Abstract

Previous studies suggest that both motivation and task difficulty expectations activate brain regions associated with cognitive control. However, it remains an open question whether motivational and cognitive determinants of control have similar or dissociable impacts on conflict processing on a neural level. The current study tested the effects of motivation and conflict expectancy on activity in regions related to processing of the target and the distractor information. Participants performed a picture–word interference task in which we manipulated the size of performance-dependent monetary rewards (level of motivation) and the ratio of congruent to incongruent trials within a block (level of conflict expectancy). Our results suggest that motivation improves conflict processing by facilitating task-relevant stimulus processing and task difficulty expectations mainly modulate the processing of distractor information. We conclude that motivation and conflict expectancy engage dissociable control strategies during conflict resolution.

INTRODUCTION

An important precondition for successful goal-directed behavior is the operation of cognitive control processes, which allow for adjustment of information processing in response to changing task demands. For example, if we are distracted by noise outside the window while working in our office, we are able to focus more attention on the current work and ignore the distracting noise. One important aim of research on cognitive control is to reveal possible determinants of such control adjustments. Previous research has suggested that cognitive control can be modulated by the level of motivation as well as by cognitive factors such as expected task difficulty. However, their specific influence is still a matter of current research.

The assumption that expectations on task difficulty may activate cognitive control processes is supported by behavioral data on the role of expectations for conflict processing: It has been found that the amount of conflict in interference tasks is reduced when participants expect a high proportion of incongruent trials relative to congruent trials in a block (Funes, Lupianez, & Humphreys, 2010; Kane & Engle, 2003; Carter et al., 2000). This finding has been interpreted as evidence for the assumption that the expectancy of upcoming conflicts activates control processes, which improve task preparation and reduce conflict processing costs. Neuroimaging studies have shown that expected task difficulty may correlate with activity in frontal regions including the dorsal ACC (dACC) and the medial superior frontal cortex (Wilk, Ezekiel, & Morton, 2012; Dosenbach et al., 2007; Carter et al., 2000). These regions are involved in adjusting attention and/or response selection processes to guarantee successful behavior.

Besides cognitive factors, possible motivational determinants of control have increasingly aroused the interest of research on cognitive control in the recent years. Action goals of high motivational salience, for example, because of the prospect of reward, can lead to adjustments of the cognitive strategy to achieve the desired action goal (Kim, 2013). Similar to expected task difficulty, high motivation is supposed to trigger cognitive control processes because an enhanced level of motivation was found to result in reduced interference effects in tasks requiring conflict processing (Lindstrom, Mattsson-Marn, Golkar, & Olsson, 2013; Soutschek, Strobach, & Schubert, 2013; Padmala & Pessoa, 2011). This finding is explained by the assumption that the prospect of reward activates control processes biasing the selection of task-relevant relative to distracting information during conflict processing. On a neural level, increased motivation was found to be associated with both the enhancement of tonic activity and the suppression of transient activity in regions associated with cognitive control such as the dACC, the lateral pFC, and the striatum (Krebs, Boehler, Egner, & Woldorff, 2011; Padmala & Pessoa, 2011; Jimura, Locke, & Braver, 2010; Kouneiher, Charron, & Koechlin, 2009).

Although most neuroimaging studies investigating the impact of motivation or expectancy on cognitive control focused their analyses on prefrontal regions of the control network, there is evidence that motivation and expectancy modulate activity also in posterior brain regions like the parietal cortex or the fusiform gyrus (Krebs, Boehler, Roberts, Song, & Woldorff, 2012; Wilk et al., 2012; Padmala & Pessoa, 2011). For example, Krebs et al. investigated the neural correlates of motivation and task difficulty expectations in a cued attention paradigm in which participants performed a perceptual discrimination task. Participants were presented circles with two gaps of different sizes at the top and bottom of each circle. The task was to decide which of the two gaps was larger. They found both motivation and task difficulty expectations to improve behavioral performance and to correlate with activity in parietal regions related to attention control. This suggests that motivation and task difficulty expectations modulate perceptual and stimulus categorization processes to improve task performance. The current study tested how motivation and task difficulty expectations modulate activity in posterior stimulus-specific brain regions in the context of an interference paradigm. Contrary to the paradigm used by Krebs et al., interference paradigms require resolving conflicts between task-relevant and distractor information. Importantly, such conflicts can be resolved by two different processing strategies: Attention control can either strengthen the focus on the target stimulus to improve task-relevant processing, or it can inhibit the processing of the distractor information (Soutschek, Taylor, Muller, & Schubert, 2013; Polk, Drake, Jonides, Smith, & Smith, 2008; Egner & Hirsch, 2005). Thus, although the study of Krebs et al. suggests that motivation and task difficulty expectations can improve conflict processing by modulating activity in posterior brain regions related to attention processes, the question remains as to whether motivation and task difficulty expectations affect the processing of task-relevant and/or task-irrelevant information. Therefore, the purpose of the current study was to compare the effects of motivation and conflict expectancy on the fMRI activity in posterior brain regions that are associated with the processing of task-relevant stimulus information or with the processing of the distractor stimulus information. In particular, we intended to assess whether motivation and conflict expectancy modulate task-relevant stimulus processing or, alternatively, distractor processing in the related posterior brain regions. To address these issues, we applied a variant of the picture–word interference paradigm that allowed disentangling the neural correlates of the task-relevant and distractor processing: In particular, participants were instructed to respond to the gender of task-relevant face stimuli while ignoring the distractor words written across the face stimuli (Soutschek & Schubert, 2013). This task requires the resolution of conflicts between the task-relevant face and the distractor words on incongruent trials in which the gender of the face and the word do not match (e.g., a female face overlaid with the word “man”). Importantly, face processing has been related to activity in the fusiform face area (FFA), whereas the processing of visually presented words correlates with visual word form area (VWFA) activity (Stelzel, Brandt, & Schubert, 2009; Egner & Hirsch, 2005; Dehaene, Le Clec, Poline, Le Bihan, & Cohen, 2002; Kanwisher, McDermott, & Chun, 1997). In the context of the picture–word interference paradigm, FFA and VWFA activity are related to task-relevant and distractor processing, respectively. Consequently, the current task design allowed us to assess (and compare) the impacts of motivation and conflict expectancy on target- and distractor-related activity by conducting ROI analyses in the FFA and the VWFA.

In addition to our main question, we also investigated the effects of motivation and conflict expectancy on brain regions related to cognitive control. In the context of interference tasks, the dACC and the lateral pFC have often been associated with cognitive control and conflict processing (Egner & Hirsch, 2005; Kerns et al., 2004; Botvinick, Braver, Barch, Carter, & Cohen, 2001). Therefore, a further goal of the current study was to investigate whether motivation and conflict expectancy have similar or dissociable effects on these neural correlates of cognitive control.

METHODS

Participants

Twenty healthy right-handed volunteers (13 women, mean age = 25.68 years, SD = 3.83 years, range = 21–32 years) with normal or corrected-to-normal vision participated in the experiment after obtaining informed consent according to the declaration of Helsinki. They received €10/hr and additionally performance-dependent monetary rewards. Three further volunteers had been removed from the data set because of excessive head movement.

Main Experiment

Participants performed a Stroop-like picture–word interference task in a mixed blocked and event-related fMRI design. We presented participants pictures of female and male faces with neutral emotional expression from the FACES database (Ebner, Riediger, & Lindenberger, 2010). The face stimuli were overlaid by the words “FRAU” or “MANN” (German for “woman” and “man,” respectively) in red capital letters, resulting in congruent (e.g., “FRAU” across a female face) and incongruent (e.g., “MANN” across a female face) picture–word combinations. Participants were instructed to ignore the distractor words and to respond only to the gender of the presented faces by pressing a left (for male faces) or right response key (for female faces) with the index and middle finger of their right hand, respectively.

On each trial, a stimulus for the picture–word interference task was presented for 500 msec on the center of the screen, followed by the presentation of a fixation cross for 1500 msec in which the response had to be executed (Figure 1). At the start of each block, we presented two instruction cues for 4000 msec, informing participants whether they could win or lose money in the following block (motivation cue) and how many trials of the block would be congruent (conflict expectancy cue). In more detail, a motivation cue of “0 cent” (low motivation) indicated that participants could not win a reward in the following block, whereas a cue of “50 cent” (high motivation) indicated the chance to either win or lose 50 cents, depending on the performance (RTs and errors) in the block. The conflict expectancy cues of “70%” (low conflict expectancy) and “30%” (high conflict expectancy) indicated whether either 70% or 30%, respectively, of the trials in the block would be congruent (note that “low conflict expectancy” means that participants expected a low number of incongruent trials because the cue indicated that more congruent than incongruent trials would be presented, vice versa for “high conflict expectancy”). We instructed participants to use the cues to adjust their performance. Following the instruction cues, 10 trials of the picture–word interference task were presented. Trials were separated by pseudorandomly varying intertrial intervals (2, 4, or 6 sec) in which a fixation cross was presented on the screen center. After a block of 10 trials, a feedback informed participants concerning their mean RT, their number of errors in the block, and whether they had lost or won money. In high motivation blocks, participants gained 50 cents only if both the mean RT and the number of errors were below specific thresholds, which were dynamically adjusted after each block, depending on the individual participant's performance. In more detail, the start values (i.e., the thresholds for the first block) were set to 550 msec and two errors for RT and error threshold, respectively. After each block, RT and error thresholds were updated by computing the mean of the previous thresholds and the mean RTs or number of errors in the current block using the formula RT threshold block n + 1 = (RT threshold block n + mean RT block n)/2.

