1st experiment

Support
1

Hi Chris,

a few hours before your eMail arrived, I got a talk with Stefan, about the example experiments. Good timing :smiley:

He make suggestions for two simple experiments. The first one is is simple orientation discrimination:

A black screen. The subject press the “startTrial”-key.

A fixation-spot appear (but no fixation is required) for a (random/fixed) time (1000 ms):

The screen get black for ca. 500ms (fixed time) and then one of the following three lines will shown for a short time (lets say 250ms):

Then the screen get black and the subject must decide if the line is inclined to the left (answer key-1) or the the right (answer-key 2).
After the the subject press the answer-key the next trial begin (if possible with a random incline(left, zero, right)).

I hope it is not to much work, for building an experiment
 if so, please feel free to contact me (eMail, Skpe, whatEver).
On Monday I will write a description of a second experiment (with eyepositions).

Is it possible, that i can take a first glance on the code? I’m really very inquisitive how the code looks like, and how to implement a plug-in :slight_smile:

Gruß aus Göttingen
& thank you for your work
Ralf

2

Attachments:

3

Hi Ralf,

Thanks for the quick response. Apparently, you and Stefan read my mind :slight_smile:

Unfortunately, I may not have time to get to this today, and I’m going to be out on vacation all of next week (March 29 to April 2). However, putting this together for you will be a top priority when I get back to the office on April 5.

In the meantime, feel free to check out the code, all of which is on GitHub:

The mw_examples repository contains some example experiments you can check out. I’d also recommend watching Dave Cox’s screencast, which shows you how to set up and run a basic experiment:

http://public.coxlab.org/mw/Tutorial1.mov

If you want to start playing with MWorks (which I recommend), you can install a nightly build:

https://mworks.tenderapp.com/faqs/installation/install-mworks

Cheers,
Chris

4

Hi Ralf,

How is your MWorks experimentation going? Have you tried using the nightly build?

I discussed your first demo experiment with Jim. For the most part, it should be easy to implement with MWorks. However, MWorks currently doesn’t support receiving input from a keyboard. This is something we can implement, although it may require writing a new plugin. Alternatively, there’s an existing plugin that supports input via USB gamepads. (I believe the main target is Microsoft’s SideWinder gamepad: Microsoft SideWinder - Wikipedia .) Would that be an acceptable substitute for a keyboard?

Also, when you have a chance, it’d be great if you could send a description of your second experiment (the one with eye positions). We’re still trying to get a handle on what new features need to go into the next release, so I’d like to find out sooner rather than later what we need to add to support your experiments.

Thanks,
Chris

5

Hi Ralf,

Here’s an MWorks implementation of your first example experiment. It uses a game controller instead of a keyboard. To start each trial, the subject presses the “A” button. To indicate whether the line is inclined to the left or right, the subject presses the left or right trigger button, respectively. The experiment runs through 12 trials; the line incline for each trial is random.

You can inspect the experiment by opening the file in MWEditor. Most of it should be relatively straightforward. However, the way the random selection is implemented might be a little confusing, so I’ll try to explain.

The states and transitions for each trial are encapsulated in the “New Task System” item. If you open the “Show line” state and inspect the “Queue line” action, you’ll see that the stimulus to enqueue is defined as an expression:

Lines[stimulus_index]

“Lines” is a stimulus group that contains the three different line stimuli. Which line gets displayed is determined by the value of the variable “stimulus_index”.

You may have noticed that “New Task System” is contained within “New Replicator”. A range replicator is a paradigm component that creates multiple copies of its contents by varying the value of a variable. In this case, “New Replicator” creates three copies of “New Task System”, the first of which has “stimulus_index” set to 0, the second 1, and the third 2. Therefore, each of the three copies of “New Task System” displays a different line incline (zero, left, and right, respectively).

Finally, “New Replicator” is contained within “New Trial”, which is configured to draw 12 samples from its contents (it’s contents being the three different copies of “New Task System”), using the random with replacement selection scheme.

Hopefully this will all make sense, but please feel free to ask me any questions you have. (We can chat over email or IM – whatever is easier).

Cheers,
Chris

6

Attachment: ralf_example_1.xml (10.3 KB)

7

Hi Ralf,

One detail I neglected to mention: The specific game pad model supported by MWorks is the Microsoft SideWinder Plug & Play Game Pad. Although this model is no longer manufactured, it looks like you can buy a used one via Amazon. Alternatively, we can update the plugin to work with (or write a new plugin for) a currently-available controller.

Chris

8

Hi Chris,

we already bought a MS compatible model and it work without any problem :). I will ask my college for the exact name


Gruß Ralf

ps. Sorry for any delay. I’m on Puerto Rico up to the end of the month


9

Hi Chris,

my colleague send me the correct name of our GamePads: http://www.amazon.com/Logitech-963335-0403-Precision-Gamepad/dp/B00030AX3Q/ref=sr_1_1?ie=UTF8&s=videogames&qid=1273837735&sr=8-1

Works fine


Gruß aus PR
Ralf

10

Hi Ralf,

we already bought a MS compatible model and it work without any problem :slight_smile:

Great! I’m glad that the plugin works with more recent hardware. Thanks for the link.

Chris