Two types of backward crosstalk: Sequential modulations and evidence from the diffusion model

A spurious correlation between difference scores in evidence-accumulation model parameters

Evidence-accumulation models are a useful tool for investigating the cognitive processes that give rise to behavioural data patterns in reaction times (RTs) and error rates. In their simplest form, evidence-accumulation models include three parameters: The average rate of evidence accumulation over time (drift rate) and the amount of evidence that needs to be accumulated before a response becomes selected (boundary) both characterise the response-selection process; a third parameter summarises all processes before and after the response-selection process (non-decision time). Researchers often compute experimental effects as simple difference scores between two within-subject conditions and such difference scores can also be computed on model parameters. In the present paper, we report spurious correlations between such model parameter difference scores, both in empirical data and in computer simulations. The most pronounced spurious effect is a negative correlation between boundary difference and non-decision difference, which amounts to r = - .70 or larger. In the simulations, we only observed this spurious negative correlation when either (a) there was no true difference in model parameters between simulated experimental conditions, or (b) only drift rate was manipulated between simulated experimental conditions; when a true difference existed in boundary separation, non-decision time, or all three main parameters, the correlation disappeared. We suggest that care should be taken when using evidence-accumulation model difference scores for correlational approaches because the parameter difference scores can correlate in the absence of any true inter-individual differences at the population level.

Temporal aspects of two types of backward crosstalk in dual-tasks: An analysis of continuous mouse-tracking data

A common explanation for processing limitations in dual-tasking is the existence of a bottleneck during response selection, meaning that the selection of responses can only occur serially for different tasks. However, a large body of data shows that features of a (secondary) Task 2 can already influence the processing of a (primary) Task 1. Such effects are referred to as backward crosstalk effects (BCEs). In the present study, two types of such BCEs were investigated: the compatibility-based BCE, which depends on the dimensional (often spatial) overlap between task features, and a BCE based on a go/no-go task in Task 2 (no-go BCE). Joining a line of research that suggests different mechanisms for these two types of BCEs, we investigated them using a mouse-tracking setup. Time continuous analyses revealed that the compatibility-based BCE triggered a spatial activation of the Task 2 response early during Task 1 processing, whereas the no-go BCE triggered an inhibitory effect in the case of a no-go Task 2, which spills over to Task 1 execution. This occurred, however, earlier in the time course than expected. The results are discussed with regard to recent models of dual-task processing.

Given the option, people avoid incongruent responses in a dual-tasking situation

While past work on how people can optimize dual-tasking has focused on strategic timing (i.e., when to select responses), little is known about the extent to which people can optimize dual-tasking by taking care of which responses they select. Here we test whether spatial (in)congruency influences response selection in free-choice trials. In two experiments, we combined two visual-manual tasks with spatial stimulus- and response characteristics: Participants responded to the stimulus words "left" and "right" in a forced choice task and responded "up", "down", "left" or "right" with an arrow-key to either a free choice prompt or an X located at the respective position. In Experiment 1, participants reduced the proportion of incongruent pairs of responses (i.e., left in one and right in the other task). In Experiment 2, we found that such flexibility in response selection also holds in more constrained environments: Within runs of four trials the free-choice options were continuously reduced based on the responses already given. The combined results of Experiments 1 and 2 suggest that response selection in free choice trials is driven by performance optimization beyond response conflict.

Two types of between-task conflict trigger respective processing adjustments within one dual-task

In dual tasking, two different kinds of between-task conflict occur. Because in both cases, Task 2 characteristics affect Task 1 performance, they are commonly referred to as backward crosstalk effects (BCE): One with a conflict at the response selection stage when Task 1 and Task 2 have dimensional overlap (the compatibility-based BCE) and one with a conflict at the motor execution stage when response inhibition resulting from a Task 2 no-go-trial interferes with simultaneous response execution in Task 1 (the no-go BCE). Recent research suggests that these BCEs differ not only in their underlying cognitive processes, but also in how cognitive control is regulated. Here, we investigated whether both can be produced in a single dual-task set up and whether they trigger their respective processing adjustments (i.e., a sequential modulation). In three experiments, participants categorized numbers as smaller or larger than 5 in Task 1. In Experiments 1 and 2, numbers were responded to irrespective of numerical size (go-response) as Task 2. Further, dimensional overlap was provided by (non)corresponding size information in both stimuli, which was strengthened in Experiment 2 by presenting S1 and S2 in the same/different color in compatible/incompatible trials, respectively. In Experiment 3, participants were required to perform a number-size categorization also in Task 2, establishing strong dimensional overlap by activating the same or different response categories in both tasks. In all three experiments, the number 5 served as the no-go stimulus in Task 2 to induce a no-go BCE on Task 1. By and large, our results show that both types of between-task conflicts not only occur within the same type of BCE, but they also trigger their respective sequential modulation.

