Effective test taking: Skills-acquisition versus anxiety-reduction techniques

Abstract

A skills-acquisition treatment for test anxiety was compared with 2 anxiety-reduction conditions, cue-controlled relaxation and meditation, and a practice-only group in terms of improved test performance. 50 highly test-anxious Ss representing the most anxious 28% of 232 undergraduates administered the Achievement Anxiety Test were randomly assigned to 1 of 4 experimental conditions. The 3 treatment groups received 5 90-min training sessions while the practice-only group was given an equal amount of time in practice on a posttreatment analog test (AT). Results indicate that the skills-acquisition group was superior to the other 3 conditions in terms of GPA and AT performance. The skills-acquisition group reported greater knowledge of effective test-taking skills and less attentional interference during testing than the other groups. Measures of heart rate and pulse transit time during the AT showed no significant group differences. It is suggested that the term "test anxiety" be retired and that inadequate test performance be reconceptualized as ineffective test taking. (35 ref)

Full text

Journal
of
Consulting
and
Clinical Psychology
1980,
Vol.
48, No. 4,
431-439
Effective
Test Taking: Skills-Acquisition
Versus Anxiety-Reduction Techniques
Karl
Kirkland
and
James
G.
Hollandsworth,
Jr.
University
of
Southern Mississippi
A
skills-acquisition
treatment
for
test anxiety
was
compared
with
two
anxiety-
reduction
conditions, cue-controlled relaxation
and
meditation,
and a
practice-
only
group
in
terms
of
improved test
performance.
Fifty
highly
test-anxious sub-
jects representing
the
most anxious
28% of 232
undergraduates were randomly
assigned
to one of the
four
experimental conditions.
The
three treatment groups
received
five
90-min
training sessions while
the
practice-only group
was
given
an
equal
amount
of
time
in
practice
on a
posttreatment analogue test. Results indi-
cated that
the
skills-acquisition group
was
superior
to the
other three conditions
in
terms
of
performance
on the
analogue
test
as
well
as
grade point average.
In
addition
it was
determined that
the
skills-acquisition group reported greater
knowledge
of
effective
test-taking skills
and
less attentional
interference
during
testing
than
the
other
groups.
Measures
of
heart rate
and
pulse transit time dur-
ing
the
analogue testing situation failed
to
reveal
significant
group
differences.
It
is
suggested that
the
term test anxiety
be
retired
and
that inadequate test per-
formance
be
reconceptualized
as
ineffective test taking.
The
term test anxiety
has
been used
for
the
last
3
decades
to
denote
a
constellation
of
behaviors that have
a
debilitating
effect
on
academic performance (Pagano
&
Katahn,
1972).
The
nature
of
test anxiety
has
been
characterized
as
multifaceted
and
inclusive
of
task-irrelevant
cognitions, heightened
physiological
arousal,
and
inefficient
study
behavior (Spielberger, Anton,
&
Bedell,
1976).
As
the
term implies, most
researchers
have conceptualized test anxiety
in
terms
of
excessive arousal involving anxiety-based
behaviors. Accordingly, systematic
desensi-
tization,
a
technique designed
to
inhibit
anxiety-evoking
imagery
and
excessive
physiological responding,
has
been
a
major
treatment approach used
to
reduce
test
anx-
iety.
However,
the
appropriateness
of
relax-
ation-based procedures
as the
treatment
of
choice
has
been questioned
for at
least
two
The
authors
wish
to
thank Thomas
Desporte,
Richard
Kazelskis,
and
Barry Pipkin
for
their help with this
study.
Requests
for
reprints should
be
sent
to
James
G.
Hol-
landsworth, Department
of
Counseling Psychology,
University
of
Southern Mississippi,
Box
8272 Southern
Station, Hattiesburg, Mississippi 39401.
reasons. First,
the
ability
of
systematic
desensitization
and
other anxiety-reduction
techniques
to
improve test
performance,
the
ultimate
criterion,
has
only been weakly
demonstrated.
It has
been noted that
67% of
those studies involving systematic desensi-
tization have failed
to find
differences
between treatment
and
control conditions
on
performance
measures
of
test anxiety
(Denny,
1978). Second, several investigators
have questioned whether high levels
of
physiological arousal
are in
fact
related
to
less
effective
test-taking behavior
(Deffen-
bacher,
1978). Hollandsworth, Glazeski,
Kirkland,
Jones,
and Van
Norman (1979),
for
example, demonstrated that high
and
low
test
anxiety students were very similar
in
terms
of
their physiological arousal. These
results were later
replicated
on a
more
extensive
scale
by
Glazeski
and
Hollands-
worth (Note
1).
Holroyd, Westbrook, Wolf,
and
Badhorn (1978) also found
that
high
and
low
anxious students could
not be
reliably
differentiated
on the
basis
of
physiological
arousal.
These
findings
lend support
to
Wine's
(1971) attentional model
of
test anxiety
and
suggest that perhaps
the
term itself
is
mis-
Copyright
1980
by the
American
Psychological
Association,
Inc.
0022-006X/80M804-0431$00.75
431

432
KARL
KIRKLAND
AND
JAMES
HOLLANDSWORTH,
JR.
leading.
Rather than being viewed
as an
anx-
iety-related disorder,
ineffective
test
per-
formance
can be
conceptualized
as a
skills
deficit
(Kirkland
&
Hollands
worth,
1979).
The
corresponding emphasis when taking this
perspective would
be on the
acquisition
of
effective
test-taking behaviors rather than
on
inhibiting
an
undesirable
set of
responses
(McFall,
1976).
In
fact, given
the
proposal
by
several investigators that arousal
may be
highly
facilitative while taking
a
test (Def-
fenbacher,
1978;
Hollands
worth
etal.,
1979),
the
most
appropriate
treatment
may be one
that
(a)
focuses
on
providing
effective
test-
taking
skills
and
(b)
attempts
to
utilize
arousal
as a
means
of
enhancing test per-
formance.
