the dopamine system, which again could explain the early confusion over the role of dopamine. However, as mentioned above, mice that are genetically unable to produce dopamine still do like things!
So, that pleasurable feeling you get when playing a game? Dopamine is probably not the cause. In the same fashion, if someone tells you their game is designed to maximise dopamine delivery then this does not necessarily mean their game will be fun or enjoyable to play.
Learning to play
If dopamine isn’t related to pleasure, then what does it do? Well, another hypothesis, which became popular in the 1990’s, is that dopamine helps animals learn how and where to get rewards (a very useful thing to remember in games but also in life in general). This hypothesis arose when scientists started noticing that dopamine activity appeared to increase before a reward was delivered and therefore could be helping animals predict the arrival of a future reward [6-8]. That is to say that dopamine was produced when an animal saw a stimulus (such as a light coming on) that had been previously linked to getting the reward and, therefore, the dopamine release was predictive of the reward coming and not a reaction to the reward itself (as it would be if it was just about ‘liking’ that reward or getting pleasure from it). There also seems to be an increase in the activity in the dopamine system when a reward is unpredictable (like a random loot drop in a game). That dopamine activity increased most when an animal was expecting or learning about an unpredictable reward appears to make sense if dopamine is about learning. After all, if a reward appears to be unpredictable then you should pay more attention/try and learn about what signals the reward so you can work out how to better obtain that reward in the future .
Again, this evidence for dopamine’s role in learning looked pretty good [7-8]. However, once again mutant mice have shaken up this idea. In a quite clever study in this area, scientists at the University of Washington showed that not only do mice that cannot produce dopamine still ‘like’ rewards but that they were also capable of learning where a reward was . Specifically, these mutant knock-out mice could still learn that a reward (food) was in the left hand side of a T shaped maze, although they did so only after being given caffeine. The addition of caffeine to the mice is unrelated to dopamine production but was needed because without this mice that cannot produce dopamine don’t do much of anything. As you can see for yourself in this video with a normal and a dopamine deficient mouse the mutant mouse tends to just sit there. In fact, these mutant mice do so little that they will die from not eating and drinking enough unless given regular shots of a drug that effectively restores their dopamine function for a day or so .
In addition to the above experiment, it also appears that mice that have more dopamine than normal do not demonstrate any advantages when it comes to learning . However, as mentioned, the fact that dopamine deficient mutant mice will essentially starve to death means that dopamine must do something. But if dopamine isn’t for pleasure from rewards, and isn’t for learning about rewards (although argument here still exists), then what does dopamine do?
Wanting (desiring, needing) to play
It turns out, as far as where science is currently at (and remember science does, and should, change as new evidence is found), that it seems that dopamine is most clearly related to wanting a reward [4-6, 15, 17]. This is not wanting as it would perhaps commonly be used in terms of a subjective feeling or cognitive statement like “Oh, I want to finish Saints Row IV tonight” but rather as a drive, a desire, or a motivation to get a reward. So, this is not about a feeling of ‘liking’ and pleasure, instead what we are talking about is a feeling of a need or drive to do something. Subjectively this may be like when you just have to take one more turn in Civilization (or start playing and then 5 hours later realise you are still going) or when you have just get a few more loot drops in Diablo before you stop for the night. Indeed in the literature, when discussing the results of experiments on mice, some researchers suggest that dopamine creates a ‘magnetic’ attraction or compulsion towards obtaining a reward [3, 6]. Indeed, it could be argued that the evidence for dopamine being involved in ‘learning’ is in fact just a sign of ‘wanting’ being directed towards an uncertain reward, which then motivates learning to occur as a side effect (i.e. if I want something, I am likely to try and learn how to get it).
Again, here, we can look at mutant mice to confirm the role of dopamine in ‘wanting’. In mice that cannot produce dopamine, their motivation to move towards and work for rewards (which, remember, they do like and have learnt how to get) is deficient [3, 17]. This means that these mice do like sugary water, and they have learnt that the sugary water comes from the drinking tube on the right; however, they just aren’t motivated to walk over there and drink it . Conversely, mutant mice that have more dopamine than is normal have been shown to be more motivated to gain rewards, both in terms of how fast they approach rewards and how much effort they will expend to get the reward [3, 10]. Also remember those rats with electrodes in their brains working away at the expense of everything else just to get more stimulation? Well that can also be explained in terms of ‘wanting’ the simulation rather than ‘liking’ it [1, 2]. Kind of like how someone with OCD will wash their hands over and over and over again, even though they often get no pleasure from this act (and in fact it can be quite distressing).
