Search

Attentional control – also referred to as cognitive control, executive control, executive attention, or executive functions – is defined as a person’s ability to maintain and implement a goal and goal-relevant information in the face of distraction (von Bastian et al., 2020). Recent research has put forward the difficulty of measuring attentional control as a general construct leading to success to different tasks and situations (e.g., Draheim et al., 2019; Hedge et al., 2018; Karr et al., 2018; Rey-Mermet et al., 2018, 2019, 2020). This difficulty has been mainly explained by the reliability and validity issues that the tasks used to measure attentional control are facing. One way to overcome these issues may be to improve how attentional control is extracted from reliable and valid tasks. For example, attentional control has been considered as a cognitive process necessary to perform working-memory (WM) tasks correctly (Engle, 2002; Engle & Kane, 2003; Kane et al., 2001; see also Miyake et al., 2000; Miyake & Friedman, 2012). In these studies, structural equation modeling (SEM) was used to model attentional control in two ways. On one side, attentional control was assessed as the residual variance of WM after controlling for short-term memory (STM; see, e.g., Engle et al., 1999). Thus, attentional control was conceptualized as the construct that is left once short-term memory has been extracted. On the other side, attentional control was modelled as the common variance shared by all WM and STM tasks (see, e.g., Kane et al., 2004). Here, attentional control was estimated as the common factor across all WM and STM tasks, and the loadings from the WM tasks were higher than those from the STM tasks. Thus, attentional control was conceptualized as a construct that is required in all tasks but this requirement is higher in the WM tasks. So far, there is no systematic comparison of both ways of modelling attentional control. To our knowledge, only two studies have estimated attentional control in both ways (Hale et al., 2011; Rey-Mermet & Rothen, 2022). However, in both studies, special samples were tested (i.e., participants were aged between 20 and 89, and participants had a morning or evening chronotype, respectively). Thus, their results may be confounded by age and chronotype. Moreover, across these studies, there was no systematic investigation of which model fitted best the data, how reliable attentional control was assessed, and how robust the results are. Therefore, it remains unknown which way is best suited to model attentional control.

In memory research, different types of memory – such as priming, recognition, and recall – have been put forward, (e.g., Baddeley et al., 2020). Priming refers to the memory through which the presentation of an item facilitates the processing of a subsequent item. Recognition refers to the memory through which an item is identified as having been previously encountered. Recall refers to the memory through which a previously learned item is brought into mind. Typically, recall and recognition measures have been emphasized as relying on a common process (e.g., Unsworth & Brewer, 2009; Yonelinas, 2002). Previous research has also suggested that priming and recognition measures might rely on a common underlying process (e.g., Berry et al., 2008; Henke, 2010; Jacoby, 1991; Wagner & Gabrieli, 1998; Yonelinas, 2002). Both lines of research have so far only focused on two different types of memory, thereby ignoring the full spectrum of different types of memory. Therefore, it is unknown whether a common process between priming and recognition measures can be observed once the more established shared process between recognition and recall measures has been taken into account. Most previous research supporting the assumption of a common underlying process between priming and recognition measures has focused on experimental manipulations, such as, for example, those involving attention or study duration (see Yonelinas, 2002, for an overview). So far, to our knowledge, only one study has used a correlational approach to investigate whether priming, recognition, and recall are associated (Wang & Yonelinas, 2012). In that study, participants first performed a study phase in which they decided whether the word was concrete or abstract. In a second step, they performed a recognition task in which they were asked to decide whether the word was presented in the study phase or not. Half of the words were presented in the study phase. For each word, participants additionally performed subjective confidence ratings about the memory strength (i.e., a receiver operating characteristic confidence rating paradigm in Experiment 1a and a remember/know paradigm in Experiment 1b). These confidence ratings were used to categorize recognition performance as priming- vs. recall-based. In a final step, participants performed a free association task as a priming task. In this task, participants were presented with words and they were asked to respond verbally with the first strongly associated word that came to mind. Similar to the recognition task, half of the words were associated to the words presented in the study phase. In this case, the idea is that participants responded more frequently with the words presented in the study phase. The results revealed moderate to strong correlation between the priming measures and the priming-based recognition measures (i.e., .47 in Experiment 1a and .59 in Experiment 1b). However, the correlations were small and not significant between the priming measures and the recall-based recognition measures (i.e., -.15 in Experiment 1a and .29 