The Backward Crosstalk Effect Does Not Depend on the Degree of a Preceding Response Conflict

A common observation in dual tasking is a performance decrement in one or both tasks compared with single tasking. Besides, more specific interference occurs depending on certain characteristics of the two tasks. In particular, even Task 1 performance is often improved when responses in both tasks are compatible (e.g., both require left responses) compared to when they are incompatible: the compatibility-based backward crosstalk effect (BCE). Similar to what is observed for conflict tasks, the BCE is sequentially modulated: It is larger following compatible than following incompatible trials. Previous work has attributed this observation to adaptation effects triggered by response conflict arising during incompatible trials. In two experiments, we assessed sequential modulations following trials with different degrees of such a response conflict. In contrast to our expectations, a clear and sizeable sequential modulation was observed even under conditions where no BCE, and thus no empirical sign of an objective response conflict, was present in the previous trial. Therefore, our results show sequential modulations even without prior response conflict, which is not the (sole) trigger of sequential modulations accordingly. We discuss these results with regard to other potential triggers such as the subjective experience of conflict or difficulty, episodic retrieval, and repetitions of response combinations.

S1-R2 and R1-R2 Backward Crosstalk Both Affect the Central Processing Stage

A frequent observation in dual-task experiments is that performance in Task 1 is influenced by conceptual or spatial overlap with features of Task 2. Such compatibility-based backward crosstalk effects (BCEs) can occur when overlap exists between the responses of two tasks-the R1-R2 BCE-or between the stimulus in Task 1 and the response in Task 2-the S1-R2 BCE. The present study investigated whether the S1-R2 BCE has a perceptual locus, and by implication, whether the two BCEs have a common processing locus or different ones. To this end, we applied the additive factors logic and manipulated the duration of the Task 1 perceptual stage. The results argue against a perceptual locus for both BCEs. As a possible explanation, we suggest that the R1-R2 BCE and the S1-R2 BCE have their locus within a capacity-limited central stage, but that they arise from different processes within this stage. The R1-R2 BCE influences Task 1 response selection, whereas the S1-R2 BCE influences Task 1 stimulus classification. A plausible though post-hoc model is presented within the Discussion.

Two sources of task prioritization: The interplay of effector-based and task order-based capacity allocation in the PRP paradigm

When processing of two tasks overlaps, performance is known to suffer. In the well-established psychological refractory period (PRP) paradigm, tasks are triggered by two stimuli with a short temporal delay (stimulus onset asynchrony; SOA), thereby allowing control of the degree of task overlap. A decrease of the SOA reliably yields longer RTs of the task associated with the second stimulus (Task 2) while performance in the other task (Task 1) remains largely unaffected. This Task 2-specific SOA effect is usually interpreted in terms of central capacity limitations. Particularly, it has been assumed that response selection in Task 2 is delayed due to the allocation of less capacity until this process has been completed in Task 1. Recently, another important factor determining task prioritization has been proposed—namely, the particular effector systems associated with tasks. Here, we study both sources of task prioritization simultaneously by systematically combining three different effector systems (pairwise combinations of oculomotor, vocal, and manual responses) in the PRP paradigm. Specifically, we asked whether task order-based task prioritization (SOA effect) is modulated as a function of Task 2 effector system. The results indicate a modulation of SOA effects when the same (oculomotor) Task 1 is combined with a vocal versus a manual Task 2. This is incompatible with the assumption that SOA effects are solely determined by Task 1 response selection duration. Instead, they support the view that dual-task processing bottlenecks are resolved by establishing a capacity allocation scheme fed by multiple input factors, including attentional weights associated with particular effector systems.

Parallel and serial task processing in the PRP paradigm: a drift-diffusion model approach

Even after a long time of research on dual-tasking, the question whether the two tasks are always processed serially (response selection bottleneck models, RSB) or also in parallel (capacity-sharing models) is still going on. The first models postulate that the central processing stages of two tasks cannot overlap, producing a central processing bottleneck in Task 2. The second class of models posits that cognitive resources are shared between the central processing stages of two tasks, allowing for parallel processing. In a series of three experiments, we aimed at inducing parallel vs. serial processing by manipulating the relative frequency of short vs. long SOAs (Experiments 1 and 2) and including no-go trials in Task 2 (Experiment 3). Beyond the conventional response time (RT) analyses, we employed drift–diffusion model analyses to differentiate between parallel and serial processing. Even though our findings were rather consistent across the three experiments, they neither support unambiguously the assumptions derived from the RSB model nor those derived from capacity-sharing models. SOA frequency might lead to an adaptation to frequent time patterns. Overall, our diffusion model results and mean RTs seem to be better explained by participant’s time expectancies.