Looking at humans, if we examine patients with Parkinson’s disease then there are also studies that show that some of these patients demonstrate increased ‘reward wanting’ and have compulsion problems when given a drug that enhances dopamine production. For example, such patients have been reported to go on obsessive shopping sprees and demonstrate other ‘manic’ type behaviour .
Furthermore, if we go back and look at reports of people who, like the previously mentioned rats , had direct (self) electrical stimulation of their so called ‘pleasure centres’ in their brain (usually for questionable medical reasons), then we see subjective reports of increased sexual desire or motivation to perform various activities (or just to press the button to self-stimulate, which they would do thousands of times). However, we do not actually see clear reports from these people of increased pleasure, sexual or otherwise (these accounts are mostly from the 70’s & 80’s where this kind of thing was going on, and are summarised in  if you are interested). In one highly ethically questionable example, the researchers actually hired a female prostitute at the request (although, one could easily question if this was a true, ethically acceptable, request) of an electrically self-stimulating man who was being ‘treated’, in part, for homosexuality, along with depression, drug abuse, and epilepsy . So, in these human accounts we see suggestions of what looks like a wanting response to self-stimulation of dopamine related brain areas but not necessarily a liking response (i.e. the subjects expressed increased desire but not necessarily increased pleasure). Although, it should be pointed out that if you were depressed, and then suddenly started feeling motivated to do things again that this may, as a side effect, increase your mood .
The upshot of all of the research mentioned above is that it appears that dopamine is not directly about pleasure (or learning) but rather it is about motivation or, if you want to be more sinister, compulsion. So, when those podcasters I was listening to said that a game was compulsive because it was giving them their dopamine shot, they may have been right. However, dopamine was not directly responsible for also making the game they were talking about fun (please note, I am not so serious that I expect videogame podcasters to be exact about this kind of thing, rather I am just using them as a convenient example).
What does this mean for games?
It is very popular at the moment to attach the term ‘neuro’ to almost anything. In academia this has led to a, justified in my opinion, neuroskeptic movement that is calling for, well, more evidence. However, there is no doubt that this ‘neurofication’ is popular with people. In fact, there is even research  showing that, at the moment, people seem to have a bias towards believing that data that is presented to them in a neuroscience-like fashion (i.e. via an image of a brain scan) is more scientifically valid than the same data presented to them in a more mundane fashion (i.e. via a bar graph).
So, what does knowing all this neuroscience mean for those who are making games? Well, from a strictly pragmatic and applied perspective it could be argued it means very little for every day game design. The neuroscience I have presented here is mostly pure, not applied, science and comes from the perspective that we already know that certain rewards and ways of delivering reward are particularly motivating and pleasurable (and may or may not have anything to do with dopamine to different extents). For example:
- Rewards that are unpredictable (loot drops) are generally more motivating than rewards that are predictable (100 xp per monster) [21-23].
- Rewards should be meaningful, e.g. food is not particularly motivating for most people if you are already full, or if you are in a relatively visually sparse setting then new, unusual, stimuli will attract your attention more readily .
- People tend to have a preference for immediate rewards and feedback and are not so motivated by delayed rewards and feedback. This preference for immediate gratification is strongest when young, but persists throughout life [24-26].
- Learning to get and want a certain reward is enhanced by immediate feedback about what behavioural response produced that reward. Uncertainty about what behaviour produced the reward will often lead to trial-and-error type exploration, which will be more likely to continue if further rewards arrive [23, 27, 28].
- If people perceive they are progressing towards a reward, even if that progress is artificial/illusionary, they are more likely to be motivated to obtain the reward (just one more turn…) .
- Similarly, people tend to report that they will work harder to keep what they have rather than to gain something they don’t yet possess .
- People have somewhat of a bias towards large numbers. Therefore to some extent will prefer, and be more motivated by, a system where they earn 100 xp per monster and need 1000 xp to level up over a system where they earn 10 xp from a monster and need 100 xp to level up [29, 31].