Figure 1. 

Schematic illustration of a block of the main experimental task. Participants were instructed to respond to the gender of the presented face stimulus. At the start of a block, instruction cues displayed the proportion of congruent trials in the current block (“70%” vs. “30%”; conflict expectancy cue) and whether or not participants would receive performance-dependent monetary gains/losses after the block (“0 cent” vs. “50 cent”; motivation cue). Ten trials (trial duration: 2 sec) of the picture–word interference task were presented in each block, which were separated by intertrial intervals (ITIs; 2 sec, 4 sec, or 6 sec) during which a white fixation cross presented on the center of the screen. At the end of the block, participants received a short feedback indicating whether they had lost or won money. Blocks were separated by intervals of 10 sec in which a fixation cross was displayed.

Figure 1. 

Schematic illustration of a block of the main experimental task. Participants were instructed to respond to the gender of the presented face stimulus. At the start of a block, instruction cues displayed the proportion of congruent trials in the current block (“70%” vs. “30%”; conflict expectancy cue) and whether or not participants would receive performance-dependent monetary gains/losses after the block (“0 cent” vs. “50 cent”; motivation cue). Ten trials (trial duration: 2 sec) of the picture–word interference task were presented in each block, which were separated by intertrial intervals (ITIs; 2 sec, 4 sec, or 6 sec) during which a white fixation cross presented on the center of the screen. At the end of the block, participants received a short feedback indicating whether they had lost or won money. Blocks were separated by intervals of 10 sec in which a fixation cross was displayed.

Scales

Participants filled out a questionnaire including the Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) scales (Carver & White, 1994; German version by Strobel, Beauducel, Debener, & Brocke, 2001), which are thought to measure individual attitudes toward reward and punishment. Because in the current study we did not differentiate between the effects of potential rewards (activation system) and punishments (inhibition system), a general reward sensitivity score was determined by summing up the individual scores of the BIS/BAS subscales.

Localizer Experiment

Before the main experiment, participants performed a localizer task to determine the individual ROIs in the FFA and the VWFA. The FFA and the VWFA have consistently been found to be activated in tasks requiring face and word recognition, respectively (Stelzel et al., 2009; Dehaene et al., 2002; Kanwisher et al., 1997). ROI analyses in the FFA and the VWFA allowed us to examine the effects of motivation and conflict expectancy on task-relevant and on distractor processing, respectively, in the picture–word interference task of the main experiment. To localize the FFA, participants performed a face recognition task in which, similar to the main experimental task, they had to respond to the gender of presented female and male faces by pressing a right or left response key, respectively (face task). In contrast to the main experimental task, however, the faces were presented without distractor words. To localize the VWFA, participants were instructed to categorize the gender of German female and male first names, which were presented in white capital letters (word task). In both tasks, a stimulus for the face or the word task was displayed for 1000 msec. Responses were to be executed during stimulus presentation or the subsequent presentation of a fixation cross for 1000 msec. Each block included 10 trials, six blocks of each task were presented in alternating order, separated by baseline trials of 10 sec in which a fixation cross was presented.

fMRI Procedure

All images were acquired with a 3-T Siemens Tim Trio MRI scanner equipped with a fast gradient system for EPI at the Berlin Center for Advanced Neuroimaging (Berlin, Germany). We used a 12-channel head coil and stabilized participants with cushions to restrict head motion comfortably. Functional images were acquired in five runs (one run for the localizer task and four runs for the main task), using a whole-brain one-shot gradient-echo, echoplanar sequence (echo time = 25 msec, matrix size = 64 × 64, field of view = 24 cm, flip angle = 78°, repetition time = 2 sec). Each functional volume consisted of 32 axial slices with 3 mm thickness and 0.75 mm interslice gap. We also acquired a structural T1-weighted 3-D MP-RAGE scan at the end of the experiment (matrix size = 256 × 256, slice thickness = 1.0 mm, flip angle = 9°). In addition, we acquired field maps using the same slice prescription as for functional scans. Anatomical images were used for the normalization of the functional data to the Montreal Neurological Institute (MNI) atlas space.

fMRI Data Analyses

All analyses were performed with SPM 8 (www.fil.ion.ucl.ac.uk/spm). For the analysis of the main experimental task, the functional data set of each participant was first motion corrected, unwarped, slice-timing corrected, and then coregistered to the anatomical image. Following segmentation, we spatially normalized the data into standard MNI space. Finally, data were smoothed with an 8-mm FWHM Gaussian kernel and high-pass filtered during statistical analysis. Temporal correlations were calculated by applying the general linear model for serially autocorrelated data. For the analysis of the localizer task, the same procedure was adopted, however, without the slice-timing correction.

We analyzed the localizer task as a block-related design with two covariates for the face and word tasks The main experimental task was analyzed with an event-related design including the factors Motivation, Conflict expectancy, and Congruency (instead of analyzing the motivation- and conflict expectancy-related activity in a block-related design). Importantly, this procedure allowed us to examine interactions between the factors Motivation, Conflict expectancy, and Congruency. In addition, a block-related effect of conflict expectancy would be confounded with the effects of congruency because the ratio of congruent and incongruent trials varied between low conflict expectancy and high conflict expectancy blocks, whereas an event-related design allows disentangling the effects of conflict expectancy and congruency. The design included eight covariates to model each combination of the factors Motivation (low vs. high), Conflict expectancy (low vs. high), and Congruency (congruent vs. incongruent). In addition, we included regressors for error trials, instruction cues, feedback, and the fixation periods between the blocks as well as the movement parameters as covariates of no interest.

We then estimated regression parameters and temporal derivatives in every voxel for each participant and entered the parameter estimates in a between-subject, random effects analysis to obtain statistical parametric maps for one-sample t test. To protect the whole-brain analysis against false-positive activations, we identified brain activations showing significant contrasts of parameter estimates by a double-threshold approach, that is, by combining a voxel-based threshold with a minimum cluster size (Forman et al., 1995). This nonarbitrary voxel cluster size was determined on the basis of a Monte Carlo simulation (10,000 iterations) determined with AFNI's AlphaSim tool (Ward, 2000) as implemented in the REST matlab toolbox. To ensure an overall image-wise false-positive rate of 5%, we determined the minimal cluster size for an individual voxel height threshold of t > 3.61 (p < .001, uncorrected). This resulted in a cluster threshold of 96 voxels. Activations exceeding this double threshold are therefore considered to be activated at an experiment-wise threshold of p < .05, corrected for multiple comparisons.

ROI Analyses

To test our hypothesis regarding the effects of motivation and conflict expectancy on task-relevant and irrelevant processing, we determined the FFA and the VWFA in the localizer experiment in each participant individually. Using MarsBaR software (marsbar.sourceforge.net/), we first defined the ROIs for the right FFA and the left VWFA by determining the peak voxels of the contrasts (face task > 0) and (word task > 0) in the fusiform/parahippocampal gyri, respectively. Then, we applied spherical masks with a diameter of 8 mm centered at the individual peak voxels. Using these masks, the percent signal changes were extracted from the FFA and VWFA ROIs of the individual participants in the main experiment for the combinations of the factors Motivation, Conflict expectancy, and Congruency. The extracted data were statistically analyzed by repeated-measures ANOVAs and paired-samples t tests.

In addition to the FFA and VWFA analyses, we conducted further ROI analyses to examine the impact of motivation and conflict expectancy on activity related to conflict processing. These ROIs were defined as spherical masks (diameter = 8 mm) at prefrontal local peak activity maxima derived from the main effect of Congruency (incongruent > congruent) in the whole-brain analysis (for details, see below). As for the FFA and VWFA analyses, we then estimated percent signal changes for the combinations of the factors Motivation, Conflict expectancy, and Congruency. These ROI-based values were used to illustrate the effects of motivation and conflict expectancy on conflict processing in these ROIs without performing any additional statistics with regard to the contrast that was used to determine to ROIs (Kriegeskorte, Simmons, Bellgowan, & Baker, 2009). In addition, we also tested the effects of motivation, conflict expectancy, and their interactions with congruency on signal changes in these ROIs. It is important to emphasize that the statistical tests we computed for the ROIs were independent of the selection criterion for the ROIs (i.e., incongruent > congruent).

