Condition Consistency Checking

Consistency Checking

o In the case of ScriptAlyzeR and MovAlyzeR experiments, the consistency of each trial with specified criteria can be verified.

o The result of the verification is stored in the Consistency Error file. Inconsistent trials will receive a red trial mark and will not be included in the Summarize and Analysis steps. Use the default setting for minimal or no consistency checking.

o Stroke Description: See the stroke description in the next section.

o Minimum Strokes: The minimum number of strokes per trial.

NOTE: This is also the number of strokes that is actually processed for consistency checking.

If there are fewer number of strokes than the minimum value, there are two options during the processing of the trials:

1. If in Experiment settings (for the experiment corresponding to the trials) > processing > consistency checking, the option 'Discard trial if number of strokes is out of range' is ON, then the trial is discarded.

2. Else, the error is ignored and a warning is generated in the output file.

o Maximum Strokes: Maximum number of strokes in a trial.

If there are a greater number of strokes than the maximum value, there are two options during the processing of the trials:

1. If in Experiment settings (for the experiment corresponding to the trials) > processing > consistency checking, the option ''Discard trial if number of strokes is out of range'' is ON, then the trial is discarded.

2. Else, the additional strokes are ignored and a warning is generated in the output file.

o Strokes to Skip: The number of initial strokes to be skipped for the processing of a trial per condition.

EXAMPLE: If there is a trial with 15 strokes, where you want to have the first 3 strokes skipped during consistency checking, then set Strokes to skip = 3. The resulting consistency checking and analysis will be done only for strokes 4-15. The feature extraction file will still contain data for all the strokes.

o Stroke Length (cm): The user can specify the stroke length of a target in a trial per condition. This is used for a condition with only one stroke per trial.

o Range Length (cm): Range length is used to specify +/- value for the stroke length.

- Conversion Calculator: You can use this tool to convert from inches to centimeters (and vice versa) and also degrees to radians (and vice versa).

NOTE: If the recorded stroke length for any stroke is not within the specified +/- range stroke length, the trial is discarded.

o Stroke direction (radians): The user can specify the target direction of a stroke in a trial per condition. Certain directions can be specified by using the stroke descriptions, for example, '/' for 45 degrees.

If you need to use any other target directions other than the ones that can be specified by a stroke description, you have to use this option. For example, 120 degrees +/- 5 degrees.

o Range direction (radians): Range Length is used to specify +/- value for the stroke direction.

NOTE: Setting the length and direction values zero will remove the restrictions.

EXAMPLE: Set stroke length = 5 cm, range length = 1 cm, stroke direction = 1.57 (90 degrees), and range direction = 0.26 (~15 degrees).

In a trial in which there is one stroke of length = 5.2 cm and direction= 95 degrees then the trial is NOT discarded. Consider another trial, in which there is stroke of length = 6.5 cm and direction = 90 degrees. This trial is discarded during consistency checking (Process/reprocess trials) as the stroke length is out of the specified range.

o Check for target sequence: This option allows the consistency checking on targets and discard trials that do not meet the specified target sequence. The required target order is set in the imperative stimulus for the condition > properties > target use/sequence list.

EXAMPLE: If the order of targets is 1,2,3,4 and during the trial the order produced is 2,1,3,4, the trial is discarded and the first occurences of the error is reported.

NOTE: Since the targets reached are registered during recording and this is a new feature from Version 3.0, trials recorded with versions prior to 3.0 cannot be tested for target sequence.