Participants 101

Twenty-nine participants (23 females) between 19 and 26 years of age (mean 22.0) with normal or corrected-to-normal vision participated in the study. Participants were recruited at the University of Konstanz, and received a base fee of 7 € and a performance-dependent bonus (mean: 7.25 €, range: 2.00 – 11.40 €). Eye-blink artifacts were excessive in the EEG data recorded from four participants and we therefore excluded these from our analysis. The study was conducted in accordance with institutional guidelines and informed consent was acquired from all participants.

II.3.3.2 Stimulus Material

The test stimuli consisted of 288 Chinese and Chinese-looking characters. Each character comprised a height of 2.5° to 2.9° visual angle and a width of 2.7° to 2.9° visual angle at a viewing distance of 60 cm, and was presented in white color on a black

background. For each participant, the stimuli were randomly divided into 24 sets of six items with each item consisting of a pair of characters. Within each item, each character was presented 0.5° visual angle left or right from the screen center. The position of each character (left or right) was randomly determined for each presentation. The feedback stimuli and validity stimuli comprised a height of 1.2° visual angle and a width of 3° visual angle, which were presented centrally on the screen. For each participant, stimuli were combined to 144 pairs that were randomly assigned to 36 blocks so that every block included four pairs.

II.3.3.3 Design and Procedure

Each item was presented in a learning trial and a test trial. In learning trials,

participants had to guess which one of the two characters (left or right) was associated with a reward. The procedure of a learning trial is illustrated in Figure 12. After presentation of an item, responses had to be given within five seconds by pressing one of two keys on a German standard keyboard (Y with left index finger for a left response, M with right index finger for a right response). Following the response (or after five seconds), the item disappeared and the screen remained blank for 200 ms. Then, the feedbackstimulus was presented for 600 ms. A feature mask followed after another 200 ms of blank screen. This feature mask was a mixture of the words “RICHTIG” and “FALSCH” and was presented for 600 ms. Another blank screen was replaced after 200 ms by the validity cue, which was presented for 600 ms.

Afterwards a final blank screen appeared for 1950 to 2450 ms. Participants were told to use this time period for further feedback processing and learning until a new trial started.

Figure 12. Sequence of events in a typical trial during the learning phase. The stimulus consisted of two Chinese-looking characters. After participants pressed a response key, the initial feedback was presented followed by a feature mask and finally the validity cue that indicated whether the initial feedback had been valid or invalid.

The initial feedback consisted the German word RICHTIG (engl. right) or FALSCH (engl. false) that was presented centrally in white color. If no response was given within the

time limit, the German word VERPASST (engl. missed) was presented. During the learning phase, these misses were associated with a malus of 10 Euro Cents. The validity cue consisted of either the German word KORREKT (engl. correct) or “INKORREKT” (engl. incorrect) depending on validity condition. Note that the correctness of the response could only be deduced by taking both, the feedback and the validity cue, into consideration. The validity cue indicated whether the preceding feedback was correct or incorrect. Therefore, for instance, the combination of “right” and “incorrect” meant that one had chosen the wrong character. The validity of the initial feedback depended on the experimental condition: it was valid in either 62.5% (unreliable feedback) or 87.5% (reliable feedback) of the cases. Crucially, participants were made aware of the feedback’s validity by a prompt at the beginning of each learning phase. This prompt stated that the feedback in the following learning block would be valid in 62.5% or 87.5% of the cases, respectively.

In test trials, the same items were presented again, but now, correct and wrong responses were associated with a win of 10 Euro Cents or a loss of 10 Euro Cents,

respectively. To ensure that the stimulus rather than the response side was learnt, the positions of the characters were changed relative to that in the learning trials in half of the items. The procedure was the same as in learning trials with one exception: A feedback was presented for 800 ms that indicated the amount of win (“+10 cent”) or loss (“-10 cent”) depending on the correctness of the response. Again, a feedback indicating a miss was given in case of a late response. During the test phase, these misses were associated with a loss of 30 Euro Cents.