- A predictor for a reward can serve/become a replacement for that reward in terms of behavioural response (e.g. getting points in a game becomes associated with having fun and points can therefore become a motivating reward in themselves) [21-23, 29, 32, 33].
- People tend to dislike rewards that are delivered in a way that is perceived to be controlling [22, 34-36].
- Feelings of mastery, self-achievement, and effortless high performance appear to be quite rewarding, if somewhat more difficult to achieve than other types of reward [35-37].
As such, the neuroscience research I have discussed isn’t, primarily, aimed at working out how to make a reward motivating or pleasant but rather at understanding why it is so at a physiological level. Or if it does take an applied view, it is usually about using drugs or direct brain stimulation to get results.
As such, from a practical perspective in games, then looking at behaviour (such as the masses of data being collected all the time on player behaviour by metrics or even your own small scale playtests) for game design directions is likely to be more valuable than looking to neuroscience for answers. Indeed, it is likely that even if you could record the exact dopamine activity of every player that interacted with your game that it would not really produce a substantially different design outcome than just looking at what they do (i.e. their behaviour). One exception could be that a theoretical neurological approach may be able to detect if a player was ‘wanting’ to play your game without consciously realising it (something that may indeed be possible) but even in this case you could still see the same outcome in their future behaviour without having to worry about the (complicated and costly) neurology of it. All this said, if you are interested in knowing what your games may be doing to peoples brains, or perhaps you are working in serious games and want to see if games can improve (or worsen) brain function. Then, here, neuroscience can be valuable. But please, be neuroskeptical!
One possible application of the research I have outlined here is, I guess, that because at the moment people seem biased towards accepting explanations with a ‘neuro’ component as being more scientific , then you could argue that talking about games in this fashion could be a viable marketing strategy. Be careful though. At the moment people appear to think dopamine is related to pleasure. As such, they may not mind games being publicised or talked about as being designed to maximise a dopamine response. However, if the public perception changes and it becomes even clearer that dopamine is about wanting and motivation, not pleasure, then such messages become more sinister. In that, the message changes from “this game is designed to be fun, so you will want to play it” to “this game is designed to make you want to play it, even if, perhaps, you don’t like or enjoy doing so”. Furthermore, there is some evidence that people who are already stressed (physically and/or mentally) or lacking in stimulation are more vulnerable to the motivational effects of dopamine (this makes evolutionary sense, as if you are in a bad position at the moment you should be motivated to go out and take risks to try and improve things – but in modern life this tendency can sometimes be harmful) . The upshot of this is that if you design your game to really push dopamine buttons and tie that into some kind of monetization (or even if you are just asking people to give you their time), then you may have to take the risk that those aspects will be working best on those who are less able to defend themselves and may also be the least able to pay (and you can’t even necessarily suggest that at least you are giving them a fun time because liking, while often correlated with, is not needed for wanting). Aside from any moral feelings this may or may not produce for you, such ‘dopamine designed’ games (which would most likely be games that rely on uncertain reward systems with good direct feedback systems on behaviour) may attract the eye of governments and lead to regulation (as, it could be argued, it already has in Japan with the restriction of certain ‘gambling’ components in mobile games).
A complicated matter
Before I finish up, it should be noted that the brain is a complex subject matter. When neuroscience is talked about in the media there are often references to brain areas being ‘pleasure centres’ or certain neurotransmitters doing very specific things. But that is not the reality but rather a side effect of trying to tell a clear story. In reality physiology often follows a many-to-one or many-to-many pattern. Which is to say it is not usually a case that X causes Y, but rather that X, Z, B, and/or C can, if circumstances are right, cause Y or even that any of X, Z, B, and/or C can cause any of Y, J, A, K, and/or U, depending on the situation. Furthermore, since neurotransmitters don’t exist in isolation you often have instances where changes in one neurotransmitter will affect one or more other neurotransmitters. Dopamine is good example of this as dopamine is a precursor for another neurotransmitter, norepinephrine. So while dopamine may be involved in wanting a reward other neurotransmitters many also produce this effect. Or maybe dopamine only produces wanting when certain other situations are met and has different effects at other times.
