Viewing Trials

File

Extension

How to view

Notes

Raw data per trial

.HWR

Right click on the trial > View Numerical Data > View Raw Data

The raw data shows the x, y and z position per sampling point in three columns, including the pen-up points. The z-coordinate for the pen-up points will be zero.

Raw data per trial
(GripAlyzeR)

.FRD

Right click on the trial > View Numerical Data > View Raw Data

The raw data shows the x (Lower grip), y (Upper grip) and z (Load force) data of the sampling points in three columns.

Processed data per trial

.TF

Right click on the trial > View Numerical Data > View Processed Data

View Time Function (TF) Data shows the numerical and calibrated time functions and several important parameters.

Segment data per trial

.SEG

Right click on the trial > View Numerical Data > View Segment Data

Displays the number of segmentation points and the time (sec) at which they occur.

Extracted data per trial

.EXT

Right click on the trial > View Numerical Data > View Extracted Data...

View Extract File shows the sequence of extracted features for each trial and segment (stroke = 1 segment, for no submovement and = 3 segments with submovement - MovAlyzeR ONLY).

Note: This file contains details of all the trials of the same condition as the trial being viewed. Navigate by trial number.

Consistency errors per trial

.ERR

Right click on the trial > View Numerical Data > View Consistency Error Data

View Error File shows the reason for a consistency error compared to the target movement. The criteria can be set in the Stroke Description when creating the experiment conditions.

Note: This file contains details for all the trials in the same condition as trial being viewed. Navigate by trial #.

Consistent data per condition

.CON

Right click on any trial for that condition > View Numerical Data > View Consistency Data

Consistent data file shows the same information as the .EXT feature extraction file only for trials that passed consistency checking.

To view the data corresponding to all the good trials per condition, go to Experiment settings for that experiment > Summarize > Select the subjects and then open the .CON file.

Note: This file will contain details for all the trials in the same condition as trial being viewed. Navigate by trial #.

Discontinuities per trial

.DIS

Right click on trial > View Numerical Data > View Discontinuities

Discontinuity is a sudden jump in the recorded data. The jump can be in time or distance between samples.

Word Extraction data per condition

.WRD

Right click on trial from which words were extracted > View Numerical Data > View Word Extraction Data

Word Extraction can be used to automatically extract individual words from longer handwriting paragraphs. The .WRD file has the following format

Word number: Sample numbers in original .HWR file | Reason for being detected as a new word

# tvyzero

Number of segmentation points and the time in sec at which each point occurs.

The time in seconds corresponding to the zero-crossing velocity points.

# tayzero
(MovAlyzeR ONLY)

Number of submovement points and the time in sec at which each point occurs.

If you do a sub-movement analysis, there is another set of data for the sub-movement points. The sub-movement points correspond to the point after the peak velocity for a segment.

tseg(sec)
(GripAlyzeR ONLY)

Always shows five segmentation points. The load force as a function of time is segmented into 4 phases or strokes ( 5 segmentation points)

Refer to segmentation procedure for detailed description of the segmentation method

Start time relative to the start of the recording  

Start time is the start time per stroke, that is the time in seconds from the beginning of the recording (= Sample 0) till the beginning of the stroke (= Beginning segmentation point). StartTime together with Duration of the stroke will tell when exactly the stroke was performed. When setting the start of the recording at any rate, the recording will being when the imperative stimulus begins. StartTime of Stroke 1 is Reaction Time (RT).

Straightness error or the normalized standard deviation from a straight line. The estimation is calculated as follows:

~ Estimate the length of the straight line [cm]
~ Fit a straight line
~ Estimate the (perpendicular) distances of each point to the fitted line [cm]
~ Estimate the standard deviation of the distances [cm]
~ Divide by the length of the line between begin and end [cm]

FORMULA: StraightErr = (1/length) * sqrt (Sum (y(t) - yd(t))**2 / N)

REFERENCE: Contreras-Vidal, J.L., Teulings, H.L., & Stelmach, G.E. (1998). Elderly subjects are impaired in spatial coordination in fine motor control. Acta Psychologica, 100, 25-35.

Surface or the area of the loop enclosed by the previous and present stroke in cm**2. The surface is not normalized. If the the crossing does not occur within the previous stroke, although a loop has been formed, the loop area will be zero.

So, if within an apparent loop a segmentation point is detected, the loop may fail to be detected. If data are insufficiently filtered, extremely small loops are sometimes detected where no loops should be detected. This feature can be verified by measuring (in cm on paper) the long and the short axes of ellipse-like loops. The loop area should be 3.1416 * (0.5 * short axis) * (0.5 * long axis).

Ratio of penup duration to total duration

The value ranges from 0 to 1. 0 corresponds to no pendown and 1 corresponds to all pendown and the values in between show the ratio.

Dysfluency measure, theoretically independent of stroke duration and size and marginally dependent upon stroke shape. Normalization can only be done per stroke and is not meaningful for multiple strokes or for submovements. Normalized Jerk is unitless as it is normalized for stroke duration and size. It is primarily meant for goal-directed movements but can also be applied in repetitive, reciprocal strokes, drawing stroke sequences, or handwriting stroke sequences. Strokes can be segmented by vertical/tangential velocity zero crossings or by absolute velocity minima at points of sharp curvature. Normalized jerk is dependent upon the choice of the low-pass filter. Normalized jerk increases with increasing duration because the limited-bandwidth motor system has more options to generate dysfluencies. A maximally smooth, straight harmonic strokes yield a value of 7.75 (pi**3/2**2) and perfectly circular, constant-velocity strokes yield a value of 10.96 (pi**3/2**1.5).

PROCEDURE:
o Lowpass filter, estimate velocity [cm/s] and jerk [cm/s**3]
o Segment into stroke from from near-zero velocity to the next near-zero velocity
o Estimate stroke size [cm]
o Estimate stroke duration [s]
o Squared jerk[cm**2/s**6]
o Integrate over time [cm**2/s**5]
o Divide by distance**2
o Multiply by duration**5
o Square root (to reduce dynamic range and to make it proportional with jerk).
o Divide by 2 (convention)

FORMULA: sqrt (0.5 * Sum (jerk(t)**2) * duration**5 / length**2).

RELATED MEASURES:
o Stroke Duration
o Number of acceleration peaks per stroke
o Frequency of secondary submovement per stroke
o Relative duration of secondary submovement
o Relative size of secondary submovement

REFERENCE: Teulings, H.L., Contreras-Vidal, J.L., Stelmach, G.E., and Adler, C.H. (1997). Coordination of fingers, wrist, and arm in Parkinsonian handwriting. Experimental Neurology, 146, 159-170. 

'OK' if the trial is good

Otherwise, reason for which the trial was discarded.