Spatial Learning

Question 1

SIMPLE T-MAZE

Answer 1

• rat must learn that left arm contains food
• runs down to cross
• must learn to choose left > right
• learning possibilities include:
  1. SMELL
  2. PAVLOVIAN CONDITIONED APPROACH
  3. INSTRUMNTAL RESPONSE
  4. MAP

Question 2

T-MAZE: SMELL

Answer 2

• follows food/previous run odour
• can be tested via:
  1. extinction test (food taken out of trained side)
  2. swapping T-maze arms (food in right arm)
• results = rats still solve problem

Question 3

T-MAZE: PAVLOVIAN CONDITIONED APPROACH

Answer 3

• learns to approach beacon near goal
• tested via removing (deletion)/altering configuration (spatial arrangement) of landmark
• if removing correct arm beacon hurts performance = Pavlovian

Question 4

T-MAZE: INSTRUMENTAL RESPONSE

Answer 4

• learns to turn left
• tested via rotating maze 180 degrees to differentiate between response based on place VS instrumental learned
• dismissed if show control by going right post rotation VS control by place; both outcomes happen
• BUT are they using beacon associated w/food?

Question 5

T-MAZE: MAP

Answer 5

• knows goal location; finds path to it
• resistant to most tests against other explanations (ie. food odour/Pavlovian/instrumental)

Question 6

RADICAL ARM MAZE: TESTS

Answer 6

• abundant evidence of rats typically solving maze via external landmarks
• rotation tests & landmark deletion/rearrangement studies point display this

Question 7

RADICAL ARM MAZE: ROTATION TEST

Answer 7

• rats forced to 4 arms; maze rotated 45 degrees (landmarks rotated relative to maze)
• now rats offered choice between unvisited arm (didn’t go down -> got food -> BUT now visited) VS visited arm (went down -> no food -> BUT now unvisited)
  RESULTS
• rats choose visited arm
• controls barely went to visited arms (v reliable via Chi-squared)

Question 8

RADICAL MAZE: SUZUKI, AUGERIONOS & BLACK (1980)

Answer 8

• cylindrical testing chamber w/discrete landmarks at each arm
• rats followed landmark rotation w/respect to maze BUT rearrangement relative to one another (landmark transposition) hugely worsens study/test performance
• suggests rats use landmark configuration to define locations > beacon usage

Question 9

RADICAL MAZE: O’KEEFE ET AL. (1978)

Answer 9

• animals use map to navigate
• mechanisms for constructing/using map = located in hippocampus
• cues used in combination to designate given region so any 2 cues define it (ie. card/buzzer/light)
• losing all 4 = place cell no longer fires
• BUT losing 2 = fine

Question 10

MORRIS WATER MAZE: BLOCKING

Answer 10

RODRIGO ET AL. (1997)
- problem for cognitive map hypothesis
- 2 groups trained in water maze to find platform on dif landmark set basis (ABC/ABCX)
- 1st group = landmark X added to ABC + ^ training
- tests w/ABC/ACX reveal how well rats learned to use X to find platform in conjunction w/other landmarks
- logic = conditioning (blocking group won’t learn X)/cognitive maps (opposite) predict dif outcomes

Question 11

MORRIS WATER MAZE: BLOCKING RESULTS

Answer 11

RODRIGO ET AL. (1997)
- blocking group w/ABC = v good > controls
- BUT ABX test (needs X to find platform) = < controls
- NOT predicted by cognitive map theory
- does NOT mean CMs/place cells don’t exist
- just same conditioning principles applying to learning tone/shock also apply to spatial learning aka. Pavlovian

Question 12

SUMMARY

Answer 12

• give rat problem of finding food in maze -> learns to do so w/any method available
• BUT does use landmarks & < configurations
• BUT learning still subject to same learning phenomena discussed (aka. conditioning)
• 1 principle set governs learning