|
|
hi,
could anyone give me some advice on the following question, please?
in my experimental design (psychoacoustics), i'm collecting
six repeated values of each parameter.
in a first step, i calculate the median of these six repeats
to determine the median performance of each individual taking part
in the experiment.
as my research is concerned with age dependency effects, i'd
furthermore like to determine group performance values.
is it admissible to compute those group medians (e.g., for N 20- to
30-year-olds) from the (N) individual medians of all the N group members,
or had i better to calculate them based on the raw data, i.e., from
the 6N original measured values?
thanks for support!
sonja.
|
Locked
