|
|
|
EDAS II (Eye movement Data Analysis System II) provides three
kinds of eye movement data analysis -
- Text. With text, EDAS II can be used for fixation
sequence analysis and object-based analysis. The analysis of reading
data yields a wide range of word-specific movement measures such
as the size saccades to words, saccade duration, landing position,
skipping rate, regression probability, number of passes through
a word total number of fixations, and various fixation dependent
measures (including first fixation duration, second fixation duration,
gaze duration, total viewing duration, and more).
- Image. With images, EDAS II can be used to obtain
a large number of oculomotor indexes for various to be analyzed
regions. These oculomotor indexes include the size of the saccade
to a region, the duration of the following fixation, the number
of fixations within a region, cumulated viewing time for a region,
the re-viewing of a region after other regions have been viewed,
and more.
- iText - Image based text which is especially usefully for multiple
language studies. Every experiment consists of multiple trials,
each trail has multiple images and each image might have multiple
line text.
|
|
|
|
The
following is a snapshots from EDAS II single line text analysis
|
|
|
|

|
|
|
|
Figure
1. Single Line Text Analysis
|
|
|
|
EDAS II
is well suited for the analysis of multiline data (Text/iText).
With two dimensional recordings fixations often appear above or
below a particular line as in Figure 2 below. EDAS II records but
automatically corrects for intra-line deviations. EDAS II also corrects
systematic vertical shifts as can be seen in Figure 3. Here line
one and line 3 reading are accompanied by systematic upward-directed
drift. Although the vertical difference between the first and last
fixation on line two (fixation 15 and 25, respectively) is larger
than the inter-line distance, fixation 25 is accurately assigned
to line two rather than line one.
|
|
|
|

|
|
|
|
Figure
2. Multiple Line Text Analysis1
|
|
|
|

|
|
|
|
Figure
3. Multiple Line Text Analysis2
|
|
|
|

|
|
|
|
Figure
4. Output spreadsheet for Text analysis (Partial)
|
|
|
|
Features of EDAS II
- Visualization - A graphical depiction of the
sequence of fixations/response during object viewing for each
trial.
- Powerful - Xml based experiment definition can
be extended easily. A very large set of measures including some
rarely reported measures such as pupil size, saccade duration,
and the size micro position-changes during individual fixations.
- Automatic coding of a number of low-level object properties
(trial number, word number in the sentence, word length, punctuation).
- Flexible condition files that map one to five different
experimental conditions on each to be analyzed object.
- Convienient import feature for Text analysis - Generate XML Experiment files from
eye-link source file.
- Intelligent correct for vertical drift when more than one line
is read. EDAS II will parse two horizontal scans with a high degree
of accuracy onto two consecutive lines, even when there is substantial
vertical drift so that horizontal recordings for the two lines
partially overlap.
- Batch process - Provided for both Text and iText analysis which saves lots of
time for re-analyzing experiment response data files for more new output columns.
- Large number of output columns - There are 198 columns output for Text analysis
and 161 columns output for iText analysis.
- Spreadsheet creation of object-based and sequence based eye
movement data.
- Manual correction/reject option for individual trials with corrupted
data.
|
|
|
|
More
samples for multiple line text analysis
|
|
|
|

|
|
|
|
Figure
4. Skimming
|
|
|
|

|
|
|
|
Figure
5. Careful Reading
|
|
|
|
Comments
from our customers
|
|
|
|
"I've been using EDAS II for analyzing a large-scale reading
Experiment: 30 subjects were each reading about 220 pages of text
with each page containing up to 10 lines of text. The EDAS II program
was immensely helpful when analyzing this data set. It has an algorithm
that automatically assigns fixations to words and letters, and it
also automatically corrects for many problems with drift or suboptimal
calibration. The graphical display of eye movements on the stimulus
material provides enormous help in getting a feeling for the data
and allows for easy and convenient by-hand corrections that are
necessary in such experiments (e.g., to mark bad trials, correct
results from the algorithm, ...). Multiple aspects of the data can
be displayed: e.g., fixated line number, fixation number, now also
only a selected subset of fixations can be displayed. Very importantly,
the developers of EDAS II have been VERY (!!!) helpful and constructive
in working together to make EDAS II an even better Software for
analyzing data from eye movement experiments. Not only did they
provide extensive support with setting up the software and constructing
the relevant files to get the analysis running. They were also highly
responsive to include new features into the software when such features
seemed useful to us from an applied perspective: e.g., manual responses
can now easily be displayed with the eye movement data; also, only
subsets of fixations can be displayed which can be a big help if
many fixations lie in one area of the display; and many more and
smaller changes were made in response to our suggestions. Lastly,
EDAS II readily provides a large number of output variables, like
gaze durations, single fixation durations, and many many more. Also
here, the developers were very very helpful in defining new variables
specifically for our purposes. Doing the data analysis for our dataset
without EDAS II would have demanded a very high amount of programming
work (e.g., in Matlab; in particular to deal with bad trials/fixations),
but was quite convenient and feasible through the use of EDAS II.
For this reason, and due to the good experiences that I've had with
the software and the developer team, I advise to use EDAS II for
analyzing data from eye movement experiments, in particular for
multi-line reading experiments." - Daniel Schad, Ph.D candidate
at Psychology Department, Posdam University, Germany
|
|
|
|
"I've used Entroware's EDAS II program for the analysis of eye
movement data that we collected with different types of Eyelink
systems. The program is perfect for my needs: it provides a graphical
interface for the annotated viewing of raw data, it allows me to
select -per mouseclick- trial data that are to be accepted, to be
edited, or to be rejected, it allows me to import experimental condition
files into to be analyzed data sets, and it yields two excel-type
outputs: one in which fixation order determines the ordering of
output data and one in which the order of words in the text determines
the ordering of output data. The word-order spreadsheet includes
a large number of conventional indexes that are frequently used
in the study of reading (such as incoming saccade properties, first
fixation duration, gaze duration, total viewing duration, skipping,
outgoing saccade properties) and some that are rarely used (e.g.,
pupil size during word viewing, intrafixation movement dynamics).
Best of all, EDAS II has been modified in response to user-specific
demands, giving it the feel of a custom-tailed program. For my purposes,
it's a great research tool with superb support." - Albrecht
Inhoff, Professor at Psychology Department, SUNY/Binghamton, United
States
|
|
|