Another complication is that it may be possible that the activation (or suppression) of brain areas that produce a neurotransmitter, such as dopamine, can have effects completely separate from the specific neurotransmitter that you are examining (which is why, at the start of this blog I said I was limiting myself to just discussing the action of the neurotransmitter and not of the brain regions themselves). Furthermore, binding sites (where neurotransmitters fit and work) can often be activated by multiple different chemicals (they just may be more strongly attracted to one neurotransmitter), meaning that in the absence of dopamine perhaps another chemical binds a these sites. Another complication would be that dopamine appears to be simply released during repetitive motor movements (such as where you play Rock Band or are just tapping away on buttons on a controller), which may or may not have anything to do with rewards .
In this blog, I have mentioned a few other neurotransmitters by name (e.g. opioids and cannabinoids) but I have focused purposefully on the role of dopamine and its role in ‘wanting’ rewards (it is likely that dopamine does more than this but ‘wanting’ seems to be its main role in terms of rewards). As such, I hope that this blog has been an interesting read and that maybe you have learned one or two new things.
However, there are many more neurotransmitters could play a role in how people react to games. Serotonin, for example, is likely to be involved . Also, many games are more than purely about reward. They also often involve social aspects (oxytocin and vasopressin are the neurotransmitters currently getting the most attention here), competition, skilled performance, negative, as well as positive, emotions, punishments, and many other factors. Indeed it seems that games are thankfully, much like the human brain, a complex subject matter. So, if this kind of subject interests you then read and learn about it (check out the references below, the majority of which are open access and not locked behind paywalls). But be aware of the current human positive bias towards neuro-related subjects and instead try to be neuroskeptical.
Thanks to Professor Dr. Oliver Tucha and Dr. Lara Tucha of the department of Clinical Neuropsychology, the University of Groningen, for providing peer-review on this post.
1. Olds, J., & Milner, P. (1954). Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain. Journal of Comparative and Physiological Psychology, 47, 419–427.
2. Berridge, K.C. (2003). Pleasures of the brain. Brain and Cognition, 52, 106-128.
3. Berridge, K.C. (2005). Espresso Reward Learning, Hold the Dopamine: Theoretical Comment on Robinson et al. (2005). Behavioral Neuroscience, 119 (1), 336 – 341.
4. Berridge, K.C. & Kringelbach, M.L. (2008). Affective neuroscience of pleasure: reward in humans and animals. Psychopharmacology, 199, 457 – 480.
5. Berridge, K.C., Robinson, T.E., & Aldridge, J.W. (2009). Dissecting components of reward: ‘liking’, ‘wanting’, and learning. Current Opinion in Pharmacology, 9, 65 – 73.
6. Berridge, K.C. (2012). From prediction error to incentive salience: mesolimbic computation of reward motivation. European Journal of Neuroscience, 35, 1124 – 1143.
7. Schultz, W. (2007). Behavioral dopamine signals. TRENDS in Neurosciences, 30(5), 203 – 210.
8. Schultz, W. (2010). Dopamine signals for reward value and risk: basic and recent data. Behavioral and Brain Functions, 6(24), 1-9.
9. Cannon, C. M., & Palmiter, R. D. (2003). Reward without dopamine. Journal of Neuroscience, 23, 10827–10831.
10. Peciña, S., Cagniard, B., Berridge, K. C., Aldridge, J. W., & Zhuang, X. (2003). Hyperdopaminergic mutant mice have higher “wanting” but not “liking” for sweet rewards. Journal of Neuroscience, 23, 9395–9402.
11. Sienkiewicz-Jarosz, H., Scinska, A., Kuran, W., Ryglewicz, D., Rogowski, A.,Wrobel, E., Korkosz, A., Kukwa, A., Kostowski, W. & Bienkowski, P. (2005) Taste responses in patients with Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry, 76, 40–46.
12. Wickelgren, I. (1997) Neuroscience: getting the brain’s attention. Science, 278, 35–37.
13. Eikemo, M. (2012). The Role of the Opioid System in Reward Responsiveness. PhD Thesis, University of Oslo.
14. Siviy, S.M. & Panksepp, J. (2011). In search of the neurobiological substrates for social playfulness in mammalian brains. Neuroscience and behavioral reviews, 35, 1821-1830.
