**********************************************************
**********************************************************

PA3 README

  1. EXPERIMENTAL DESIGN
  2. RUNNING PA3
  3. ANNOTATING THE DATA
  4. ANALYSIS

**********************************************************
**********************************************************


1. EXPERIMENTAL DESIGN

The PA2 experiment was designed to be used specifically during
electrophysiological recordings. This experiment uses interference
versus no interference, forward versus backward recall, and item
versus associative encoding to try and elicit differences in brain
activation.

It is based on the older pypa2 design.


2. RUNNING PA3

*********
PA3 Update: looking at old data on rhino it appears both the easy
and regular lists have been used in the past. For example:
/data1/eeg/group2/BW002/behavioral/pa2/data/21/subdat.csv
/data1/eeg/group2/BW002/behavioral/pa2/data/22/subdat.csv
*********

First check to see that the appropriate links for your language are 
made in the txt/ directory. For example, if running in Germany, 
'nounpool.txt' should be a link to 'nounpool_german.txt', and 'intro.txt'
should be linked to 'intro_german.txt'. Likewise, 'nounpool.txt' should
point to the desired noun pool file.
Once the proper links have been configured, make sure that 'config.py' 
has the configuration you desire.

Then type,

   pythonw pypa3.py -s SUBJECT_ID

and enter the subject identifier in place of SUBJECT_ID. You may quit 
the experiment between trials by pressing the ESC-F1 keys together. 
The experimenter is required to hit a key after each subject response 
to proceed to the next trial.


3. ANNOTATING THE DATA

Annotate the files in 'data/SUBJECT_ID/session_RUN' as you normall would.
The wordpool file is whatever 'txt/nounpool.txt' points to.

If the user says nothing during the entire file, mark a vocalization at
the very end of the file.

4. ANALYSIS

  pythonw post-process.py -s SUBJECT_ID

The behavioral results of the experiment are contained within the 
subdat.csv file. 

The EEG results can can be found in 
data/SUBJECT_ID/session_RUN/eeg.eeglog and 
data/SUBJECT_ID/session_RUN/session.log .

The analyze_subject.m script can be run from within matlab to assess
the performance of an individual subject. This script is well
documented.