Alt 6: Attention and workload Flashcards
What is workload?
Mental workload (referred to as just workload) is a general term used to describe the cost of accomplishing task requirements:
Workload varies as a function of
task demands placed on the human operator
AND
the operator’s capacity to meet those demands
Task demands influenced by factors:
Task Difficulty
Task Priority
Situation
Operator capacity
Individual operator differences:
Each operator has a supply of mental resources
Mental resources shape how we cope with task
demands
excessive workload:
overload
sustained low workload:
underload
Underload can cause:
(1) Increased fatigue
(2) Lower task engagement
(3) Lower vigilance
(4) Generally reduces operators capacity to deal
with the unexpected (spare capacity)
Why measure workload?
Workload prediction
To maintain workload within acceptable limits
- Safety
- Change management: Know in advance what effect changes will have
Why measure workload?
Equipment assessment
Optimize the system to suit operator
- Designers may want to know what workload will be at the design, production, or usage stages
- Identify problem areas where performance may decline
Why measure workload?
Employee selection
Choose the right operators, or further training
- Can train to improve some capacity, but only up to a point
There are three major categories of workload measures:
- Performance
- Subjective
- Physiological
Performance measures
System designers are typically concerned with performance, some might argue workload is only important if it affects performance
Primary task performance
Common general measures: Accuracy, Response time, Signal detection
Task specific measures: Mouse movements, Altitude deviation, Braking time,
Primary Task performance measurement advantages
Workload reflected directly by performance outcome
Non-invasive and non-interfering
Tracks changes in workload dynamically. (i.e., as performance proceeds)
Uncontaminated by memory issues
Primary Task performance measurement disadvantages
Mental workload is not the only thing that can influence performance
Doesn’t take effort into account
Might not always have the technology or interface
available to monitor performance
Secondary task performance
Purpose: To measure the spare capacity not being used by the primary task
Usually, an increase in primary task workload is associated with a decrease in secondary task perf.
Potential measures including: Rhythmic tapping task, Probe reaction tasks, Auditory monitoring
Secondary task performance measurement advantages
The same secondary task can be used with very different primary tasks
There are a collection of available (and validated) secondary tasks
Provide a measure of space capacity
Useful when there is little primary task data available
(e.g. monitoring jobs, baggage scanner)
Secondary task performance measurement disadvantages
Only relevant when secondary task taps into the same resources as the primary task
It adds in another task to the work environment (could increase workload)
Can interfere/interrupt primary task
Subjective measures
Where you ask someone to quantify their personal experience of workload
Three variables that are useful to categorise different subjective rating techniques:
Dimensionality (single or multiple)
Evaluation (absolute or relative)
Immediacy (when queried?)
Subjective measures advantages
Cheap and simple to administer
No need to interfere with primary task
Reasonably good reliability and validity
NASA-TLX has a huge literature available for comparison
Subjective measures disadvantages
People aren’t always aware of workload
People find it difficult not to baseline their workload
Not sensitive to moment-to-moment changes
Physiological measures
The assumption is that biological functions can reflect mental work done.
Popular Physiological measures include:
Cardiovascular
Ocular
Brain activity
Cardiovascular measures
Heart-rate variability shown to increase with workload
Particularly relevant to emotional stress
Potential issues:
• Need to establish a baseline
• Quite a range due to individual differences
• Only sensitive when stress is involved
Ocular measures
Pupil diameter increases with workload
Visual scanning patterns and pathways change with workload
Length of time spent fixating on items changes with workload
Brain activity measures
It is intuitive that brain activity will reflect mental workload
Brain activity sensitive to different kinds of cognitive demands (e.g. memory, vigilance, flexibility, etc.)
Measures: Electroencephalogram (EEG), Event- related potentials (ERP), and Cerebral blood flow
Physiological measures advantages
Provide a relatively continuous measure over time
Objective
Many different techniques available
Physiological measures disadvantages
High cost
Often obtrusive
Cannot be used to determine how the operator feels about the task (can be important)
Multiple Resource Theory
If two (or more) tasks are similar they will draw on a common resource pool. This will affect performance negatively If the two (or more) require different resources, then performing the tasks at the same time will have little negative consequence
Multiple resource theory assumptions (3)
- There is a common, finite pool of resources that can be flexibly allocated to tasks
- Allocating more resources to a task speeds up the rate of processing of that task
- The quantity of resources made available can be increased or decreased
Multiple resource theory
Stages of processing:
Perception (I can see a car stopped ahead…)
• Cognition (…I predict that car will be a problem…)
• Response (…slows down)
Multiple resource theory
Codes of processing
(relevant in all stages)
Spatial (e.g. pointing out a car)
Verbal / Linguistic (e.g. describing a car)
Multiple resource theory
Modalities
(only relevant for perception stage) Auditory perception (e.g. listening to radio) Visual perception (e.g. looking for a sign)
Multiple resource theory
Visual channels
(only relevant in visual modality) Focal channel (focus of attention, e.g. car in front) Ambient channel (peripheral vision, e.g. cars on either side)
Driver safety
Driving provides an example of a practical task domain where competition for limited resources is common
