Dual-task performance

Question 1

According to the mobile phone report (2002), order stopping distances from the smallest to the largest. Is the handheld vs. hands-free difference supported by independent studies? Why?

Answer 1

Normal, have drunk alcohol, talking on hands-free device, talking on handheld device. No, because the limitation is cognitive rather than motor in origin

Question 2

Welford (1952) suggested that dual-tasking is difficult given our single channel, limited capacity processor. Whilst supported by Cherry (1953) with very similar tasks of dichotomous listening, Allport (1952) found that…, supporting the…

Answer 2

The ability to sight read music was the same when performed on its own vs. at the same time as performing a verbal shadowing task. Multiple resources theory i.e. separate pools of attention for different types of task

Question 3

Broadbent argued that it was impossible to perform two tasks simultaneously due to a cognitive ___. What we call multitasking actually involves rapid switching between tasks. The switch cost can be reduced by…

Answer 3

Bottleneck. Expertise at performing one or both of the two tasks

Question 4

Posner (1971) partially refutes multiple resources theory by finding that T1 (task 1) and T2 (task 2) performance deteriorated, the closer together in time the two tasks were performed despite…. What were T1 and T2?

Answer 4

T1 involving a verbal, visual input and T2 involving a nonverbal, auditory input and outputs being made on different hands (but both being in the motor domain!). T1: is this 2nd letter the same or different from the 1st letter? T2: is this 2nd tone the same or different from the 1st tone?

Question 5

The “first” and “second” labels on the letter RT graph indicate…

Answer 5

The first and second TONES

Question 6

How did McLeod (1978) demonstrate that to remove the performance cost substantially, outputs as well as inputs must be different domains? This supports MRT

Answer 6

By finding a reduction in performance cost when responses to the auditory task were made manually and responses to the visual task were made in the auditory domain (by saying BOP)

Question 7

What did McLeod (1978) find specifically re: performance costs?

Answer 7

That although dual task performance was always slower, the drop in performance did not depend on the timing of the task I.e. the usual time-locked nature of the performance cost disappeared

Question 8

Triesman (1973) provided further support for MRT by finding that recall and word detection was more accurate when streams were presented…

Answer 8

Bimodally. I.e. one visually onscreen and one in the auditory domain rather than both visually or both in the auditory domain

Question 9

Shallice (1985) investigated whether the auditory input logogen could be dissociated from the articulatory (auditory) output logogen. This is the same as asking whether…

Answer 9

Speech perception mechanisms can be dissociated from speech production mechanisms

Question 10

Wickens’ (1984) box model has been hugely influential in cognitive ___. It states that when simultaneously performed tasks ___ for the same resources performance is impaired. Does it explain the danger of driving at the same time as being on the phone well?

Answer 10

Ergonomic, compete. No

Question 11

There are 4 axes on Wickens’ (1984) model. From left to right these are…

Answer 11

Input code (spatial vs. verbal), input modalities (visual vs. auditory), stage (encoding, central processing, response) and output (manual vs. vocal)

Question 12

Kinsbourne (1971) required participants to balance a dowel on their left or right index finger whilst speaking or in silent. It was found that…. Why was a dual task benefit found? Does the box model account for distinct hemispheric resources?

Answer 12

The left hemisphere/ right hand showed a dual task deficit because talking also demands the LH, whereas the right hemisphere/ left hand showed a dual task “benefit”! Because talking arouses the brain. No

Question 13

What is Kinsbourne’s (1978) neural noise H1?

Answer 13

Dual task limitations on performance arise when tasks draw upon neural areas which are close and highly inter-linked. Degree of task interference is the inverse of the “functional distance” between each task’s cerebral control centre

Question 14

Beilock (2002) presented 2 examples in which T1 performance was boosted by simultaneously carrying out T2. What were they? Participants were…

Answer 14

Performing an auditory task whilst putting a golfball or dribbling a football. Highly practised sports people

Question 15

How did Wickens (1984) account for the driving + talking exception and similar others?

Answer 15

By arguing that resource pools leak and so become not entirely separate when under great pressure

Question 16

Describe the procedure of a typical psychological refractory period (PRP) experiment by Pashler (1992). What is the stimulus onset asynchrony (SOA)?

Answer 16

Participants respond to S1 (a bar) with their left hands and S2 (a letter) with their right hands. SOA is the time between presentations of S1 and S2

Question 17

The performance cost in PRP experiments is to ___, not ___

Answer 17

R2, R1

Question 18

What is the PRP?