RESULTS

Behavioral Data

Mean RT and Error Data

For the RT analysis, all error trials were excluded from the data set. A 2 (Motivation) × 2 (Conflict expectancy) × 2 (Congruency) repeated-measures ANOVA revealed a significant main effect of Motivation, F(1, 19) = 39.07, p < .001, ηp2 = 0.673, indicating that mean RTs were faster in high motivation (507 msec) than in low motivation (573 msec) blocks. The main effect of Congruency, F(1, 19) = 66.10, p < .001, ηp2 = 0.777, pointed to the occurrence of a congruency effect, with slower RTs in incongruent (556 msec) than in congruent trials (525 msec). Importantly, we found the magnitude of the congruency effect to be reduced in high motivation (26 msec) compared with low motivation (34 msec) blocks, Motivation × Congruency, F(1, 19) = 4.84, p < .05, ηp2 = 0.203. In addition, the congruency effect was also smaller in high conflict expectancy (23 msec) compared with low conflict expectancy (37 msec) blocks, Conflict expectancy × Congruency, F(1, 19) = 4.49, p < .05, ηp2 = 0.191. Because the three-way Motivation × Conflict expectancy × Congruency interaction did not reach statistical significance, F < 1, p > .72, the effects of Motivation and Conflict expectancy on the congruency effect were, on a behavioral level, independent of each other (Figure 2).

Figure 2. 

Mean RTs (in msec) as a function of the factors Motivation (low motivation vs. high motivation), Conflict expectancy (low conflict expectancy vs. high conflict expectancy), and Congruency (congruent vs. incongruent). Error bars depict the standard errors of the mean.

Figure 2. 

Mean RTs (in msec) as a function of the factors Motivation (low motivation vs. high motivation), Conflict expectancy (low conflict expectancy vs. high conflict expectancy), and Congruency (congruent vs. incongruent). Error bars depict the standard errors of the mean.

A 2 (Motivation) × 2 (Conflict expectancy) × 2 (Congruency) repeated-measures ANOVA on error rates revealed a significant main effect of Conflict expectancy, F(1, 19) = 7.27, p < .05, ηp2 = 0.277, reflecting the fact that participants committed less errors in low conflict expectancy (4.9%) compared with high conflict expectancy (5.5%) blocks. The significant effect of Congruency, F(1, 19) = 31.38, p < .001, ηp2 = 0.623, pointed to the occurrence of a congruency effect with higher error rates in incongruent (7.3%) compared with congruent trials (3.0%). This error congruency effect was reduced in high conflict expectancy (3.6%) compared with in low conflict expectancy (6.2%) blocks, Conflict expectancy × Congruency, F(1, 19) = 6.76, p < .05, ηp2 = 0.262 (Table 1). No further effect passed the statistical threshold, Fs < 1, ps > .36.

Table 1. 

Mean RTs (in msec) and Error Rates (in %) as a Function of the Factors Motivation (Low Motivation vs. High Motivation), Conflict Expectancy (Low Expectancy vs. High Expectancy), and Congruency (Congruent vs. Incongruent)

Low MotivationHigh Motivation
Low Conflict ExpectancyHigh Conflict ExpectancyLow Conflict ExpectancyHigh Conflict Expectancy
Congruent 
RTs 552 (13) 564 (15) 492 (10) 499 (11) 
Errors 3.4 (0.8) 3.6 (0.9) 2.7 (0.5) 2.4 (0.5) 
 
Incongruent 
RTs 594 (18) 590 (15) 524 (13) 519 (11) 
Error 9.7 (1.4) 6.9 (1.5) 8.7 (1.4) 6.3 (1.5) 
Low MotivationHigh Motivation
Low Conflict ExpectancyHigh Conflict ExpectancyLow Conflict ExpectancyHigh Conflict Expectancy
Congruent 
RTs 552 (13) 564 (15) 492 (10) 499 (11) 
Errors 3.4 (0.8) 3.6 (0.9) 2.7 (0.5) 2.4 (0.5) 
 
Incongruent 
RTs 594 (18) 590 (15) 524 (13) 519 (11) 
Error 9.7 (1.4) 6.9 (1.5) 8.7 (1.4) 6.3 (1.5) 

Standard errors of the mean are in brackets.

Distributional Analysis

Note that the finding of a decreased congruency effect in the high motivation compared with the low motivation condition may be confounded with the faster response speed in the high motivation condition. Previous studies showed the magnitude of the Stroop effect to decrease with faster responses (Pratte, Rouder, Morey, & Feng, 2010). Thus, the differences in conflict processing between the high motivation and low motivation conditions may be, in theory, a consequence of the different RT levels in these conditions and not of motivational effects on cognitive control. To provide more conclusive support for the assumption that motivation improves conflict processing by enhancing the level of cognitive control independently on the response speed, we compared the congruency effects in the motivation conditions by means of a distributional analysis. Distributional analyses graph the size of the congruency effect for different RT levels (Soutschek, Schwarzkopf, et al., 2013; Ridderinkhof, 2002), allowing us to assess the impact of motivation on conflict processing at comparable RT levels in the low motivation and the high motivation condition.

The distributional analysis was conducted as follows: We rank-ordered all RTs of each participant separately for the different levels of the factors Motivation and Congruency and divided them into four bins of equal size (quartiles). Next, we calculated first the mean RT ((RTcongruent + RTincongruent)/2) and then the congruency effect (RTincongruent − RTcongruent) for each quartile, separately for low motivation and high motivation trials. The congruency effect in the first quartile of low motivation blocks, for example, was calculated by subtracting the mean RT in the first quartile for congruent trials from the mean RT in the first quartile for incongruent trials in the low motivation condition. The results of this distributional analysis are illustrated in Table 2, in which congruency effects for every quartile are depicted as a function of mean response speed in the corresponding quartile. In a similar manner, we also conducted a distributional analysis for the error congruency effect in which, as for the RT analysis, the congruency effect in the error rates is calculated separately for the different RT quartiles. First, we again subdivided the RTs in both correct and incorrect response trials into quartiles, separately for all levels of the factors Motivation and Congruency. Then, we calculated the error congruency effect in each quartile by subtracting the error rate in congruent trials from the error rate in incongruent trials in the corresponding quartiles, separately for low motivation and high motivation blocks. Finally, error congruency effects were plotted against the mean response speed in the corresponding quartile (Table 2).

Table 2. 

Distributional Analyses of RT and Error Congruency Effects

QuartileLow MotivationHigh Motivation
1st2nd3rd4th1st2nd3rd4th
Mean RT 451 (12) 524 (11) 587 (15) 723 (26) 412 (10) 472 (10) 520 (11) 619 (16) 
RT congruency effect 16 (3) 27 (5) 37 (5) 59 (9) 12 (3) 22 (3) 27 (4) 43 (6) 
Error congruency effect 6.9 (1.3) 5.0 (1.4) 3.4 (1.1) 2.4 (1.6) 6.6 (1.8) 5.5 (2.0) 1.9 (1.0) 3.3 (1.1) 
QuartileLow MotivationHigh Motivation
1st2nd3rd4th1st2nd3rd4th
Mean RT 451 (12) 524 (11) 587 (15) 723 (26) 412 (10) 472 (10) 520 (11) 619 (16) 
RT congruency effect 16 (3) 27 (5) 37 (5) 59 (9) 12 (3) 22 (3) 27 (4) 43 (6) 
Error congruency effect 6.9 (1.3) 5.0 (1.4) 3.4 (1.1) 2.4 (1.6) 6.6 (1.8) 5.5 (2.0) 1.9 (1.0) 3.3 (1.1) 

Congruency effects and mean RTs are reported as a function of the factors Motivation (low motivation vs. high motivation) and Quartile. Standard errors of the mean are in brackets.

As suggested by Table 2, mean RTs in the second quartile of low motivation blocks are comparable to mean RTs in the third quartile of high motivation blocks. In line with this observation, a paired-samples t test showed no significant differences between mean RTs in the second quartile of low motivation blocks (524 msec) and mean RTs in the third quartile of high motivation blocks (520 msec), p = .62, whereas mean RTs between all other quartiles differed from each other, p < .05. Although the RT congruency effects did not differ between the second quartile of low motivation blocks and the third quartile of high motivation blocks, t(19) < 1, the congruency effect in error rates was significantly reduced in the third quartile of high motivation blocks (1.9%) compared with the second quartile of low motivation blocks (5.0%), t(19) = 2.28, p < .05. Thus, the distributional analysis of error congruency effects shows that motivation improves conflict processing even when controlling for motivational effects on response speed, whereby a facilitating effect of motivation on conflict processing was evident only in error rates, not in RTs (see also Veling & Aarts, 2010).

fMRI Data

ROI Analyses

FFA

The main goal of the current study was to test how motivation and conflict expectancy modulate activity in posterior brain regions related to task-relevant and task-irrelevant information processing. For that purpose, we conducted ROI analyses for the FFA and the VWFA as neural correlates of task-relevant and distractor processing, respectively, in the picture–word interference paradigm. We analyzed BOLD signal changes in the FFA with a 2 (Motivation) × 2 (Conflict expectancy) × 2 (Congruency) repeated-measures ANOVA (for details, see Table 3). In the FFA, the BOLD signal was significantly increased on high motivation relative to low motivation trials, F(1, 19) = 4.54, p < .05, ηp2 = 0.193, suggesting that high motivation enhanced the processing of the task-relevant face stimulus (Figure 3A). The factors Conflict expectancy and Congruency did not show significant effects, both Fs(1, 19) < 1.65, ps > .21; thus, conflict expectancy and congruency had no main effects on target-related activity in the FFA. However, the significant Motivation × Conflict expectancy × Congruency interaction, F(1, 19) = 7.88, p < .05, ηp2 = 0.293, reflects the fact that activity differences between incongruent and congruent trials in the FFA were modulated by an interaction of the factors motivation and conflict expectancy (Figure 3B): When conflict expectancy was low, we found no evidence for differences in FFA activity between incongruent and congruent trials, independent of the level of motivation, ts < 1, ps > .34. However, we found significantly increased FFA activity on incongruent relative to congruent trials when participants could anticipate the occurrence of conflicts (high conflict expectancy) and, in addition, the level of motivation was low, t(19) = 2.20, p < .05. In contrast, there were no activity differences between incongruent and congruent trials under conditions of both high conflict expectancy and high motivation, t(19) = 1.11, p > .28. Taken together, this shows that high conflict expectancy leads to increased FFA activity during the processing of incongruent relative to congruent trials only in low motivation, but not in high motivation, trials.

