Once an experiment has been configured in Habit, you can test and/or run the experiment. To do this, select the experiment in the main Habit window and click the Run button (shown right):
Before running the experiment Habit will first check your experiment’s configuration. You must indicate whether you are running a real subject or simply testing. You must also specify the order(s) for stimuli in each configured phase of the experiment.
Checking your experiment’s configuration
Habit will test your configuration to make sure the experiment can run. The following tests are performed:
If any of these tests fails the experiment will not run, and you will be presented with a dialog (shown right):
The Show Details button will explain what settings must be corrected (shown right):
Specifying stimuli order for each phase
After all configuration errors (if any) are corrected, Habit asks that you specify, for each phase of the experiment, the order in which the stimuli are presented (shown right):
For each phase you can select one of two types of stimuli orders. The default order simply uses the stimuli in the same order that they were entered (the same order you see if you view the stimuli settings for this experiment/phase). If you added pre-specified orders to your experiment’s configuration, you may click the Select radio button, and the combo box at the right will be enabled, allowing you to select one of the orders you created.
You may also click the Randomize checkbox (using either Default or Select orders) and choose a randomization method. With Simple Randomization, each stimulus is chosen at random from the set of configured stimuli for the phase. With Block Randomization, stimuli are chosen in such a way that all stimuli will be presented once before any stimulus is shown a second time.
Note that when you configure more stimuli than there are trials for a given phase, your subject will not be presented with all stimuli. In this case it is preferable to configure specific orders so that experimenters may simply select an order at run time.
Also, during the habituation phase the subject may become habituated before seeing all stimuli. Recall that for the habituation phase the number of trials is a maximum; habituation phase will stop once the subject’s looking meets the habituation criteria.
Test run or Actual Subject?
The run settings dialog also allows you to indicate whether you are running an actual subject. If so, uncheck the Test Run? checkbox and enter information to identify the subject data files. You may also enter the name of the observer and comments or notes. The subject ID (or the word “Test”) will become part of the results filenames.
Starting the experiment and data collection
Once the run settings are determined, the Run button is enabled. When you click the Run button, the experiment’s Control Panel is displayed (shown right):
The dialog shows the name Habit will attach to data files. The name consists of the experiment name, the subject ID (or “TEST”), and the date and time. Once you click the Start Trials button the first configured phase will commence and stimuli will be displayed. The various parts of the dialog will display status information on the progress of the experiment as the experiment proceeds.
Habit requires input, either from the keypad of the computer running the experiment, or from an eye tracker (eye tracker integration is not yet available), to determine the subject’s looking behavior. The subject’s looking behavior will, in turn, determine when trials start and end, whether the subject habituates, etc.
If your experiment uses an attention getter, then it will be displayed prior to each trial.
The Control Panel will indicate that an attention getter is playing (shown right):
Habit will play the attention getter until you indicate Attention by hitting the <Keypad-Enter> key or the <Space> bar.
➡ On the Mac keyboard, the keypad has an Enter key which is distinct from the Return key. Habit requires either the <Enter> or <space>; the <Return> key is intercepted by the window system and cannot be used by Habit.
After you indicate that the subject’s attention is directed towards the screen(s), Habit will display the stimulus for the current trial. The Control Panel will indicate that a stimulus is playing:
TODO: The information displayed in the Control Panel is controlled by the control bar options under the General settings in your experiment configuration. This behavior is currently not implemented, and the Control Bar Options are ignored. This will be corrected in an update to Habit soon.
If you are using keypad look detection (as opposed to an eye tracker), you should indicate when the subject is looking by pressing the keypad keys:
<Keypad Enter> or <space> : Attention - proceed with trial
<Keypad 4> : Subject is looking at the left stimulus
<Keypad 5> : Subject is looking at the center stimulus
<keypad 6> : Subject is looking at the right stimulus
Conversely, indicate that the subject has stopped looking by releasing the key. Habit will record the time when the key was pressed and released and will process the data accordingly.
Note that Habit will ignore looking keys that are not appropriate for the type of stimulus layout you have specified for the experiment. In other words, if your experiment is set to use a single stimulus, Habit will expect that you use the “5” key when the subject is looking at the stimulus; Habit will ignore the “4” and “6” keys.
A trial may be skipped by pressing the Next Trial button on the Control Panel. This button should be used with caution: If you hit Next Trial during the display of a stimulus, then the trial is skipped, not repeated. This can skew your results and can make habituation impossible.
When the experiment is complete, Habit displays a summary page (see below). The results are automatically saved; the summary is a convenience. Note that you can print the summary page as well. The page also indicates whether the subject habituated (see “habituation success”) or not. This summary page can be reproduced later as well (see Viewing Habit Results).