Observational Learning Notes

Observational Learning

Outline:

• Why is observational learning interesting?

• What defines imitation, and how does it differ from other forms of observational or social learning?

• Theories of imitation.

• Applications of observational learning.

Behavioral Perspective on Imitation

Imitation can be viewed as a generalized operant, similar to rule-governed behavior. This perspective suggests that imitation is part of a broader set of learned behaviors. Operant and pavlovian conditioning both require experience and trial and error to acquire new behaviors.

The Interest in Observational Learning

Observational learning allows the acquisition of new behaviors without prior experience. It may contribute to humans being more efficient learners compared to other animals. Imitation has been identified as a method of acquiring behavior that isn't solely reliant on operant and Pavlovian conditioning. Early studies suggested imitation was more common in humans than in other animals, but this is debatable and depends on the definition of imitation.

Differentiation of Imitation

It's important to differentiate true imitation from other forms of learning that may resemble it but can be explained by simpler mechanisms like operant conditioning, Pavlovian conditioning, or unlearned responses.

Implications of Imitation

Imitation facilitates the transmission of information and cultural norms. It may be a major way in which cultural norms are acquired, instead of direct teaching and experience. Imitation seems to facilitate the acquisition of new behavior faster than shaping or repeated trials in Pavlovian conditioning.

Limitations of Imitation

While imitation facilitates fast acquisition of behavior, it has limitations. For example, learning a thumbs-up is easy through observation, mastering the piano requires skill levels unattainable through observation alone. Skill level is needed to acquire behavior using imitation.

Focus of the Discussion

The discussion focuses on forms of observational learning that may not meet the criteria for true imitation and categorizes learning by observing others into observational learning and true imitation. The terms "observational learning" and "social learning" are used differently across ethology and psychology literature.

Definition of True Imitation (Thorndike)

True imitation is defined as learning to perform an act by seeing it done. The behavior acquired must be new to the observer and not explainable by other known mechanisms such as Pavlovian conditioning, unconditioned responses, or operant conditioning.

Identifying True Imitation

Indicators of true imitation include behavior change being enduring and the behavior occurring in the absence of the model. By concentrating on the screen, the lecture attempts to have the viewers yawn to illustrate a potential point, even against social expectations. It can be difficult to suppress the desire to laugh when others are laughing, even without knowing what they're laughing about.

Yawning and Laughter as Examples

Yawning and laughter might appear to be imitation because people are doing the same thing. However, they are not examples of true imitation because the behavior is not new. These are unconditioned responses to unconditioned stimuli. Performing the behaviors at the same time as other people is useful from an evolutionary standpoint. This phenomenon is called social contagion.

Prairie Dog Example

An example of prairie dogs chewing may not be true imitation because the behavior is not new. It may be occurring only in the presence of the model.

Monkey Food Preference Study

Researchers taught monkeys in one tribe to eat only red-colored food and monkeys in another tribe to eat only blue-colored food. The monkeys would ignore the color they have not been conditioned to eat. This indicates that there is something new in terms of learning by observation as well. The behavior was present whether the little trained monkeys were present, or absent.

Orangutan Example

An orangutan learns to crack a coconut by observing its caregiver, demonstrating potential true imitation or failure to do so.

Observational Learning Types

Other types of observational learning that do not meet the criteria for true imitation include social contagion, social facilitation, response facilitation, local or stimulus enhancement, and goal emulation. Vicarious Pavlovian conditioning has been covered elsewhere.

• Observer: One who learns by observing.

• Model: Demonstrator of the behavior to be learned.

• Conspecific: A member of the same species.

Social Contagion

Social contagions are behaviors of social origin. They appear to be unconditioned responses to unconditioned stimuli (e.g., yawning, laughing, smiling).

Social Faciliation

Social facilitation occurs when the behavior is already in the subject's repertoire. The model prompts the behavior. Stopping and looking is an example, where people will also look up because someone is peering intently upwards. Buskers do this in order to create earnings. Chickens pecking is also an example. The behavior is already known, and the effect is often temporary.

Response Facilitation

Response facilitation involves a general increase in species-specific behavior in the presence of conspecifics. The others do not have to be doing the behavior to produce the effect, but many individuals do it because it is a species-typical behavior. When one chicken is present with a group, it will start pecking regardless if the others are pecking. Talking with a group of people could also be a human equivalent. It's essential to design experiments where these factors don't obscure the actual imitation.

Species Similarity for Response Facilitation

Species must be the same to get this response effect from each other. Response facilitation is a common confound in early imitation studies, where researchers used typical behaviors instead of teaching an animal a new thing. Social facilitation and response facilitation are sometimes mixed up in literature.

Local or Stimulus Enhancement

Local or stimulus enhancement occurs when the model makes a location or aspect of a stimulus more salient to the observer. This shortens the time it takes for the observer to learn the behavior, but they haven't learned the behavior by seeing it done. This could be operant conditioning, or Pavlovian conditioning invoking attention to a stimulus making it learning through trial and error. Seabirds diving for fish may be an example of this. A bird diving attracts them to that location, and they get their fish. Local Enhancement is very specific, locally attracting you to an area, while Stimulus enhancement drawing attention to a stimulus, which Pavlovian conditioning plays a part. Bumblebees perform the behavior if in these set-ups:

• Nectar is available with a block of wood.

• Pretend Bumblebee.

• Actual Bumblebee.

During the experiment, bumblebees were trained to go the pink or yellow flowers for the experiment. The bumblebees acquired in this setting that actual cue and the block of wood cue did similar the same amount of foraging (Local Enhancement). As a result, the bees essentially learn that color relates to food. The bumblebees learned to approach where they have been. If approach is an unconditioned response to nectar, then approaching where they've been. Approach also enhances the color, then the color can elicit the bee to approach. The bees also had to have socializing behaviors to produce these actions in them. This effects does not need to have imitation, but can use pavlovian and operant conditioning.

Goal Emulation

In goal emulation, observers do not copy the model. What they do is observe the end effects. This may result in them being motivated to trial and error to get the same effects. The way they get it may not look the same way as how the observer demonstrates it. Therefore, it allows the person to get the same goal, just a different way.

Examples of True Imitation

Examples include:

• Food Washing Monkeys; monkeys washed their food after seeing one drop the food in the water.

• Milk Stealing Birds, where birds pecked milk caps from houses around the UK.

Experiment Design

Experiments should be designed to eliminate other explanations. For example, the model's behavior can't make an aspect of the environment more salient. The animal should not be immediately reinforced after the action.

Quail Study

A quail study aimed to distinguish true imitation from other observational learning methods. Quail models trained to press treadle with their feet or beak. Observer quails watched, and their behavior was observed. The quails that watched the pressing the feet pressing one, pressed more with their feet, and the ones that watched the watched the quails that did it with their beak did it with their beak. These results suggest true imitation can occur in non-human animals.

Theoretical Ideas

Theories: The straight behavioral view is that imitation is an example of a generalized operant. Future lectures may look at applications as well as other explanations.