Reward and motivation Flashcards
What is referred to as an reward in cognitive Neuroscience?
A reward is a stimulus that elicits an approach behavior. Consequently, rewards can also be understood as reinforcers or unconditioned incentives.
What is referred to as a punishment in Cognitive Neuroscience?
A stimulus that elicits avoidance behavior.
What two types of reinforcers are discussed? Give an example for each and explain how they relate to each other.
There are primary reinforcers (e.g. food) and there are secondary reinforcers (e.g. money). Secondary reinforcers go back to primary reinforcers (e.g. you can buy food with money)
What is understood as “negative reinforcement”?
The reinforcement of behavior due to the removal of consequence (e.g. an unpleasant experience).
Can rewards be state-dependent? Explain using an example of choice.
Yes, they can be. For instance water is valued more strongly as a reward when one is thirsty, than when one is not.
Name the main dopaminergic pathways in the brain and what function they are generally linked to.
There is the nigrostriatal pathway (from the substantia nigra, to the striatum and then to the basal ganglia), which is central to motor control.
And there are the mesolimbic and mesocortical pathways (from the ventral tegmental area to the nucleus accumbens and the prefrontal cortex). These pathways are believed to be essential to reward processing.
Where is dopamine synthesized in the mesolimbic and mesocortical pathways?
Dopamine is synthesized in the cell bodies of the VTA. Axons of these VTA cells then travel to their destinations and release dopamine in their respective brain regions. Dopamine is not produces in these later brain regions!
How was the link between dopamine and the reward system first studied?
In an experiment by Romo & Schultz (1990), a monkey grabbed into a box without seeing its contents. It would either touch a wire or a food item. In the cases where the monkey touched food items, there would be a surge of activity in dopamine neurons of the midbrain, precisely in the substantia nigra.
This suggests a link between dopamine neurons and a reward value.
What key brain region is commonly studied, when researching rewards? Where is it located and part of what system is it?
It is believed that the nucleus accumbens is a key brain region, related to reward. It is part of the mesolimbic and mesocortical pathways and sits at the junction of the globus pallidus and caudate nucleus.
Another important region to reward processing in the brain is the orbitofrontal cortex (OFC). Explain why this region could be related to reward processing and what a lesion study in monkeys can tell us about the OFC’s function.
The OFC is a zone of multimodal convergence. This means that information on different modalities is integrated into this region. This makes sense as reward properties should be stable across modalities (e.g. a pleasantly smelling AND tasting cake).
In a study by Butter (1969), a group of monkeys received a sham lesion and another had an ablation of the OFC. Those monkeys that had an OFC ablation would now choose food items and non-food items in more similar frequencies than those who had only a sham lesion.
This suggests, that the OFC is important to reward processing. However, it must be noted that different explanations for such findings are possible (e.g. impaired discrimination of stimuli due to the lesion?)
How could the OFC - as a multimodal zone of convergence that is important to reward processing - be identified in a neuroimaging study?
Rolls (2000) showed, that by presenting pleasant stimuli of different modalities, the measurement of activity always overlapped in the OFC.
According to O’Doherty et al. (2001), what region of the brain codes for reward, and what region codes for punishment?
A region in the medial OFC codes for reward (increased activity for an increased award)
A lateral region of OFC codes for punishment (significant correlation with punishment)
Explain Pawlows classical conditioning paradigm using the example of an alarm clock and a pizza in the oven.
Before one has ever made pizza in the oven, the alarm clock is a neutral stimulus (NS), that elicits no conditioned response (NCS). The Pizza itself however is an unconditioned stimulus (US) that by itself elicits an unconditioned response (UR). That would be watering of the mouth for example.
If we now repeatedly pair the sound of the alarm clock with the pizza the alarm clock will become a conditioned stimulus (CS) that on its own can elicit a conditioned response (CR). We would hear the alarm and would respond with a watering mouth, even when there is no pizza around.
How is dopamine related to the primary reward anticipation vs. the primary reward receipt? Explain this with a neuroimaging study discussed in class (O’Doherty, 2002).
In a 2002 paper O`Doherty showed participants abstract stimuli that cue glucose, salt, or a neutral-tasting reward. After a delay period, participants tasted the reward for a certain period before swallowing it.
Using an fMRI two results could be found:
In the anticipation-period participants responded with increased activity in the midbrain and striatum to the glucose anticipation in relation to the salt anticipation.
After the reward was given activity in the insula could be measured, as this region represents certain body parameters, such as olfactory sensation or heart rate.
This suggests that in classical conditioning anticipation can be coded in the midbrain.
A paper by O’Doherty et al. (2001) could show differential activation of the lateral and medial OFC for reward and punishment. Another paper by van Duren et al. showed diffuse activation of dopamine neuron populations in the OFC. How can this difference in findings be explained? Argue on the basis of a 2010 Haynes study presented in class.
In the paper presented participants had to respond to red vs. green and clockwise vs. counter-clockwise rotating stimuli. The expected reward for each stimulus was reported and then correlated with the actual reward. In this type of classical conditioning task, an fMRI was used to measure the OFC.
It was found that certain dopamine neuron subpopulations coded with more activation for higher rewards and some subpopulations for lower rewards. The microstructure of the OFC shows more complex differentiation than just lateral vs. medial OFC differences.
Using this depiction to explain how temporal discounting works. What does the discounted value reward represent? How does the best-fit hyperbolic model work?
On the left side, we see how the frequency of SS choices relates to the relative worth of an SS reward to the (constant) LL reward. The different colors plot the different delays between SS and LL.
To make this clearer, let’s say you receive either 5 marshmallows now or 10 in 2 seconds. We would now look on the x-axis at how large the relative reward of SS is to the absolute reward of LL. In this case, we would have 5/10 = 50%. Now we take a look at what value is corresponded to that 50% on the y-axis. As we have a2 second delay we choose the black curve and see that the frequency of SS choices is almost 0%. This means, that at this short delay almost always LL is preferred.
Of course, if SS is 100% it is equal to LL. We will then always choose the SS reward.
The best-fit hyperbolic model (right side) is a mathematical model that allows us to plot, how large the SS value must be relative to the LL value to result in a 50% choice frequency (equal choice frequency). This is plotted against the delay of the LL reward.
The longer the delay gets, the lower SS must be, so it will be chosen in exactly 50% of the trials.
