Seeing as I’m too busy/lazy to get into a routine of writing original blog posts at present, I thought I’d post some essays that might be of interest or relevance or amusement or boredom, which I wrote whilst at university.
This is an essay I wrote in my third year of my undergraduate degree. The essay question is taken from a sample exam paper for the course PSYC3111 Human-Computer Interaction, in which 40 minutes are allocated for each exam essay. I wrote this essay in 45 minutes (shame on me and my tardiness) (and hence the embarrassingly short reference list).
“People can often use one of multiple strategies for sharing effort across tasks. Using an example, describe factors that may influence how people choose between different strategies for interleaving attention between multiple tasks presented on different devices.”
Human multitasking refers to the process of executing two or more tasks simultaneously by a single individual. There are costs associated with task interruptions and switching between tasks, for instance as demonstrated by studies showing people are slower to resume a primary task the longer it is suspended due to a secondary task (Hodgett & Jones, 2006). In particular, an area where the implications of multitasking are highly critical is driving, while dialling numbers on a mobile phone as a secondary task. Drivers are shown to interleave attention at task boundaries of lower cognitive load, or right before task boundaries if a given task requires particular high cognitive demand. However, when given the choice, drivers are found to use their mobile phone regardless of the demand of the road – implying that mobile device use is prioritised over driving safety.
It has been found that a general strategy when multitasking is to switch between multiple tasks at task boundaries, i.e. when the cognitive load is minimised. Payne, Duggan and Neth (2007) gave participants two tasks: both involved presentations of letters in randomised order with the instruction to generate as many words as possible in 10 minutes. In one task, it was easy to generate words, and in the other, it was difficult to generate words. Participants could only look at one set of randomised letters at a time, but they were free to switch between the easy and difficult sets. It was found that instead of focusing on one task only, participants regularly switched between the two. This switching was dependent on the amount of reward received from each task, such that participants were more persistent with a set if they continued to generate new words in that set. More critically, participants were found to switch after completed subgoals, i.e. after they had generated a word. Arguably, this was because cognitive load is minimised at subgoal completion, implying that people adopt a rational strategy of efficient information processing.
The Payne et al. study may not be particularly strong in terms of ecological validity, as simple word generation tasks are not frequently encountered in everyday life outside board games such as Scrabble. Also, as the two tasks were the same in all but level of difficulty, it is not clear whether the implications can be extended to cases where the separate tasks are substantially more different in nature – specifically, when multitasking occurs on separate devices. Nevertheless, Payne and colleagues demonstrated that even in an isolated laboratory setting, people tend to share time and effort between tasks in order to optimise reward.
More complicated and substantially different than simple word generation tasks, is driving whilst using mobile phones, which has further supported the view that people use subgoal completions to switch tasks. Brumby, Salvucci and Howes (2009) had participants using a driving simulator while entering a 10-digit US phone number of the format XXX-XXX-XXXX on a mobile phone. The same phone number was used repeatedly throughout the experiment to mimic the real life experience of dialling familiar phone numbers. The study revealed that when participants were instructed to focus on steering, they spent longer dialling the number; whereas when they were instructed to focus on dialling, the lateral deviation of the car from the road increased. Mirroring the findings of Payne et al. regarding interleaving at subgoals, participants tended to pause dialling and focus on steering at phone number chunk boundaries, i.e. as the number was 215-895-7634, participants would switch their attention back on the road after having dialled 215 and then return to dial 895. In addition to the strategy of reducing cognitive load, the results suggest that participants sacrifice total time spent on both tasks in favour of closely monitoring the primary task of driving. Simultaneously, they are performing the secondary task of dialling, effectively extending the time it takes them to complete dialling.
Besides reducing cognitive load by switching at subgoal completions, multitasking strategies seem to be influenced by performance objectives. Interestingly, when a UK phone number with only one chunk boundary (XXXXX-XXXXXX) was used in a study by Janssen and Brumby (2010), participants tended to interleave the dialling task before the task boundary and direct their attention to the steering. This qualifies the theory that people strictly use subgoal boundaries as cues for task interleaving. Rather, Janssen and Brumby’s study implies that people choose a strategy whereby they actively shift their attention under cognitive load in order to meet specific performance objectives – proper steering, in this case. The studies by Brumby and colleagues are, however, limited. The participant samples could be improved: in the US phone number studies, only 8 participants were used, with 2 being female. This is hardly a sufficient number for reaching any strong conclusions. More importantly, the driving simulator did not incorporate driving activities other than steering, such as braking or accelerating. This was therefore a very simplified representation of real-life driving behaviour, which is more complex and arguably involves a higher level of multitasking. Additionally, these studies did not fully incorporate the real-life aspect of free choice of when to do which task, as participants were explicitly instructed to perform both tasks together.
