Pepperberg

Learning Approach

studies social learning theory and operant conditioning

social learning

(Psychology)

attention, retention, and reproduction

operant conditioning

(Psychology)

vicarious reinforcement and aversive stimuli

symbolic representation

(Psychology)

can Alex use vocal labels to understand sameness and difference between two objects?

Background

• Alex already knew 80 objects based on shape, color, and material, and can request them

• all symbolic representation studies have been done with chimpanzees and in complex situations

• Pepperberg believed object differentiation was not limited to chimpanzees

Aim

to find out if Alex can use labels to demonstrate symbolize understanding of same and different

Hypothesis

A parrot, Alex, can use vocal labels for understanding concepts of same and different

Independent Variable

objects (novel and familiar)

Dependent Variable

response to the question “same/different?”

Method

• laboratory

• experimental case study with an animal

• repeated measures because Alex is being exposed to both independent variables

Data

quantitative - tabulated right and wrong answers; any data over 33% shows he didn’t guess

Familiar Apparatus

• color - red, blue, green

• shapes - squares and triangles

• material - wood and rawhide

Novel Apparatus

any color, shape, or material other than what is familiar

Participant

Alex, an African grey parrot purchased by Pepperberg at 1 year old in 1976, who lived until age 31 in 2007; opportunity sample

Control #1

When it comes to training Alex, two researchers (not Pepperberg) will be in charge

Control #2

When it comes to testing Alex, researchers other than the trainers will train Alex

Control #3

Another researcher not involved in the experiment will pair objects together

Control #4

Pepperberg will never directly look at Alex during the experiment

Control #5

The amount of time spent teaching is short

General Design Part 1

Alex could already name at least five colors, seven shapes, and four materials

General Design Part 2

Alex has some comprehension as to what color or what shape

General Design Part 3

Alex will only ever be assessed on color, shape, and material

General Design Part 4

The task is to identify characteristics compared to another object

General Design Part 5

The novel stimuli is used for transfer information

General Design Part 6

To make sure Alex couldn’t memorize combinations, Pepperberg made the number of possible stimulus pairs too big

General Training Step 1

There will be two research assistants. Research Assistant 1 is the trainer, and Research Assistant 2 is the model. Research Assistant 1 will present an object, and ask questions to Research Assistant 2. If they answer right, they get the object, and if they get it wrong, it is taken away, hidden from sight, and brought back later.

General Training Step 2

The two researchers flip roles, so Researcher 1 answers the questions and Researcher 2 asks them, but if Alex answers the question correctly first, he gets the object, or can ask for a different object

General Training Step 3

Once Alex is trained on all his familiars, he moves on to Same/Different Training

Same/Different Training Step 1

The researchers use the same Model-Rival technique, but instead of one object presented at a time, the trainer presents two at a time, and the question “what same/what different” is asked

Same/Different Training Step 2

Once they go through all the pairs, the roles are swapped, and Alex is added in. Since Alex can’t say the word “material,” instead he is taught to say “mah-mah”

Testing Step 1 (familiar objects)

Over 26 months, in any given week, Alex will be tested 1 to 4 times. A new secondary trainer will present the items to Alex. They will present 2 objects to Alex, and will ask “what same/what different?” If Alex correctly answers, he gets the object or another one of his choosing, and if he incorrectly answers, he is told “No!,” the object is taken away, and it is brought back later. They keep doing this until he gets them all right. During the trial phase, Alex will complete other activities to prevent boredom. The questions “what same?” and “what different?” will always be alternating . Once Alex got a pair correct, the same pair and question could never be asked again

Testing Step 2 (transfer objects)

Alex will be presented with objects not part of his familiars. The novel stimuli would be placed in the lab several days prior to being presented so as not to freak him out

Testing Step 3 (use of probes)

When certain pairs are put together, they will have multiple “what same” or “what different” answers

Data

quantitative - correct responses

Finding #1

Alex scored a 69/99 (69.7%) on the first-response trials for the familiar objects

Finding #2

Alex scored a 99/129 (76.7%) on the total trials for familiar objects

Finding #3

Alex scored a 79/96 (82%) on the first-response on the first-response trials for transfer objects

Finding #4

Alex scored a 96/113 (85%) on the total trials for transfer objects

Finding #5

Alex scored a 55/61 (90%) on the total trials for transfer objects

Conclusion #1

Alex has symbolic representation; he can categorize and comprehend same/different

Conclusion #2

Guessing never played a role in the results

Conclusion #3

Symbolic representation isn’t only for primates

Ethics

None, because the participant is an animal

Strength #1

The procedure is highly detailed and standardized

Strength #2

The experiment is highly controlled

Strength #3

It’s a case study that lasted over 2 years, so it provided lots of in-depth data

Strength #4

It solidifies the use of learning theory techniques in animals

Weakness #1

The tasks of a parrot being trained for symbolic representation and playing with objects lacks mundane realism

Weakness #2

Alex himself is too unique to be generalized to other African grey parrots

Weakness #3

The laboratory setting lowers ecological validity

Use of Animals

(Issues & Debates)

Numbers: only one animal is used

Rewards: Alex is rewarded for correctness, and isn’t deprived

Housing: Alex was put in his usual cage, without other parrots, but he gets social interaction

Nature v. Nurture

(Issues & Debates)

Alex went through extensive training protocol, so this study supports nurture

Application

(Issues & Debates)

• to training animals to perform certain tasks

• to training service and working dogs