Lecture 19- Vigilance

Question 1

What is vigilance?

Answer 1

• The ability to maintain focus and remain alert over long periods.
• This is vital in many real world tasks where you have to detect small changes that might result in danger (e.g. screening bags or for cancer, checking the water as a life guard) or for quality checking (e.g. proof reading)

Question 2

What kind of stimuli might a vigilance task invovle?

Answer 2

Either auditory or visual/ both

Question 3

What is not a vigilance task? (how to identify in a list of options)

Answer 3

A typist checking for errors as they type: not a vigilance task because they are engaged in the task of typing as they go. They not observing output from something else where it would be hard to pay attention.

Question 4

How does human factors relate to vigilance tasks?

Answer 4

How can we redesign the vigilance tasks to make us as humans better at them (overall we are pretty shit)

Question 5

How has the history of vigilance research played out?

Answer 5

• Originated 100 years ago where the focus was on those with attention disorders
• Now focused on normal people (shift happened after world war 2) when Norman Maxwell looked at how submariners could miss enemy ships on radars

Question 6

What is a simple test used to measure vigilance?

Answer 6

Look at a clock hand, get people to report if it skips forward

Question 7

What are four classifications for something to be a vigilance task?

Answer 7

• Monitor 1 thing + info source
• Prolonged period
• Detect low probability signal (why we become complacent and miss things)
• Signal requires response

Question 8

What is an example of a continuous performance task?

Answer 8

• Widely used
• Task is to respond to some specific stimuli on the screen e.g. press the space bar when you see the x (can make more complicated by adding additional rules)

Question 9

Draw a 2 by 2 matrix of the correct decisions and wrong decisions you can make with regards to signal detection…

Answer 9

Answers on slide

Question 10

Which type of error is worse in a deadly situation?

Answer 10

-Miss (error of omission)

Question 11

What is signal perception theory?

Answer 11

• Signal detection theory combines hit and false alarm data into a signal piece of perceptual data.
• This is referred to a D prime

Question 12

What is vigilance decrement? Why might this occur?

Answer 12

• Performance declines with increased time on a task

- Might occur because people have reduced alertness from lack of stimulation

Question 13

Are vigilance tasks demanding?

Answer 13

Yes, overall self report measures show them to be quite stressful

Question 14

In relation to vigilance how will humans perform if we need to take the reins from self driving cars when a stressful situation arrises?

Answer 14

Terrible- we will have completely loss attention on the road and it will be very hard for us to be brought back to alertness in time to act

Question 15

What did Finomore, Mathews, Shaw & Warm (2009) show about selecting individuals to perform vigilance tasks?

Answer 15

• Individual differences i.e. compulsivity doesn’t necessarily predict performance on a vigilance task
• Not all measures are created equal
• And the relationship would vary based on parameters and the context

We need to be careful when we generalize!

Question 16

What did Hattenschwiler, Mendes & Schwaninger (2019) show about security screening?

Answer 16

• It is a complex task, security officers need to identify not just weapons but anything that could be assembled into a weapon
• This could be difficult depending on angle
• 3D images help a lot in that you can rotate the image to see better

Question 17

What did Wolfe et al. find about the prevalence effect and vigilance tasks?

Answer 17

• If targets occur more often we will spend a lot more time searching when there isn’t one (motor errors occur, we think there is going to be one at any point) therefore, false alarms occur
• Reducing the target prevalence reduces false alarms but because our brain thinks they are never going to occur we are a lot faster at checking and have misses. The result of this could be deadly.
• This is a stable effect both in lab and in real life tasks (e.g. security screening)

Question 18

How can we stop the prevalence effect? What groups of people do these findings apply to?

Answer 18

• False feedback this could work: but needs to be plausible and person can’t go back and check
• Insert fake targets: quality control, infers that they do their job well if they respond appropriately

These measures get rid of the complacency that nothing is ever going to happen. Apply to both individuals and trained professionals