It is not good to talk: conversation has a fixed interference cost on attention regardless of difficulty

Melina A Kunar,Louise Cole,Angeline Cox,Jessica Ocampo

doi:10.1186/s41235-018-0124-5

Melina A Kunar, Louise Cole + Show 2 more

Open Access

https://doi.org/10.1186/s41235-018-0124-5

Copy DOI

Abstract

It is well-documented that telephone conversations lead to impaired driving performance. Kunar et al. (Psychon Bull Rev 15:1135–1140, 2008) showed that this deficit was, in part, due to a dual-task cost of conversation on sustained visual attention. Using a multiple object tracking (MOT) task they found that the act of conversing on a hands-free telephone resulted in slower response times and increased errors compared to when participants performed the MOT task alone. The current study investigates whether the dual-task impairment of conversation on sustained attention is affected by conversation difficulty or task difficulty, and whether there was a dual-task deficit on attention when participants overheard half a conversation. Experiment 1 manipulated conversation difficulty by asking participants to discuss either easy questions or difficult questions. The results showed that there was no difference in the dual-task cost depending on conversation difficulty. Experiment 2 showed a similar dual-task deficit of attention in both an easy and a difficult visual search task. Experiments 3 and 4 showed that in contrast to work using a dot tracking and choice reaction time task (Emberson et al., Psychol Sci 21:1383–1388, 2010), there was little deficit on MOT performance of hearing half a conversation, provided people heard the conversations in their native language. The results are discussed in terms of a resource-depleted account of attentional resources showing a fixed conversational-interference cost on attention.

Highlights

As we go about our daily lives we often perform several tasks simultaneously
Within-participants analysis of variance (ANOVA) for mean correct Reaction time (RT) showed there to be a main effect of condition, F(2, 58) = 17.1, p < 0.01, ηp2 = 0.371
Planned t tests showed a significant difference between the no conversation and difficult conversation condition, t (29) = 5.1, p < 0.01, d = 0.872, where RTs were slower in the difficult conversation condition compared to the no conversation condition, and a significant difference between the no conversation and easy conversation condition, t (29) = 4.5, p < 0.01, d = 0.718, where RTs were slower in the easy conversation condition compared to the no conversation condition

Summary

Background

As we go about our daily lives we often perform several tasks simultaneously. Sometimes these two tasks are simple and can be performed together (e.g., humming while walking). It has been found that performing a more difficult word generation task impaired simulated driving and MOT performance compared to performing an easier speech shadowing task (see Strayer & Johnston, 2001 and Kunar et al, 2008, respectively) Evidence from these studies are in favour of a difficulty-dependent account, stating that the more demanding a task, the more cognitive resources it uses, leaving fewer attentional resources available for subsequent tasks. To preview the results, when the type of conversation was controlled for by asking questions in both conditions, a similar dual-task deficit on MOT performance occurred with both the difficult and the easy conversations, in line with a resource-depleted account of attentional resources.

Results and discussion