Fagen et al. (2012) Flashcards
What is operant conditioning
- rewards learning by consequence of rewards or punishment
- negative = punishment
- positive = reward
Learning approach
- learning as a result of conditioning
(classical/operant) - social learning
(observations and limitations)
Positive and negative reinforcement
Positive reinforcement = favorable events/outcomes that are presented after the behavior
Negative reinforcement = involves removing something that is unfavorable after display of behavior (removing something considered unpleasant/negative)
Punishment
Where an adverse event/outcome is presented which causes a decrease in the behavior it follows
Positive punishment
Adding something to the behavior where it’s less likely to reoccur
Negative punishment
Removing the stimulus to behavior
Reinforcement schedules
Time and frequency of reinforcement that influences how new behaviors are learned and old behaviors are modified
Continuous reinforcement
- involves delivering a reinforcement every time a response occurs
- this is used best during initial stages of learning to create strong association between behavior and response
Partial reinforcement
- conditioning process where behavior is reinforced only a portion of the time, rather than every time it occurs
- more resistant for behavior to go extinct
- time: fixed-interval, variable-interval
- response: fixed-ratio, variable-ratio
Fixed-ratio schedule
- fixed to be given to a certain number of responses that have been made
- eg: after every 5 responses, a reinforcement will be made
Variable-ratio schedule
- interval of time isn’t always the same, but centers around the same average length of time
- only provided after a variable number of correct responses have been made
- eg: sometimes 3, 4, 5 responses and then make an average
Animal ethical guidelines
- numbers
- housing
- replacement
- species and strain
- pain and distress
- procedure
- rewards
Fixed interval schedule
- occurs only after a certain interval of time has gone
- this means it’s predetermined
- interval of time is always the same
- eg: every 3 secs time has elapsed
Variable interval schedule
- interval is not always the same
- centers around same average length of time
Primary reinforcers
A reward that fulfils a biological need
- eg: food, drink, shelter
What is Secondary Positive Reinforcement (SPR)?
When a secondary reinforcer, like a sound marker, is used and then followed with administration of a primary reinforcer, like food
Advantage of SPR: ability to shape captive animal’s behavior to improve their health and well-being
Secondary reinforcers
They are learned and worked via association with primary reinforcers
- eg: money, good grades, tokens, etc
Background of the study
Investigate the use of SPR training to teach elephants to reliably/voluntarily engage in trunk washing procedure designed to maintain their well-being
Aim of the study
The aim of the study is to investigate if free-contact, traditionally trained elephants can be trained to participate in trunk wash by using positive reinforcement
Sample
- 5 female elephants
–> 4 juveniles (5-7 yrs old, born at stable)
–> 1 adult (in her 50s) - all housed at same elephant stable in Nepal
- not pregnant
- all traditionally trained
- free contact with their mahouts
- none of elephants had experience with SPR
- mahouts volunteered
Elephant sample
- elephants spent most of day grazing in jungle (mahouts controlled them)
- rest of day: spent in stable with leg chained
- diet: fresh grass, grain, nutritional supplements, water during grazing hours
Research method
- controlled observation
- structured observation
- behavior checklist used to record elephants’ responses
Procedure
- primary reinforcer = chopped banana
- secondary reinforcer = a short whistle
- indoor sessions = 7:30am - 10am and 4pm to 7pm
- conducted by trainer and mahout was present for safety
- mahout stood to the side and didn’t speak or signal to elephants
- elephants had a choice: they could choose not to engage with session by turning/walking away from trainer
Training for trunk wash
- elephant puts end of trunk in trainer’s hand
- allow trainer to instill saline/sterile water into trunk
- lift trunk upward so fluid runs to base of trunk
- hold fluid there before lowering tip of trunk into collection container
- blow sample out
Purposes of training
- teach elephants to perform voluntary trunk wash, actively moving their trunks in response to a cue
- capture: waiting for animal to perform behavior naturally, then ‘capturing’ it by marking with reward
- lure: for non-natural behavior, as animal is ‘lured’ into certain body position by placing reward into certain place
- shaping: after starting either capture/lure, rewards are then only given for behaviors that are ‘best’
What are the behavioral tasks?
- Trunk here: lure and shaping
- Trunk up: lure and shaping
- Bucket: lure and shaping
- Blow: capture and shaping
- steady: shaping
3 ways data was collected
- session times
- number of offers
- performance tests
What is behavior chaining?
