Dear Researchers,
The 2023 Fall Study Grammar Update is almost here! This update contains quite a few new features and improvements that we are really excited about!
Please note:
FindingFive servers (US and EU, in their respective local times) will be updated on the early morning of August 14, 2023 to include the following changes:
1. Reaction time data for Keypress Responses now calculated from previous keypress!
Previously, reaction time data for all keypresses recorded in a Keypress Response were calculated from the onset of the response (when the response is rendered on the trial). Based on feedback, we have decided to instead calculate the reaction time of each successive keypress relative to the preceding keypress, with the first keypress reaction time calculated from response onset.
In other words, if a participant were to strike a key every 500ms, the old reaction time data would look like this: 500, 1000, 1500, etc. It will now look like this: 500, 500, 500, etc.
2. Choice Responses now support dropdown selection!
It is now possible to organize the options of a choice response in a dropdown menu by setting the “mode” property to “dropdown”. Dropdown mode supports feedback and resetting of the choice response, but does not presently support the “multi_select” parameter (this will be made available in a future release).
A big thank you to Dr. Ming Xiang of University of Chicago for requesting this feature!
3. Tokenized Text Stimuli now support bidirectional self-paced reading!
It is now possible for participants to move both forward and backward through tokenized text stimuli. Simply set the new “bidirectional” property to true. This feature is particularly useful for backward-masked and singleton self-paced reading setups.
Please note: This feature is only possible if “self_paced” is set to true. In addition, please see the new “key_reverse” property which supplements the “key_advance” property to support custom key bindings.
A big thank you to Dr. Jeffery Green of BYU Dept. of Linguistics for requesting this feature!
4. Changes to Tokenized Text Stimulus instructions
With the addition of bidirectional tokenized text stimuli, we are deprecating the old “key_advance_instruction” property in favor of a new “hint” property. This aligns key advance instructions across tokenized text and audio stimuli.
Please note: Studies employing the old property should continue to work for now, but we do not actively support deprecated features.
5. A bug fix affecting the recording of Response Mode in the results CSV
A bug affecting how “response_mode” was recorded in the results.csv file of a completed session or study preview from grammar version 3.5.0 has been fixed. The response mode is now correctly recorded.
Questions?
If you have any questions regarding this grammar update, please leave your comments below. Thanks!
The 2023 Fall Study Grammar Update is almost here! This update contains quite a few new features and improvements that we are really excited about!
Please note:
- Do you use the Keypress Response type? If so, please pay special attention to item #1, as it may affect any scripts you have for processing reaction time data.
- As always, you MUST preview your study in full to update to the latest grammar version in order to use these features.
FindingFive servers (US and EU, in their respective local times) will be updated on the early morning of August 14, 2023 to include the following changes:
1. Reaction time data for Keypress Responses now calculated from previous keypress!
Previously, reaction time data for all keypresses recorded in a Keypress Response were calculated from the onset of the response (when the response is rendered on the trial). Based on feedback, we have decided to instead calculate the reaction time of each successive keypress relative to the preceding keypress, with the first keypress reaction time calculated from response onset.
In other words, if a participant were to strike a key every 500ms, the old reaction time data would look like this: 500, 1000, 1500, etc. It will now look like this: 500, 500, 500, etc.
2. Choice Responses now support dropdown selection!
It is now possible to organize the options of a choice response in a dropdown menu by setting the “mode” property to “dropdown”. Dropdown mode supports feedback and resetting of the choice response, but does not presently support the “multi_select” parameter (this will be made available in a future release).
A big thank you to Dr. Ming Xiang of University of Chicago for requesting this feature!
3. Tokenized Text Stimuli now support bidirectional self-paced reading!
It is now possible for participants to move both forward and backward through tokenized text stimuli. Simply set the new “bidirectional” property to true. This feature is particularly useful for backward-masked and singleton self-paced reading setups.
Please note: This feature is only possible if “self_paced” is set to true. In addition, please see the new “key_reverse” property which supplements the “key_advance” property to support custom key bindings.
A big thank you to Dr. Jeffery Green of BYU Dept. of Linguistics for requesting this feature!
4. Changes to Tokenized Text Stimulus instructions
With the addition of bidirectional tokenized text stimuli, we are deprecating the old “key_advance_instruction” property in favor of a new “hint” property. This aligns key advance instructions across tokenized text and audio stimuli.
Please note: Studies employing the old property should continue to work for now, but we do not actively support deprecated features.
5. A bug fix affecting the recording of Response Mode in the results CSV
A bug affecting how “response_mode” was recorded in the results.csv file of a completed session or study preview from grammar version 3.5.0 has been fixed. The response mode is now correctly recorded.
Questions?
If you have any questions regarding this grammar update, please leave your comments below. Thanks!