PosterPDF Available

Equivalence of a Continuous EEG Discrimination Task to Standard Operant Control Training

Authors:

Abstract

Biofeedback is commonly believed to train increased awareness and voluntary control over physiological processes that would otherwise remain unconscious and involuntary (Frederick, in press; Olson, 1987; Plotkin, 1981). Brener (1974) argued that repeated pairing of external feedback with internal afferents related to the response lead to the awareness and learning of a “response image,” which allowed control of the response without external feedback. However, relatively little research has examined the relationship between awareness and control of physiological states. Most clinical and experimental biofeedback research has emphasized operant control, by rewarding subjects for producing desired physiological states. Discrimination training is another form of operant conditioning, that rewards subjects for increased awareness, that is, for correctly reporting their physiological state. For instance, the first historical report of operant conditioning of the EEG was a discrimination task in which subjects were prompted to guess whether they were in a high or low alpha amplitude state, and immediately informed if their response was correct (Kamiya, 2011). Since similar kinds of self-monitoring and evaluation have been found to facilitate learning of motor skills (Boutin, Blandin, Massen, Heuer, and Badets, 2014; Kolovelonis, Goudas, & Dermitzaki, 2011), we hypothesized that combining discrimination training with standard neurofeedback would facilitate learning of operant control of EEG alpha (8–12 Hz) amplitude. However, Heim, Dunn, Klein, Powers and Frederick (2016) argued that Kamiya’s “discrete” discrimination paradigm resulted in very slow learning because it provides relatively few trials per minute. The present study examined the effect of a continuous discrimination task (CDT) on learned operant control of EEG alpha. In the CDT, participants manipulated a controller indicating their subjective rating of their alpha amplitude on a 10-point scale. The changing pitch of a tone (presented about 120/min) represented the absolute difference between the participant’s rating and their actual alpha amplitude. One group (n = 9) received seven 40-minute sessions of standard operant control training to increase and decrease alpha in eight alternating 5-minute runs. The CDT-mixed group (n = 9) received only half of that training per session, instead performing the CDT during the 11–20th and 31–40th minutes of each session. Performance in the alpha operant control task was defined as the percent average difference between the increase and decrease conditions. The CDT group performed nonsignificantly higher than the control group (mean session improvement over first session baseline 14.0% vs. 6.8%, one-tailed t(16) = 0.822, p = 0.22, d = 0.39). Although these results do not demonstrate that the CDT facilitates learning of voluntary control of EEG alpha, the fact that the CDT can be substituted for standard operant control training during half of the session time without a deficit in learning suggests that these two tasks are functionally equivalent and may have substantial similarity in the skills required. While results in this sample of mostly normal subjects show a small effect size, the CDT may have some clinical utility to provide additional motivation to clients who show deficient attention toward standard neurofeedback tasks.
Equivalence of a Continuous EEG Discrimination
Task to Standard Operant Control Training
Jon A. Frederick1, Casey J. Klein2, Andrew S. Heim2, Kelli N. Dunn2, and Julia Schreiber2
1
SCldS Ui i SCldMN
2
Middl
TS
Ui i
Mf b TN
Biofeedback is commonly believed to train
increased awareness and voluntary control over
physiological processes that would otherwise remain
idil (Fdiki
Methods Discussion
1
S
t.
Cl
ou
d
S
tate
U
n
i
vers
i
ty,
S
t.
