DMDX
COMPONENTS
|
timedx |
Sets
and checks software and hardware features for running dmdx |
| dmdx |
Presents
stimuli and records responses |
| utility
programs |
The
following three programs prepare data for analysis:
UnloadAZK: unload data from different computers into a single
data file
Analyze: initial processing of raw data, trim data, assign items
to conditions, compute reaction time and error means for subject and item,
Concatenate: collaps data from different presentation lists of
the same experiment into a single data set.
The following three programs convert file formats used by DMDX and
an earlier DOS version DM
azk2dtp:converts .azk output files to binary .dtp
format
dtp2azk:converts .dtp files to .azk format
img2bmp:converts dmtg .img files to Windows .bmp files
|
DMDX
PROCEDURES
|
A.
Download, install, & test dmdx |
1.
If your computer does not have Directx 5.0 or later, download Directx 7.0
into the folder Program files\dierectx\setup\; double click on it; it will
install itself.
2. Download dmdx package; create a dmdx folder under Program files, unzip
dmdx into this folder;
3. Run Timedx;
4. Test and learn what dmdx can do by running file "features.rtf"
in the demos folder. |
| B.
Write item files |
1. Use Wordpad for
writing item files and save them in rich text format (.rtf).
2. If you use two or more counterbalanced lists, name them as ch1a, ch1b,
etc.
Use file transparent file names so that you remember which file is for
what purpose. My .rtf file names begin with my initial jn, then two letters
for the project, then experiment #, then presentation list if counterbalanced.
For example, jnmf1a, jnmf1b, jnmf2a, jnmf2b.
3. If you are testing both native speakers and nonnative speakers, they
should have separate item files. If there are more than one group of non-native
speakers, e.g., Spanish ESL students and Chinese ESL students, you also
need to have separate item files for them. This way, the data from each
group will be saved in separate data (.azk) files. You can always combine
them into a single data set at a later stage. You don't want to have data
from different groups in the same .azk file and then have to separate
them manually or even worse forget which data is from whom.
|
| C.
Run experiment |
1. Start dmdx,
2. Locate the item file,
3. Begin the experiment,
4. Save the data (data will be added to the same .azk file every time
when you test a participant)
Before you recruit participants and before they arrrive,
make sure
a. all your item files are ready by testing yourself from the beginning
to the end, and they are already on the computer you will use,
b. get the consent form ready and you have enough copies,
c. get the instructions ready,
d. get the language learning background questionnaire ready if you test
non-native speakers,
e. reserve the lab.
Once the participant is in the lab, make sure
a. you ask him or her to read and sign the consent form first,
b. you ask them to read the instructions and give them opportunites
to ask questions,
c. you ask them to complete the questionnaire if there is any,
After the participant is done,
a. save the data,
b. complete exit interview or debriefing if necessary,
c. answer all the questions they have,
d. pay the participants and ask them to sign on the payment form,
e. or give them credit in the GSU
experiment scheduling system.
|
| D.
Prepare data for analysis |
1. Combine data files
for the same presentation list on different computers;
2. Initial processing
of data (compute subject and item means for different conditions);
3. Collapes data from different presentation lists of the same experiment.
|
| E.
Simplified Flowchart of DMDX Programs and
Files | Detailed Flowchart |
|
ANALYZE
DATA
|
|