Failed replication: The “brain drain” effect of Ward et al. (2017), the purported effect on cognitive performance due to the mere presence of one's smartphone

Reexamining the “brain drain” effect: A replication of Ward et al. (2017). Ana C.Ruiz Pardo, John Paul Minda. Acta Psychologica, Volume 230, October 2022, 103717. https://doi.org/10.1016/j.actpsy.2022.103717

Abstract: The present study was a pre-registered direct replication of Ward et al.'s (2017) second experiment (OSF pre-registration found at: https://osf.io/5fq4r). This replication assigned both smartphone location (on desk, in pocket/bag, or outside of the testing room) and smartphone power (on, or off) for a total of six conditions. Participants completed an automated operation span (OSpan) task, a Cue-Dependent Go/No-Go task, and the smartphone attachment and dependency inventory. It was hypothesized that performance on an attention-demanding task (i.e., the OSpan task) would be worse for those in closer proximity to their smartphone (on desk) and that those with greater smartphone attachment and dependency would have a larger “brain drain” effect. Using the same tasks and conditions as in Ward et al.'s (2017) second experiment, the present study found that the “brain drain” effect did not replicate: there was no difference between smartphone location conditions on performance on either the o-span task or the go/no-go task. These findings demonstrate that the mere presence of one's smartphone may not be enough to affect cognitive performance. Understanding these effects is crucial in a time where smartphones are a basic necessity.

Keywords: Smartphone presenceAttentionResponse inhibitionSmartphone dependency

4. Discussion

Smartphones provide an easy and effective method of communicating with the world right at our fingertips. The rising prevalence of smartphones (Pew Research Center, 2019) has prompted research including possible behavioural addictions (WHO, 2015) and how these might affect cognitive abilities. Although there are many benefits to using a smartphone in terms of communication, the present study investigated how smartphones affect performance on cognitively demanding tasks. This was done by reexamining the “brain drain” effect (i.e., those who were in closer proximity to their smartphone performed worse on a cognitively demanding task, which is moderated by smartphone reliance) found by Ward et al.'s (2017) second experiment. The three main hypotheses (i.e., location effect, power effect, and moderation effect) from Ward et al. (2017) were evaluated in the present study.

4.1. The OSpan Task and the Cue-Dependent Go/No-Go task

There were no significant main or interaction effects of smartphone location on performance on OSpan absolute score. There was a significant main effect of cue type and an interaction effect of cue type and smartphone location on omission errors in the Cue-Dependent Go/No-Go task (Bezdjian et al., 2009). However, this effect was explored with tests of simple main effects and found no significant effect of smartphone location for either cue type. Overall, the present study did replicate Ward et al.'s null effect on the Cue-Dependent Go/No-Go task performance. More notably, however, the present study's findings failed to replicate Ward et al.'s main effect concerning performance on the OSpan task (Unsworth et al., 2005). Therefore, the “brain drain” effect was not replicated in the present study. The smartphone power effect hypothesis was supported: there was no significant difference between power conditions (i.e., powered ON vs. OFF) on performance for both tasks. This was a replication of Ward et al.'s findings.

4.2. Factor analysis of the smartphone attachment and dependency inventory

Findings from a principal components analysis on the smartphone attachment and dependency inventory (Ward et al., 2017) partially supported the two-factor findings from Ward et al. (i.e., smartphone dependence and emotional attachment), but also added a third factor: smartphone distractibility.

4.3. Moderation analysis on OSpan Score

Finally, the moderation effect did not replicate: smartphone dependency, emotional attachment, and distractibility were not significant moderators of OSpan performance. In contrast with Ward and colleagues, emotional attachment showed a trend for those in the desk condition, where higher emotional attachment predicted lower OSpan performance. It should be noted that this analysis was completed as a pre-registered analysis and was exploratory in nature. Overall, the present study demonstrated that the “brain drain” effect may not be a replicable effect of smartphone presence on cognition. Possible reasons for this are given.

4.4. Failure to replicate the “brain drain” effect

A stark difference in performance was observed between the present study's OSpan performance and in Ward et al.'s (2017) second experiment. This was one of the critical results in Ward et al., because they described the OSpan as a difficult working memory task intended to be sensitive to a decrease in cognitive capacity. They argued that this difficulty difference was the reason why they found an effect on OSpan performance but not on the Cue-Dependent Go/No-Go (Bezdjian et al., 2009) performance, and indeed this was the locus of the “brain drain” effect. However, participants in our study did not find the OSpan as challenging and the presence of their own smartphone on the desk did not seem to interfere with their performance on the task. Not only was the mean-difference in OSpan performance for the present study much smaller than for Ward et al. but also, the average performance between the present study and Ward et al. implies that participants in the present study did not find the OSpan task as challenging as in Ward et al.'s study. This difference was also seen when compared to Ward et al.'s first experiment, where average OSpan performance was lower than a score of 34. These differences may explain why participants in our experiment did not experience a “brain drain” in their performance: the task did not diminish participant’s available cognitive capacity. In fact, the present study showed participants with perfect performance on both the math and letter recall components and, consequently, there was a possible ceiling effect. This defeated the purpose of the OSpan as a more difficult cognitive task. Therefore, to determine the underlying mechanisms behind smartphones' impact on cognition, future work should use reliable and normed cognitive tasks. The Cambridge Brain Sciences (CBS; Hampshire et al., 2012) test battery, for example, evaluates a broad range of cognitive abilities such as selective attention, response inhibition, reasoning, and working memory. These short cognitive tests have been used across different populations (Wild et al., 2018) to test people across three main components (i.e., short-term memory, reasoning, and verbal ability) with varying difficulty levels. Therefore, using this test battery could examine how smartphone presence affects an overview of cognitive aspects and could explain why the present study did not replicate the “brain drain” effect.

Another limitation to consider in the present study was the measure for smartphone reliance. In order to directly compare the present study to Ward et al.'s second experiment, the smartphone attachment and dependency inventory (Ward et al., 2017) was used to measure participant’s smartphone attachment and dependency (i.e., reliance). However, current research typically uses additional measures to measure things such as nomophobia (i.e., the fear of being without one's phone or the internet; (Yildirim & Correia, 2015) and smartphone involvement (Walsh et al., 2010). Although the use of the smartphone attachment and dependency inventory (Ward et al., 2017) allowed the present study to directly compare findings to Ward et al.'s second experiment, measuring smartphone reliance based on only one scale limited the present study. Therefore, future research should expand on other measures of smartphone reliance.

Additionally, it should be noted that the present study focused on a North American population to compare directly to Ward et al.'s original study. However, as smartphone prevalence emerges globally and differently across countries (Silver, 2019), future research should consider comparing different countries' smartphone use.