Cl
ou
d
,
MN
,
2
Middl
e
T
ennessee
S
tate
U
n
i
vers
i
t
y
,
M
ur
f
rees
b
oro,
TN
Introduction Results
These results fail to support the hypothesis that
continuous discrimination training would facilitate
learning of voluntary control of EEG alpha. It is,
th l f i t r t th t th CDT r p did t
EEG was recorded using a BrainMaster Atlantis
amplifier and BrainMaster 3.7i software. A 150 sec
recording was taken at the start of each session to establish
Performance in the continuous discrimination task was
defined as the average absolute difference between the
subjective rating and the actual alpha amplitude on a ten
Prfr i th lph pr t tr lt k
unconsc
i
ous an
d
i
nvo
l
untary
(F
re
d
er
i
c
k
,
i
n press;
Olson, 1987; Plotkin, 1981). Brener (1974) argued
that repeated pairing of external feedback with
internal afferents related to the response lead to the
awareness and learning of a “response image,” which
allowed control of the response without external
feedback. However, relatively little research has
examined the relationship between awareness and
none
th
e
l
ess, o
f
i
n
t
e
r
es
t
th
a
t
th
e
CDT
g
r
ou
p
did
no
t
perform worse (they performed nonsignificantly
better) than the control group, despite substituting
half of the training time with the CDT. This
observation suggests one of two possible
interpretations: (1) there are diminishing returns for
training sessions longer than 20 minutes; or (2) the
CDT trains similar skills or facilitates learnin
g
of
a baseline and measure the peak alpha frequency (PAF; the
highest median amplitude between 8–12 Hz. The alpha
band was defined as a 5 Hz band surrounding the PAF.
For instance, if PAF = 9, the alpha band was 7–11 Hz.
Participants sat in a reclining chair with their eyes closed
throughout each session.
One group of received seven 40-minute sessions of
standard operant control training to increase and decrease
o
n
sca
e.
P
e
rf
o
r
mance
i
n
th
e a
lph
a o
p
e
r
an
t
con
tr
o
l
t
as
k
was defined as the percent average difference between the
increase and decrease conditions. Subjects improved in
both tasks over the course of seven sessions (figures 1 and
2). The average within-subject correlation between the
two tasks over seven sessions was –0.23 (one-tailed one-
sample t(8) = 2.72, p = .026, d = .90. However, the
between sub
j
ect correlation between avera
g
e session
examined
the
relationship
between
awareness
and
control of physiological states.
Most clinical and experimental biofeedback
research has emphasized operant control, by
rewarding subjects for producing desired
physiological states. Discrimination training is another
form of operant conditioning, that rewards subjects
fi d hif l
g
voluntary control just enough that the CDT is a
functionally equivalent form of training.
Discrimination training may have benefits other than
facilitation of voluntary control (Frederick, in press),
such as increasing client motivation and engagement
in the session; greater insight about the subjective
correlates of brainwave states; and different measures
fith ftiiFillthi
standard
operant
control
training
to
increase
and
decrease
alpha in eight alternating 5-minute runs. The experimenter
adjusted a threshold to maintain 15–30% time in reward
based upon a 60 sec event trend window. Nine subjects in
this group completed 7 sessions, and an additional 6
subjects completed 1–6 sessions.
The CDT-mixed group received only half of that
training per session, instead performing the CDT during
jg
performance in the two tasks was not significant (r =
0.26, df = 7).
The CDT group performed nonsignificantly higher than
the control group in the voluntary control task (mean
session improvement over first session baseline 14.0% vs.
6.8%, one-tailed t(16) = 0.743, p= 0.23, d= 0.35).
Performance of the CDT had no within-session effect
pf fth l t tltkThti th
f
or
i
ncrease
d
awareness, t
h
at
i
s,
f
or correct
l
y
reporting their physiological state. For instance, the
first historical report of operant conditioning of the
EEG was a discrimination task in which subjects were
prompted to guess whether they were in a high or low
alpha amplitude state, and immediately informed if
their response was correct (Kamiya, 2011).
o
f
assess
i
ng
th
e success o
f
t
ra
i
n
i
ng.
Fi
na
ll
y,
thi
s eyes-
closed CDT paradigm provided feedback with a nine
musical note scale and input with two keys indicating
higher and lower. A different paradigm providing
visual feedback and using a joystick or dial
potentiometer input may interface better with human
factors influencing learning and performance.
the 11
20th and 31
40th minutes of each session. In the
CDT, participants manipulated a controller indicating their
subjective rating of their alpha amplitude on a 10-point
scale. The changing pitch of a tone (presented about
120/min) represented the absolute difference between the
participant’s rating and their actual alpha amplitude. Nine
subjects in this group completed 7 sessions, and an
additional 7 subjects completed 1
6 sessions.
on
p
er
f
ormance o
f
th
e vo
l
un
t
ary con
t
ro
l
t
as
k
.
