Psychophysics

Psychophysics is a branch of psychology that investigates the relationship between physical stimuli and an observer's sensation, which is represented by the so-called psychophysical function. Studying this psychophysical function thus is the main interest in this field. By applying mathematical models as is done in the branch, specific and falsifiable predictions for data can be made. In the following, we focus on the Indecision Model of Psychophysical Performance by Miguel A. Garcia-Perez and Rocio Alcala-Quintana (2017))

Experimental setup

The Indecision Model of Psychophysical Performance focuses on dual-presentation tasks in which the observer is confronted with two stimuli: standard and test. In so-called detection tasks only one of the two stimuli contains sensory material; this is the test. The observer then has to identify this sensory material. In discrimination tasks on the other hand side, both stimuli presented are detectable for the observer but have to be compared by a certain attribute (e.g., 'Which distance is larger?'). Within modalities, stimuli can differ in more than one aspect (e.g., length and rotation).\ It also is important to select the adequate response format. The classical, binary response format forces the observer to decide between two response alternatives. This leads to more or less random answers when the observer is unsure. In contrast, ternary response formats offer the opportunity to express uncertainty by adding a neutral response ('I don't know' or I cannot tell the difference between the stimuli'). The Indecision Model of Psychophysical Performance was developed primarily for the ternary response format, but can also be applied to the binary response format.

Theory and parameters

For every individual and every modalitity, a psychophysical function $$\mu(x)$$ (with x denoting the physical magnitude of the sensory evidence) is assumed to exist. The exact shape of the psychophysical function has been debated intensively and is very likely to differ for different modalities. The following parameters influence the shape of the psychophysical function $$\mu(x) = log(1 + 2*\exp{\frac{x - \alpha}{\beta}})$$. When standard and test differ in more than one attribute, two psychophysical and psychometric functions result. Alpha and beta then have to be specified for both standard and test.

- alpha: detection boundary

- beta: refers to the slope

The percieved sensation of a stimulus is largely influenced by its physical magnitude.

- standard level: (fixed) level of the standard

- test levels: magnitude of test stimuli

Judgements of sensations are performed on the basis of subjective criteria and thus are considered a probabilistic process. In the Indecision Model of Psychophysical Performance this is modeled by so-called psychometric functions, illustrating the likelihood for each response category dependent on the percieved magnitude.

- delta_width: the width of the range of subjective magnitudes for which the observer is undecided

- bias: decisional bias; asymmetry in the 'undecided'-range

This shiny app provides an interactive way of simulating data from dual-presentation tasks. It allows the user to set the parameter values for one observer with a fixed standard and several test levels. The user is able to select the response format (binary or ternary) as well as the task type (detection or discrimination). From the simulated data, the psychometric and the psychophysical function are estimated under the Indecision Model of Psychophysical Performance. Estimates for the model parameters are provided.