Treatments focusing
on
attentional skills
have been reported
in the
literature. Mei-
chenbaum
(1972)
found
that
a
self-instruc-
tional training procedure, when combined
with
systematic desensitization,
was
supe-
rior
to
desensitization alone
in
terms
of
improving performance
in a
testing
situation.
Little
and
Jackson (1974) used attentional
training
with seventh-
and
eighth-grade stu-
dents. Their performance measures, how-
ever, appeared
to be
equally
affected
by all
three treatments (attentional training, relax-
ation training,
and
attentional training plus
relaxation)
and a
placebo-expectancy
con-
dition.
Consequently,
no
conclusions
can be
drawn
from
their study
in
terms
of
perform-
ance.
Studies
in the
literature
that
report
changes
in
performance measures have
been
the
exception rather than
the
rule.
Denny
(1978)
has
noted that
of the
studies
in
which
performance measures have been
used,
the
treatments
involving
cognitive
coping
techniques have demonstrated
the
greatest success.
The
superiority
of the
studies that involve training subjects
to
emit
adaptive self-statements would
further
sup-
port Wine's attentional model
of
test
anxiety.
This
same
pattern
of
favorable
results
for
cognitively
based
techniques
is
found
when
examining
studies that report
significant
increases
in
academic
grades,
the
most rele-
vant
criterion
of
all. Denny
and
Rupert
(1977),
Holroyd (1976),
and
Meichenbaum
(1972)
all
found
that treatments
involving
active, cognitive coping strategies were
superior
to
traditional, somatically based
relaxation conditions
in
this respect.
Although
one
study
has
found
that both
a
desensitization
and
a
modeling treatment
resulted
in
better
final
course grades than
either
a flooding or
control condition
(Home
&
Matson,
1977),
the
preponderance
of
evidence suggests that treatments
de-
signed
to
increase
the use of
adaptive cog-
nitive
responses
may be the
single most
important
treatment component. Neverthe-
less,
the
cognitive
coping
strategies
noted
above
all
emphasize
the
reduction
of
physio-
logically
cued anxiety. Even Holroyd's
attention-focusing
procedure sought
to
eliminate anxiety-engendering thoughts
in
addition
to
increasing
the
frequency
of
anx-
iety-relabeling statements
and
attention-
focusing
self-instructions.
The
question that
arises
is, How
important
was the
anxiety-
reduction
focus
of
these treatments
in
effecting
these impressive results?
This
question,
coupled
with
the findings
that suggest
a
reappraisal
of the
role
of
phys-
iological arousal
in
test-taking situations,
leads
one to ask how
much
emphasis,
if
any,
one
needs
to
place
on the
anxiety compo-
nent
of
test-taking behavior.
The
present
study
was
designed
to
investigate this ques-
tion
by
evaluating
the
relative effectiveness
of
a
skills-acquisition approach
to
test
taking,
which placed
no
emphasis whatso-
ever
on
anxiety
reduction,
as
compared with
two
anxiety-reduction
techniques,
cue-con-
trolled relaxation
and
noncultic meditation,
both
of
which have been reported
in the
literature
as
appropriate treatments
for
test
anxiety
(Counts,
Hollands
worth,
&
Alcorn,
1978;
Fabick,
1977).
This study
was
also designed
to
observe
the
effects
of
different
treatments
on
differ-
ent
variables. Current thinking suggests that
there
is a
high
degree
of
response
specificity
and
patterning
of
bio-behavioral systems
(Davidson, 1978; Davidson
&
Schwartz,
1976).
For
example
one
cannot assume that
attentional measures
are
equally responsive
to
both self-instructional training
and
muscle
relaxation, since
one is a
cognitive
and the
other
a
somatic intervention. Thus, while
the
test-taking skills treatment
was
expected

EFFECTIVE TEST TAKING
433
to
result
in
superior
test
performance,
cue-
controlled, deep muscle relaxation
was an-
ticipated
to
affect
somatically
based mea-
sures
to a
greater
degree
than
the
mind-
calming
technique,
meditation, which
in
turn
was
expected
to
produce more
of an
effect
on
the
more cognitively based variable.
Method
Subjects
Undergraduate students
(N =
232)
in
introductory
sociology
and
psychology courses
at a
large south-
eastern university were administered
the
Achievement
Anxiety
Test (Alpert
&
Haber,
1960). Several authors
have
recommended
the use of
this instrument
in
screen-
ing
for
test
anxiety (Huck
&
Jacko,
1974;
Kirkland
&
Hollandsworth,
1979).
A
difference
score (debilitative
minus
facilitative
score)
was
computed
for
each
subject.
From this
population,
subjects scoring
in the
upper
28%
of
the
distribution
(high
debilitative/low
facilitative test
anxiety)
were contacted
and
invited
to
participate
(n
=
65).
Five subjects declined
and 60
subjects agreed
to
participate
in the
experiment. Random assignment
of
these subjects
following
pretreatment assessment
resulted
in an n of 15 for
each
of the
four
experimental
conditions.
Attrition during
the
experiment
was
evenly
distributed
across
the
four
conditions
and
resulted
in
final
n
s of 13
each
for the
skills-acquisition
and
cue-con-
trolled relaxation groups
and
12
for
both
the
meditation
and
practice groups.
Measures
Cognitive-attentional
measures. Four self-report
measures
of
cognitive
and
attentional interference were
administered
to all
subjects.
Particular
care
was
taken
in
selecting measures that would
reflect
the
relative
degree
of
involvement
of the
cognitive
and
attentional
response systems.
To
obtain
measures
of
cognitive
response
involve-
ment,
the
following
measures were used:
the
Worry
scale
of the
Worry-Emotionality
Scale
(WES-W;
Liebert
&
Morris,
1967),
the
Cognitive
scale
of the
Cognitive-Somatic
Anxiety Questionnaire (CSAQ-C;
Schwartz, Davidson,
&
Goleman, 1978),
and the
Cog-
nitive
Interference Questionnaire
(CIQ;
Sarason, Note
2).