Th
a
t
i
s,
th
e
performance during minutes 21–30 compared to minutes
1–10 of the voluntary control task (mean of 7 sessions)
was 6.1% lower in the control group, and 6.0% lower in
the CDT group (who performed the CDT during minutes
11–20).
Acknowledgement
Since similar kinds of self-monitoring and
evaluation have been found to facilitate learning of
motor skills (Boutin, Blandin, Massen, Heuer, and
Badets, 2014; Kolovelonis, Goudas, & Dermitzaki,
2011), we hypothesized that combining discrimination
training with standard neurofeedback would facilitate
learning of operant control of EEG alpha (8–12 Hz)
li d
dik i
References
Boutin, A., Blandin, Y., Massen, C., Heuer, H., & Badets, A. (2014).
Conscious awareness of action potentiates sensorimotor learning. Cognition,
133(1), 1–9.
Brener, J. (1974). A general model of voluntary control applied to the
phenomena of learned cardiovascular change. In Obrist, P., Black, A. H.,
Brener, J., & DiCara, L. V. (Eds.), Cardiovascular psychophysiology. Chicago:
additional
7
subjects
completed
1
6
sessions.
Acknowledgement
We are grateful to BrainMaster Technologies, Inc. for
providing outstanding technical support along with the
Atlantis EEG system used in this research.
amp
li
tu
d
e. However, Fre
d
er
i
c
k
, He
i
m, Dunn, Powers,
and Klein (in press) argued that Kamiya’s “discrete”
discrimination paradigm resulted in very slow learning
because it provides relatively few (about 3) trials per
minute. Therefore, in the present study, we examined
the effect of a continuous discrimination task (CDT),
which provided about 120 trials per minute, on
learned operant control of EEG
alpha.
We
Aldine.
Frederick, J. A. (2012). Psychophysics of EEG Alpha State Discrimination.
Consciousness and Cognition, 21(3), 1345–1354.
Frederick, J. A. (in press). EEG state discrimination and the phenomenal
correlates of brainwave states. In Collura, T. F. & Frederick, J. A. (Ed.),
Handbook of Clinical QEEG and Neurotherapy. New York: Routledge Taylor &
Francis.
Frederick, J. A., Heim, A. S., Dunn, K. N., Powers, C. D., & Klein, C. J. (in
press). Generalization of skills between operant control and discrimination of
EEG alpha. Consciousness and Cognition.
Kamiya, J. (2011). The first communications about operant conditioning of
th EEG
Jlf
Nthp
15
(1) 65
73
www.postersession.com
learned
operant
control
of
EEG
alpha.
We
hypothesized that participants who performed the
CDT during half of the session time would show
superior performance in controlling their alpha
amplitude compared to participants who received
only the standard neurofeedback control task. Figure 2. Percent difference between alpha enhance and
inhibit conditions, relative to first session baseline.
.
Figure 1. Average absolute difference between subjective
rating and actual alpha amplitude on a ten point scale.
th
e
EEG
.
J
ourna
l
o
f
N
euro
th
era
p
y
,
15
(1)
,
65
73
.
Kolovelonis, A., Goudas, M., & Dermitzaki, I. (2011). The effect of different
goals and self-recording on self-regulation of learning a motor skill in a
physical education setting. Learning and Instruction, 21(3), 355–364.
Olson, P. (1987). Definitions of biofeedback. In Schwartz, M. S. (Ed.),
Biofeedback: A practitioner’s guide (pp. 33–38).New York: The Guilford Press.
Plotkin, W. B. (1981). A rapprochement of the operant-conditioning and
awareness views of biofeedback training: The role of discrimination in
voluntary control. Journal of Experimental Psychology: General, 110(3), 415–428.
ResearchGate has not been able to resolve any citations for this publication.