in Experiment 1b). In a subsequent experiment, participants were asked to perform a cued-recall task instead of the priming task. In that task, they had to respond only with a previously studied word that was associated with the presented word. The results showed substantial correlations between recall and recognition measures, irrespective of the confidence ratings. Overall, these findings were interpreted as supporting the assumption of a common underlying process between priming and recognition measures. However, there are some challenges in this study. First, although priming, recognition and recall tasks were investigated, these tasks were performed pairwise in distinct experiments. Thus, it does not allow us to determine the impact of recall on the common process observed between priming and recognition measures. Second, the sample sizes were small for a correlational approach (i.e., 49 participants in Experiment 1a, 24 participants in Experiment 1b, and 36 participants in Experiment 2; see Brysbaert, 2019, for current standards in sample sizes). Moreover, the samples only consisted of undergraduate students. In addition to decreasing the generality of the results, this might have decreased the magnitude of some correlations (e.g., the correlations between priming and recall-based recognition measures). The reason is that testing only undergraduate students limits the variability that could be observed in the different processes. Third, the validity of confidence ratings is widely debated, making unclear whether it is legitimate to assess recognition in this particular way (see, e.g., Dunn, 2008; Fiacconi, 2022; Kantner & Dobbins, 2019). Fourth, the reliability of the measures used by Wang and Yonelinas (2012) was not reported. This is, however, critical because the reliability of measures limits the maximum magnitude of a correlation that can be observed between two measures (Parsons et al., 2019). Accordingly, a correlation between two measures cannot be larger than the average reliability of these measures. However, previous research reported low to very low reliability estimates for the priming measures (see, e.g., Buchner & Wippich, 2000; Meier & Perrig, 2000; Waechter et al., 2010). Therefore, it is quite surprising to observe such moderate to strong correlations between a priming measure and a priming-based recognition measure. The goal of the present study is to determine with a correlational approach whether a common process between priming and recognition can be established once the common process between recognition and recall has been taken into account. We will overcome the challenges from Wang and Yonelinas (2012) in four ways. First, we will bypass the challenge of ignoring the full spectrum of memory types by conducting two experiments in which all participants will perform priming tasks, recognition tasks, and free-recall tasks with either words, pseudowords or pictures as material. Second, we will overcome the challenge of small undergraduate samples by having large sample sizes consisting of participants from the general population. Third, we will bypass the challenge of measuring priming- and recall-based recognition with confidence ratings by using structural equation modeling (SEM). That is, following previous research (Brehmer et al., 2020; Quamme et al., 2004; Unsworth & Brewer, 2009; Vann et al., 2009; Yonelinas et al., 2002, 2007), we will model the common process between recall and recognition as the common variance between free-recall and recognition measures. Similarly, we will model the common process between priming and recognition as the common variance between priming and recognition measures. Fourth, we aim to increase the reliability of the priming measures by following the recommendations from Buchner and Wippich (2000). That is, we opted for tasks with clear instructions and responses. To this end, in the first experiment, we will ask participants to perform a word identification and a picture clarification task. In the word-identification task, participants are asked to identify a word that is presented very briefly. In the picture clarification, participants are presented with pictures in their most defragmented versions, and they are asked to identify the object represented in the picture. In case they identify the object correctly, a new picture will be presented. In case they do not identify the object correctly, they are presented with the same picture in a less defragmented version. In both priming tasks, half of the stimuli are presented in a previous study phase, whereas the other half are not. The priming effect will thus be assessed as the performance difference between those stimuli that were presented in the study phase (i.e., “old” stimuli) and those stimuli were not previously presented (i.e., “new” stimuli). In the second experiment, we ask participants to perform a repetition-priming task with words, pseudowords, and pictures as materials. In this task, two stimuli are presented in succession. The first stimulus is presented very briefly, and participants are asked to read the word or pseudoword or to name the picture. After a short delay, the second stimulus is presented until the participants perform a simple decision (i.e., a decision about the pleasantness or about the size of the object). We opt for a procedure in which the stimulus repeats in 25% of the trials. The reason is that Waechter et al. (2010) reported acceptable reliability estimates for a similar procedure. In this experiment, the priming effect will be assessed as the performance difference between trials in which the stimulus repeats and those in which it differs.

