Psychology 1

PREVIOUS: Methods and Statistics

Our mind generally learns like this: We pick up information through our senses,form internal representations of what that information means, reflect on that knowledge, which guides our behavior. Afterwards, we communicate that knowledge to others.

A traditionalist view of learning and association involves only external stimuli, an organism’s responses, and its behavior regarding that stimuli.

Reflex: Simplest connection between stimulus and behavioral response, involving the individual muscles. This does not involve learning.

Taxis: Similar to reflexes, but they involve the entire body. For example, moths flying towards light, or me when I see a cockroach.

Instincts: stereotyped patterns of action and responses to specific stimuli. Species specific–imprinting on geese (how baby geese will follow the first living thing it sees moving).

These are all sort of automatic responses to stimuli. For example, when the doctor smashes your kneecap with their hammer.

There is evidence that behavior evolves among species, but this is extremely slow. At the same time, environmental changes are extremely fast. So this brings us to the question: How can we adapt our behavior quickly??

Learning is defined as the relatively permanent change in behavior, resulting from experience.

Classical Conditioning

The classic example of classical conditioning is Pavlov’s Dogs. He noticed that dogs naturally salivate to the smell of meat powder–yay food! So Pavlov set up an experiment to see if it was possible to get the dog to salivate at different stimuli–ie the sound of a bell.

On the first trial, we might see something like this:

After a couple more trials, where we ring the bell and then give the food, the dog may started to associate the sound of the bell with food, and begin salivating, even before the presence of food.

When the trials continue, the dog will begin salivating even without the presence of food.

In this situation:

  • Unconditioned Stimulus: Food
  • Unconditioned Response: Salivation to food
  • Conditioned Stimulus: Bell
  • Conditioned Response: Salivation to bell

The process of associating the CR to the CS is called acquisition,and it is reinforced by the US. At the same time, if the dog hears the bell but is not given food, it begins to lose this association in a process called extinction. After extinction, if we run the trials for bell->salivation again, the dog will learn much faster (reacquisition/savings in relearning).

You will also begin generalizing, creating associations between things that look like the conditioned stimulus. For example, if you were stung by a bee, you would fear bees, but also hornets or wasps, since they resemble bees.

Operant Conditioning

Also known as instrumental conditioning, this is a type of learning where reinforcement is the key player in learning. Thorndike’s puzzle box was essentially a cage for a cat, which included some switch that would open the door. Naturally, the cat wouldn’t know how to get out and start biting at the bars or the door, but eventually it would trigger the switch. Thorndike began charting the time needed to escape, using the same box and the same cat, and the time would decrease over trials.

The reinforcement of getting out of the box, is the process of increasing the future probability of the most recent response. On the contrary, punishments can decrease the likelihood of a response.

Law of Readiness: States (hunger, …) arouse behavior and set conditions and reward.

Law of Effect: Responses to reward are strengthened and responses without reward are weakened.

Law of Exercise: Associations between stimuli and responses are strengthened by practice.

Different distributions of reward can also change the learning curve: fixed/variable ratios of reinforcement and fixed/variable intervals for the reinforcement.

The end result? In classical conditioning, the organism learns to predict environmental events. In operant conditioning, the organism learns to control events.

Animals will learn regardless of reinforcement, but the presence of reinforcement makes the process faster. ie: Rats running through maze and creating a mental map. Group 1 was regularly rewarded with food while Group 2 wasn’t rewarded until trial 10. Group 2 still made it through the maze and made fewer errors with each successive trial, but once the reward was introduced, there was a huge drop in the number of errors made.

There is the case of the Rhesus Monkeys where the monkeys displayed no difference in performance whether they were rewarded for escaping a puzzle room or not. They were simply curious–an intrinsic motivation. Even if the monkeys were hungry (extrinsic motivation), they solved the puzzle just as quickly as others.

Observational Learning

This is the cognitive view of learning. Rather than seeing learning as a relatively permanent change in behavior, this view suggests that learning is a change in knowledge, resulting from experience. In classical and operant conditioning, the experiences to the organism were directly to the organism. It turns out that learning can happen just by watching other organisms. You don’t need to have the experience yourself.

Observational Fear Conditioning in wild rhesus monkeys towards snakes was thought to have learned through evolution and was an innate trait. However, nerds noticed that this was not the case in monkeys that were born and raised in the labs–there was no fear to snakes. They concluded that you needed to have an experience of the deadly snake–but this didn’t make sense either because you’d probably be dead if you were bitten by a deadly snake. So monkeys must’ve learned the fear of snake from observing other animals.

However, they tried an experiment where the the adult monkey fears a snake, but the offspring only sees a flower. The offspring didn’t fear flowers–so snake fear isn’t fully innate, but it is more prepared than something like a flower.

Bobo Doll: See Crash Course video.

Language is also an example of observational learning.

Cognitive Social Learning Theory: Learning is through expectations, not behavior. We can learn either by response consequences (trial error, reinforcement) or observationally (examples or precept/teaching.) This essentially says that our learning is also passed down generation after generation, so we don’t necessarily have to evolve. We learn from our cultures and previous members of our species. But if everyone learned from each other’s experiences, there would be no cultural change, leading to cognition, memory, and thinking.

NEXT: Sensation and Perception