15. Trezza, V., Baarendse, P.J.J., & Vanderschuren, L.J.M.J. (2010). The pleasures of play: Pharmacological insights into social reward mechanisms. Trends in Pharmacological Sciences, 31, 463-469.
16. Anselme, P. (2013). Dopamine, motivation, and the evolutionary significance of gambling-like behaviour. Behavioural Brain Research, In Press.
17. Robinson, S., Sandstrom, S.M., Denenberg, V.H., & Palmiter, R.D. (2005). Distinguishing Whether Dopamine Regulates Liking, Wanting, and/or Learning About Rewards. Behavioral Neuroscience, 119, 1, 5-15.
18. O’Sullivan, S.S.,Wu, K., Politis, M., Lawrence, A.D., Evans, A.H., Bose, S.K., Djamshidian, A., Lees, A.J. & Piccini, P. (2011) Cue-induced striatal dopamine release in Parkinson’s disease-associated impulsive-compulsive behaviours. Brain, 134, 969–978.
19. Heath, R. G. (1972). Pleasure and brain activity in man. Deep and surface electroencephalograms during orgasm. Journal of Nervous and Mental Disease, 154(1), 3–18.
20. McCabe, D.P. & Castel, A.D. (2008). Seeing is believing: The effect of brain images on judgements of scientific reasoning. Cognition, 107, 343 - 352. -
21. Skinner, B. F. (1953). Science and human behavior. Macmillan.
22. Skinner, B. F. (1971). Beyond freedom and dignity. Springer.
23. Skinner, B. F. (1974). About behaviourism. Knopf, New York.
24. Green, L., Fry, A.F., & Myerson, J. (1994). Discounting of Delayed Rewards: A Life-Span Comparison. Psychological science, 1994, 5, 33 – 36. -
25. Hariri, A. R., Brown, S. M., Williamson, D. E., Flory, J. D., de Wit, H., & Manuck, S. B. (2006). Preference for immediate over delayed rewards is associated with magnitude of ventral striatal activity. The Journal of Neuroscience, 26(51), 13213-13217.
26. McClure, S. M., Laibson, D. I., Loewenstein, G., & Cohen, J. D. (2004). Separate neural systems value immediate and delayed monetary rewards. Science, 306(5695), 503-507. -
27. Abrahamse, W., Steg, L., Vlek, C., & Rothengatter, T. (2005). A review of intervention studies aimed at household energy conservation. Journal of Environmental Psychology, 25(3), 273-291.
28. Lehman, P. K., Geller, E. S., & Bolderdijk, J. (In Press). Applications of social psychology to increase the impact of behaviour-focused intervention. In L. Steg, B. Buunk & K. E. Keizer (Eds.), Applied social psychology.
29. Bagchi, R., & Li, X. (2011). Illusionary progress in loyalty programs: Magnitudes, reward distances, and step-size ambiguity. Journal of Consumer Research, 37(5), 888-901.
30. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica: Journal of the Econometric Society, 263-291.
31. Pelham, B. W., Sumarta, T. T., & Myaskovsky, L. (1994). The easy path from many to much: The numerosity heuristic. Cognitive Psychology, 26, 2, 103 – 133.
32. Thorndike, E. L. (1911). Animal intelligence: Experimental studies. Transaction Pub.
33. Hsee, C. K., Yu, F., Zhang, J., & Zhang, Y. (2003). Medium maximization. Journal of Consumer Research, 30(1), 1-14.
34. Brehm, J. W. (1966). A theory of psychological reactance. London.
35. Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6), 627.
36. Deci, E. L., Koestner, R., & Ryan, R. M. (2001). Extrinsic rewards and intrinsic motivation in education: Reconsidered once again. Review of Educational Research, 71(1), 1-27.
37. Csíkszentmihályi, Mihaly (1990). Flow: The Psychology of Optimal Experience. New York: Harper & Row, Publishers.
38. Egerton, A., Mehta, M.A., Montgomery, A.J., Lappin, J.M., Howes, O.D., Reeves, S.J., Cunningham, V.J., & Grasby, P.M. (2009). The dopaminergic basis of human behaviors: A review of molecular imaging studies. Neuroscience and behavioral reviews, 33, 1109-1132.