Data Generation Wizard

Generating Data

o Known data patterns can be generated using File > Data Generation Wizard.

o Several predefined patterns can be selected.

o To customize a pattern, switch to Custom.

o Patterns are added to an experiment (that can be simultaneously defined) with groups, conditions, subjects as successive trials. Excess trials can be deleted.

Predefined patterns exist for a Custom pattern, One Stroke, Two Loops, Four Loops, Eight Loops, or a Diagonal Line

For the Custom settings you can control the following settings

o Stroke Velocity Profile:

~ Harmonic: Sine wave pattern

~ Minimum Jerk: Quintic spline per stroke (6*t**5 - 15*t**4 + 10*t**3). This jerk measure used to minimize is the integral of the squared 3rd time derivative. Jerk signifies here the total changes in force -- 2nd time derivative). The minimum jerk pattern is maximally smooth.

o Device Settings: Simulating the noise of the tablet. Other tablet settings are borrowed from the Device Settings

~ Noise Amplitude (cm): Add random (white) noise to each sample. Noise amplitude is the maximum amplitude. The corresponding SD is noise amplitude / sqrt(12) = amplitude * 0.289. There is virtually no quantization as the noise has 2000 steps from peak to peak. The quantization Noise levels of digitizers are 0.001 to 0.1 cm.

~ Round Raw Data: The output data are calculated in digitizer coordinates, using the device settings of resolution and sampling frequency. However, the calculated data are not integers as received from the digitizer but rational (real) data. By rounding the data to integers, quantization noise can be simulated, consistent with the digitizer resolution.

o Total Movement: Both x and y data use the following settings (See also extracted features per trial).

~ Start Time (s): Time to start the movement

~ Stroke Duration (s): Duration per up or down stroke

~ Number of Strokes: Number of consecutive up and down strokes

~ Trail Time (s): Time between the end of the movement and the pen lift. The pen pressure remains above the minimum pen pressure during the movement.

o X-Movement Component of the Pattern: Settings related to the horizontal movement component.

~ Position Start (cm): x-position where the movement starts

~ Size of Stroke (cm): x-size of the strokes

~ Start Phase (deg): Phase of the x negative cosine wave. 90 Degrees means that the horizontal component goes first left for half a stroke, then right for the second half of the stroke, the further left for the first half of the second stroke, then right again for the second part of the second stroke. An upright circle, ellipse, or loop result. To produce a forward slatned ellips, use 60 degrees.

~ Velocity (cm/s): Constant x-velocity comparable to the left-to-right translation of the hand during writing. This turns circles into loops.

o Y-Movement Component of the Pattern: Settings related to the vertical movement component.

~ Position Start(cm): y-position where the movement starts

~ Size of Stroke(cm): y-size of the strokes

~ Start Phase(deg): Phase of the y negative cosine wave. 0 degrees means that the vertical component starts at Position Start and increases to Position Start + Size of Stroke

Submovement Generation

Submovements can be generated in 2 ways, by a sudden standstill of the movement during or by a fluent variation of the speed of time.

~None: No simulated submovement added to stroke data.

~Delay around end of primary submovement(s): Method 1 requires specification of a nonzero duration in seconds where x and y movement are suddenly halted during each stroke. The total stroke duration will increase by this delay. The brief stand-still of movement in the raw data will result in an aritficial separation between the primary submovement and the secondary submovement.

~Sinusoid time warping amplitude (0=No warping, 1+=Warping reversal): The "speed of time" is fluently changing from faster to slower during each stroke, depending upon the warping amplitude. The total stroke duration will remain the same. When set to 1, there will be an infinitesimal time that the time and thus the movement stands still. Values larger than 1 will result in a reversal of movement.

~Duration of primary submovement(s): This duration needs to be shorter than the total stroke duration. The duration is calculated from the start time of the stroke. The moment the movement stands still is centered around the end of the primary submovement.

o All data analyses can be performed and the accuracy of MovAlyzeR can be verified using data with known underlying movement features.

o Data generation can be used to estimate the effects of random noise and/or digitizer quantization noise upon the accuracy of the experimental design, including number of trials and number of subjects.

Select Experiment, Group, Condition and Subject to which the generated data will be appended as a trial.