Item-specific proportion congruency (ISPC) modulates, but does not generate, the backward crosstalk effect

When both tasks in a psychological refractory period (PRP) paradigm have compatible manual responses, a compatibility benefit in RT can often be observed on Task1 performance, in apparent violation of a strict traditional response selection bottleneck model. This compatibility-based backward crosstalk effect (BCE) has been generally attributed to automatic activation of Task2 response information, in parallel with attended Task1 performance. This paper tests a potential alternative mechanism of the BCE. Item-specific proportion congruency (ISPC) effects are previously well demonstrated, where learning of associations between stimuli and task conflict (e.g., that particular Stroop items are typically incongruent) allows rapidly and automatically elicited control adjustments in performance. Similar proportion manipulations have recently been shown to modulate the BCE in dual-task performance. If participants could similarly learn associations between particular pairs of stimuli and resulting response conflict in a PRP task, this kind of mechanism could produce relative speeding versus slowing of Task1 RT on response compatible versus incompatible trials. This pattern of data directly describes the BCE, and represents a potential alternative mechanism that does not require any response crosstalk, and would reinforce a stricter view of the response selection bottleneck model, if true. Over two experiments, we demonstrate that while the BCE is sensitive to ISPC-like effects based on Task1 conflict contingencies, the BCE is insensitive to relationships between particular pairs of stimuli and associated conflict. While ISPC effects can modulate the BCE, they do not generate the BCE. These findings reinforce the current Task2 parallel response activation account of the BCE.

Introspection about backward crosstalk in dual-task performance

The present study investigated participants’ ability to introspect about the effect of between-task crosstalk in dual tasks. In two experiments, participants performed a compatibility-based backward crosstalk dual task, and additionally provided estimates of their RTs (introspective reaction times, IRTs) after each trial (Experiment 1) or after each pair of prime and test trials (Experiment 2). In both experiments, the objective performance showed the typical backward crosstalk effect and its sequential modulation depending on compatibility in the previous trial. Very similar patterns were observed in IRTs, despite the typical unawareness of the PRP effect. In sum, these results demonstrate the reliability of between-task crosstalk in dual tasks and that people’s introspection about the temporal processing demands in this complex dual-task situation is intriguingly accurate and severely limited at the same time.

Mood state and conflict adaptation: an update and a diffusion model analysis

The present study investigated the affective modulation of conflict adaptation. In a first step, we conducted a direct replication of a previous study (Schuch & Pütz, 2018). Positive vs. negative mood state was induced by a success-failure manipulation (between-groups, N = 40 per group). In a subsequent task-switching experiment, the congruency sequence effect was assessed in task repetitions and task switches, measuring conflict adaptation within tasks and between tasks, respectively. We found conflict adaptation (averaged across task repetitions and task switches) to be enhanced in negative mood. We did not replicate our previous finding of enhanced conflict adaptation in task switches in positive mood. In a second step, we combined the replication data with the original data set, yielding a larger database with N = 80 per mood group. Using diffusion modeling, we explored the affective modulation of conflict adaptation in task repetitions. Conflict adaptation was reflected in drift rate, consistent with the idea that response conflict triggers an increase in processing selectivity, thereby attenuating the influence of the irrelevant stimulus dimension. Conflict adaptation was also reflected in boundary separation, suggesting that response conflict on the previous trial triggered an increase in response caution. The mood manipulation did not seem to affect processing selectivity (as captured by drift rate) but affected the setting of response caution (as captured by the boundary separation parameter), with faster and more error-prone responding in the negative than positive mood group. We discuss theoretical implications of these findings, and also briefly consider the affective modulations of other cognitive control measures.

Sequence Knowledge on When and What Supports Dual-Tasking

The constraints in overlapping response selection have been established in dual-tasking studies with random sequence of stimuli and responses as well as random stimulus onset asynchrony (SOA). While this approach makes it possible to control for advance activation of upcoming stimuli or responses, it leaves open whether such preparatory processing can indeed influence dual-task performance. We investigated whether and how the sequence of stimuli and responses and the sequence of SOAs can be learned and used under dual-tasking. In each trial, participants (N = 28 in Experiment 1 and N = 30 in Experiment 2) were first presented with a random two-choice task followed by a four-choice Serial Reaction Time Task (SRTT), presented in a sequence of length four (position sequence). The SOA (timing) sequence also had length four. In test phases, one or both of the sequences were randomized. Results showed that both position and timing sequences were learned and supported dual-task performance, suggesting that predictive processing with respect to timing and identity of stimuli and responses can help to circumvent the response selection bottleneck constraints. Furthermore, in contrast to previous work on acquisition of interval sequences in single tasking, we found that the sequence of what (i.e. stimulus) and the sequence of when (i.e. interval between two tasks) contributed independently to performance.