Human Machine Systems Assignment

Human Machine Systems Assignment

• Inspectors Clouseau and Morse: Monitor weld quality at a factory. Clouseau's neural activity distributions are shown for noise and signal plus noise (top figure), and Morse's activity is shown in the lower figure.

Inspector Sensitivity

• Sensitivity: Clouseau or Morse? One has better ability to detect flaws. The question asks which inspector is better at identifying flaws in welds.

Criteria and d' Value

• Criteria: Morse (neutral) and Clouseau (liberal). A vertical line represents the criterion on each graph.
• ß < 1: Determine which inspector has a ß value less than 1.
• d' Value (Clouseau): Calculate Clouseau's d' value given a miss rate of .20 and a correct rejection rate of .80, using the provided table or formulas.

Payoff Conditions

• Session One: $1 for detecting a faulty weld, no penalty for false alarms.
• Session Two: $1 for correct detection, $1 penalty for false alarms.
• Session Three: No payment for correct detections, $1 penalty for false alarms.

ROC Curves

• ROC Curve Construction: Create ROC curves for Clouseau and Morse using the provided data for hit rate and false alarm rates across the three sessions. Plot the curves on the same axes, labeling the axes.
• Observer Differences: Did both observers' performance change equally under the different payoff conditions or were there significant differences?

Inspector Smith's Sensitivity

• Shallow Slope: Inspector Smith's ROC curve on normal probability paper has a shallow slope (less than 45 degrees).
• Quantifying Sensitivity: Explain how to quantify Inspector Smith's sensitivity given this slope. Explain the reasoning behind your approach

Optimal Beta Value (Morse)

• Payoff: $10 for a hit, $1 for a false alarm.
• Optimal Beta: Calculate the optimal beta value for Inspector Morse using appropriate values. Two P(N)/P(S) scenarios are provided to guide you.