magpie3-causal-implicature-variation

Variations on experiments, building on work by Gershman & Ulman, investigating PACE effects (Preferences in Assignment of Cause and Effect) found in non-causal expressions like "A is associated with B".

Setup

First, set up the project with npm install

To run the project locally, run npm run serve

Live version

Try a live version:

• association pilot 1
• association pilot 2
• association pilot 3
• association pilot 4
• association pilot 5
• association pilot 6
• association pilot 7

Experiment information

Exp 1

• test items:
  • "A is associated w/ B"
  • "A and B are associated"
• 200 participants from Prolific
• paid 30 pence for 2 min estimated work time
• requirements
  • location: UK or US
  • first language: English
  • approval rate: 90% at least
  • number previous submissions: at least 50

Exp 2

• test items
  • "As for A, it is associated w/ B"
  • "As for B, A is associated with it"
• 200 participants from Prolific
• paid 30 pence for 2 min estimated work time
• requirements
  • location: UK or US
  • first language: English
  • approval rate: 90% at least
  • number previous submissions: at least 50
  • no participation in previous experiment

Exp 3

• test items
  • "Drug A is associated with B"
  • "A is associated with drug B"
  • "Disease A is associated with B"
  • "A is associated with disease B"
• 400 participants from Prolific
• same as for Exp 2

Exp 4

• test items:
  1. observation condition ("I was surprised to see A in the samples")
  • "A is associated with B"
  • "B is associated with A"
  2. avoidance condition ("I was surprised that you removed A from the samples")
  • "A is associated with B"
  • "B is associated with A"
• 400 participants from Prolific
• same as for Exp 2

Exp 5

• similar to Exp3, but different kind of topic-hood manipuation
• test items:
  • "What can you tell me about A?" - "A is associated with B"
  • "What can you tell me about A?" - "B is associated with A"
• 200 participants from Prolific
• same as for Exp 2

Exp 6

• test items:
  • "Having A is associated with B"
  • "Getting A is associated with B"
  • "A is associated with having B"
  • "A is associated with getting B"
• 400 participants from Prolific
• same as for Exp 2

Exp 7

• test items:
  • "A is associated with B"
• preceded by two trial of generating alternative descriptions for "A causes B" with A or B in subject position
• 100 participants from Prolific
• same as for Exp 2
• only for Desktop users (unlike all previous), due to typing