Table 3. 

Signal Changes (in %) of FFA and VWFA Activity as a Function of the Factors Motivation (Low Motivation vs. High Motivation), Conflict Expectancy (Low Expectancy vs. High Expectancy), and Congruency (Congruent vs. Incongruent)

Low MotivationHigh Motivation
Low Conflict ExpectancyHigh Conflict ExpectancyLow Conflict ExpectancyHigh Conflict Expectancy
FFA 
Congruent 0.52 (0.05) 0.51 (0.06) 0.54 (0.06) 0.60 (0.05) 
Incongruent 0.52 (0.05) 0.56 (0.05) 0.57 (0.05) 0.58 (0.06) 
 
VWFA 
Congruent 0.38 (0.05) 0.41 (0.05) 0.36 (0.05) 0.40 (0.05) 
Incongruent 0.43 (0.05) 0.40 (0.05) 0.44 (0.05) 0.38 (0.06) 
Low MotivationHigh Motivation
Low Conflict ExpectancyHigh Conflict ExpectancyLow Conflict ExpectancyHigh Conflict Expectancy
FFA 
Congruent 0.52 (0.05) 0.51 (0.06) 0.54 (0.06) 0.60 (0.05) 
Incongruent 0.52 (0.05) 0.56 (0.05) 0.57 (0.05) 0.58 (0.06) 
 
VWFA 
Congruent 0.38 (0.05) 0.41 (0.05) 0.36 (0.05) 0.40 (0.05) 
Incongruent 0.43 (0.05) 0.40 (0.05) 0.44 (0.05) 0.38 (0.06) 

Standard errors of the mean are in brackets.

Figure 3. 

Signal changes (in %) for the FFA and the VWFA. In A, FFA signal changes are plotted as a function of motivation. B and D depict congruency effects in signal changes (i.e., incongruent − congruent) of FFA and VWFA activity, respectively, as function of motivation and conflict expectancy. Error bars indicate the standard error of the mean (C = congruent; I = incongruent). Individual differences in FFA (C) and VWFA (E) signal changes between high motivation and low motivation trials plotted against the BIS/BAS sum score for each participant.

Figure 3. 

Signal changes (in %) for the FFA and the VWFA. In A, FFA signal changes are plotted as a function of motivation. B and D depict congruency effects in signal changes (i.e., incongruent − congruent) of FFA and VWFA activity, respectively, as function of motivation and conflict expectancy. Error bars indicate the standard error of the mean (C = congruent; I = incongruent). Individual differences in FFA (C) and VWFA (E) signal changes between high motivation and low motivation trials plotted against the BIS/BAS sum score for each participant.

The FFA analysis reported above revealed that high motivation increased activity in brain regions related to task-relevant stimulus processing. To examine whether the modulatory effect of motivation on FFA activity depended on the individual participants' attitude toward reward incentives (measured with the BIS/BAS sum score), we calculated the contrasts between high motivation and low motivation trials (i.e., percent signal change high motivation − percent signal change low motivation) for each participant and correlated them with the individual BIS/BAS sum scores. In fact, we found a significant positive correlation between the individual attitudes toward rewards and the effects of motivation on FFA activity, r = 0.522, p < .05. This result suggests that the enhancement of target-related activity in the FFA was the more pronounced the more participants described themselves as sensitive to potential rewards or punishments (Figure 3C).

VWFA

Similar to the FFA analysis, we computed an ANOVA to examine the impact of motivation, conflict expectancy, and congruency on BOLD signal changes in the VWFA as indicator of distractor processing. This ANOVA revealed a trend toward higher BOLD responses to incongruent relative to congruent stimuli, F(1, 19) = 3.84, p < .07, ηp2 = 0.168, suggesting that distractor processing tended to be enhanced on incongruent compared with congruent trials. These activity differences between incongruent and congruent trials were modulated by the level of conflict expectancy, Conflict expectancy × Congruency, F(1, 19) = 5.21, p < .05, ηp2 = 0.215 (Figure 3D). Post hoc t tests revealed that the BOLD signal in the VWFA was increased in incongruent relative to congruent trials only when conflict expectancy was low, t(19) = 2.60, p < .05, but not when conflict expectancy was high, t(19) < 1, p > .38. Thus, although distractor processing was enhanced on incongruent (in which target face and distractor word did not match) relative to congruent trials when participants did not expect the occurrence of incongruent stimuli, high conflict expectancy suppressed the differences between incongruent and congruent trials in distractor-related VWFA activity. Neither the main effect of motivation nor any interaction effect passed the statistical threshold, Fs(1, 19) < 1, ps > .39.

Although we found no evidence for an effect of motivation on VWFA activity, we again tested, as for the FFA analysis, whether motivation-related differences in VWFA activity correlated with the individual reward sensitivity scores. In contrast to the FFA analysis, however, we found no significant correlation between the individual attitudes toward rewards and the (high motivation − low motivation) contrast in the VWFA, r = −0.014, p > .95 (Figure 3E).

Between-ROI analysis

The reported ROI analyses of FFA and VWFA signal changes suggest dissociable effects of conflict expectancy and motivation on target- and distractor-related activity: Although there was evidence for an effect of motivation only on FFA but not VWFA activity, we found a Conflict expectancy × Congruency interaction (independent of the level of motivation) only in the VWFA. To provide further support for such a potential dissociation, we conducted a between-ROI analysis by conducting an ANOVA on signal changes in the FFA and the VWFA including the factors Conflict expectancy, Motivation, and Congruency and the additional factor ROI (FFA vs. VWFA). For this analysis, we z-standardized the signal changes in the FFA and the VWFA to control for systematic differences between the hemodynamic response functions in different brain regions (Handwerker, Ollinger, & D'Esposito, 2004). The ANOVA yielded a significant Conflict expectancy × Congruency × ROI interaction, F(1, 19) = 17.63, p < .001, ηp2 = 0.481, which is consistent with the observation that a significant Conflict expectancy × Congruency interaction was evident only in the VWFA but not in the FFA. No further effect passed the statistical threshold, F(1, 19)s < 4.36, ps > .05.

Furthermore, to substantiate the finding of the single ROI analyses that high motivation increased activity only in the FFA, but not in the VWFA, we conducted a further analysis in which we tested whether the effects of motivation on FFA and VWFA activity depended on the participants' reward sensitivity. Such a relationship between the effects of motivation and individual reward sensitivity scores is suggested by the FFA analysis, which had shown that the motivational effect on FFA activity increases with high individual reward sensitivity scores. Thus, motivation may have dissociable effects on FFA and VWFA activity only in participants with high reward sensitivity. To test this, we split the distribution of BIS/BAS sum scores at the median and conducted an ANOVA with the factors Motivation and ROI and the between-subject factor Reward Sensitivity (low sensitivity vs. high sensitivity) on FFA and VWFA signal changes. The Motivation × ROI × Reward Sensitivity interaction approached significance, F(1, 19) = 4.10, p < .06, ηp2 = 0.185, supporting the assumption that the effect of motivation on FFA and VWFA activity depends on the individual reward sensitivity. In fact, paired-samples t tests showed that high motivation, relative to low motivation, increased FFA activity only in the high sensitivity, t(9) = 2.86, p < .05, not in the low sensitivity group, t < 1, whereas there was no effect of motivation on VWFA activity neither in the low nor the high sensitivity group, ts < 1. Taken together, the results of the between-ROI analyses suggest conflict expectancy and motivation to have dissociable effects on FFA and VWFA activity.

Regions Associated with Cognitive Control

Next, we tested how motivation and conflict expectancy modulated the activity in brain regions that are related to cognitive control processes during conflict processing. To identify regions associated with conflict control processes, we determined which regions showed significantly increased activity on incongruent relative to congruent trials during the main experimental task. The (incongruent > congruent) contrast revealed significant activity in the dACC and in the left middle/inferior frontal gyrus, two regions that have been related to cognitive control processes in interference tasks (for further significant clusters, see below and Table 5). Although the dACC has been hypothesized to be part of a core network in cognitive control and performance monitoring (Holroyd & Yeung, 2012; Niendam et al., 2012; Botvinick, 2007; Dosenbach et al., 2007), the middle/inferior frontal gyrus may be related to reactive control adjustments and response inhibition processes (Neubert, Mars, Buch, Olivier, & Rushworth, 2010; Kouneiher et al., 2009; Forstmann et al., 2008; Kerns et al., 2004; Miller & Cohen, 2001).

