PcrceptuaI and Motor Skillr, 1968, 27, 947-950.
@ Southern Universities Press 1968
INHIBITION A N D T H E FACILITATING EFFECT OF NOISE O N
INTERFERENCE TASKS1
B. KENT HOUSTON
Uniusr~ityof Texas at Austin
Summary.-3 groups of 14 college students performed the Digit Symbol
Test (DST) and Embedded-figures Test (EFT) under one of 3 conditions:
varied noises to be ignored, varied noises requiring attention for a spoken com-
mand, and no noise. I t was hypothesized that performance o n DST, but not
EFT, would be enhanced in the condition of noises to be ignored. The results
supported the hypothesis. A case is made for explaining the findings in terms
of an interaction between inhibitory processes.
Houston (Houston, 1967; Houston & Jones, 1967) presents the hypothesis
that, if a person has to inhibit responding to one source of stimulation, it would
be easier if he had to inhibit responding to another source of stimulation at the
same time. For example, if there are interfering cues in a task to which a
person must inhibit responding, it should be easier if continuous noise is present
to which the person must also inhibit listening. The notion is that an interaction
between the inhibitory processes occurs such that one facilitates the other.
To test this notion, Houston and Jones (1967) conducted an experiment
to determine whether the presence of noise which S was told to ignore would
facilitate performance on a task which required him to inhibit response to dis-
tracting cues (viz., the Stroop Color-word test2). It was found that noise did
facilitate performance on the task involving inhibition (the Stroop) but did not
do so on a similar task which did not require this.
The study described here is an attempt to test the generalizability of the
previous study's findings by examining the effects of noise on a different set of
tasks, namely, the Digit Symbol Test (DST) described by Karp (1962, 1963)
as involving "overcoming a distracting context" and the Embedded-figures Test
(EFT) which involves "overcoming an embedding context." In Karp's factor
analytic study, DST loaded highly (.61) on a factor identified as involving the
capacity ". . . to manipulate or locate items surrounded by a matrix of irrelevant
items which serve to distract the subject from performance of the task." The
EFT loaded highly (.73) on a factor identified as involving the capacity to
manipulate or locate items which are embedded in configurations which serve
to "break up" the target items by embedding their parts in other, more compel-
ling gestalts. DST and EFT are similar then in that S's task is to locate or ma-
nipulate items within a matrix of other stimuli. The two tasks differ, however,
'This investigation was supported by United States Public Health Research Fellowship M H
15,757-03 and United States Public Health Research Grant H D 01092-01. T h e author
wishes to thank Eleanor Maccoby for her invaluable aid in completing this study, Thomas
Jones for running Ss, and Ray ~ u l r yfor his constructive commentary-on the manuscript.
'The experimental tasks are described more fully in the earlier report (Houston & Jones,
1967 ).
948 B. K. HOUSTON
in the extent to which the stimuli of this matrix are distracting and must be ig-
nored for effective performance.
The inhibition-interaction hypothesis predicts that the inhibitory process
involved in ignoring noise will facilitate performance on a task which requires
inhibiting response to distracting cues (viz., D S T ) but will not d o so for a simi-
lar task which does not require inhibition (viz., E F T ) . It would not be expected
that other kinds of noise conditions-one, for instance, involving vigilance for a
cue and subsequent performance of a response to that cue-would facilitate per-
formance o n either task.
T o assess the specificity of the effects of "noise inhibition" o n a task re-
quiring inhibition, performance o n the two experimental tasks, D S T and EFT,
was compared across three conditions: noise allowing S to maintain an in-
hibitory set toward the sounds, noise requiring vigilance for a cue within the
sounds, and quiet. If the hypothesis were correct, a specific interaction should
be observed: the noise-inhibition condition should enhance performance on the
task involving inhibition ( D S T ) but not performance on the other task ( E F T ) ,
and the vigilance noise condition should not facilitate performance on either
task.
PROCEDURE
Forty-five college students were randomly assigned to three groups: inhi-
bition-noise ( I N ) , vigilance-noise ( V N ) , and quiet (Q). I n the two noise
groups, I N and V N , Ss performed DST, EFT, and two tasks reported elsewhere
(Houston & Jones, 1967) while tape recorded noises were delivered to them
through earphones at an average sound intensity level of approximately 78 db.
T h e noises, which were the same in both I N and V N , consisted of a variety of
familiar sounds, e.g., trains, dripping water, etc., and unfamiliar sounds, e.g.,
electronic music, gibberish, etc. In IN, Ss were instructed to completely ignore
the noises. I n V N , they were told to tap the table with a pencil whenever they
heard the spoken word "tap" on the tape. The word "tap" was interspersed ran-
domly but occurred on an average of about once every 11 sec. In Q, Ss per-
formed the experimental tasks without the auditory tape.
The D S T was the W A I S Digit Symbol subtest (Wechsler, 1 9 5 8 ) . The
score for this task was the number of correct substitutions made during 9 0 sec.
The EFT was a version developed by Jackson, Messick, and Meyers ( 1964). I t
consists of 16 complex patterns each o n one side of a page and a simple figure,
which is to be located within the complex pattern, on the obverse side. T h e
score was number of figures correctly located in 10 min.
T h e order in which DST and the two previously reported tasks were per-
formed was counterbalanced over Ss and EFT was always performed last. T h e
latter was done because EFT takes so much longer to complete than the other
tasks, and the possibility that EFT might affect performance on the other tasks if
it preceded them seemed more likely than the possibiliry that the order or num-
ber of tasks preceding EFT might affect performance o n it.