Research on more realistic multitasking behaviour whilst driving incorporating driver-initiated secondary task distractions has suggested that people do not adopt rational strategies of task interleaving. Instead of using a driving simulator, Horrey and Lesch (2009) had participants drive a real vehicle around a predetermined test track. The drivers became fully familiar with the demands of different sections of the road, narrow sections being of highest demand and shoulder sections being of lowest demand. Whilst in the track, participants were instructed to have a phone conversation, read a text message, find an address or recover an object from the floor of the car. A rational multitasking strategy would be to perform the secondary tasks at driving sections of low demand. This would both minimise time, effort and cognitive load as well as maximise task precision and success. Nevertheless, participants were shown not to defer interruptions by targeting low-demand driving sections, but rather perform the secondary tasks regardless of driving conditions. Indeed, it was found that they were more likely to initiate a phone call while pulling onto the roadway after having remained parked. The overall behaviour led to frequent errors in steering. Contrary to previous research, then, this study demonstrated that people tend to disregard predetermined performance objectives and environmental demands in favour of mobile device use.
Horrey and Lesch’ conclusions are arguably stronger than those of Brumby and colleagues because of the more naturalistic experimental situations. The sample size was also larger, with 20 participants and an equal gender balance. Limitations should still be held in mind, including the participants’ awareness that the test track was closed and thereby minimised risk of accidents. Also, their only motivation to perform secondary tasks was to complete them within a given deadline, whereas in everyday life, people have intrinsic motivations such as needing to make a phone call for personal reasons. Nevertheless, the study highlights the strong preference mobile devices have in our daily activities and how multitasking strategies often irrationally revolve around prioritising these.
In conclusion, people seem to adopt rational multitasking strategies in certain conditions, but not in others. In simple laboratory settings and highly controlled driving simulations, it has been demonstrated that people will rationally task interleave at subgoal completion. This is rational as cognitive load is minimised and people can monitor a primary task such as driving while simultaneously pursuing a secondary task of dialling a phone number. However, in more naturalistic driving experiments where people are given free choice of whether and when to engage in secondary tasks, it appears that people forsake safety benefits of performing secondary tasks at low cognitive load (i.e. at low environmental demands) in favour of mobile devices. In terms of wider, societal implications, then, public awareness campaigns directed towards banning mobile phone use while driving seem futile, as people will still persist in using mobile phones. Rather, these multitasking behaviours should be accepted and safer methods of incorporating mobile phone use in traffic should be developed. A potentially promising area is the development of audio-based mobile interfaces, whereby it could be hypothesised that performance on driving and dialling separately will improve due to the tasks being in different modalities.
Brumby, D.P., Salvucci, D.D., & Howes, A. (2009=. Focus on driving: How cognitive constraints shape the adaptation of strategy when dialing while driving. In Proceedings of the SIGHI Conference on Human Factors in Computing Systems, CHI ’09 (pp. 1629-1638). New York: ACM Press.
Hodgett, H.M., & Jones, D.M. (2006). Interruption of the Tower of London task: Support for a goal activation approach. Journal of Experimental Psychology: General, 135, 103 – 115.
Horrey, W.J., & Lesch, M.F. (2009). Driver-initiated distractions: Examining strategic adaptation for in-vehicle task initiation. Accident Analysis & Prevention, 41, 115-122.
Janssen, C.P., & Brumby, D.P. (2010). Strategic adaptation to performance objectives in a dual-task setting. Cognitive Science, 34, 1548-1560.
Payne, S.J., Duggan, G.B., & Neth, H. (2007). Discretionary task interleaving: Heuristics for time allocation in cognitive foraging. Journal of Experimental Psychology: General, 136, 370-388.