Activity in the dACC was significantly reduced in high conflict expectancy relative to low conflict expectancy trials, F(1, 19) = 4.74, p < .05, ηp2 = 0.199. We found no main effect of motivation nor Motivation × Congruency or Conflict expectancy × Congruency interactions, Fs(1, 19) < 1.15, ps > .29, on dACC activity. However, the analysis revealed a significant Motivation × Conflict expectancy × Congruency interaction, F(1, 19) = 10.08, p < .01, ηp2 = 0.347, indicating that conflict expectancy effects on activity differences between incongruent and congruent trials were modulated by motivation: In particular, we found that high conflict expectancy, relative to low conflict expectancy, reduced the activity differences between incongruent and congruent trials only when the level of motivation was high t(19) = 2.74, p < .05, but not when motivation was low, t(19) < 1.24, p > .23 (Figure 4A). This shows that high conflict expectancy reduced conflict-related dACC activity only when participants were highly motivated but not in conditions of low motivation.

Figure 4. 

Signal changes (in %) for the dACC (A) and the middle/inferior frontal gyrus (B), plotted as a function of motivation, conflict expectancy, and congruency. The ROIs for the middle/inferior frontal gyrus and the dACC were defined by the (incongruent > congruent) contrast in the whole-brain analysis. Error bars indicate the standard error of the mean.

Figure 4. 

Signal changes (in %) for the dACC (A) and the middle/inferior frontal gyrus (B), plotted as a function of motivation, conflict expectancy, and congruency. The ROIs for the middle/inferior frontal gyrus and the dACC were defined by the (incongruent > congruent) contrast in the whole-brain analysis. Error bars indicate the standard error of the mean.

In the middle/inferior frontal gyrus, we found no significant main effects of Motivation or of Conflict expectancy, F(1, 19) < 1, p > .72 on fMRI activity. However, we found a significant Conflict expectancy × Congruency interaction, F(1, 19) = 4.74, p < .05, ηp2 = 0.200. This interaction reflect the fact that the activation on incongruent trials was larger than on congruent trials only in situations in which conflict expectancy was low, t(19) = 4.36, p < .001, but not when conflict expectancy was high, t(19) = 1.39, p > .18 (Figure 4B). No further effects were significant.

An overview over the effects of motivation, conflict expectancy, and congruency on the different behavioral and neural measures is provided by Table 4.

Table 4. 

Overview over the Significant Effects of Motivation, Conflict Expectancy, and Congruency on RTs, Errors, and Signal Changes in the FFA, VWFA, dACC, and Middle/Inferior Frontal Gyrus

RTsErrorsFFAVWFAdACCM/IFG
Congruency p < .001 p < .001 ns ns   
Motivation p < .001 ns p < .05 ns ns ns 
Conflict expectancy ns p < .05 ns ns p < .05 ns 
Motivation × Congruency p < .05 ns ns ns ns ns 
Conflict expectancy × Congruency p < .05 p < .05 ns p < .05 ns p < .05 
Motivation × Conflict expectancy ns ns ns ns ns ns 
Motivation × Conflict expectancy × Congruency ns ns p < .05 ns p < .01 ns 
RTsErrorsFFAVWFAdACCM/IFG
Congruency p < .001 p < .001 ns ns   
Motivation p < .001 ns p < .05 ns ns ns 
Conflict expectancy ns p < .05 ns ns p < .05 ns 
Motivation × Congruency p < .05 ns ns ns ns ns 
Conflict expectancy × Congruency p < .05 p < .05 ns p < .05 ns p < .05 
Motivation × Conflict expectancy ns ns ns ns ns ns 
Motivation × Conflict expectancy × Congruency ns ns p < .05 ns p < .01 ns 

Note that we did not test for a congruency effect in the dACC and middle/inferior frontal gyrus ROIs because these regions were defined by the (incongruent > congruent) contrast in the whole-brain analysis. M/IFG = middle/inferior frontal gyrus.

Whole-brain Analysis

Finally, we conducted a whole-brain analysis to investigate the neural correlates of congruency, motivation, and conflict expectancy. We found an enhanced BOLD response on incongruent compared with congruent trials in dACC, left middle/inferior frontal gyrus, brainstem, and superior parietal lobule (Table 5), consistent with findings of previous studies (Kerns et al., 2004; Peterson et al., 2002). High motivation increased the BOLD signal relative to low motivation blocks in the caudate body/dorsal striatum, the left caudate tail, and the precuneus, whereas decreased activity in high motivation compared with low motivation blocks occurred in the orbitofrontal cortex, the middle temporal gyrus, and the culmen. Although we found no significant main effect of Conflict expectancy on fMRI activity, the interaction between Conflict expectancy and Congruency revealed significant activation clusters in the right precentral, left lingual gyrus, left caudate body, right fusiform gyrus, and the left brainstem. In these clusters, the increase of activity in the incongruent compared with the congruent trials (i.e., congruency effect) was reduced in high relative to low conflict expectancy blocks. In addition, we found several clusters in which motivation modulated significantly the interaction between congruency and conflict expectancy, that is, in the left superior frontal gyrus, in the left orbitofrontal cortex, middle and inferior temporal gyrus, occipital lobe, and cerebellum. In these regions, the observed reduction of activity differences between incongruent and congruent stimuli in high conflict expectancy compared with low expectancy trials was more pronounced in high motivation than in low motivation conditions.

Table 5. 

Anatomical Location and MNI Coordinates of the Peak Activations for the Main Effects of Congruency (Incongruent > Congruent) and Motivation (High Motivation > Low Motivation; Low Motivation > High Motivation) as well as the Conflict Expectancy × Congruency and the Motivation × Conflict Expectancy × Congruency Interactions

RegionHemBAMNI CoordinatesCluster Sizet
xyz
Incongruent > Congruent 
Anterior cingulate cortex 32 −9 14 40 173 7.26 
Middle/inferior frontal gyrus 6/9 −42 11 28 114 5.65 
Brainstem −6 −19 −8 205 5.47 
Parietal cortex 40/7 −30 −55 49 151 5.61 
 
High Motivation > Low Motivation 
Caudate body 32 −18 23 13 423 7.76 
32 24 41 603 6.44 
Caudate tail/precuneus 31/19 −21 −34 19 177 5.08 
 
Low Motivation > High Motivation 
Orbitofrontal cortex L/R 32 38 −14 601 8.22 
Middle temporal gyrus 39 −45 −76 28 318 6.95 
19 57 −64 19 269 6.64 
Culmen −9 −55 188 5.36 
 
Conflict Expectancy × Congruency 
Precentral gyrus 4/6 42 −22 64 318 5.48 
Caudate body −6 −1 13 156 5.53 
Brainstem −3 −22 173 6.18 
lingual gyrus 18 −9 −91 −11 175 6.11 
Fusiform gyrus 19 30 −79 −8 172 5.36 
 
Motivation × Conflict Expectancy × Congruency 
Middle/superior frontal gyrus −24 29 49 274 5.89 
Orbitofrontal cortex 10 −6 62 −8 131 5.98 
Middle temporal gyrus 21 54 −17 129 4.76 
21/22 −63 −43 −8 281 5.24 
Inferior temporal gyrus 19/39 54 −70 375 5.70 
cerebellum/occipital lobe 19 −39 −79 37 739 6.36 
RegionHemBAMNI CoordinatesCluster Sizet
xyz
Incongruent > Congruent 
Anterior cingulate cortex 32 −9 14 40 173 7.26 
Middle/inferior frontal gyrus 6/9 −42 11 28 114 5.65 
Brainstem −6 −19 −8 205 5.47 
Parietal cortex 40/7 −30 −55 49 151 5.61 
 
High Motivation > Low Motivation 
Caudate body 32 −18 23 13 423 7.76 
32 24 41 603 6.44 
Caudate tail/precuneus 31/19 −21 −34 19 177 5.08 
 
Low Motivation > High Motivation 
Orbitofrontal cortex L/R 32 38 −14 601 8.22 
Middle temporal gyrus 39 −45 −76 28 318 6.95 
19 57 −64 19 269 6.64 
Culmen −9 −55 188 5.36 
 
Conflict Expectancy × Congruency 
Precentral gyrus 4/6 42 −22 64 318 5.48 
Caudate body −6 −1 13 156 5.53 
Brainstem −3 −22 173 6.18 
lingual gyrus 18 −9 −91 −11 175 6.11 
Fusiform gyrus 19 30 −79 −8 172 5.36 
 
Motivation × Conflict Expectancy × Congruency 
Middle/superior frontal gyrus −24 29 49 274 5.89 
Orbitofrontal cortex 10 −6 62 −8 131 5.98 
Middle temporal gyrus 21 54 −17 129 4.76 
21/22 −63 −43 −8 281 5.24 
Inferior temporal gyrus 19/39 54 −70 375 5.70 
cerebellum/occipital lobe 19 −39 −79 37 739 6.36 

All further effects did not show significant activations.

Hem = Hemisphere; L = left; R = right; BA = Brodmann's area.