The
directions
for
the
CSAQ
were slightly reworded
to
make
it a
situationally specific measure
of
test anx-
iety.
A
measure
of
attentional interference,
the
Atten-
tional
Interference
Scale
(AIS),
developed
by
Glazeski
and
Hollandsworth (Note
1), was
administered
follow-
ing
the
anagram task.
Somatic
measures.
Two
self-report
measures
reflecting
somatic responding were administered
to all
subjects before
and
after treatment.
These
measures
were
the
Emotionality
scale
of the
Worry-Emotionality
Scale
(WES-E;
Liebert
&
Morris,
1967)
and the
Somatic scale
of the
Cognitive-Somatic
Anxiety Ques-
tionnaire
(CSAQ-S;
Schwartz
et
al.,
1978).
In
addition
heart
rate
and
pulse
transit
time
were
monitored during
the
anagram task using
a
Cyborg BL907 pulse wave
velocity
and
heart rate instrument. Heart rate (HR)
was
chosen
as a
dependent measure because
it has
been
widely
studied
and is
closely associated with emotion-
ality
(Lang, 1977; Valins, 1966). Pulse transit time
(PTT),
a
recent development
in the
assessment
of
car-
diovascular activity,
has
been
found
to be
consistently
related
to
systolic blood pressure and, like
HR,
appears
to be
sensitive
to
sympathetic
influences
on the
myo-
cardium
(Gribbin,
Steptoe,
&
Sleight, 1976;
Obrist,
Light, McCubbin, Hutcheson,
&
Hoffer,
1979).
Measures were obtained
by
placing
an
acoustic
motion
sensor
(low frequency microphone) over
the
radial
artery (distal portion, anterior surface)
and a
second sensor over
the
superficial temporal
artery.
HR,
reported
in
beats
per
minute,
was
averaged over every
five
beats
and was
obtained
by
detecting
the
pulse wave
at the
radial site.
Differences
in
arrival times
at the two
sensing
sites resulted
in
transit times that were updated
with every
heart
beat.
In
that
transit time
is
inversely
dependent
on
pressure alterations, reductions
in
blood
pressure were indicated
by a
lengthening
of the
PTT.
Readings
for
both
HR and PTT
were transmitted
directly
to a
Cyborg
Q740
digital panel printer, which
generated output
at
10-sec
intervals.
These values were
averaged over
1-min
blocks
for
analysis.
The
last
2
min
of
the
20-min adaptation period were averaged
to
pro-
vide
a
baseline value
for
each measure.
Difference
scores were calculated
by
subtracting this baseline
from
each
of the 10
1-min block
values
occuring
during
the
anagram task itself.
Test-taking
skills
measure.
The
Exam Behavior
Scale
(BBS)
of
Brown's (1975)
Effective
Study Test
was
administered
to all
subjects before
and
after
treatment.
Kirkland
and
Hollandsworth (1979)
found
that exam
behavior,
as
reflected
by
this
scale,
was a
better
pre-
dictor
of
academic performance than measures
of
other
study
behaviors
or
test anxiety.
Performance
measures.
Three performance mea-
sures
were
included
in the
present
study. Cumulative
grade point averages (GPAs) were collected
from
the
University
Registrar
for the
term preceding treatment
and for the
term during which treatment occurred,
All
subjects
were administered
the
Otis-Lennon Mental
Abilities
Test,
Advanced Level (O-L; Otis
&
Lennon,
1968) before (Form
J) and
after (Form
K)
treatment.
In
addition,
all
subjects were administered
a
difficult
10-
item
anagram
test
following treatment. Five-
and
six-
letter anagrams were preselected
for
difficulty
through
pilot
testing
prior
to the
study with
students
who
were
not
subjects
in
this
study.
Treatments
Two
master's level psychology students served
as
trainers
for the
three experimental conditions. These
individuals
received
training
from
the first
author
and
remained unaware
of the
purposes
and
hypotheses
of
the
study
until
after
treatment sessions
had
been termi-
nated.
The
three treatment conditions
uniformly
utilized
the
training components suggested
by
Meichenbaum's
(1977)
stress inoculation training,
for
example, educa-

434
KARL
KIRKLAND
AND
JAMES HOLLANDSWORTH,
JR.
tion,
practice,
and
application phases.
The
amount
of
practice
was
held
constant
across
all
four
conditions
so
that
at
posttesting
the
practice-only group
had as
much
practice with
the
anagram task
as the
treatment groups.
Each
of the
three treatment groups
met
with
one of
the
trainers
for five
90-min
sessions. Identical training
formats
were maintained across experimental condi-
tions
and
involved
the
presentation
of a
rationale
for
treatment,
practicing
the
respective techniques,
and
application
of
skills
using
two
practice anagram
tests.1
Skills
acquisition.
Training
in
skills
acquisition
defined
test anxiety
as a
specific skills
deficit.
As
such, train-
ing
focused
on
deficits
in
three
specific
areas: effective
test-taking
strategies, adaptive self-instructional state-
ments,
and
attentional-control skills.
No
emphasis
was
placed
on
arousal
or
anxiety reduction. Rather,
it was
stressed that
the
acquisition
of the
above
skills
would
allow
the
person
to use his or her
arousal
effectively
to
increase
performance.
Strategies
for
effective
test taking were based
on the
work
of
Robinson (1946)
and
included such methods
as
surveying
the
length
of the
test, seeing
if
certain sections
count
more
or
require more time than
others,
answering
only
one
item
at a
time,
and
marking harder items
to
return
to
later.
Self-instructional
training
was
based
on
the
findings
of
Hollandsworth
et
al.
