Performance of the Effort-Control task provides two metrics. A first metrics is the motivational break-point that described the maximal amount of effort (balloon outlines) over the expected reward outcomes (numbers of tokens) that lead a subject to avoid choosing that option over the alternative option (1). The break-point computation also entails an estimate of the subjective value of the chosen side as an ‘expected utility’ metric that reflect the ratio of expected benefits (amount of tokens) and costs (number of balloon outlines). As a secondary metrics the Effort Conrol task allows estimating the ‘motivational vigor’ as the average duration of pumping up a balloon normalized by the number of outlines/pumps needed to explode it (2).

Performance of the Flex-Learning task provides five metrics. The speed of ‘object-value learning’ (1) corresponds to the time to reach criterion performance defined as the first trial leading on average to 75% correct performance over a forward looking 10-trial window. Learning speed indexes cognitive flexibility and comprises multiple subfunctions that this and other tasks tap into. Comparison of the learning speed for blocks with different gains indexes reward sensitivity (2) and for different losses indexes loss sensitivity (3). Flex-Learning allows quantifying error monitoring for avoiding perseveration of choosing a non-rewarded object (4), using a previous-trial analysis that measures the accuracy in the trials after an error trial (EC_n analysis). The EC_n analysis allows quantifying a perseveration score that is higher when a subject shows a shallow rise or no improvement of performance after errors. Comparing of the perseveration score in blocks with the same versus novel objects than in the previous block estimates the vulnerability to ‘pro-active interference’ from the previous blocks target object (5).

Performance of the Effort-Control task provides two metrics. A first metrics is the motivational break-point that described the maximal amount of effort (balloon outlines) over the expected reward outcomes (numbers of tokens) that lead a subject to avoid choosing that option over the alternative option (1). The break-point computation also entails an estimate of the subjective value of the chosen side as an ‘expected utility’ metric that reflect the ratio of expected benefits (amount of tokens) and costs (number of balloon outlines). As a secondary metrics the Effort Conrol task allows estimating the ‘motivational vigor’ as the average duration of pumping up a balloon normalized by the number of outlines/pumps needed to explode it (2).

Performance of the Flex-Learning task provides five metrics. The speed of ‘object-value learning’ (1) corresponds to the time to reach criterion performance defined as the first trial leading on average to 75% correct performance over a forward looking 10-trial window. Learning speed indexes cognitive flexibility and comprises multiple subfunctions that this and other tasks tap into. Comparison of the learning speed for blocks with different gains indexes reward sensitivity (2) and for different losses indexes loss sensitivity (3). Flex-Learning allows quantifying error monitoring for avoiding perseveration of choosing a non-rewarded object (4), using a previous-trial analysis that measures the accuracy in the trials after an error trial (EC_n analysis). The EC_n analysis allows quantifying a perseveration score that is higher when a subject shows a shallow rise or no improvement of performance after errors. Comparing of the perseveration score in blocks with the same versus novel objects than in the previous block estimates the vulnerability to ‘pro-active interference’ from the previous blocks target object (5).

Performance of the Maze-Learning task provides four metrics. The average number of choices to complete a maze, normalized by the path length of the maze quantifies the speed of ‘spatial learning’ (1). The proportion of rule-breaking errors indexes ‘error monitoring for following a good strategy’ (2), while the proportion of erroneously repeating choosing the same incorrect tile reflects repetition errors used to evaluate ‘error monitoring for avoiding perseveration’ (3). Each maze is repeated with its unique context background and start-/end-position at 30-50 min after the initial learning of the maze. The difference in learning speed of the repeated versus initial maze evaluates ‘spatial memory’ (4).

Performance of the THR task is measured by the number of correct selections, and ensures subjects learn to touch and release visually displayed stimulus for a controlled duration.

Performance of the Delayed-Match-Sample Working Memory task provides three main metrics. The temporal delay that causes the first statistically significant decrement in accuracy relative to the shortest (0.25ms) delay estimates the ‘temporal persistence’ of working memory (1). The difference in average performance in trials with versus without distractor objects during the delay estimates the degree of ‘WM-interference control’ (2), which differs from the metrics ‘perceptual interference control’ that quantifies the accuracy decrease when the sample and test objects are multidimensional as opposed to one-dimensional.

Performance of the WWW task provides four metrics. The average number of trials required to complete the four-order sequence corresponds to the ability of ‘object sequence learning’ (1), which involves forming object-object associations and chunking them into a sequence. Errors differ depending on whether an object from earlier positions is erroneously chosen again (repetition errors), or whether subjects chose the wrong object at a particular temporal position (slot errors). The proportion of repetition errors to all errors signifies the ability of ‘error monitoring/perseveration’ (2), while the proportion of slot errors indexes error monitoring/strategy following (3). Object sequences of the WWW task are accompanied by one distractor object that is irrelevant for the sequence, but which is similar to either the second or third object of the sequence it appears in. The number of distractor errors at the ordinal position at which the distractor was similar to the correct object is divided by all distractor errors to quantify how likely the distractor is confused with the correct object that specific to the ordinal position, i.e. it measures ‘temporal distractor interference’ (4). This ordinal positioning distractor effect is similar to the symbolic distance effect used in the literature to infer the learning of sequential relationships.