DISCUSSION

Impact of Motivation and Expectancy on Task-relevant and Distractor Processing

The purpose of the current study was to examine the effects of motivation and conflict expectancy on the processing of task-relevant and distracting stimulus information in interference tasks. Previous studies suggest that both motivation and conflict expectancy trigger cognitive control processes that facilitate conflict resolution (Wilk et al., 2012; Krebs et al., 2011; Padmala & Pessoa, 2011; Carter et al., 2000). Our behavioral results support the assumption that motivation and conflict expectancy improve conflict resolution because congruency effects were reduced in the high motivation relative to the low motivation condition as well as in the high conflict expectancy relative to the low conflict expectancy condition. The current imaging data provide important insights into how motivation and conflict expectancy improve conflict processing. This is because the ROI analyses in FFA and the VWFA allowed us to examine the effects of motivation and conflict expectancy on the activity in posterior brain regions that are related to the processing of the task-relevant target and task-irrelevant distractor information.

The results of the current study suggest dissociable effects of motivation and conflict expectancy on activity in posterior brain regions related to target and distractor processing: The increased FFA signal changes in high motivation compared with low motivation trials indicates that participants focused more attention on the face information when they could receive performance-dependent monetary rewards in a block compared with blocks in which they could not win or lose money. In other words, the present findings show that motivation facilitates task-relevant stimulus processing, consistent with existing evidence (Pessoa & Engelmann, 2010; Engelmann, Damaraju, Padmala, & Pessoa, 2009; Kobayashi, Lauwereyns, Koizumi, Sakagami, & Hikosaka, 2002). In addition, we found that motivation modulated the impact of conflict expectancy on conflict processing in the FFA because high conflict expectancy increased FFA activity on incongruent relative to congruent trials only in the low, but not high, motivation condition. This finding suggests high conflict expectancy improves task-relevant processing only if task-relevant processing is not already facilitated (at its most) by a high level of motivation. Interestingly, the effect of motivation on FFA activity correlated significantly with participants' reward sensitivity scores, indicating that individual differences in the attitude toward reward and punishment allow predicting the facilitatory effect of motivation on task-relevant processing: The more the potential reward was considered as attractive, the more attention was directed on task-relevant information. However, the data provided no evidence for an impact of motivation on distractor processing in the VWFA or for a correlation between individual reward sensitivity and motivational effects in the VWFA. Thus, enhanced motivation affects only the processing of task-relevant but not of distractor information in the current study. Note that one previous study found an effect of motivation on distractor processing (Padmala & Pessoa, 2011); in contrast to our experiment, motivation was manipulated trial-wise instead of block-wise in that study, which may have led to a different strategy to obtain the reward. Importantly, however, this shows that our data do not allow concluding that motivation modulates task-relevant and not distractor processing in all task contexts. Further studies are needed to clarify under which conditions motivation modulates task-relevant or distractor processing.

In contrast to motivation, conflict expectancy modulated distractor-related neural activity because we found higher VWFA signal changes on incongruent relative to congruent trials only when conflict expectancy was low, but not when conflict expectancy was high. The observed activity differences between incongruent and congruent trials in low conflict expectancy conditions may reflect the fact that participants oriented the attention to the distractor word in incongruent trials when they (i.e., the participants) did not expect the occurrence of conflict trials. The data suggest that high conflict expectancy, instead of inhibiting distractor processing per se (which would result in reduced VWFA activity in high conflict expectancy compared with low conflict expectancy trials), specifically modulates this orientation of attention to the distracting information in incongruent relative to congruent trials. However, our findings indicate also an effect of conflict expectancy on target-specific activity because FFA activity was enhanced on incongruent relative to congruent trials only in the low motivation–high conflict expectancy condition. The neural data suggest that, when participants expect a high proportion of incongruent trials in a block (and when their level of motivation is low), they may not only suppress the processing of distractor information but also focus more attention on the task-relevant information on incongruent trials. Taken together, the ROI analyses suggest that motivation and conflict expectancy improve conflict resolution by dissociable strategies: Motivation enhanced selectively target-related, but not distractor-related, neural activation in high reward sensitivity individuals, whereas conflict expectancy modulated, in addition, also distractor-related activity in posterior stimulus-specific brain regions.

Interestingly, a previous study investigating the impacts of conflict-triggered control on activity in brain regions related to task-relevant and irrelevant processing in an experimental paradigm similar to the picture–word interference task we used reported an effect of conflict-triggered control only on task-relevant but not on distractor processing (Egner & Hirsch, 2005; but see Polk et al., 2008). In contrast to the current study, which investigated conflict expectancy effects on proactive control processes, the study of Egner and Hirsch examined reactive adjustments of control. Whereas the term “proactive control” refers to a sustained and anticipatory activation of cognitive control processes, “reactive control” refers to the transient, stimulus-driven updating of control efforts (Braver, 2012). In their study, Egner and Hirsch examined control processes that are activated in response to the experience of conflicts in the preceding trial; these conflict-driven adjustments of control are supposed to be independent of proactive conflict expectations (Jimenez & Mendez, 2013; Duthoo & Notebaert, 2012). Thus, proactive and reactive conflict-triggered control processes may facilitate conflict resolution in different ways (Funes et al., 2010): Whereas reactive control adjustments appear to enhance attention toward task-relevant stimulus information, proactive control processes triggered by conflict expectancy may mainly suppress distractor processing. It is important to note, however, that the manipulation of conflict expectancy in the current experiment may be confounded with stimulus-driven effects. Because of contingency learning, participants might be able to associate incongruent stimuli in high conflict expectancy blocks with attentional filters that allow inhibiting distractor processing (Bugg & Crump, 2012). Note, however, that previous studies provided clear evidence for an effect of conflict expectancy on interference processing even when controlling for stimulus contingencies (Soutschek, Strobach, et al., 2013; Bugg, Jacoby, & Toth, 2008). In addition, in the current experiment, the number of 10 trials per block may have been too small for contingency learning. Nevertheless, the possibility of contingency learning effects cannot be excluded completely. Thus, we conclude that the current data show that motivation and conflict frequency activate different control strategies, but the effects of conflict frequency may—besides conflict expectancy—partly depend also on stimulus contingencies.

Impact of Motivation and Conflict Expectancy on Conflict-related Regions

A further goal of the current study was to examine the effects of motivation and conflict expectancy on brain regions associated with conflict processing. These regions were defined by the (incongruent > congruent) contrast. In line with previous studies, we found the dACC and the middle frontal gyrus to be associated with conflict processing: The dACC is thought to be part of a core network of cognitive control involved in task set implementation and performance monitoring (Holroyd & Yeung, 2012; Alexander & Brown, 2011; Dosenbach et al., 2007; Botvinick et al., 2001), whereas the middle frontal gyrus has been related to reactive control processes (Kouneiher et al., 2009; Dosenbach et al., 2007; Kerns et al., 2004). In the middle frontal gyrus, we found no evidence for an effect of motivation; however, conflict-related activity was reduced in high relative to low conflict expectancy trials. Thus, high conflict expectancy appears to allow participants to prepare for upcoming conflicts and to resolve them more efficiently, in line with the assumption that conflict expectancy leads to a switch from a reactive to a proactive control strategy (Krug & Carter, 2012; Wilk et al., 2012). This interpretation is in line with previous studies ascribing the middle frontal gyrus a role in conflict resolution (Egner & Hirsch, 2005; Kerns et al., 2004).

A more complex result pattern was found in the dACC: dACC activity was reduced in high conflict expectancy relative to low conflict expectancy trials, and in addition, the Motivation × Conflict expectancy × Congruency interaction indicated that high conflict expectancy suppressed the activity differences between incongruent and congruent trials only under conditions of high motivation. One possible interpretation of this result pattern can be formulated in terms of the conflict monitoring account of Botvinick et al. (2001), which ascribes the dACC a functional role in performance monitoring and conflict detection. Conflict detection processes (indicated by the higher dACC activity on incongruent than congruent trials) may be engaged more strongly when participants are in a highly motivated state and, simultaneously, rather unprepared to the occurrence of conflicts. In contrast, when participants expect mainly incongruent trials in a block, then less performance monitoring may be required on incongruent trials because the cognitive system is already well prepared to the occurrence of conflicts. Interactive effects of motivation and conflict expectancy on dACC activity have already been found by Krebs et al. (2012) and are supportive for the assumption that the dACC plays a crucial role in integrating motivational, emotional, and cognitive information in higher-level control processes (Shenhav, Botvinick, & Cohen, 2013; Holroyd & Yeung, 2012; Alexander & Brown, 2011; Botvinick, 2007).

