$[MethodName<>Trial Design]{

'$[Name]' Method.

} $[MethodName=Trial Design]{

Trial Conditions.

}

$[MethodName=Constant Stimuli] { A method of constant stimuli was used to measure performance as function of the $[NameLevel] level (dependent variable). There were $[NbLevels] levels between $[MinLevel] and $[MaxLevel] according to a $[SelectionLevel] sampling, each presented $[TrialsLevel] times. Performance was fitted using a $[FittingDirection] $[FittingName] psychometric function (chance level of $[ChanceLevel]% and miss rate of $[MissRate]%), and a $[NameLevel] threshold was derived for a performance criterion of $[Criterion]%. } $[MethodName=Staircase] { A staircase method was used to estimate a $[NameLevel] threshold$[NBCONDITIONS?0]{. The } $[NBCONDITIONS?1] { $[NBCONDITIONS<>1]{and several instances ($[NBCONDITIONS]) of the staircase were interleaved to investigate the role of the stimulus parameters defined above. In each staircase, the } $[NBCONDITIONS=1]{The } } dependent variable ($[NameLevel]) was reduced after $[NbGood] consecutive correct responses and increased after $[NbWrong] wrong response(s), corresponding to a criterion of $[THRESHOLDCRITERION]% correct responses. The initial $[NameLevel] value was randomly selected in the range $[InitValue]±$[RangeValue]. $[MaximumMode=Clamp] { The maximum $[NameLevel] value was always limited to $[MaximumValue]. } $[MinimumMode=Clamp] { The minimum $[NameLevel] value was always limited to $[MinimumValue]. } $[StepSize=Relative] { The reduction rate in $[NameLevel] was $[DecRateBeforeFirstReversal]% before the first reversal and $[DecRateAfterFirstReversal]% after the $[FirstReversal]th reversal, while the increase rate was always $[IncRate]%. } $[StepSize=Absolute] { The $[NameLevel] level was reduced by $[DecRateBeforeFirstReversal] before the first reversal and $[DecRateAfterFirstReversal]% after the $[FirstReversal]th reversal, while the increase step size was always $[IncRate]. } Each session was terminated after $[LastReversal] reversals, and the threshold was computed from the mean of the last $[NBREVERSALSTHRESHOLD] reversals. } $[MethodName=Bayesian] { The Ψ Bayesian method (Kontsevich & Tyler, 1999) was used to estimate the psychometric slope and threshold for the stimulus $[NameLevel] (dependent variable). A Bayesian adaptive estimation of psychometric slope and threshold requires two basic components: 1) a probability density function defined over a two-dimensional space of psychometric parameters (α and β), and 2) a one-dimensional space of possible stimuli $[NameLevel]. The method's basic goal is to accelerate the estimation of the psychometric parameters by efficiently searching the stimulus space for $[NameLevel] levels that improve the information gained over the psychometric parameter space on each trial (Kontsevich & Tyler, 1999). The ranges of possible psychometric parameters are: $[ALPHAMIN] to $[ALPHAMAX] with $[ALPHASAMPLES] samples for α and $[BETAMIN] to $[BETAMAX] with $[BETASAMPLES] samples for β. The possible ranges for stimuli were $[LEVELMIN] to $[LEVELMAX] with $[LEVELSAMPLES] samples for $[NameLevel]. The parameter and stimulus spaces are defined on $[SelectionLevel=Uniform]{linear}$[SelectionLevel=Log 10]{log} grids. The psychometric function was a $[FittingDirection] $[FittingName] (chance level of $[ChanceLevel]% and miss rate of $[MissRate]%), and a $[NameLevel] threshold and slope were derived for a performance criterion of $[Criterion]%. $[NbTraining] training trials were run before starting the actual data collection which consists of $[NbTrials] trials. } $[MethodName=Trial Design] { The experimental protocol described below was run for $[NBCONDITIONS<>1] { the following $[NBCONDITIONS] conditions: $[BlockDesign] These conditions were run in a $[ConditionOrder=Preorder]{predefined} order. } $[NBCONDITIONS=1]{several conditions defined to investigate the $[BlockDesign]} $[BlockRepeat<>1]{The condition block was repeated $[BlockRepeat] times.} Different trials for each condition were $[TrialOrder=Random]{randomized}. $[KeepTrialSequence=1]{But same trials were run consecutively.} }