(1979) that
identified
two
basic kinds
of
self-instructions
for the
effective
test
taker: on-task statements
(e.g.,
"I
have plenty
of
time;
read
the
questions carefully")
and
positive self-
evaluations
("I
will
perform well
on
this
test because
I
am
well
prepared").
Attentional-control training
was
based
on the
appropriate
use of
self-instructions
in
relation
to the
task
as
suggested
by
Wine (1971)
and
Meichenbaum
(1972, 1977). Thus,
the
direction
of
attention
was
mediated
in
training
by the
appropriate
use
of
self-instructions
such
as "I
will
think about that
later;
now
back
to the
test."
Cue-controlled
relaxation.
Subjects
in the
cue-con-
trolled
relaxation condition were presented
with
a
rationale
for
treatment that characterized test anxiety
solely
in
terms
of
excessive autonomic arousal.
As a
result,
the
treatment they were
to
learn
was
described
as
being
designed
for the
control
of
excessive
somatic
reactions. Training
in
cue-controlled relaxation
was
based
on the
model
of
Russell
and
Sippich (1973, 1974),
wherein
subjects learned
to
pair
a cue
word with exha-
lation
as
part
of a
progressive, deep muscle relaxation
procedure.
The cue
word calm
was
used
by all
subjects.
Meditation.
Subjects
in the
meditation
condition
were given
a
rationale
for
treatment that characterized
meditation
as a
technique that deals directly with
the
problem
of
academic
underachievers,
that
is,
intrusive
cognitive
anxiety
or
worry over performance. Medita-
tion
was
presented
as a
technique
for
calming
the
mind.
Training
in
meditation
was
based
on the
model
of
Benson's
(1975)
relaxation response technique.
The
word calm also served
as the
mantra
or
neutral device
for
the
meditation group.
Practice
only.
Subjects
in the
practice-only condi-
tion
received
no
instructions
and
were
not
exposed
to
any
form
of
systematic
treatment
during
the
experi-
mental
period. Practice-only subjects, however,
did
engage
in
practice anagram tests that were identical
to
those used
in the
application phase
of
treatment
for the
experimental
conditions. Therefore,
all
groups received
identical
amounts
of
exposure
to the
anagram task, which
was
used
at
posttreatment
as a
performance
measure.
Experimental
Design
In
accordance with
the
points noted
in the
introduc-
tion,
several
a
priori hypotheses were formulated. First,
it
was
hypothesized
that
in
terms
of the
test-taking
skills
and
performance
measures,
the
skills-acquisition group
would
be
superior
to the
practice-only condition
and
the
two
anxiety-reduction treatments combined.
As far
as the
cognitive-attentional measures were
concerned,
it
was
hypothesized that
the
group using meditation,
a
cognitive
treatment,
would
experience
less
attentional
interference
than that using
the
relaxation condition,
a
somatic treatment.
On the
other hand,
it was
hypothe-
sized
that
the
relaxation condition group would exhibit
less somatic reactivity than
the
meditation group
as
determined
by
using self-report
and
physiological mea-
sures.
All
pre-
and
posttreatment measures (O-L, WES,
CIQ,
CSAQ,
BBS,
and
GPA) were analyzed using
analysis
of
covariance
with
the
premeasure
as the
covar-
iate.
The
post-only measures
(HR,
PTT,
AIS
items
and
anagram performance score) were analyzed using
analysis
of
variance.
A
priori
/
tests
were
used
for
the
planned comparisons, whereas post
hoc
compari-
sons were made using
the
Newman-Keuls
method.
The
level
of
significance
was set at .05 in all
cases.
Results
Test-Taking
Skills
Results
indicated
that
the
skills-acquisi-
tion
group
reported
significantly
greater
knowledge
of
effective
test-taking
behavior
as
measured
by the
BBS
than
either
the
practice-only
group,
t(46)
=
3.775,
p <
.005,
or
the
anxiety-reduction
treatments
com-
bined,
t(46)
=
4.402,
p <
.0001.
Post
hoc
comparisons
revealed
that
the
skills-acquisi-
tion
group
was
superior
to the
three
other
groups,
which
did not
differ
among
them-
selves,
F(3,
45) =
2.851,
p
<
.05.
The
means
and
standard
deviations
for all of the de-
pendent
measures
are
reported
in
Table
1.
Cognitive-Attentional
Measures
Planned
comparisons
between
the
cue-
controlled
relaxation
and
meditation
groups
for
the
cognitive-attentional
measures
(WES-W,
CSAQ-C,
and
CIQ)
did not
gen-
erate
significant
results
except
for the
aver-
age
score
from
the
CIQ. (Individual
items
from
the CIQ and AIS
were
not
used
in the a
priori
comparisions.)
It was
found
that,
as
1
Copies
of the
treatment manuals
are
available from
the
authors
on
request.

EFFECTIVE TEST TAKING
435
predicted,
subjects
in the
meditation
group
reported fewer interfering thoughts
on the
average than
did the
cue-controlled relaxa-
tion group,
f(46)
=
1.830,
p <
.05.
Post
hoc
comparisons
for the
cognitive-
attentional measures revealed that
all
three
treatment groups were superior
to the
prac-
tice-only condition
in
terms
of
cognitive
interference
as
measured
by the
WES-W,
/r(3f
45)
=
4.380,
p <
.009; that there were
no
significant
differences
in
terms
of the
CSAQ-C
(p >
.05);
and
that
the
skills-
acquisition group reported
significantly
fewer
interfering thoughts than
the
practice-
only
group
but not
significantly
fewer than
the
other
two
treatment
groups,
which
in
turn
did not
differ
significantly
from
each
other
or
from
the
practice-only condition,
F(3,
45) =
5.309,
p <
.004.
When
individual items
from
the
CIQ
and
AIS
were analyzed,
it was
found
that three
items
from
both
the CIQ and AIS
generated
significant
F
ratios.