In addition, our results replicated findings of previous studies showing effects of motivation and conflict expectancy on the orbitofrontal cortex, the superior frontal gyrus, and the striatum. The orbitofrontal cortex has been related to a default mode network (Raichle et al., 2001), but also to value judgment and selection of higher-order goals during task performance (Kim, 2013). The current observation that high motivation reduced orbitofrontal activity is consistent with previous findings showing that orbitofrontal neurons suppressed their firing rates with increasing size of a potential reward (Wallis & Miller, 2003). The superior frontal gyrus has been associated with the implementation of task sets (Wilk et al., 2012; Shi, Zhou, Muller, & Schubert, 2010; Dosenbach et al., 2007), suggesting that motivation and conflict expectancy act in concert to modulate task control processes. This may indicate that additional cognitive resources can be activated in response to high task difficulty expectancy when the task goal is additionally associated with a potential reward that seems to be worth the effort (Krebs et al., 2012). In addition, we found both motivation and conflict expectancy to be associated with activity in the caudate body, with motivation activating more dorsal parts of the striatum than conflict expectancy. Both the dorsal and ventral parts of the caudate body have been related to the processing of reward expectations (Liljeholm & O'Doherty, 2012; Aarts, van Holstein, & Cools, 2011; Harsay et al., 2011). Task difficulty-dependent activation in the striatum has already been found by previous studies (Krebs et al., 2012; Boehler et al., 2011), suggesting a general role of the striatum in adjusting processing to changing task contexts, which can be driven both by extrinsic (e.g., monetary rewards) and intrinsic (e.g., task difficulty expectations) motivational factors.

Summary

Summarizing, the current data suggest that motivation and conflict expectancy have both common and dissociable effects on conflict processing. Dissociable effects of motivation and conflict expectancy were found in the FFA and the VWFA because motivation selectively modulated activity in the FFA, whereas conflict expectancy also affected distractor-related activity in the VWFA (but note that the between-ROI analysis suggests the effects of Conflict expectancy to be more pronounced in the VWFA than in the FFA). This suggests that motivation improves conflict processing mainly by directing enhanced attention to task-relevant information, whereas conflict expectancy also suppresses the impact of distractor information. Interestingly, the motivational effects on task-relevant processing varied as a function of individual reward sensitivity scores. Brain regions that were affected both by motivation and conflict expectancy include, among others, dACC and superior frontal gyrus, hence regions that have been related to performance monitoring and cognitive control processes, respectively (Holroyd & Yeung, 2012; Alexander & Brown, 2011; Botvinick, 2007; Dosenbach et al., 2007). In these regions, motivation and conflict expectancy modulated neural activity nonlinearly, that is, the facilitatory effects of conflict expectancy on conflict processing were more pronounced under conditions of high than of low motivation. Together, these two mechanisms improve conflict processing to a level, which cannot be achieved by any of the two manipulations in isolation. Although both motivation and conflict expectancy trigger proactive control processes and modulate activity in a common control network, they seem to engage different conflict processing strategies and have partly dissociable effects on posterior brain regions related to target and distractor processing.

Acknowledgments

Reprint requests should be sent to Alexander Soutschek, Institute for Psychology, Humboldt-Universität zu Berlin, Unter den Linden 6, 10009 Berlin, Germany, or via e-mail: alexander.soutschek@hu-berlin.de.

