7b Leipold

Question 1

What is a common assumption about how people form judgements in probabilistic environments?

Answer 1

They use probabilistic cues of the environment to form judgements

Question 2

Beifly explain the key assumptions of the Brunswick Lens Model

Answer 2

People use probabilistic cues to make judgements about uncertain situations
the cues have ecological validities that determine how well they can predict a criterion
cues differ in what weight a person gives them in making a judgement
achievement is the correlation between criterion & judgement

Question 3

Illustrate the key assumptions of the Brunswick Lens Model with a real-world example

Answer 3

Doctor has to give a diagnosis. Patient explains different complaints. The doctor makes judgement how much the complaints match symptoms that could belong to a diagnosis. For the doctor, different symptoms (probabilistic cues) have different weights (cue utilization) in making a disgnosis (judgement) e.g. person has a slow heart beat & sweats a lot -> not a high blood pressure

Question 4

Explain what the G parameter means in the context of the Lens Model Equation

Answer 4

matching-parameter
measure of matching performance
systematic component explained by cues
true correlation between judgement & criterion value if there was no noise in the observed data

Question 5

Explain what the C parameter means in the context of the Lens Model Equation

Answer 5

nonlinear-component
systematic component, not explained by the cues
correlation between the residuals, interpreted as cues that were not included in the model

Question 6

How is the G parameter in the Lens Model Equation calculated?

Answer 6

r(klein: YeDachYjDach)
YDach = predicted values

Question 7

How is the C parameter in the Lens Model Equation calculated?

Answer 7

r(klein:ZeZj)
Z = residuals
e -> environment
j -> judgement

Question 8

Briefly describe three possible sources of bias that can occur when the G parameter of the lens model equation is calculated using participant-wise regression models?

Answer 8

-Ignorance of the clustered data structure: false idea that every person judges the same scenario using the same set of cues
-overfitting: assumption that participants have common mean & are evenly distributed among it, data is first overfitted & then averaged
-item-effects overlooked e.g. cue works well for one item but not for another
-unknown cues cannot be accounted for (indicated by high C)