Post
hoc
comparisons
revealed
that
the
skills-acquisition group
reported
that they thought less frequently
about
the
level
of
their ability (CIQ Item
7)
than
the
other
three groups, which
did not
differ
from
each other, F(3,
45) =
4.871,
p
<
.005.
In
addition
the
skills-acquisition
and
relaxation groups
reported
that they
thought less
frequently
about
how
much time
was
left
(CIQ Item
4)
than
the
practice-only
group,
but
they were
not
significantly
differ-
ent
from
the
meditation group, which
in
turn
was not
significantly
different
from
the
prac-
tice-only
group,
F(3,45)
=
5.664,p
<
.003.
The
third
CIQ
item
("I
thought about
how I
should work more carefully") generated
a
significant
F
ratio,
F(3,
45) =
3.578,
p <
.02,
but
failed
to
differentiate between
groups when
the
post
hoc
procedure
was
used.
In
terms
of the
individual items
from
the
AIS,
the
skills-acquisition
and
relaxation
groups
reported
that
they
thought
less
fre-
quently about
how
hard
each
item
was
(AIS
Item
1)
than
did the
meditation group. How-
ever,
these
two
groups were
not
signifi-
cantly
different
from
the
practice-only
group, which
in
turn
was not
significantly
different
from
the
meditation group, F(3,
46)
=
3.666,
p <
.02.
The
skills-acquisition
group
reported
that they thought
less
fre-
quently than either
the
meditation
or
prac-
tice-only groups about
how
poorly they were
doing
(AIS Item
4). The
skills-acquisition
group,
however,
was not
significantly
dif-
ferent
on
this item
from
the
relaxation group,
which
in
turn
was not
significantly
different
from
the
other
two
groups,
F(3,46)
=
4.321,
p
<
.04.
Finally,
the
skills-acquisition group
reported that their degree
of
concentration
on
each item (AIS Item
7) was
higher than that
of
the
practice-only group. However
the
skills-acquisition group
was not
significantly
different
on
this item
from
the
relaxation
and
meditation
groups,
both
of
which
in
turn
did
not
differ
significantly
from
the
practice-
only
condition,
F(3,
46) =
2.861,
p <
.05.
The
means
and
standard deviations
for
these
items
are
reported
in
Table
1.
Somatic Measures
Planned comparisons between
the
cue-
controlled relaxation
and
meditation groups
for
the
somatic measures
(WES-E,
CSAQ-
S, HR, and
PTT)
did not
generate
significant
results except
for the
CSAQ-S.
It was
found
that,
as
predicted,
the
meditation group
reported higher levels
of
somatic anxiety
during
testing than
the
relaxation group,
?(46)
=
1.781,
p <
.05. However, analysis
of
variance
of the
CSAQ-S
as
well
as the
WES-E
failed
to
yield
a
significant
F
ratio.
Difference
scores
for HR and PTT
were
analyzed using
a
Groups
x
Minute-Interval
repeated measures design. Neither
signifi-
cant group main
effects
nor
Group
x
Inter-
val
interaction
effects
were
found
for
either
physiological measure. However,
a
signifi-
cant interval
effect
was
noted
for
PTT, F(9,
414)
=
2.121,
p <
.03, indicating
a
signifi-
cant increase over baseline
in
blood pres-
sure (reduction
in
PTT) during
the
1st
minute
followed
by a
return
to
baseline during
the
2nd
minute
and
stabilizing
for the
remaining
8
minutes
at a
level indicating
a
somewhat
reduced
blood
pressure.
This
finding
sug-
gests
the
presence
of a
mobilization-adapta-
tion
response (cf.
Blatt,
1964)
on the
part
of
most subjects
as
they worked
on the
ana-
gram
task.
Performance
Measures
Planned comparisons between
the
skills-
acquisition group
and the
practice-only

436
KARL
KIRKLAND
AND
JAMES HOLLANDSWORTH,
JR.
group
and the two
anxiety-reduction groups
combined
yielded
significant
results
for two
of
the
three performance measures.
Although
significant
/
values were
not
obtained
for the
O-L,
the
skills-acquisition
group
was
found
to be
superior
to
the,
prac-
tice-only
group,
f(46)
=
3.775,
p <
.005,
and
anxiety-reduction groups combined,
t(46)
=
4.402,
p <
.0001,
on the
anagram
task. Perhaps more importantly,
t
tests
on
the
adjusted
GPA
group means revealed
that
the
skill-acquisition group
was
superior
to
both
the
practice-only group,
t(46)
=
4.959,
p <
.0001,
and the
anxiety-reduction
groups combined,
?(46)
=
3.122,
p <
.005.
Post
hoc
comparisons
following
the
sig-
nificant
F
ratios generated
by the
anagram
task,
F(3,
46) =
8.094,
p <
.002,
and
GPA,
F(3,
34) =
6.943,p
<
.001, revealed that
in
each
case
the
skills-acquisition
group out-
performed
the
three other groups. These
groups,
in
turn,
did not
differ
significantly
from
each other.
Discussion
The
skills-acquisition treatment, which
viewed
test anxiety
as a
skills
deficit
rather
than
an
anxiety-based disorder
and
which
involved
training
for
effective
test-taking
behaviors, resulted
in
significant
improve-
ments
in
academic performance, whereas
the two
anxiety-reduction techniques, cue-
controlled relaxation
and
meditation,
and a
practice-only condition
did
not.
On the
ana-
gram test,
a
task with which subjects
in all
Table
1
Pre-
and
Postgroup
Means, Standard Deviations,
and
Post
Hoc
Comparisons
for
the
Dependent Variables
Treatment condition
S-A"
CCRa
Mb
P-O"
Measure
M
SD
M
SD
M
SD
M
SD
Test-taking
skills
EBS
Pre
Post1'
Cognitive-attentional
WES-W
Pre
Postc
CSAQ-C
Pre
Postc
CIQ
avg.