- This is when strings of behaviors are created through behavior chaining
Behavior chaining procedure
- elephant taught to blow consistently into bucket
- elephant rewarded for blows made in contact with bucket
- elephant was taught to string other behaviors together in small sequences to ensure smooth transitions
- separate behaviors were paired in different combinations and then practiced in multiple behavior sequences that had various sections of full trunk-wash chain
- trainer continued to mark the behaviors at appropriate times and then followed up with primary reinforcer at end of sequence
- trainer strung together all behaviors:
–> trunk here with short steady
–> trunk up with longer steady
–> then bucket and blow
Desensitization & counterconditioning
- syringe and sample fluid were introduced using desensitization and counterconditioning methods
- new potentially negative experience was introduced and paired with reward to make experience less disliked for elephant
- trainer introduced syringe in trunk-here position
- elephant was rewarded for staying in trunk-here position when syringe was presented
- touching outside of trunk tip with syringe was transitioned to touching inside of nostril and then inserting syringe tip into nostril
- increasing amounts of fluid introduced into trunk from syringe
- fluid ranged from 1-15ml
- started with a drop and building up in small increments to reach tolerance to 60ml
Saline and water
- elephants started on 0.9% saline as sample medium
- then transitioned to plain water for training purposes
- each elephant was offered water to drink at beginning of each training session to reduce likelihood that elephant would drink solution and interfere with completing tasks
- one elephant had preference for saline water and would continue to drink saline after rejecting drinking water
- she was transitioned to use water instead of saline as sample medium and success rate improved
- there was no prescribed time spent at each stage in training process
Results: number of sessions
- 4 juvenile elephants all successfully learned trunk wash in time available for study
- adult elephant, elephant 5, didn’t successfully learn trunk wash
Elephant 1
–> passed test after 30 min training sesh
–> mean = 12.42 mins
Elephant 2
–> 25 training sesh
–> mean = 10.29 mins
Elephant 3
–> 35 training sesh
–> mean = 13.27 mins
Elephant 4
–> 35 training sesh
–> mean = 11.11 mins
Elephant 5
–> was never tested on trunk wash
–> she didn’t learn all the necessary components
Results: individual tasks
- elephant 5:
–> never passed her blow into bucket, desensitization to syringe, and steady tests - elephant 2 and 4:
–> never passed steady tests, despite passing full trunk-wash test
Result: training time
- 257 mins for elephant 2
- 451 mins for elephant 4
- mean total training time:
–> 378 mins all elephants - elephants who successfully passed trunk wash test (1-4):
–> 367 mins - each training sesh:
–> mean duration of 12 mins all elephants
Overall passing rate
- session 35 test, only 3-4 elephants were being trained
- mean % correct never reached 100%
- all individuals reached default score of 90%
- by time full trunk wash was being passed, elephants could only reach a score greater than 90% for full trunk wash, steady, and trunk down tasks
Training methodology and sequences
TRUNK HERE:
- elephant places send of trunk into trainer’s hand to allow saline/water into trunk using lure and shaping methods
TRUNK UP:
- elephant lifts trunk up to allow saline/water to run down to trunk base using lure and shaping techniques
BUCKET:
- end of elephant’s trunk is then placed into bucket in preparation to blow.
- used lure in bottom of bucket, then removed, marked, and rewarded
BLOW:
- elephant gives out powerful exhale to blow out solution for sample collection, while using capture technique
STEADY:
- elephant holds position she had just been previously asked to do using shaping technique.
- behavior was reinforced by repetitive marking of hold with primary reinforcer
SPR technique
Primary reinforcer:
- chopped banana
Secondary reinforcer:
- short whistle blow
Training methodology
Morning sesh:
- 7.30am - 10.00am
Afternoon sesh:
- 4.00pm - 7.00pm
1) Voluntary method used for training trunk wash
2) elephant put end of trunk into trainer’s hand
3) trainer instills saline/sterile water into trunk
4) lifts trunk upwards so fluid runs to base of trunk
5) hold fluid then lower trunk into collection container
6) elephant blows sample out
What’s capture, lure, and shaping
Capture:
- behavior that elephant does without training or is similar to spontaneous behavior
Luring:
- behaviors that aren’t natural to elephant and is used to bring elephant to trainer’s needs by using reward (banana)
Shaping:
- behavior that’s only seen as best or closest to what’s desired from elephant is rewarded
Results: overall passing rate
- elephants’ performance improved from mean success rate of 39.0% after 10 sessions to 89.3% after 35 sessions
- mean % never reached 100%
- all individual behaviors within sequence received default score of 90%
- 90% or greater, but never 100%
Conclusions
- SPR great tool for captive management programs
- it’s easy to train juvenile, free contact, traditionally trained elephants in Nepal who have no prior experience with SPR technique
- teaching new tasks is an efficient process
- elephants respond reliably
How was performance test operationalized?
80% = entire trunk wash
90% = individual tasks
Improvements
- to make it both male and female
- more in number
- equal number of juveniles and adults
- all age groups
- elephants from other countries
- have 3rd party available to evaluate/video tape sessions available for review and scoring
- set stricter criteria regarding time position held with exact positioning measures
Evaluation methodology
STRENGTHS:
- quantitative data
- qualitative data
- high reliability
(controls - amount of saline used, number of sessions taken)
- high ecological validity
WEAKNESSES:
- low generalizability
- no standardized procedure
- subjective assessment criteria
Evaluation ethics
STRENGTHS:
- pain and distress
- number of elephants
- housing
- procedure
- rewards
- sample size kept to a minimum
