Identifying Independent and Dependent Variables

HAL10000

Hi, I am trying to identify the independent and dependent variables in the following experiment:

Two groups of patients (Epileptic and Normal) are exposed to varying wavelengths of light and their neural response is quantified and recorded.
I am using statistical analysis to find the wavelength at which the most significant difference between the Epileptic and Normal groups is found.

I thought that the condition of the patient (Epileptic or Normal) was the independent variable and the neural response was the dependent variable but I'm not sure whether to think of the wavelength as an independent variable or a control

bpet

It could be argued that patient's condition is the "dependent" variable (aka "response variable").

If I understand correctly, the broad goal is to use neural responses to predict patient's condition, i.e. the "independent" variables (explanatory variables) are the patient's neural response - one for each wavelength, or possibly several for each wavelength. Your statistical analysis is evaluating which single variable/group leads to a classifier that performs best on the "training" data already observed where condition is known.

From this point of view, wavelength isn't really a variable - you could say it's just part of the label for each candidate explanatory variable.

However the roles of all the variables would change in another statistical model that relates neural response to wavelength in people with and without epilepsy - here the explanatory variables would be wavelength and condition, and the response variable(s) would be neural response. This model might play an important role in other aspects of the project such as experimental design.