REFERENCES

REFERENCES
Aarts
,
E.
,
van Holstein
,
M.
, &
Cools
,
R.
(
2011
).
Striatal dopamine and the interface between motivation and cognition.
Frontiers in Psychology
,
2
,
163
.
Alexander
,
W. H.
, &
Brown
,
J. W.
(
2011
).
Medial prefrontal cortex as an action-outcome predictor.
Nature Neuroscience
,
14
,
1338
1344
.
Boehler
,
C. N.
,
Hopf
,
J. M.
,
Krebs
,
R. M.
,
Stoppel
,
C. M.
,
Schoenfeld
,
M. A.
,
Heinze
,
H. J.
,
et al
(
2011
).
Task-load-dependent activation of dopaminergic midbrain areas in the absence of reward.
Journal of Neuroscience
,
31
,
4955
4961
.
Botvinick
,
M. M.
(
2007
).
Conflict monitoring and decision making: Reconciling two perspectives on anterior cingulate function.
Cognitive, Affective & Behavioral Neuroscience
,
7
,
356
366
.
Botvinick
,
M. M.
,
Braver
,
T. S.
,
Barch
,
D. M.
,
Carter
,
C. S.
, &
Cohen
,
J. D.
(
2001
).
Conflict monitoring and cognitive control.
Psychological Review
,
108
,
624
652
.
Braver
,
T. S.
(
2012
).
The variable nature of cognitive control: A dual mechanisms framework.
Trends in Cognitive Sciences
,
16
,
106
113
.
Bugg
,
J. M.
, &
Crump
,
M. J.
(
2012
).
In support of a distinction between voluntary and stimulus-driven control: A review of the literature on proportion congruent effects.
Frontiers in Psychology
,
3
,
367
.
Bugg
,
J. M.
,
Jacoby
,
L. L.
, &
Toth
,
J. P.
(
2008
).
Multiple levels of control in the Stroop task. [Research Support, N.I.H., Extramural].
Memory & Cognition
,
36
,
1484
1494
.
Carter
,
C. S.
,
Macdonald
,
A. M.
,
Botvinick
,
M.
,
Ross
,
L. L.
,
Stenger
,
V. A.
,
Noll
,
D.
,
et al
(
2000
).
Parsing executive processes: Strategic vs. evaluative functions of the anterior cingulate cortex.
Proceedings of the National Academy of Sciences, U.S.A.
,
97
,
1944
1948
.
Carver
,
C. S.
, &
White
,
T. L.
(
1994
).
Behavioral inhibition, behavioural activation, and affective responses to impending reward and punishment: The BIS/BAS Scales.
Journal of Personality and Social Psychology
,
67
,
319
333
.
Dehaene
,
S.
,
Le Clec
,
H. G.
,
Poline
,
J. B.
,
Le Bihan
,
D.
, &
Cohen
,
L.
(
2002
).
The visual word form area: A prelexical representation of visual words in the fusiform gyrus.
NeuroReport
,
13
,
321
325
.
Dosenbach
,
N. U.
,
Fair
,
D. A.
,
Miezin
,
F. M.
,
Cohen
,
A. L.
,
Wenger
,
K. K.
,
Dosenbach
,
R. A.
,
et al
(
2007
).
Distinct brain networks for adaptive and stable task control in humans.
Proceedings of the National Academy of Sciences, U.S.A.
,
104
,
11073
11078
.
Duthoo
,
W.
, &
Notebaert
,
W.
(
2012
).
Conflict adaptation: It is not what you expect.
Quarterly Journal of Experimental Psychology
,
65
,
1993
2007
.
Ebner
,
N. C.
,
Riediger
,
M.
, &
Lindenberger
,
U.
(
2010
).
FACES—A database of facial expressions in young, middle-aged, and older women and men: Development and validation.
Behavior Research Methods
,
42
,
351
362
.
Egner
,
T.
, &
Hirsch
,
J.
(
2005
).
Cognitive control mechanisms resolve conflict through cortical amplification of task-relevant information.
Nature Neuroscience
,
8
,
1784
1790
.
Engelmann
,
J. B.
,
Damaraju
,
E.
,
Padmala
,
S.
, &
Pessoa
,
L.
(
2009
).
Combined effects of attention and motivation on visual task performance: Transient and sustained motivational effects.
Frontiers in Human Neuroscience
,
3
,
4
.
Forman
,
S. D.
,
Cohen
,
J. D.
,
Fitzgerald
,
M.
,
Eddy
,
W. F.
,
Mintun
,
M. A.
, &
Noll
,
D. C.
(
1995
).
Improved assessment of significant activation in functional magnetic resonance imaging (fMRI): Use of a cluster-size threshold.
Magnetic Resonance in Medicine
,
33
,
636
647
.
Forstmann
,
B. U.
,
Jahfari
,
S.
,
Scholte
,
H. S.
,
Wolfensteller
,
U.
,
van den Wildenberg
,
W. P.
, &
Ridderinkhof
,
K. R.
(
2008
).
Function and structure of the right inferior frontal cortex predict individual differences in response inhibition: A model-based approach.
Journal of Neuroscience
,
28
,
9790
9796
.
Funes
,
M. J.
,
Lupianez
,
J.
, &
Humphreys
,
G.
(
2010
).
Sustained vs. transient cognitive control: Evidence of a behavioral dissociation.
Cognition
,
114
,
338
347
.
Harsay
,
H. A.
,
Cohen
,
M. X.
,
Oosterhof
,
N. N.
,
Forstmann
,
B. U.
,
Mars
,
R. B.
, &
Ridderinkhof
,
K. R.
(
2011
).
Functional connectivity of the striatum links motivation to action control in humans.
Journal of Neuroscience
,
31
,
10701
10711
.
Handwerker
,
D. A.
,
Ollinger
,
J. M.
, &
D'Esposito
,
M.
(
2004
).
Variation of BOLD hemodynamic responses across subjects and brain regions and their effects on statistical analyses.
Neuroimage
,
21
,
1639
1651
.
Holroyd
,
C. B.
, &
Yeung
,
N.
(
2012
).
Motivation of extended behaviors by anterior cingulate cortex.
Trends in Cognitive Sciences
,
16
,
122
128
.
Jimenez
,
L.
, &
Mendez
,
A.
(
2013
).
It is not what you expect: Dissociating conflict adaptation from expectancies in a Stroop task.
Journal of Experimental Psychology: Human Perception and Performance
,
39
,
271
284
.
Jimura
,
K.
,
Locke
,
H. S.
, &
Braver
,
T. S.
(
2010
).
Prefrontal cortex mediation of cognitive enhancement in rewarding motivational contexts.
Proceedings of the National Academy of Sciences, U.S.A.
,
107
,
8871
8876
.
Kane
,
M. J.
, &
Engle
,
R. W.
(
2003
).
Working-memory capacity and the control of attention: The contributions of goal neglect, response competition, and task set to Stroop interference.
Journal of Experimental Psychology: General
,
132
,
47
70
.
Kanwisher
,
N.
,
McDermott
,
J.
, &
Chun
,
M. M.
(
1997
).
The fusiform face area: A module in human extrastriate cortex specialized for face perception.
Journal of Neuroscience
,
17
,
4302
4311
.
Kerns
,
J. G.
,
Cohen
,
J. D.
,
MacDonald
,
A. W.
, III,
Cho
,
R. Y.
,
Stenger
,
V. A.
, &
Carter
,
C. S.
(
2004
).
Anterior cingulate conflict monitoring and adjustments in control.
Science
,
303
,
1023
1026
.
Kim
,
S. I.
(
2013
).
Neuroscientific model of motivational process.
Frontiers in Psychology
,
4
,
98
.
Kobayashi
,
S.
,
Lauwereyns
,
J.
,
Koizumi
,
M.
,
Sakagami
,
M.
, &
Hikosaka
,
O.
(
2002
).
Influence of reward expectation on visuospatial processing in macaque lateral prefrontal cortex.
Journal of Neurophysiology
,
87
,
1488
1498
.
Kouneiher
,
F.
,
Charron
,
S.
, &
Koechlin
,
E.
(
2009
).
Motivation and cognitive control in the human prefrontal cortex.
Nature Neuroscience
,
12
,
939
945
.
Krebs
,
R. M.
,
Boehler
,
C. N.
,
Egner
,
T.
, &
Woldorff
,
M. G.
(
2011
).
The neural underpinnings of how reward associations can both guide and misguide attention.
Journal of Neuroscience
,
31
,
9752
9759
.
Krebs
,
R. M.
,
Boehler
,
C. N.
,
Roberts
,
K. C.
,
Song
,
A. W.
, &
Woldorff
,
M. G.
(
2012
).
The involvement of the dopaminergic midbrain and cortico-striatal-thalamic circuits in the integration of reward prospect and attentional task demands.
Cerebral Cortex
,
22
,
607
615
.
Kriegeskorte
,
N.
,
Simmons
,
W. K.
,
Bellgowan
,
P. S.
, &
Baker
,
C. I.
(
2009
).
Circular analysis in systems neuroscience: The dangers of double dipping.
Nature Neuroscience
,
12
,
535
540
.
Krug
,
M. K.
, &
Carter
,
C. S.
(
2012
).
Proactive and reactive control during emotional interference and its relationship to trait anxiety.
Brain Research
,
1481
,
13
36
.
Liljeholm
,
M.
, &
O'Doherty
,
J. P.
(
2012
).
Contributions of the striatum to learning, motivation, and performance: An associative account.
Trends in Cognitive Sciences
,
16
,
467
475
.
Lindstrom
,
B. R.
,
Mattsson-Marn
,
I. B.
,
Golkar
,
A.
, &
Olsson
,
A.
(
2013
).
In your face: Risk of punishment enhances cognitive control and error-related activity in the corrugator supercilii muscle.
PLoS One
,
8
,
e65692
.
Miller
,
E. K.
, &
Cohen
,
J. D.
(
2001
).
An integrative theory of prefrontal cortex function.
Annual Review of Neuroscience
,
24
,
167
202
.
Neubert
,
F. X.
,
Mars
,
R. B.
,
Buch
,
E. R.
,
Olivier
,
E.
, &
Rushworth
,
M. F.
(
2010
).
Cortical and subcortical interactions during action reprogramming and their related white matter pathways.
Proceedings of the National Academy of Sciences, U.S.A.
,
107
,
13240
13245
.
Niendam
,
T. A.
,
Laird
,
A. R.
,
Ray
,
K. L.
,
Dean
,
Y. M.
,
Glahn
,
D. C.
, &
Carter
,
C. S.
(
2012
).
Meta-analytic evidence for a superordinate cognitive control network subserving diverse executive functions.
Cognitive, Affective & Behavioral Neuroscience
,
12
,
241
268
.
Padmala
,
S.
, &
Pessoa
,
L.
(
2011
).
Reward reduces conflict by enhancing attentional control and biasing visual cortical processing.
Journal of Cognitive Neuroscience
,
23
,
3419
3432
.
Pessoa
,
L.
, &
Engelmann
,
J. B.
(
2010
).
Embedding reward signals into perception and cognition.
Frontiers in Neuroscience
,
4
,
17
.
Peterson
,
B. S.
,
Kane
,
M. J.
,
Alexander
,
G. M.
,
Lacadie
,
C.
,
Skudlarski
,
P.
,
Leung
,
H. C.
,
et al
(
2002
).
An event-related functional MRI study comparing interference effects in the Simon and Stroop tasks.
Brain Research, Cognitive Brain Research
,
13
,
427
440
.
Polk
,
T. A.
,
Drake
,
R. M.
,
Jonides
,
J. J.
,
Smith
,
M. R.
, &
Smith
,
E. E.
(
2008
).
Attention enhances the neural processing of relevant features and suppresses the processing of irrelevant features in humans: A functional magnetic resonance imaging study of the Stroop task.
Journal of Neuroscience
,
28
,
13786
13792
.
Pratte
,
M. S.
,
Rouder
,
J. N.
,
Morey
,
R. D.
, &
Feng
,
C.
(
2010
).
Exploring the differences in distributional properties between Stroop and Simon effects using delta plots.
Attention, Perception, & Psychophysics
,
72
,
2013
2025
.
Raichle
,
M. E.
,
MacLeod
,
A. M.
,
Snyder
,
A. Z.
,
Powers
,
W. J.
,
Gusnard
,
D. A.
, &
Shulman
,
G. L.
(
2001
).
A default mode of brain function.
Proceedings of the National Academy of Sciences, U.S.A.
,
98
,
676
682
.
Ridderinkhof
,
K. R.
(
2002
).
Micro- and macro-adjustments of task set: Activation and suppression in conflict tasks.
Psychological Research
,
66
,
312
323
.
Shenhav
,
A.
,
Botvinick
,
M. M.
, &
Cohen
,
J. D.
(
2013
).
The expected value of control: An integrative theory of anterior cingulate cortex function.
Neuron
,
79
,
217
240
.
Shi
,
Y.
,
Zhou
,
X.
,
Muller
,
H. J.
, &
Schubert
,
T.
(
2010
).
The neural implementation of task rule activation in the task-cuing paradigm: An event-related fMRI study. [Research Support, Non-U.S. Gov't].
Neuroimage
,
51
,
1253
1264
.
Soutschek
,
A.
, &
Schubert
,
T.
(
2013
).
Domain-specific control mechanisms for emotional and nonemotional conflict processing.
Cognition
,
126
,
234
245
.
Soutschek
,
A.
,
Schwarzkopf
,
W.
,
Finke
,
K.
,
Hennig-Fast
,
K.
,
Muller
,
H. J.
,
Riedel
,
M.
,
et al
(
2013
).
Interference control in adult ADHD: No evidence for interference control deficits if response speed is controlled by delta plots.
Acta Psychologica
,
143
,
71
78
.
Soutschek
,
A.
,
Strobach
,
T.
, &
Schubert
,
T.
(
2013
).
Motivational and cognitive determinants of control during conflict processing.
Cognition & Emotion
,
28
,
1076
1089
.
Soutschek
,
A.
,
Taylor
,
P. C.
,
Muller
,
H. J.
, &
Schubert
,
T.
(
2013
).
Dissociable networks control conflict during perception and response selection: A transcranial magnetic stimulation study.
Journal of Neuroscience
,
33
,
5647
5654
.
Stelzel
,
C.
,
Brandt
,
S. A.
, &
Schubert
,
T.
(
2009
).
Neural mechanisms of concurrent stimulus processing in dual tasks.
Neuroimage
,
48
,
237
248
.
Strobel
,
A.
,
Beauducel
,
A.
,
Debener
,
S.
, &
Brocke
,
B.
(
2001
).
Eine deutschsprachige Version des BIS/BAS-Fragebogens von Carver und White.
Zeitschrift für Differentielle und Diagnostische Psychologie
,
22
,
216
227
.
Veling
,
H.
, &
Aarts
,
H.
(
2010
).
Cueing task goals and earning money: Relatively high monetary rewards reduce failures to act on goals in a Stroop task.
Motivation and Emotion
,
34
,
184
190
.
Wallis
,
J. D.
, &
Miller
,
E. K.
(
2003
).
Neuronal activity in primate dorsolateral and orbital prefrontal cortex during performance of a reward preference task.
European Journal of Neuroscience
,
18
,
2069
2081
.
Ward
,
B.
(
2000
).
Simultaneous inference for fMRI data. AFNI 3dDeconvolve Documentation
.
Milwaukee, WI
:
Medical College of Wisconsin
.
Wilk
,
H. A.
,
Ezekiel
,
F.
, &
Morton
,
J. B.
(
2012
).
Brain regions associated with moment-to-moment adjustments in control and stable task-set maintenance.
Neuroimage
,
59
,
1960
1967
.

Author notes

*

Shared senior authorship.