Pre
Post'
CIQ 3
Pre
Post
CIQ 4
Pre
Post'
CIQ 7
Pre
Post0
AIS
1
Post
AIS
4
Post
AIS
7
Post
17.46
18.83a
9.08
6.22a
15.15
11.68
2.75
1.69a
2.69
1.60a
3.31
2.04a
2.39
1.58a
2.69a
2.77a
4.08a
3.10
2.67
4.22
1.66
4.29
2.91
.93
.58
1.14
.84
1.49
.88
1.21
.84
1.03
1.17
1.19
16.31
17.46b
11.31
7.3
la
12.31
13.33
2.65
2.27a,b
2.77
2.42a
3.31
2.27a
2.15
2.70b
3.00a
3.62a.b
3.54a,b
2.58
2.15
3.75
3.26
4.14
3.65
.66
.51
.97
1.00
1.44
1.12
1.10
.82
1.22
1.12
.97
16.75
16.62b
9.67
7.21a
16.58
12.41
2.59
2.02a,b
3.33
1.71a
3.67
2.86a,b
2.50
2.45b
4.17,
4.08b
3.67a,b
1.69
1.97
3.82
3.39
5.06
3.42
.76
.55
1.37
.80
1.31
.76
1.04
.87
1.11
1.16
.65
16.17
16.90b
9.08
10.14,,
13.58
13.17
2.51
2.30b
2.83
2.52a
3.58
3.38b
1.75
2.75b
3.42a,b
3.92b
2.92b
2.33
2.96
2.60
2.99
4.05
3.41
.59
.43
1.07
.96
.95
.86
.72
.85
1.31
1.31
1.08
(table continued)

EFFECTIVE
TEST
TAKING
437
Table
1
(continued)
Measure
Treatment
condition
S-A"
CCRa
M"
P-Ob
M
SD
M
SD
M
SD
M
SD
Somatic
WES-E
Pre
Post0
CSAQ-S
Pre
Postc
HR
Post"
PTT
Poste
Performance
O-L
Pre
Post0
ANA
Post
GPA
Pre
Post0
9.23
6.08
14.85
10.53
.28
2.85
49.31
51.29
6.08a
2.26
2.42a
1.80
2.41
5.55
5.11
7.91
15.83
16.21
16.23
2.66
.77
.70
9.08
6.64
12.69
11.39
-1.21
-2.02
46.69
48.18
3.92C
2.38
2.29,,
2.92
2.24
3.17
2.27
5.13
13.52
11.43
13.26
2.33
.38
.34
8.75
7.23
14.33
13.23
.41
.84
52.42
49.77
2.50b
2.20
2.36b
2.77
2.65
4.11
2.69
8.33
8.62
14.10
15.42
2.20
.60
.53
9.42
8.90
14.08
12.44
3.80
4,62
43.58
49.89
2.17b
2.33
2.23b
1.98
5.17
4.82
2.84
9.69
14.01
15.55
15.50
1.47
.61
.67
Note.S-A
=
skills-acquisition;
CCR =
cue-controlled
relaxation;M
=
meditation;
P-O =
practice-only;
BBS
=
Exam Behavior
Scale;
WES-W
=
Worry-Emotionality
Scale,
Worry
scale;
CSAQ-C
=
Cognitive-Somatic
Anxiety Questionnaire, Cognitive
scale;
CIQ
Avg.
=
average
score
from
Cognitive Interference
Questionnaire;
CIQ
3,4,7
= CIQ
Items 3,4,7;
AIS
1,4,7
=
Attentional Interference
Scale
Items 1,4,7;
WES-E
=
WES, Emo-
tionality
scale;
CSAQ-S
=
CSAQ, Somatic
scale;
HR =
heart
rate;
PTT =
pulse
transit
time;
O-L =
Otis-
Lennon Mental Abilities
Test;
ANA =
anagrams
test;
GPA =
grade
point
average.
Subscripts
indicate
post
hoc
comparison following
significant
F
test;
means with different subscripts
are
significantly different from
each
other,
"
n
= 13.
b
«
=
12.°
Adjusted means.
d
Difference
score
in
beats
per
minute.
e
Difference
score
in
milliseconds.
four
conditions
had had
equal amounts
of
practice,
the
skills-acquisition subjects
solved
significantly
more anagram problems
under
stress-inducing testing conditions than
did
subjects
in the
other three groups.
The
failure
to
obtain performance differences
on
the
Otis-Lennon
Mental Abilities Test
may
suggest
that skills training should
be
designed
as
much
as
possible
to
meet
the
specific
demands
of the
test
to be
taken,
and
that
one
cannot assume that these skills will
automatically
generalize
to
tests employing
different
formats
and
response strategies.
This
suggestion
would,
at first,
seem
to
con-
tradict
the
speculation that self-instructional
training
tends
to
enhance generalization
of
training
effects.
This apparent contradiction
may
be
resolved, however, when
it is
con-
sidered that
the
cognitively based treatments
reported
in the
literature appear
to be
more
effective
when dealing with
the
more per-
vasive
forms
of
anxiety (Goldfried, 1977)
than
with
specific
cognitive operations
involving
symbol manipulation
and
calcula-
tions. Just
as
Denny
(1978)
has
noted
the
need
for
more work
to
determine
the
"domains
of
effectiveness"
of
various pro-
cedures,
it may be
that
in the
present study,
the
skills-acquisition strategy
was
able
to
demonstrate increases
in
performance
on
specific
tasks
at the
expense
of
generaliza-
bility.
Further inspection
of the
results obtained
from
the
other dependent variables indicated
that
the
skills-acquisition group reported
significantly
more knowledge
of
effective
test-taking behaviors
at
posttesting than
the
other three groups did.
In
addition, subjects
in
the
skills-acquisition group reported that
they
thought
less
frequently
about their level
of
ability, less often about
how
much time
was
left,
less frequently about
how
hard

438
KARL
KIRKLAND
AND
JAMES HOLLANDSWORTH,
JR.
each item was,
and
less often about
how
poorly they were doing. Also they reported
that they were
able
to
concentrate more
on
each
item.