Circadian arousal has been assumed to affect maintenance (i.e., the ability to retain access to information) and interference control (i.e., the ability to avoid being distracted by irrelevant information). Accordingly, maintenance and interference control should be best when participants are tested at on-peak times compared to off-peak times. However, the evidence – in particular for interference control – is mixed. Whereas some studies found an impact of circadian arousal, others did not. There are multiple reasons for these mixed results. First, most studies were statistically underpowered (i.e., they included small sample sizes in a between-subjects design). Second, although different tasks were used for each construct across the studies, most studies included only one task. Third, in most tasks used to assess interference control, it is unclear to what extend these tasks measure interference control in a reliable and valid way. In the present study, we aim to test the impact of circadian arousal on maintenance and interference control by examining a large sample (more than 400 participants) in a within-subjects design (i.e., all participants will be tested at on- and off-peak times). In order to asses both constructs reliably and in a task-independent manner, our participants will be asked to perform tasks used to assess short-term memory (STM, i.e., our ability to retain access to a limited amount of information) and working memory (WM, i.e., our ability to retain access to a limited amount of information and manipulate these information in the service of complex tasks). Maintenance will be measured at the latent-variable level by establishing a common factor from all STM and WM tasks. Interference control will be measured at the latent-variable level by separating WM-specific variance from WM and STM maintenance. The impact of circadian arousal will be then assessed using a linear second-order latent growth curve model.

Interference tasks such as the Simon, Flanker, and Stroop tasks have largely been used to investigate cognitive processing, with an emphasis on executive control processes such as the ability to inhibit irrelevant information (i.e., inhibition) and disengage attention from the previous trial (i.e., attentional switching). In the present study, we aim to examine the theoretical issues concerning changes in speed of processing, executive control processes, and modified brain activity patterns underlying physiological ageing, mild cognitive impairment (MCI) and Alzheimer’s disease (AD). To this end, we will carry out a meta-analysis of studies that used the mentioned interference tasks and that tested cognitive changes in healthy older adults as well as in older adults with MCI and AD. In addition, we will review the neurophysiological findings obtained by means of electroencephalogram (EEG), event-related brain potentials (ERP), and functional magnetic resonance imaging (fMRI) to shed light on several theoretical issues concerning changes in speed of processing, inhibitory control, and modified brain activity patterns underlying physiological ageing and AD. We conclude by discussing to what extent the results of the present study provide support for neurocognitive theories explaining cognitive and neural changes associated with physiological ageing, MCI, and AD.

