Information Technology Reference
In-Depth Information
Fig. 3. Comparison of the experimental question for the CE method and the bisection method
Furthermore, identical numerical values were presented in both conditions. The left
number was constantly zero, while the right number included 15 different
values that were placed around 1000. The inner number varied between seven catego-
ries, containing the exact arithmetic mean C of the two outer numbers as well as
ranges of 50 units (C+50, C-50), 150 units (C+150, C-150) and 300 units (C+300, C-
300) from the center. All numerical values were multiplied by the factors 1, 10 and
100. Thus, not only values within the range of 1000 resulted, but additional values
within the range of 10000 and 100000 were also induced.
Color codification and response configuration of the YES/NO answers to the
right and left index finger were randomized among the subjects. The duration of the
experiment was approximately one hour per session. The adjustment of the two
conditions and the different number values during the experiment were distributed
randomly.
In order to examine the stimulus presentation of the inner number, a fragmentation
into the seven categories and the following YES/NO responses was conducted. The
NO answers, in this respect, represent negative deflections from the center position C,
and the YES answers positive deflections.
For analyzing the behavioral data only the averaged YES answers were taken into
account, given that the NO answers are equivalent.
2.3 EEG Recording and Analysis
The electroencephalogram was recorded from 29 thin electrodes mounted in an elastic
cap and placed according to the international 10-20 system. The EEG was re-
referenced offline to the mean activity at the left and right mastoid. In order to enable
the offline rejection of eye movement artifacts, horizontal and vertical electrooculo-
grams (EOG) were recorded using bipolar montages. All channels were amplified
(bandpass 0.05 - 30 Hz) and digitized with 4 ms resolution, impedances were kept
below 10 kΩ.
EEG-data with eye blinks were corrected using a blind source separation method
[19]. EEG-periods with pulse artifacts were excluded from the data set. Following the
artifact procedure, stimulus-locked bins were calculated (epoch length 900 ms, base-
line 100 ms) for each subject and the following conditions (see table 1).
Search WWH ::




Custom Search