For
each
of
these variables,
the
skills
group
was
found
to be
superior
to at
least
one
other
condition, whereas none
of
the
remaining treatments produced results
that were superior
to
those
of the
skills group
on
any
measure. Consequently, these
find-
ings
not
only demonstrate that
the
skills-
acquisition group outperformed
the
other
groups
but
they also give
us
some indication
of
how
this performance
was
achieved.
The
results
do
provide some support
for
differential
treatment
effects.
However,
these weak
findings,
which suggest that
the
relaxation treatment group experienced less
somatic anxiety
and the
meditation group
fewer
interfering thoughts during testing,
are
of
relatively
less
importance when
it is
con-
sidered that
the
skills-acquisition group
did
equally
as
well
as
either group
on
both
of
these measures.
The
failure
to find
significant treatment
effects
for the
various physiological mea-
sures
can be
interpreted
as
further
confirma-
tion
of the
relative unimportance
of
somatic
reactivity
as a
factor
in
ineffective
test tak-
ing.
In
other
words,
a
lack
of
change
on the
physiological variables
did not
obstruct sig-
nificant
gains
in
performance
by
subjects
receiving
the
skills-acquisition training.
Also, this failure
to find
differences cannot
be
attributed
to the
insensitivity
of the
mea-
sures, given that
PTT
clearly demonstrated
a
significant
interval
effect
suggesting
a
mobilization
and
adaptation response dur-
ing
testing
for
most subjects.
The
failure
of the
cue-controlled relaxa-
tion
treatment
to
affect physiological
responding
may be
explained
by the
reports
at
debriefing that
the
technique
was too
unwieldy
to
use.
In
fact, unlike subjects
in
the
skills-acquisition group, relaxation sub-
jects
reported that they
did not use
their
training
strategy during testing. This
may
also explain
why the
present study failed
to
obtain
performance gains
for
this treatment
in
contrast with
the
successful
use of
this
approach
by
Counts, Hollandsworth,
and
Alcorn (1978).
The
most obvious difference
between
the two
uses
of the
same procedure
is
that
in the
present study subjects were
trained
in
groups, whereas
in the
earlier
report subjects were trained individually.
Thus,
it
might
be
said that cue-controlled
relaxation
or
meditation treatments would
have resulted
in
significant
performance
gains
for
these groups
as
well.
But why
bother?
If, in
fact,
ineffective
test taking
is
an
attention-focusing skills
deficit,
why not
train
for
that specifically
and
de-emphasize
the
role
of
physiological
arousal?
As
Counts
and
his
colleagues suggested, training
in the
use of
subvocalized
cue
words
may be
nothing
more than teaching
the use of
task-
orienting
self-instructions.
Perhaps
it is
time
to
give
the
phrase test anxiety
a
respectful
burial
and
talk about inadequate test per-
formance
in
terms that more accurately
describe what
it is,
namely,
ineffective
test
taking.
Reference
Notes
1.
Glazeski,
R. C., &
Hollandsworth,
J. G. An
investi-
gation
of the
role
of
physiological arousal
in
test
anxiety.
Paper presented
at the
13th
annual
conven-
tion
of the
Association
for
Advancement
of
Behavior
Therapy,
San
Francisco, December 1979.
2.
Sarason,
I. G.
Cognitive
Interference
Questionnaire.
Unpublished
instrument,
1976.
(Available
from
I. G.
Sarason, Department
of
Psychology,
Nl-25,
Uni-
versity
of
Washington,
Seattle,
Washington
98195.)
References
Alpert,
R.,
&
Haber,
R.
Anxiety
in
academic
achieve-
ment
situations. Journal
of
Abnormal
and
Social
Psychology,
1960,67,
207-215.
Benson,
H. The
relaxation
response.
New
York: Mor-
row,
1975.
Blatt,
S. J.
Patterns
of
cardiac arousal
during
complex
mental
activity.
In R. J.
Harper,
C. C.
Anderson,
C. M.
Christensen,
& S. M.
Hunka
(Eds.),
The
cog-
nitive
processes: Readings.
Englewood
Cliffs,
N.J.:
Prentice-Hall,
1964.
Brown,
W. F.
Effective
Study
Test'.
Manual
of
direc-
tions.
San
Marcos,
Tex.:
Effective
Study Materials,
1975.
Counts,
D.
K.,
Hollandsworth,
J.
G.,
&
Alcorn,
J. D.
Use of
electromyographic
feedback
and
cue-con-
trolled
relaxation
in the
treatment
of
test
anxiety.
Journal
of
Consulting
and
Clinical
Psychology,
1978,
46,
990-996.
Davidson,
R. J.
Specificity
and
patterning
in
biobehav-
ioral
systems:
Implications
for
behavior change.
American Psychologist, 1978,
33,
430-436.
Davidson,
R.
J.,
&
Schwartz,
G. E. The
psychobiology
of
relaxation
and
related states:
A
multi-process
theory.
In D. I.
Mostofsky
(Ed.),
Behavior control
and
modification
of
physiological
activity.
Engle-
wood
Cliffs,
N.J.: Prentice-Hall, 1976.

EFFECTIVE
TEST
TAKING
439
Deffenbacher,
J. L.
Worry, emotionality,
and
task-gen-
erated interference
in
test anxiety:
An
empirical test
of
attentional theory. Journal
of
Educational Psy-
chology, 1978,
70,
248-254.
Denny,
D. R.
Self-control
approaches
to the
treatment
of
test anxiety.
In I. G.
Sarason
(Ed.),
Test anxiety:
Theory,
research,
and
applications.
Hillsdale,
N.J.:
Erlbaum, 1978.
Denny,
D.
R.,
&
Rupert,
P. A.
Desensitization
and
self-
control
in the
treatment
of
test anxiety. Journal
of
Counseling Psychology,
1917,24,
272-280.
