Hi PostGrads,
I have a behavioural dataset with huge amounts of missing data.
Essentially I have 2 groups (clinical and control, between subjects factor); and 1 within subjects factor of condition (call it 'Cond'; there are 5 variants). All subjects experience all conditions. During each condition they are allowed to make a response, but only if they want.
The DV of interest is the RT between Cond onset and the optional response if made, so this data is categorical.
My question is, this dataset is full of missing information. A certain subject may never make this response, or may make it only in some conditions.
How do I analyse this? Would I replace all the missing data with the time of the entire condition (eg the max for the condition)? It is my intuition that this would not give an accurate picture of the data.
Best wishes
Fiona McGruer
|