Answer 18

The time during which the response for S1 is being selected and therefore no further responses for no further stimuli can be selected

Question 19

If S1 ___ ___ is prolonged (takes up more time than expected), the onset of S2 response selection and so RTs to S2 will be delayed. The bottleneck lies at the stage of ___ ___, not in…or…

Answer 19

Response selection. Stimulus perception. Response execution

Question 20

RT on the y axis is plotted against ___ on the x axis in Posner’s (1992) study. The plot for R1 shows a ___ line. The plot for R2 shows a line…

Answer 20

SOA. Straight. Decreasing less and less as the SOA increases until eventually it is also straight

Question 21

If S1 response execution is prolonged, S2 RTs are/not prolonged. If S2 response selection is prolonged, S2 RTs are/not prolonged. If S2 perceptual processing is prolonged, S2 RTs are/not prolonged. Why?

Answer 21

Not because S2 response selection can still be conducted on time. Are because response execution could have been performed sooner (S1 response execution is independent). Not because S2 response selection could not be performed any sooner because the response selection bottleneck was not free

Question 22

Pashler (1992) refutes the box model by finding a PRP even when…. Pashler also found that the PRP casued by the response selection bottleneck cannot be reduced by…

Answer 22

Input modalities differ. Automatisation of T1 or T2 I.e. practise!

Question 23

Give 2 findings of how the PRP can be overcome. How is it overcome in each e.g.?

Answer 23

By using the same stimuli for T1 and T2 I.e. Pps make multiple responses to the same stimuli (green or red? Horizontal or vertical?). Because response selection is only required once. By making R1 and R2 ideomotor-compatible with the stimuli e.g. Say A when you hear A. Because RS occurs within perception

Question 24

Is there a performance cost even when one of the two tasks does not require attention? How did Joseph (1997) conclude this?

Answer 24

Yes, in contrast to Triesman’s beliefs. Pps attended to T1 - reporting white targets in a stream of black distractors - whilst performing T2 - detecting feature targets in a circular array. T2 should not have required attention but a “% correct” cost was still found. The accuracy cost was less and less as the lag increased

Question 25

Does practice in one task equally ameliorate the performance cost when T2 is a feature search task and when T2 is a conjunction search task? Would Triesman be pleased with Braun’s (1998) conclusion?

Answer 25

No, practice reduces (but does not eliminate) the performance cost in feature search much more than in conjunction search. Yes, with the very small performance cost after practice at feature search

Question 26

Given that drivers have much practice at driving, is there a dual-task performance cost when braking in a simulator? Which 3 DVs were measured and what was found?

Answer 26

Yes. Gas-off time (time between brake light onset and letting up the gas pedal) showed a similar curve when plotted against SOA as brake RT (time taken to press the brake). Movement time (time between letting up the gas and braking) was only slightly affected by SOA = interference targets the decision to let up the gas

Question 27

Can we eliminate the PRP with practice? How did Schumacher (2001) investigate this?

Answer 27

Yes, T1 was auditory-vocal (say 1, 2, 3 to low, medium or high tone) and T2 was visual-manual (left, middle or right finger for word beginning with, containing or ending with O). After 2000 trials the performance was almost eliminated

Question 28

What might make the PRP in Schumacher’s (2001) tasks easier to overcome with practice? Which according to Schumacher’s (2001) E2 was particularly influential?

Answer 28

Ideomotor-compatibility, contrasting input & output modalities, simultaneous presentation of tasks rather than Pashler’s serial presentation. Simultaneous presentation: when the SOA was raised above 0 and Pps were told to perform the task first, the PRP returned

Question 29

Schumacher’s 3rd experiment highlighted the large extent of individual differences in how daring people are to multi-task? What was found in terms of different cultures’ ways of scheduling? What is required to achieve perfect time sharing?

Answer 29

American Pps were more likely to exhibit daring scheduling of tasks (concurrent performance), whilst Chinese Pps were more likely to exhibit cautious scheduling of tasks (sequential performance). Adaptive executive control

Question 30

Are women or men better at multi-tasking?

Answer 30

Men

Question 31

What % of the population are supertaskers who show no dual task performance cost?

Answer 31

2.5%

Question 32

To summarise, name 3 types of dual task interference models and give e.g.s

Answer 32

1) cross-talk models e.g. the neural noise H1
2) capacity sharing models e.g. single vs. multiple resource theories
3) bottleneck models e.g. single channel theories, PRP and the response selection bottleneck

Question 33

To summarise, name 8 factors which make dual-task performance easier:

Answer 33

1) practice at one/both tasks, 2) performing tasks which draw upon different resources (input/output modalities & hemispheres) = 3) task dissimilarity, 4) high ideomotor compatibility, 5) being a supertasker, 6) increasing the SOA, 7) unspeeded (not rushed) responding, 8) linking responses to attributes of the same stimulus