Fabick,
S. D. The
relative effectiveness
of
systematic
desensitization, cognitive
modification,
and
mantra
meditation
in the
reduction
of
test anxiety (Doctoral
dissertation, West Virginia University, 1976). Dis-
sertation Abstracts International, 1977,
37,
4862A.
(University
Microfilms
No.
DBJ77-02548)
Goldfried,
M. R. The use of
relaxation
and
cognitive
relabeling
as
coping skills.
In R. B,
Stuart
(Ed.),
Behavioral
self-management: Strategies, techniques,
and
outcomes.
New
York:
Brunner/Mazel,
1977.
Gribbin,
B.,
Steptoe,
A.,
&
Sleight,
P.
Pulse wave veloc-
ity
as
measure
of
blood pressure change. Psycho-
physiology, 1976,
13,
86-90.
Hollandsworth,
J.
G.,
Glazeski,
R.
C.,
Kirkland,
K.,
Jones,
G.,
& Van
Norman,
L. R. An
analysis
of the
nature
and
effects
of
test anxiety: Cognitive, behav-
ioral,
and
physiological components. Cognitive
Therapy
and
Research,
1979,3,
165-180.
Holroyd,
K. A.
Cognition
and
desensitization
in the
group treatment
of
test anxiety. Journal
of
Consult-
ing
and
Clinical
Psychology,
1976,
44,
991-1001.
Holroyd,
K.
A.,
Westbrook,
T.,
Wolf,
M.,
&
Badhorn,
E.
Performance, cognition,
and
physiological
responding
in
test
anxiety.
Journal
of
Abnormal Psy-
chology,
1978.S7,
442-451.
Home,
A.
M.,&Matson,
J. L. A
comparison
of
model-
ing,
desensitization,
flooding,
study skills,
and
con-
trol
groups
for
reducing
test
anxiety. Behavior
Therapy,
1977,8,
1-8.
Huck,
S. W., &
Jacko,
E. J.
Effect
of
varying
the
response
format
of the
Alpert-Haber
Achievement
Anxiety
Test.
Journal
of
Counseling
Psychology,
1974,21,
159-163.
Kirkland,
K.,
&
Hollandsworth,
J. G.
Test anxiety,
study
skills,
and
academic performance. Journal
of
College
Student
Personnel,
1979,
20,
431-436.
Lang,
P. J.
Research
on the
specificity
of
feedback
training: Implications
for the use of
biofeedback
in
the
treatment
of
anxiety
and
fear.
In J.
Beatty
& H.
Legewie
(Eds.),
Biofeedback
and
behavior.
New
York: Plenum
Press,
1977.
Liebert,
R.
M.,
&
Morris,
L. W.
Cognitive
and
emo-
tional components
of
test
anxiety:
A
distinction
and
some initial data. Psychological
Reports,
1967,
20,
975-978.
Little,
S.,
&
Jackson,
B. The
treatment
of
test anxiety
through attentional
and
relaxation training. Psycho-
therapy:
Theory,
Research,
and
Practice, 1974,
//,
175-178.
McFall,
R. M,
Behavioral training:
A
skill
acquisition
approach
to
clinical
problems.
In J. T.
Spence,
R. C.
Carson,
& J. W.
Thibaut
(Eds.),
Behavioral
approaches
to
therapy.
Morristown,
N.J.: General
Learning
Press,
1976.
Meichenbaum,
D. H.
Cognitive modification
of
test-
anxious
college students. Journal
of
Consulting
and
Clinical
Psychology,
1972,39,
370-380.
Meichenbaum,
D. H.
Cognitive—behavior
modifica-
tion:
An
integrative approach.
New
York: Plenum
Press, 1977.
Obrist,
P.
A.,
Light,
K.
C.,
McCubbin,
J.
A.,
Hutche-
son,
J.
S.,
&
Hoffer,
J. L.
Pulse
transit
time: Relation-
ship
to
blood pressure
and
myocardial
performance.
Psychophysiology,
1979,
16,
292-301.
Otis,
A.
S.,
&
Lennon,
R. T.
Otis-Lennon
Mental
Ability
Test: Manual
for
Administration.
New
York:
Harcourt, Brace
&
World, 1968.
Pagano,
D.
R.,
&
Katahn,
M.
Construct
validity, dis-
confirming
evidence
and
test anxious
students.
Jour-
nal
of
Personality,
1972,40,
137-148.
Robinson,
F. P.
Effective
study.
New
York: Harper,
1946.
Russell,
R.
K.,
&
Sippich,
J. F.
Cue-controlled relax-
ation
in the
treatment
of
test anxiety. Journal
of
Behavior
Therapy
and
Experimental
Psychiatry,
1973,4,47-49.
Russell,
R.
K.,
&
Sippich,
J. F.
Treatment
of
test anx-
iety
by
cue-controlled relaxation. Behavior
Therapy,
1974,5,
673-676.
Schwartz,
G.
E.,
Davidson,
R.
J.,
&
Goleman,
D. J.
Patterning
of
cognitive
and
somatic
processes
in the
self-regulation
of
anxiety: Effects
of
meditation ver-
sus
exercise. Psychosomatic Medicine, 1978,
40,
321-328.
Spielberger,
C.
D.,
Anton,
W.
D.,
&
Bedell,
J. The
nature
and
treatment
of
test anxiety.
In M.
Zucker-
man
& C. D.
Speilberger
(Eds.),
Emotions
and
anx-
iety:
New
concepts,
methods,
and
applications.
New
York: Wiley, 1976.
Valins,
S.
Cognitive
aspects
of
false
heart-rate
feed-
back. Journal
of
Personality
and
Social
Psychology,
1966,4,
400-403.
Wine,
J. D.
Test
anxiety
and
direction
of
attention. Psy-
chological
Bulletin,
1971,
76,
92-104.
Received
October
22,
1979
•