Podcast
Questions and Answers
Which inspector has greater sensitivity (i.e., is better able to detect a flaw in the weld)?
Which inspector has greater sensitivity (i.e., is better able to detect a flaw in the weld)?
- Inspector Clouseau
- Inspector Morse (correct)
Inspector Morse has a neutral criterion, and Inspector Clouseau has a liberal criterion. Draw and label a vertical line representing the criterion in each graph. For which inspector is (\beta < 1)?
Inspector Morse has a neutral criterion, and Inspector Clouseau has a liberal criterion. Draw and label a vertical line representing the criterion in each graph. For which inspector is (\beta < 1)?
Inspector Clouseau has (\beta < 1)
If Inspector Clouseau had a miss rate of .20 and a correct rejection rate of .80, what would his d' value be? (Hint: may use attached d' table, Wickens)
If Inspector Clouseau had a miss rate of .20 and a correct rejection rate of .80, what would his d' value be? (Hint: may use attached d' table, Wickens)
d' = 1.28
Construct an ROC curve for each observer (plot both curves on the same axes). Label the axes.
Construct an ROC curve for each observer (plot both curves on the same axes). Label the axes.
Were both observers affected by the payoffs the same amount, or was one observer more affected by the payoffs? Explain.
Were both observers affected by the payoffs the same amount, or was one observer more affected by the payoffs? Explain.
The performance of a third inspector, Inspector Smith, was also monitored under different payoff conditions. His standardized ROC curve, plotted on normal probability paper, had a shallow slope (less than 45 degrees). How would you choose to quantify this inspector's sensitivity, and why?
The performance of a third inspector, Inspector Smith, was also monitored under different payoff conditions. His standardized ROC curve, plotted on normal probability paper, had a shallow slope (less than 45 degrees). How would you choose to quantify this inspector's sensitivity, and why?
If Inspector Morse was paid $10 for every hit, but charged $1 for every false alarm, compute the optimal (\beta ) value for this inspector. (a) Assume P(N)/P(S)=1 (1 pt).
If Inspector Morse was paid $10 for every hit, but charged $1 for every false alarm, compute the optimal (\beta ) value for this inspector. (a) Assume P(N)/P(S)=1 (1 pt).
Flashcards
Sensitivity
Sensitivity
The ability to detect a true signal amid noise.
Inspector Clouseau
Inspector Clouseau
An inspector with a liberal criterion for detecting weld flaws.
Inspector Morse
Inspector Morse
An inspector with a neutral criterion for detecting weld flaws.
Criterion
Criterion
Signup and view all the flashcards
d' (d prime)
d' (d prime)
Signup and view all the flashcards
Miss Rate
Miss Rate
Signup and view all the flashcards
Correct Rejection Rate
Correct Rejection Rate
Signup and view all the flashcards
ROC Curve
ROC Curve
Signup and view all the flashcards
Hit Rate
Hit Rate
Signup and view all the flashcards
False Alarm Rate
False Alarm Rate
Signup and view all the flashcards
Payoff Conditions
Payoff Conditions
Signup and view all the flashcards
Session One Rates
Session One Rates
Signup and view all the flashcards
Session Two Rates
Session Two Rates
Signup and view all the flashcards
Session Three Rates
Session Three Rates
Signup and view all the flashcards
Quantifying Sensitivity
Quantifying Sensitivity
Signup and view all the flashcards
Optimal b value
Optimal b value
Signup and view all the flashcards
P(N)/P(S)
P(N)/P(S)
Signup and view all the flashcards
Performance Monitoring
Performance Monitoring
Signup and view all the flashcards
False Alarm
False Alarm
Signup and view all the flashcards
Wickens Table
Wickens Table
Signup and view all the flashcards
Session Effects
Session Effects
Signup and view all the flashcards
Inspector Smith
Inspector Smith
Signup and view all the flashcards
Standardized ROC Curve
Standardized ROC Curve
Signup and view all the flashcards
Shallow Slope
Shallow Slope
Signup and view all the flashcards
Positive Rate
Positive Rate
Signup and view all the flashcards
Negative Rate
Negative Rate
Signup and view all the flashcards
Dynamic Adjustment
Dynamic Adjustment
Signup and view all the flashcards
Detection Theory
Detection Theory
Signup and view all the flashcards
Cognitive Bias
Cognitive Bias
Signup and view all the flashcards
Signal Detection
Signal Detection
Signup and view all the flashcards
Study Notes
Human Machine Systems Assignment
- Inspectors Clouseau and Morse monitor weld quality.
- Clouseau's neural activity distributions are shown for noise and signal+noise.
- Morse's neural activity distributions are also shown.
- Sensitivity: Clouseau is better at detecting flaws than Morse.
- Criteria:
- Morse has a neutral criterion.
- Clouseau has a liberal criterion.
- β < 1: This occurs for Morse, with β representing the criterion in their respective ROC graphs.
- Clouseau's d' value: If miss rate = 0.20 and correct rejection rate = 0.80, Clouseau’s d' value can be determined using the formula or a table.
- Payoff Conditions:
- Inspectors' performance was evaluated in three sessions with different reward structures.
- Session 1: $1 for each correct detection.
- Session 2: $1 for each correct detection, but a $1 penalty for false alarms.
- Session 3: Nothing for correct detection, but $1 penalty for false alarms.
- ROC Curves:
- Create ROC curves plotting sensitivity (hit rate) against false alarm rate for each session.
- All curves should be plotted on the same axes.
- Observer Effect of Payoffs: How much the two observers' performance was affected by the different reward structures.
- Quantifying Inspector Smith's Sensitivity: Given that Inspector Smith's ROC curve on normal probability paper has a slope of less than 45 degrees, sensitivity is quantified by measuring the area under the ROC curve; a shallow slope signals less sensitivity.
- Optimal β Value for Morse: Calculation of the optimal beta value (β) if paid $10 for every hit, and charged $1 for each false alarm.
Studying That Suits You
Use AI to generate personalized quizzes and flashcards to suit your learning preferences.
