We will use multilevel modeling for the analysis. <br> **Exclusions** We will exclude all trials from the beginning of the task where participants did not sort the object (until the first sorted object). We will also exclude participants with more than 10 objects that are unsorted or sorted neither according to their shape nor color. <br> **Analysis** We will use misclassification as a primary dependent variable. A misclassification is defined as a case where an object is classified according to its shape and not its color. Classifications according to both factors (where both criteria converge) or neither factor (errors and unsorted objects) will not be used in the analysis. <br> **Hypotheses** We will explore the following hypotheses: *Participants will take the opportunity to cheat, i.e. proportion of misclassification will be higher for objects with bribes than for objects without bribes. They will also prefer bigger bribes.* *The probability of accepting bigger bribes will be higher in the condition where participants are offered small and big bribes than in the condition where they are offered middle-sized and big bribes.* *Participants who do not accept any other bribes will neither accept the utility raising one.* *Reaction times on trials with bribes will be longer than on trials without bribes and longer for bigger bribes than for smaller bribes. Also, the reaction times on trials with the first accepted bribe will be longer than on trials with subsequently accepted bribes.* *Participants will stop taking further bribes after a certain amount of loss caused to the third party.* *Males will accept more bribes than females.* *Participants who view misclassification negatively and rejecting bribes positively will accept less bribes.* *The propensity to act corruptly will be associated with Honest-humility scale from HEXACO personality inventory.*