Pepperberg Flashcards
(44 cards)
Continuous reinforcement
When a learner receives a reward each time they perform a desirable behaviour
M/R approach
a training method based on the concept of social learning. One human acts as a trainer of the second human who acts as a model to the real participant who is watching the interaction
Category label
a short word or phrase that indicates the classification of a particular object in some way. In this study, these were: shape, colour and matter
‘same’ and ‘different’
a cognitive skill identifying what objects have and don’t have in common; allowing us to classify objects. Previously, it was thought to only be present in humans
Operational definition
when an non-directly measurable attribute, such as intelligence, is made directly measurable. In this study, Alex’s responses were only recorded as correct if it was the first vocalisation he gave
- increases reliability because researchers are likely to be more consistent in the way that they collect data
‘first trial’ response rate
the quantitative data from the study - the percentage of correct responses that occurred the first, and only time, materials were presented to Alex
Transfer of knowledge
to further test Alex’s abilities, Pepperberg wanted to test Alex on novel objects that weren’t in his labelling repertoire so he would have to apply his current knowledge of familiar objects to ones her had not seen before
Self-injurious behavours
It was noted that Alex was prone to boredom, which is not uncommon amongst this intelligent parrot species. As a result, they often hurt themselves e.g. through feather picking
What is the background?
- Non-human primates have cognitive capacity to express abstract ideas and meaningful sentences
- comprehension of same/different has been singled out as a concept not typically attributable to non-primates
- pigeons acquired a concept of same but not different
- none of the studies shown actual labeling of the relation of sameness or differences
What are the 2 aspects for non-human to demonstrate ‘same’?
- recognise two independent objects A1 A2 are both blue, and it is the SINGLE attribute that makes them the same
- recognise the ‘sameness’ can be applied and transferred not only to 2 other blue items, but for 2 NOVEL, independent green items e.g. that have nothing in common with A1 and A2
What can Alex do prior to the study?
- demonstrated capacity for categorisation
- could count up to 6
- use useful phrases like ‘I wanna go X’, ‘I want…’
- produce vocal labels for 5 colours, several shape
- could response to vocal question ‘what colour’, ‘what shape’
What Pepperberg said that Alex did not show?
Did not show whether a non-human can understand and use ABSTRACT symbolic relationships when communicating, e.g. same/different comprehension
What is the aim?
To see whether African Grey parrot could use vocal labels to demonstrate symbolic comprehension of the concepts of same and different
How is Alex housed?
- free access to all parts of the lab 8hrs a day
- water and seeds supplied continuously
- toys: key, wooden blocks are available upon request
- during sleeping hours, confined to a wired cage
What does the task have to ensure?
- the symbolic concepts tested would be more abstract
- subject given equal training on concepts of both same and different
- findings could not be dismissed as stimulus-specific associations
- first trial transfer test could be examined for their significance
In order to be correct, what does Alex need to do?
- attend to multiple aspects of two different objects
- determine, from a vocal question, whether the response would be based on sameness or difference
- determine what was same or different
- produce vocally the label for this category
How does the tested objects look like
- 2 objects that could differ with respect to 3 categories, colour, shape or material
What do most previous research use? (the weakness)
- a 2 choice design, just indicate whether pairs match or do not match
- a topographically similar response for both answers (leverpressing or keypecking)
- same pairs are identical in all dimensions and different pairs are different with respect to most
How is MR approach used in the study?
- one human acts as the trainer of a second human by presenting objects
- asking questions about these objects
- giving praise and reward for correct answers, showing disapproval of incorrect answers
- second human is the model for the bird’s responses, as a rival for the trainer’s attention
- roles of model and trainer is frequently reversed
- parrot is given opportunity to participate in these vocal exchange
Describe the training
- a trainer hold 2 objects in front of the model, and ask either ‘what’s same’ or ‘what’s different’
- correct=reward, incorrect=scolded, object momentarily removed from view, then re-presented, question repeated
- roles reversed
How long did training session last?
- 2-4 times a week
- 5min-1hr
- 9 months
What boredom behaviours were observed?
- self-injurious
- restless, cease to work, interrupt with many successive requests for other items
Examples for further trainings to prevent boredom
- number concepts
- new labels
- recognition of photos
- object permanence
What is the procedure before testing?
- tested by secondary trainer who never trained Alex on same/different
- on the previous day, all of the possible objects to be tested were listed by the principal trainer
- a student not involved in testing would choose the question from pairs for same/different
- and randomly ordered all the question
- Alex nor the principle trainer could predict questions on which topic would appear on a given day