NOISE IN INTERFERENCE TASKS 949
RESULTS
In order to ensure the comparability of the three groups, Ss were compared
on grade-point average. Q tended to have a lower GPA than the other two
groups. Since correlations were obtained between the dependent measures and
GPA ( r = -.I7 for DST, and r = -.22 for difference scores between DST
and EFT) and since Q contained 17 Ss whereas the other two groups contained
14 Ss each, the three Ss with the lowest GPAs in Q were dropped before the
dependent measures were analyzed. With Q so adjusted, there was no significant
difference in mean GPA between the final groups ( F < 1.00, df = 2/39).
For each group the mean number of correct responses for each task is given
in Table 1. By inspection it may be seen that as predicted, performance on DST
was better in I N than in Q but performance on EFT was not. Also, perform-
ance on both tasks was poorer in V N than in either Q or IN.
TABLE 1
NUMBER
O F CORRECT
RESPONSESA N D STANDARD DEVIATIONS
FOR DIGIT
SYMBOL
TESTAND EMBEDDED-FIGURES TEST FOR THREENOISEGROUPS
Tasks Groups
Quiet Inhibition Vigilance
Noise Noise
DigitSymbolTest M 68.29 72.43 64.36
SD 7.36 10.37 9.38
Embedded-figures Test M 13.71 12.79 11.64
SD 2.96 4.00 3.75
An analysis of variance was performed on the data. In an ordinary analysis,
the interaction between treatments and tasks would indicate the extent to which
the treatments differentially affected the tasks without specifying which treatment
had what effect. The prediction here, however, specifies that I N should differ-
entially affect the tasks with reference to Q but VN should not. There is a
procedure (after Winer, 1962) by which the interaction sum of squares, which
in this case has 2 df,may be partitioned into two components to test such a spe-
cific interaction. One component specifically compares the differential effect of
I N on task performance with that of Q and VN. The other component of the
interaction is a remainder term. It specifies how much of the interaction vari-
ance is not accounted for by the prediction. It is obtained by subtracting the
sum of squares for the predicted, specific comparison from the total interaction
sum of squares. A test of the specific comparison showed that it was not sig-
nificant ( F = 3.78, df = 1/39, p < .06). The remainder did not approach
significance ( F < 1.00).
An additional analysis was made to determine whether order in which the
tasks were performed has an effect on the data. No significant main effect or
interaction with treatments and/or tasks were found.
The results support the hypothesis that the presence of noise which Ss are
950 B. K. HOUSTON
told to ignore facilitates performance o n a task which requires inhibiting re-
sponse to distracting cues ( D S T ) but does not do so for a similar task which
does not require inhibition ( E F T ) . Moreover, another (vigilance) type of
noise condition does not have an enhancing effect for either task. Thus this
investigation lends guarded support for the generalizability of the findings of
the Houston and Jones ( 1 9 6 7 ) study.
While the results of this study were hypothesized as a function of an inter-
action between inhibitory processes, there exists an alternative explanation which
should be considered. Noise has been demonstrated to create physiological
arousal (Harmon, 1933; Freeman, 1939), and physiological arousal has been
hypothesized to lead to narrowing of attention (Callaway & Dembo, 1958) or a
decrease in cue utilization (Easterbrook, 1959). Such a n effect of arousal could
facilitate performance on tasks in which irrelevant elements interfere with per-
formance of a central task. It is, therefore, possible that in chis study both of the
noise conditions produced arousal which enhanced performance on D S T in IN,
but the facilitating effects of arousal in VN were vitiated by the demand for
division .of attention.
Although this explanation appears attractive, it seems unlikely considering
there is evidence that arousal generated in ways other than the presence of
noise (e.g., threat of shock) has a detrimental rather than a facilitating effect
o n DST performance (Carlson & Lazarus, 1953; Goldfarb, 1961). In light of
this, a n explanation involving interaction of inhibitions seems more credible.
REFERENCES
CALLAWAY,E., & DEMBO,T. Narrowed attention: a psychological phenomenon that
accompanies a certain physiological change. Journal o f Neurology and Psychktry,
1958, 79, 74-90.
CARLSON,
V. R., & LAZARUS, R. S. A repetition of Meyer Williams' study of intellectual
control under stress and associated Rorschach factors. Journal o f Consulting Psy-
chology, 1953, 17, 247-253.
EASTERBROOK, J. A. The effect of emotion on cue utilization and the organization of
behavior. Psychological Review, 1959, 66, 183-201.
FREEMAN, G. L. Changes in tension-pattern and total energy expenditure during adapm-
tion to "distracting" stimuli. American Journal o f Psychology, 1939, 52, 354-360.
GOLDFARB, A. Performance under stress in relation to intellectual control and self ac-
ceptance. Journal o f Consulting Psychology, 1961, 25, 7-12.
HARMON,F. L. The effects of noise upon certain psychological and physiological proc-
esses. Archives o f Psychology, 1933, No. 147.
HOUSTON,B. K. Inhibition, noise and interference tasks. Unpublished doctoral dis-
sertation, Stanford Univer., 1967.
HOUSTON, B. K., & JONES,T. M. Distraction and Stroop Color-Word performance.
Journal o f Experimental Psychology, 1967, 74, 54-56.
JACKSON,D. N., MESSICK,S., & MEYERS,C. T. Evaluation of group and individual
forms of embedded-figures measures of field-independence. Educationaj and Psy-
chological Measurement, 1964, 24, 177-191.
KARP, S. A. Kit o f selected distraction tests. New York: Cognitive Tests, 1962.
KARP,S. A. Field dependence and overcoming embeddedness. Journal of Consuking
Psychology, 1963, 27, 294-302.
WECHSLER,D. T h e meaurement and appraisal o f adult intelligence. Baltimore: Wil-
liams & Wilkins, 1958.
WINER,B. J. Statistical principlesin experimental design. New York: McGraw-Hill,
1962.
Accepted October 2, 1968.