Each participant was tested in an individual session. After the fitting of the electrode cap, participants were seated comfortably in a dimly lit, electrically shielded room. After two training blocks, the actual experiment began. There were 18 blocks with reliable (87.5%

valid) and 18 blocks with less reliable (62.5% valid) feedback. Every nine blocks, feedback validity likelihood changed. Whether the experiments started with a reliable or less reliable feedback blocks was counterbalanced between participants. Each block comprised a learning and a test phase in which four items were first learnt and then tested after a short, self-paced interruption. The order of items was randomized within each phase. There was a longer break after every six blocks. After the final block, participants were given a short questionnaire and then were paid according to their performance.

II.3.3.4 Electrophysiological Recordings

Throughout the experiment, the electroencephalogram (EEG) was recorded using a BIOSEMI Active-Two system (BioSemi, Amsterdam, The Netherlands) with 64 Ag-AgCl electrodes from channels Fp1, AF7, AF3, F1, F3, F5, F7, FT7, FC5, FC3, FC1, C1, C3, C5, T7, TP7, CP5, CP3, CP1, P1, P3, P5, P7, P9, PO7, PO3, O1, Iz, Oz, POz, Pz, CPz, Fpz, Fp2, AF8, AF4, AFz, Fz, F2, F4, F6, F8, FT8, FC6, FC4, FC2, FCz, Cz, C2, C4, C6, T8, TP8, CP6, CP4, CP2, P2, P4, P6, P8, P10, PO8, PO4, O2 as well as the left and right mastoid. The CMS (Common Mode Sense) and DRL (Driven Right Leg) electrodes were used as reference and ground electrodes. Vertical and horizontal electrooculogram (EOG) was recorded from electrodes above and below the right eye and on the outer canthi of both eyes. All electrodes were off-line re-referenced to averaged mastoids. EEG and EOG were continuously recorded at a sampling rate of 512 Hz.

II.3.3.5 Data Analysis

Trials were excluded on which participants did not respond or responded too late (mean: 0.4 trials per participants). The remaining trials were classified as correct or incorrect.

The main dependent variable was the proportion of correct trials in each condition of interest.

EEG data were analyzed using EEGLAB v6.01 (Delorme & Makeig, 2004) and custom routines written in MatLab 7.0.4 (The Mathworks, Natick, MA). The data were downsampled to 256Hz and band-pass filtered excluding activity below 1 Hz and above 30 Hz. Epochs were extracted ranging from 100 ms before and 500 ms after feedback onset.

Large artifacts were identified by computing the joint probability of each epoch and excluding epochs that deviated more than five standard deviations from the distribution mean, and by excluding epochs with abnormally peaked activity as indicated by a kurtosis that deviated from the mean kurtosis value by three standard deviations. We also excluded epochs in which activity exceeds a threshold of +/-150 µV. Further artifacts were identified by application of an automated regression method, based on Least Mean Square (LMS; Delorme & Makeig, 2004). Baseline activity was removed by subtracting the average voltage from an interval between 100 ms and 0 ms before feedback onset. Finally, epochs were averaged separately for each condition of interest.

Feedback-related brain activity was quantified by calculating amplitude differences between trials with positive and negative feedback. Peak-to-peak voltages of the FRN were

additionally quantified by first filtering the data with a 15Hz low-pass filter, to allow for a more reliable peak identification, and then calculating the difference between the most negative peak between 200 and 400 ms after feedback onset and the immediately preceding positive peak (see Frank et al., 2005). This was done separately for each participant and each condition of interest. We also applied this filter for the presentation of grand averages in our figures.

Behavioral data as well as ERP data were analyzed using repeated measures

ANOVAs. To compensate for violations of sphericity, Huynh-Feldt corrections were applied whenever appropriate, and corrected p-values (but uncorrected degrees of freedom) are reported.

II.3.4 Results