This is part of SMART-T design: there MUST be at least one stimulus of all three kinds: image, sound and movie.
For convenience, there's at least one of each included in the .zip download.
Recent Forum Posts
There is some kind of bug because of which the movie in the GUI and in the actual text file don't match up. The one in the text file is the one that will actually play.
This bug is being fixed.
in discussion Bugs and requests / Bugs and bug reports » Number of trials for criterion
In designing the conditions for quitting the trial, there is a bug due to which the last 'n' trials for the three currently implemented conditions (Total looking, First look and T-test) all have to be the same. So, for example, you cannot compute Total looking over the last 5 trials and First look over the last 8 trials. Putting different number of trials for the three conditions will typically result in the computation for the maximum of these numbers.
This bug is being fixed even as you're reading this…
in discussion Bugs and requests / Bugs and bug reports » Missing Structures and Phases (Known Bug)
Bug
Sometimes (often!) in the Structure Attributes or Quit Phase panes, all the Structures and Phases don't appear in the drop-down menu.
Workaround
Go to the Structures(/Phase) pane and add a new structure(/phase). Then go back to the Struct Attributes (/Quit Phase) and set the structure(/phase) attributes, and then go to the Structures(/Phase) pane and delete the new structure(/phase).
SMART-T is designed to be fairly flexible, and can therefore be used for paradigms beyond anticipatory measures. This thread points to a few tips that expand the kinds of paradigms one can think of.
1) Presenting a single image per trial, contingent on looking at a central fixation point
- Set a background image with a fixation image (e.g. image of a cross) at the central fixation point
- Set node1 to 1 pixel below the center, set the image size to 0 and set the node as Wait Attention
- Set node2 to the central pixel, size to 1 (as defined in the Structure pane) and optionally add a loom for the desired duration
2) Presenting a single movie per trial
- The same as before, except the target object stays at size 0 at node2, which is immediately followed by the 'reward movie'
3) Timing control
- You can introduce delays or stopping at particular nodes by adding a loom that 'grows' to 100%, for the desired duration
4) Object (dis)appearance control
- You can make an object disappear by setting the Object(Image) Size Ratio (Path pane) to 0. NB: this interacts appropriately with the looms - e.g., a 150% increase in loom of a 0 size object is still 0.
5) Timing note
- If two nodes are too close to each other, SMART-T might skip one because of temporal interpolation. To deal with this, simply change the Object(Image) Speed Ratio between the two nodes in the Path pane.
1) Randomization sets (Random set pane) are only for the sounds.
2) You can define sets of windows are correct or incorrect for each Structure - SMART-T counts a data frame as being correct if it includes a look in any of the correct windows. NB: If you have overlapping windows, SMART-T will count a single frame as contributing only a single data point (to the correct, incorrect or other set of windows).
3) Reward movies (Structures pane) always plays after the last node of the path for that structure.
4) Wait time (Phase pane) is defined by two numbers. If they are the same, SMART-T introduces a fixed ISI for that duration. If they are different, SMART-T chooses a random number between the two as the ISI duration for each trial.
