function OldNewRecogExp(subNo,hand)
% OldNewRecogExp(subNo,hand);
%
% Example of an old/new recognition experiment.
%
% This is an example of a simple old/new recognition experiment. It is split
% into two phases, a study phase and a test phase:
%
% Study phase:
%
% In the study phase, the subject is presented with 3 images of objects in
% randomized presentation order and has to learn/memorize them. Each image
% is presented for a 'duration' of 2 seconds, then the image disappears and
% the subject has to advance to the next presentation by pressing the 'n'
% key on the keyboard. The list of 'study' objects is read from the file
% 'studylist.txt'.
%
% Test phase:
%
% In the test phase, the subject is presented with test images defined in
% the file 'testlist.txt', again in randomized order. Test images are
% presented for a 'duration' of 0.5 seconds or until the subjects responds
% with a keypress. The subject has to press one of two keys, telling if the
% test image is an "old" image - previously presented in the study phase,
% or a "new" image - not seen in the study phase. The keys used are 'c' and
% 'm', the mapping of keys to response ("old" or "new") is selected by the
% input argument "hand" -- allowing to balance for handedness of subjects /
% response bias.
%
% At the end of a session (when all images in the 'testlist.txt' have been
% presented and tested), the results - both response of the subject and its
% reaction time - are stored to a file 'OldNewRecogExp_xx.dat', where 'xx'
% is the subject number given as input argument "subNo" to this script.
%
% Input parameters:
%
% subNo subject number; use subNo>99 to skip check for existing file
% hand response mapping for test phase
%
% e.g.: OldNewRecogExp(99,1);
%
% Example DESIGN:
%
% STUDY PHASE: study 3 objects
% TEST PHASE: shown 6 objects, decide whether the object is old or new
%
% This script demonstrates:
%
% - reading from files to get condition on each trial
% - randomizing conditions (for the study and test phase)
% - showing image/collecting response (response time, accuracy)
% - writing data to file "OldNewRecogExp_<subNo>.dat
%
% Please refer to other included demos for new functions of Psychtoolbox-3
% vs. the old Psychtoolbox-2.
%
% Other than a few changes how to call the Screen function,
% and the try ... catch statements, this code replicates a
% typical OS9 experiment
%
% NOTE to previous MacOS-9 users: OSX is case sensitive!!!
%
% History:
%
% 05/24/05 Quoc Vuong, PhD, University of NewCastle wrote and contributed
% it, as an example for usage of Psychtoolbox-3, Version 1.0.6.
% 03/01/08 Mario Kleiner modified the code to make use of new functionality
% added in Psychtoolbox 3.0.8.
%%%%%%%%%%%%%%%%%%%%%%%%%
% any preliminary stuff
%%%%%%%%%%%%%%%%%%%%%%%%%
% Clear Matlab/Octave window:
clc;
% check for Opengl compatibility, abort otherwise:
AssertOpenGL;
% Check if all needed parameters given:
if nargin < 2
error('Must provide required input parameters "subNo" and "hand"!');
end
% Reseed the random-number generator for each expt.
rand('state',sum(100*clock));
% Make sure keyboard mapping is the same on all supported operating systems
% Apple MacOS/X, MS-Windows and GNU/Linux:
KbName('UnifyKeyNames');
% Init keyboard responses (caps doesn't matter)
advancestudytrial=KbName('n');
% Use input variable "hand" to determine response mapping for this session.
if (hand==1)
oldresp=KbName('c'); % "old" response via key 'c'
newresp=KbName('m'); % "new" response via key 'm'
else
oldresp=KbName('m'); % Keys are switched in this case.
newresp=KbName('c');
end
%%%%%%%%%%%%%%%%%%%%%%
% file handling
%%%%%%%%%%%%%%%%%%%%%%
% Define filenames of input files and result file:
datafilename = strcat('OldNewRecogExp_',num2str(subNo),'.dat'); % name of data file to write to
studyfilename = 'studylist.txt'; % study list
testfilename = 'testlist.txt'; % test list
% check for existing result file to prevent accidentally overwriting
% files from a previous subject/session (except for subject numbers > 99):
if subNo<99 && fopen(datafilename, 'rt')~=-1
fclose('all');
error('Result data file already exists! Choose a different subject number.');
else
datafilepointer = fopen(datafilename,'wt'); % open ASCII file for writing
end
%%%%%%%%%%%%%%%%%%%%%%
% experiment
%%%%%%%%%%%%%%%%%%%%%%
% Embed core of code in try ... catch statement. If anything goes wrong
% inside the 'try' block (Matlab error), the 'catch' block is executed to
% clean up, save results, close the onscreen window etc.
try
% Get screenNumber of stimulation display. We choose the display with
% the maximum index, which is usually the right one, e.g., the external
% display on a Laptop:
screens=Screen('Screens');
screenNumber=max(screens);
% Hide the mouse cursor:
HideCursor;
% Returns as default the mean gray value of screen:
gray=GrayIndex(screenNumber);
% Open a double buffered fullscreen window on the stimulation screen
% 'screenNumber' and choose/draw a gray background. 'w' is the handle
% used to direct all drawing commands to that window - the "Name" of
% the window. 'wRect' is a rectangle defining the size of the window.