Test anxiety refers to the feelings of acute stress and worry experienced during evaluative (i.e., testing) situations (Cassady & Johnson, 2002). Thus, the distinctive feature of test anxiety is its occurrence during evaluative situations (Hembree, 1988). Test anxiety has been shown to be associated with decreased memory and learning performance (Hembree, 1988). Retrieval practice is a learning strategy in which one actively recalls information from memory and is consistently associated with memory enhancement (Roediger & Karpicke, 2006). Recent evidence suggests that retrieval practice may reduce the detrimental effects of acute stress on memory (Smith et al., 2016). Currently, there is no experimental evidence investigating the impact of learning with retrieval practice during test-like conditions. The present study investigates whether learning with retrieval practice can overcome the detrimental effects of test anxiety on memory. In the learning session, participants will learn two educational texts using a retrieval practice learning strategy for one text and a restudy learning strategy for the second text. One week later, participants will perform the recall session where they are instructed to recall as many items from both texts as possible. Prior to memory recall, half of the participants are randomly placed in a test-like condition via a set of evaluative instructions (evaluative group). In this group, participants are told that they will perform an IQ test and a memory test in order to evaluate their intelligence and memory. The word “test” is repeated in the instructions in order to further increase the stakes and pressure as during evaluative situations (see attachment in measured variables section for full instruction sets). The other half are placed in a neutral testing condition via control instructions (control group). In this group, participants are told they will perform a puzzle task and a memory task. The word “task” instead of “test” is used to not give off a sense of pressure. Throughout both learning and recall sessions, all participants will fill out a state anxiety questionnaire to capture their levels of state anxiety (i.e., the anxiety felt in the current moment). At the end of both learning and recall sessions, heart rate will be also assessed as a physiological measurement of state anxiety. Finally, at the end of the study, participants will respond to test anxiety questionnaires measuring their propensity for test anxiety. Participants will also complete questionnaires to assess their level of trait anxiety (i.e., their general propensity to feel anxious across different situations). Finally, to control for the possibility that participants have co-morbidity with depression, participants will additionally complete a questionnaire assessing depression.

Attentional control – also referred, for example, to executive functions or cognitive control – is our ability to maintain and implement a goal and goal-relevant information when facing distraction (von Bastian et al., 2020). Recent research has emphasized the difficulty of establishing attentional control as a cognitive psychometric construct (e.g., Draheim et al., 2019, 2020; Karr et al., 2018; Rey-Mermet et al., 2018, 2019, 2020; von Bastian et al., 2020). That is, multiple empirical measures of attentional control did not correlate substantially among each other. Accordingly, attentional control could not be measured as a latent variable (i.e., a factor) representing shared variance across the measures. Several reasons can explain this finding. First, attentional control does not exist as a psychometric construct (Rey-Mermet et al., 2018, 2019, 2020, 2021). Second, the tasks typically used to assess attentional control are so diverse that they may measure task-specific constructs (Rey-Mermet et al., 2018, 2019, 2020, 2021). Third, the previous attentional-control measures do not allow us to differentiate maintaining and implementing goal-relevant information (a process called “proactive control”) from the goal reactivation caused by being distracted by irrelevant information (Braver et al., 2007; a process called “reactive control”; e.g., Braver, 2012). The goal of the present study is to test these possibilities. Specifically, we distinguish between proactive control and reactive control by using the AX-CPT paradigm. Moreover, we diminish the diversity of the attentional-control measures by asking participants to perform this task with different materials (i.e., letters, words, dots, matrices, and pictures). In the letter version of this task, participants are asked to press a target response when the cue A is followed after a delay period by the probe X (i.e., when a so-called “AX trial” is presented); in all other cases, they are asked to press a non-target response. Specifically, these other cases are trials in which the cue A is followed by any other letter than X (i.e., so-called “AY trials”), trials in which any other letter than A is followed by the letter X (i.e., so-called “BX trials”), and trials in which any other letter than A is followed by any other letter than X (i.e., so-called “BY trials”). One key feature of this paradigm