Download Human Factors Psychology - Cognitive Psychology - Lecture Slides and more Slides Cognitive Psychology in PDF only on Docsity! Cognitive Psychology
Notes 14
HUMAN FACTORS
PSYCHOLOGY
Human Factors Definition: “The study of those variables that influence the efficiency with which the human performer can interact with the inanimate components of a system to accomplish the system goals.” (Proctor & van Zandt, 1994, p. 2) Docsity.com Human Factors Consequences of failure: Maps given to pilots were wrong. Only one air traffic controller instead of the required two. The controller’s back was turned during the incident. Whoever is at “fault” is less important than the fact that it was possible to make that kind of mistake. Docsity.com A Story of the Range Implications: Typical user (5): My house (1): Per year: I: 0 II: 16 III: 25 IV: 28 I: 0 II: 3 III: 5 IV: 6 Lifetime: I: 0 II: 1040 III: 1625 IV: 1820 I: 0 II: 195 III: 325 IV: 390 Docsity.com A Story of the Range Shinar and Acton (1978): One main result is that you should not have a left- side control operate a right-side burner, and vice-versa. This expectation is present in 98% of the population. Docsity.com A Story of the Range This is a shameful lack of attention on the part of the people who design ranges. 65% are less than optimal arrangements. 27% violate an assumption made by 98% of the population. 0% are error-free. Speculating about the consequences of an error here makes it even more egregious. Docsity.com A Story of the Range To repeat: …Neither she nor Father seemed to study the other one’s nature. Each insisted the other one’s nature should work in some way it didn’t. I never once saw either of them observe the other in a calm, detached spirit, to see how his or her ego operated, and how to press the right button… Telling people to “be careful” or “try harder” is not a solution. Neither is relying on a label. More examples… Docsity.com Gas Pump 2 New gas pumps similar to these cost between $13,282 and $15,588. Times 12 = $160,000 - 187,000. Docsity.com Human Factors Basic human capabilities: Perceptual Attention span Memory span Physical limitations Etc. Docsity.com Human Factors Basic human capabilities: Perceptual Attention span Memory span Physical limitations Etc. Docsity.com Human Factors When targets are rare, what effect does that have on performance? Mammography 0.3% target prevalence (Wolfe, Horowitz, & Kenner, 2005). Baggage screening: Low (not typically published). Docsity.com Attention A search task can be modeled as a signal detection task. You have background noise. A signal adds some information to this noise, shifting the distribution along the evidence dimension. The separation determines sensitivity. Docsity.com Attention A search task can be modeled as a signal detection task. Users place a criterion along the evidence distribution to decide if a signal is present. Docsity.com Attention A search task can be modeled as a signal detection task. There are four kinds of outcomes: ○ You say “yes” and there’s signal: Hit. ○ You say “yes” and there’s no signal: False alarm. ○ You say “no” and there’s signal: Miss. ○ You say “no” and there’s no signal: Correct rejection. Docsity.com Attention Wolfe et al. (2007): Bursts of higher prevalence with feedback does seem to help. Suggests a regular “refreshment” procedure to move the criterion back. Docsity.com Some Psychology-- Designers Fundamental attribution error: We tend to discount the situation (environment) when trying to understand people’s behavior (Jones & Harris, 1967). The effect is usually associated with people doing something “bad.” Violating a norm. Making a mistake. (Letting a weapon on a plane.) Docsity.com Some Psychology-- Designers Classic example: Bystander apathy. People have a tendency not to help if they’re in crowds. We can judge people’s character in response to this, but overlooking the factors that actually influence people’s behavior will inhibit understanding. Docsity.com Some Psychology--Users Unrealistic optimism: People have a tendency to rate their personal risk as lower than that of the average person (Dunning, Heath, & Suls, 2004). Knowing the actual risks does not seem to inform people’s predictions for themselves. Docsity.com Some Psychology--Users Unrealistic optimism can account for a number of things people do that override impediments to human factors errors or that contribute to these errors. Docsity.com Some Psychology--Users For example, talking on a cell phone while driving is a bad idea (hands free or not). Drunk drivers’ errors arise from different sources (e.g., aggression), but talking and driving has similar performance effects as drinking and driving. Docsity.com Some Psychology--Users Another possibility: Dunning-Kruger effect (Kruger & Dunning, 1999): “For example, recognizing whether an argument is logically sound requires a firm grasp of the rules of logic. If people do not understand the rules of logic, not only will they make logical errors, but they will also not recognize that their arguments are logically defective—or that anyone else’s argument is logically superior” (Dunning et al., 2004, p. 73). Docsity.com Some Psychology--Users Another possibility: Dunning-Kruger effect (Kruger & Dunning, 1999): In other words, the competence needed to know that your behavior is leading you awry is the competence you need to do well. Not having competence makes you unable to learn from or predict mistakes. Maybe talking and driving is caused by this. Talking degrades your driving skills and the ability to detect that your skills are degraded. Docsity.com Some Psychology--Users Another possibility: Dunning-Kruger effect (Kruger & Dunning, 1999): Human factors implication: As hard as it is, don’t let people make mistakes (e.g., my car will not show a DVD on the center console screen while driving). Docsity.com Affordances What do all of these have in common? I would argue that they all afford an action that isn’t intended. Gibson (e.g., 1950) proposed a theory of direct perception. A caricature: Light is structured into an ambient optic array. Each point in this array carries potential information. This information takes the form of affordances if a particular organism happens to be there to pick it up. Information pick-up is direct. Docsity.com Affordances Perceiving affordances: If I’m looking for a place to sit, sit-on- ableness will be perceived by me in objects that have that property. If I need something to throw, that will be afforded. It’s kind of like an automatic process in that there isn’t conscious mediation. The affordance is just there. Docsity.com Affordances Evidence: Warren (1984) had people look at various configurations of stairs to rate climbability. There’s an optimal configuration for minimum energy expenditure, people preferred stairs that fit this configuration. People could accurately perceive climbability (based on a biomechanical model) from looking at the stairs. Docsity.com Population Stereotypes Which way should a knob turn to increase volume? What’s the best dimension to use for a volume control? For a next button? People’s answers to these questions can be grouped together as population stereotypes. Loosely speaking, people who drive tractors expect a certain arrangement of controls, people who use radios have certain expectations, etc. Docsity.com Population Stereotypes Population stereotypes matter because violations can lead to error. An alarm clock that you turn off to turn on can lead to problems. A volume knob on an alarm clock that goes down to increase volume can cause problems. A control on a radio in a car that violates expectations can remove just enough attention to cause an accident. Docsity.com Population Stereotypes More/ Less More is up More is down More is left More is right 85 11 24 72 Earlier/ Later Earlier is front Earlier is back Earlier is left Earlier is right 60 36 62 34 Docsity.com