% See "help PsychRects" for help on such rectangles and useful helper
% functions:
[w, wRect]=Screen('OpenWindow',screenNumber, gray);
% Set text size (Most Screen functions must be called after
% opening an onscreen window, as they only take window handles 'w' as
% input:
Screen('TextSize', w, 32);
% Do dummy calls to GetSecs, WaitSecs, KbCheck to make sure
% they are loaded and ready when we need them - without delays
% in the wrong moment:
KbCheck;
WaitSecs(0.1);
GetSecs;
% Set priority for script execution to realtime priority:
priorityLevel=MaxPriority(w);
Priority(priorityLevel);
% run through study and test phase
for phase=1:2 % 1 is study phase, 2 is test phase
% Setup experiment variables etc. depending on phase:
if phase==1 % study phase
% define variables for current phase
phaselabel='study';
duration=2.000; % Duration of study image presentation in secs.
trialfilename=studyfilename;
message = 'study phase ...\nstudy each picture ... press _n_ when it disappears ...\n... press mouse button to begin ...';
% Screen('DrawText', w, , 10, 10, 255);
% Screen('DrawText', w, , 10, 40, 255);
% Screen('DrawText', w, , 10, 70, 255);
else % test phase
% define variables
phaselabel='test';
duration=0.500; %sec
trialfilename=testfilename;
% write message to subject
str=sprintf('Press _%s_ for OLD and _%s_ for NEW\n',KbName(oldresp),KbName(newresp));
message = ['test phase ...\n' str '... press mouse button to begin ...'];
% Screen('DrawText', w, 'test phase ...', 10, 10, 255);
% Screen('DrawText', w, str, 10, 40, 255);
% Screen('DrawText', w, '... press mouse to begin ...', 10, 70,
% 255);
end
% Write instruction message for subject, nicely centered in the
% middle of the display, in white color. As usual, the special
% character '\n' introduces a line-break:
DrawFormattedText(w, message, 'center', 'center', WhiteIndex(w));
% Update the display to show the instruction text:
Screen('Flip', w);
% Wait for mouse click:
GetClicks(w);
% Clear screen to background color (our 'gray' as set at the
% beginning):
Screen('Flip', w);
% Wait a second before starting trial
WaitSecs(1.000);
% read list of conditions/stimulus images -- textread() is a matlab function
% objnumber arbitrary number of stimulus
% objname stimulus filename
% objtype 1=old stimulus, 2=new stimulus
% for study list, stimulus coded as "old"
[ objnumber, objname, objtype ] = textread(trialfilename,'%d %s %d');
% Randomize order of list
ntrials=length(objnumber); % get number of trials
randomorder=randperm(ntrials); % randperm() is a matlab function
objnumber=objnumber(randomorder); % need to randomize each list!
objname=objname(randomorder); %
objtype=objtype(randomorder); %
% loop through trials
for trial=1:ntrials
% wait a bit between trials
WaitSecs(0.500);
% initialize KbCheck and variables to make sure they're
% properly initialized/allocted by Matlab - this to avoid time
% delays in the critical reaction time measurement part of the
% script:
[KeyIsDown, endrt, KeyCode]=KbCheck;
% read stimulus image into matlab matrix 'imdata':
stimfilename=strcat('stims/',char(objname(trial))); % assume stims are in subfolder "stims"
imdata=imread(char(stimfilename));
% make texture image out of image matrix 'imdata'
tex=Screen('MakeTexture', w, imdata);
% Draw texture image to backbuffer. It will be automatically
% centered in the middle of the display if you don't specify a
% different destination:
Screen('DrawTexture', w, tex);
% Show stimulus on screen at next possible display refresh cycle,
% and record stimulus onset time in 'startrt':
[VBLTimestamp startrt]=Screen('Flip', w);
% while loop to show stimulus until subjects response or until
% "duration" seconds elapsed.
while (GetSecs - startrt)<=duration
% poll for a resp
% during test phase, subjects can response
% before stimulus terminates
if ( phase==2 ) % if test phase
if ( KeyCode(oldresp)==1 || KeyCode(newresp)==1 )
break;
end
[KeyIsDown, endrt, KeyCode]=KbCheck;
end
% Wait 1 ms before checking the keyboard again to prevent
% overload of the machine at elevated Priority():
WaitSecs(0.001);
end
% Clear screen to background color after fixed 'duration'
% or after subjects response (on test phase)
Screen('Flip', w);
% loop until valid key is pressed
% if a response is made already, then this loop will be skipped
if ( phase==1 ) % study phase
while (KeyCode(advancestudytrial)==0)
[KeyIsDown, endrt, KeyCode]=KbCheck;
WaitSecs(0.001);
end
end
if ( phase==2 ) % test phase
while ( KeyCode(oldresp)==0 && KeyCode(newresp)==0 )
[KeyIsDown, endrt, KeyCode]=KbCheck;
WaitSecs(0.001);
end
end
% compute response time
rt=round(1000*(endrt-startrt));
% compute accuracy
if (phase==1 ) % study phase
ac=1;
else % test phase
ac=0;
% code correct if old-response with old stimulus,
% new-response with new stimulus, or study phase
if ( (KeyCode(oldresp)==1 && objtype(trial)==1) || (KeyCode(newresp)==1 && objtype(trial)==2) )
ac=1;
end
end
resp=KbName(KeyCode); % get key pressed by subject
% Write trial result to file:
fprintf(datafilepointer,'%i %i %s %i %s %i %s %i %i %i\n', ...
subNo, ...
hand, ...
phaselabel, ...
trial, ...
resp, ...
objnumber(trial), ...
char(objname(trial)), ...
objtype(trial), ...
ac, ...
rt);
end % for trial loop
end % phase loop
% Cleanup at end of experiment - Close window, show mouse cursor, close
% result file, switch Matlab/Octave back to priority 0 -- normal
% priority:
sca;
ShowCursor;
fclose('all');
Priority(0);
% End of experiment:
return;
catch
% catch error: This is executed in case something goes wrong in the
% 'try' part due to programming error etc.:
% Do same cleanup as at the end of a regular session...
sca;
ShowCursor;
fclose('all');
Priority(0);
% Output the error message that describes the error:
psychrethrow(psychlasterror);
end % try ... catch %
% only 352 lines of codes ... :)