is that AX trials are presented very frequently. Accordingly, the presentation of the cue A results in increased expectancy for the probe X and the presentation of the probe X results in a prepotent response tendency to the target response. In this paradigm, different performance indices are typically computed, such as: (1) the “d’ context” or “d’ proactive”, which is computed as a d’ between AX and BX trials, and which is rather used as a measure of proactive control (e.g., Gonthier et al., 2016; Richmond et al., 2015; Snijder, 2021; Stawarczyk et al., 2014); (2) the “d’ reactive”, which is computed as a d’ between AX and AY trials, and which is rather used as a measure of reactive control (Richmond et al., 2015; Stawarczyk et al., 2014); (3) the “BX probe interference”, which is computed as a difference between BX and BY trials, and which is rather used as a measure of reactive control (Snijder, 2021); (4) the proactive behavioral index (PBI), which is computed as [AY – BX]/[AY + BX], and which is considered as a performance index reflecting the balance between proactive and reactive control (e.g., Gonthier et al., 2016; Snijder, 2021); and (5) the “A-cue bias”, which is computed as a C criterion between AX and AY trials, and which is used to measure the bias to press the target response (e.g., Gonthier et al., 2016; Richmond et al., 2015; Snijder, 2021). Previous research has shown that the AX-CPT paradigm and, in particular, the d’ context is related to working memory (WM, i.e., our ability to maintain a limited number of goal-relevant representations; Miyake & Shah, 1999; Oberauer, 2009). For example, the d’ context was reported to be larger for participants with high WM capacity (Redick, 2014; Redick & Engle, 2011; Wiemers & Redick, 2018; but see Richmond et al., 2015, for an exception) or to correlate with WM (Ball, 2015; Boudewyn et al., 2015; MacDonald et al., 2005; Stawarczyk et al., 2014; but see Ball, 2015; and Boudewyn et al., 2015 for two exceptions including a verbal complex span as WM task and the dot pattern or the letter AX-CPT paradigm, respectively). Moreover, at a conceptual level, proactive control and WM seems very similar as they both involve the maintenance of relevant information. Therefore, a further goal of the present study is to investigate to what extend proactive control and WM are related at the latent-variable level in order to determine whether proactive control is a unique and specific construct. To this end, we will ask our participants to perform WM tasks in addition to the AX-CPT paradigms. The WM tasks will be three complex span tasks with either verbal, numerical, or spatial materials, respectively.

In the Simon task, participants perform a decision on non-spatial features (e.g., stimulus color) by responding with a left or right key-press to a stimulus presented on the left or right side of the screen. In the flanker task, they classify the central character while ignoring the flanking characters. In each task, there is a conflict between the response-relevant features and the response-irrelevant features (i.e., the location on the screen for the Simon task, and the flankers for the flanker task). Thus, in both tasks, resolving conflict requires to inhibit irrelevant features and to focus on relevant features. When both tasks were combined within the same trial (e.g., when the row of characters was presented on the left or right side of the screen), most previous research has shown an interaction. In the present study, we investigated whether this interaction is affected by a multiplicative priming of the correct response occurring when both Simon and flanker irrelevant features co-activate the correct response (Exp. 1), a spatial overlap between Simon and flanker features (Exp. 2), and the learning of stimulus-response pairings (Exp. 3). The results only show an impact of multiplicative priming.

The purpose of the present study was to determine whether responding to multiple concurrent conflicts results in a simultaneous or sequential conflict resolution. To this end, we measured event-related potentials (ERPs) in a paradigm combining a Stroop and a flanker task. In this paradigm, participants were asked to respond to the color of the central letter while ignoring the meaning of the word (Stroop task) and the color of the flanking letters (flanker task). Trials were either incongruent (i.e., inducing a conflict between two response alternatives) or congruent (i.e., inducing no response conflict) in both tasks, or incongruent in one task and congruent in the other task. The behavioral results showed a smaller Stroop congruency effect (i.e., a smaller difference between Stroop incongruent and congruent trials) for flanker incongruent than for flanker congruent trials, replicating previous findings. The ERP results showed that an early ERP component (i.e., P2) was associated with the resolution of the flanker conflict, whereas a later component (i.e., N450) was associated with the resolution of the Stroop conflict. Together, these findings emphasize a sequential organization of conflict resolution processes in the brain which is adaptive when facing multiple concurrent conflicts.