Transcript 슬라이드 제목 없음 - Korea University

IMEN 315 인간공학
6. Cognition
 people perceive, think, and remember
 three stages of human information processing system – perception,
central processing or transforming, responding
INFORMATIN PROCESSING MODELS
top-down
processing
learning
retrieval
산업경영공학과
IMEN 315 인간공학
SELECTIVE ATTENTION
 not guarantee perception, but necessary to achieve it
 four factors for the selection of channels to attend
 salience: bottom-up process, attentional capture (blindness)
 expectancy, value: top down processes
 effort
PERCEPTION
Three Perceptual Processes
1.
2.
bottom-up feature analysis
unitization: sets of features familiar (represented in LTM), more rapid
and automatic than perceptual processing
 poor bottom-up processing: degradation of visual stimulus (short glance,
tiny text, poor illumination) and auditory event (masking noise, low
intensity, unfamiliar accents)
산업경영공학과
IMEN 315 인간공학
3.
top-down processing: correct guess from expectations, based upon past
experience in LTM; associations between the perceived stimulus and
event (frequency and context)
Human Factors Guidelines in Perception
1.
2.
3.

maximize bottom-up processing
maximize automaticity and unitization
maximize top-down processing
avoid confusion; use a smaller vocabulary; create context; exploit
redundancy
WORKING MEMORY
A Model of Working Memory
 Baddeley (1986, 1990) – central executive component
 visuospatial sketch pad – analog spatial form while it is being used
 phonological loop – verbal info in an acoustic form
산업경영공학과
IMEN 315 인간공학
Limits of Working Memory
Capacity
 Around 7±2 chunks of information (Miller, 1956)
 What makes a single chunk  Familiarity based on past experience
(LTM), similar to unitization in perception
1. Chunking reduces the number of items in WM, increasing the capacity
of working memory
2. Chunking makes use of meaningful associations in LTM  retention of
the information
3. Material more easily rehearsed, more likely to be transferred to LTM
 Perceptual chunks by spatial separation
Time
 Maintenance rehearsal
 Half life in WM (Card, Moran, Newell, 1986) – 7 sec for a memory store of
three chunks and 70 secs for one chunk
산업경영공학과
IMEN 315 인간공학
Confusability and Similarity
Attention and Similarity
 WM is resource-limited
Human Factors Implications of Working Memory Limits
1.
2.
3.
4.
Minimize WM load
Provide visual echoes
Provide placeholders for sequential tasks
Exploit chunking
 Physical chunk size – 3 to 4 numbers or letters per chunk
 Meaningful sequences
 Superiority of letters over numbers
 Keeping numbers separate from letters
5. Minimize confusability
6. Avoid unnecessary zeros in codes to be remembered
7. Consider WM limits in istructions
산업경영공학과
IMEN 315 인간공학
LONG-TERM MEMORY
 Learning, training, retrieval, forgetting
 Semantic memory (memory for facts or procedures) or event memory
Basic Mechanism
Strength
 Frequency and recency of its use
Associations
WM and LTM
 Rote memory (rehearsal through simple repetition)
Forgetting
1. Weak strength due to low frequency and recency
2. Weak or few associations with other information
3. Interfering associations
 Recall, recognition
산업경영공학과
IMEN 315 인간공학
Organization of Info in LTM
 Info in LTM in associative networks (semantic network)
 The structure of the database compatible or congruent with the user’s
semantic network
Schemas and Scripts
 Schema – the knowledge structure about a particular topic
 Scripts – schemas that a typical sequence of activities
Mental Models
 schemas about dynamic systems
 Generates a set of expectancies
 Population stereotype
Cognitive Maps
 Mental representations of spatial information
 Mentally straightening
 Preferred or canonical orientation (mental rotation)
산업경영공학과
IMEN 315 인간공학
LTM Implications for Design
1.
2.
Encourage regular use of info to increase frequency and recency
Encourage active verbalization or reproduction of info that is to be
recalled
3. Standardize
4. Use memory aids
 Knowledge in the world vs. knowledge in the head
5. Careful design info to be remembered
6. Design to support development of correct mental models
 Visibility (affordance)
Episodic Memory for Events
 The personal knowledge or memory of a specific event or episode is
acquired from a single experience – very much based on visual imagery
 not always faithful “video replays”, having a number of biases
 Episodic memory process is far from perfect
산업경영공학과
IMEN 315 인간공학
 Police lineup recognition – 20% incorrect at all 3 stages of encoding,
storage, retrieval
 Cognitive interview (CI) – not recognition but recall procedure
산업경영공학과
IMEN 315 인간공학
Prospective Memory for Future Events
 Failures of prospective memory are forgetting to do something in the
future – sometimes called absentmindedness
 Reminders, checklists
SITUATION AWARENESS
 characterize user’s awareness of the meaning of dynamic changes in
their environment
 Endsley (1995) -- the perception of the elements in the environment
within a volume of time and space, the comprehension of their meaning,
and the projection of their status in the near future
 SA is distinct from performance
Measuring SA
 SA global assessment technique (SAGAT)
 subjective awareness -- metacognition
산업경영공학과
IMEN 315 인간공학
Importance of SA to Human Factors
1.
2.
3.
designing easy-to-interpret displays of dynamic systems
an important tool for accident analysis
important for training
PROBLEM SOLVING AND TROUBLESHOOTING
 troubleshooting a step within a problem-solving sequence
 troubleshooting requires a series of tests to diagnose the problem while
problem solving involves actions to implement the solution
Challenges
 heavy cognitive activity, and human performance often limited
 in troubleshooting, two or three active hypotheses in WM
 troubleshooting closely depend upon appropriate cues and test outcomes
 susceptible to attention and perceptual biases
산업경영공학과
IMEN 315 인간공학
 an important top-down processing bias in troubleshooting – cognitive
tunneling or confirmation bias
1. high system complexity
2. intermittent failures of a given system component
PLANNING AND SCHEDULING





planning may be invoked in the absence of problem solving
in dynamic systems, predicted state and command (ideal) state
sluggish (higher inertia) systems – longer range planning
the importance to planning – level 3 SA, mental model (simulation)
predictive displays
산업경영공학과
IMEN 315 인간공학
METACOGNITION AND EFFORT
 meta-knowledge or metacognition – people’s knowledge about their own
knowledge
 anticipated effort – seeking additional information related to selective
attention is also related to another metacognition
ATTENTION AND TIME-SHARING
 divide attention – do two or more things at one time
 resource demand, structure, similarity, resource allocation
Mental Effort and Resource Demand
 the relationship between singletask difficulty and dial-task
divided attention decrements -resource theory
 automaticity
산업경영공학과
IMEN 315 인간공학
Structural Similarity
 structural similarity – the similarity between key processing structures of
both tasks in a concurrently performed pair
 multiple resource theory (Navon & Gopher, 1979; Wickens, 1984, 2002) - different structures in human information processing behave as if they
were supported by multiple resources
산업경영공학과
IMEN 315 인간공학
Confusion
 similarity between items in WM; similarity-based confusion in visual
sensation;
 concurrent performance of two tasks that both have similar material
Task Management and Interruptions
 if interference, then will they both suffer? Or will one or the other be
“protected”?
 dual task performance (primary task vs. secondary task )  task
management  resource allocation
 successful time-sharing strategies – optimal switching of attention
between tasks  parallel processing vs. cognitive tunneling
산업경영공학과
IMEN 315 인간공학
Addressing Time-Sharing Overload
1.
2.
3.
4.
Task redesign
interface redesign
training
automation
산업경영공학과