Perception & Psychophysics

2003, 65 (3), 359-366

Developmental change in the

cross-modal Stroop effect
JULIE B. HANAUER and PATRICIA J. BROOKS
City University of New York, Staten Island, New York
E. M. Elliott, Cowan, and Valle-Inclan (1998) reported a cross-modal Stroop-like interference effect
in adults when an auditory distractor (a color or noncolor word) occurred simultaneously with a color
patch to be named. Response times were slower with color as opposed to noncolor distractors. To distinguish two accounts of this phenomenon, we tested 4- to 11-year-olds and adults. The suppression
hypothesis posits that the irrelevant word enters a phonological buffer and is injurious to color naming if the participant is unable to suppress its representation in time. The concurrent processing hypothesis states that interference occurs when the distractor and the color name are lexically accessed
at the same time. Our finding that the cross-modal Stroop effect occurred in young children even with
a distractor presented 500 msec in advance of the color patch favors the suppression account. Development in executive functioning may also contribute to the interference effects becoming progressively weaker with age.

The Stroop color word task (Stroop, 1935) is perhaps

the most well-known measure of selective attention. Selective attention is the ability to focus on one particular
aspect of a stimulus while defocusing other aspects of the
same stimulus (Bedi, Halperin, & Sharma, 1994). In the
Stroop task, participants are asked to name the print color
of a color word (e.g., say red when shown the word blue
printed in red). A correct response requires participants
to inhibit reading the color word, a task that is difficult
because there is a strong tendency to read the word itself,
rather than identify its print color. Individuals must selectively attend to the print color and inhibit reading the
word. According to one account (MacLeod, 1991), it
takes a greater amount of time to state the print color of
a color word, in comparison with a nonword or a noncolor
word, because reading is not entirely under conscious
control; therefore, what is read competes with the print
color to be named.
A number of researchers have modified the Stroop
procedure to see whether the interference effect is present in other modalities besides the visual modality. For
example, in Green and Barber (1981, 1983), significant
interference occurred in the auditory modality when participants were instructed to say girl if the speaker was

A portion of this paper was presented at 41st Annual Meeting of the

Psychonomic Society, New Orleans, November 2000. The authors
thank the children who participated in the study and their parents. We
thank Devin Bracco, Jose Colon, Peter Hanauer, Stephanie Landolfi,
Maya McLawrence, and Barbara Padowska for assistance with data collection and analysis and Nelson Cowan for his helpful review of the
manuscript. Please address correspondence to P. J. Brooks, Department
of Psychology,4S-112, the College of Staten Island of the City University
of New York, 2800 Victory Blvd., Staten Island, NY 10314 (e-mail:
[email protected]).

male and boy if the speaker was female. Cowan and

colleagues (Cowan, 1989; Cowan & Barron, 1987; E. M.
Elliott, Cowan, & Valle-Inclan, 1998) designed a crossmodal Stroop analogue with colors presented in the visual modality and words presented over headphones.
They found significantly slower color-naming latencies
when colors were paired with color word distractors
(e.g., seeing purple and hearing yellow), as compared
with noncolor words (e.g., seeing purple and hearing single). Cowan and Barron argued that adults could not fully
inhibit the processing of words in the auditory modality
and, therefore, could not limit lexical processing to the
visual modality. This claim, however, became controversial, since Miles and Jones (1989) and Miles, Madden, and
Jones (1989) failed to replicate Cowan and Barron. Cowan
(1989), E. M. Elliott et al. (1998), and E. M. Elliott and
Cowan (2001) claimed that the failed replications were
due to methodological errors and again demonstrated the
cross-modal Stroop effect.
The debate over the existence of the cross-modal Stroop
effect is of theoretical importance in determining the
constraints on the role of working memory in language
processing. According to the influential theory of Baddeley (1986, 1997), the phonological loop is a component of verbal working memory responsible for the temporary storage of sound patterns. Miles et al. (1989) argued
that when verbal material is being maintained in the
phonological loop, interference might occur if an item to
be remembered is displaced by another item that enters
this temporary store. Interference in the phonological
loop is most likely to occur if the phonological loops capacity has been reached and the displacing item is phonologically similar (as opposed to semantically related) to
the item to be remembered. Hence, Miles et al. did not
expect there to be a cross-modal Stroop effect, because

359

Copyright 2003 Psychonomic Society, Inc.

360

HANAUER AND BROOKS

the task does not require participants to maintain words

in the phonological loop over any period of time and,
second, on any trial, the distractor is not phonologically
related to the target color name. Cowan and Barron (1987)
argued for a broader view of the phonological loop as a
temporary speech buffer where words to be spoken may
be held prior to articulation. When there are multiple
items in this temporary store, selection of the item most
relevant to the task at hand must occur, and there may be
competition among candidate items, even in cases in
which the words are not phonologically related. In the
cross-modal Stroop task, the auditory distractor automatically enters the temporary speech buffer as it is processed. Because it takes time to trace the origin of each
item in the buffer when one selects the name of the color
patch as the intended word to be produced, interference
occurs in color naming. As the semantic relatedness of
the words in the buffer increases, the ease of selecting
the correct word decreases, leading to greater interference with color word distractors than with noncolor word
distractors.
In the E. M. Elliott et al.s (1998) study, the timing of
presentation of the color patch to be named and of the
distractor was varied by presenting the distractor either
simultaneously with or 500 msec prior to the presentation of the color. Exposure to a word 500 msec before
presentation of a visual stimulus provides sufficient time
to allow for complete processing of the word (Baddeley,
1986; Cowan & Saults, 1995). Thus, with the distractor
presented 500 msec before the color patch, E. M. Elliott
et al. predicted a cross-modal Stroop effect, because the
auditory representation of the word would still be active in
working memory. Surprisingly, they found the predicted
interference effect only with simultaneous presentation
of the color patch and the distractor. E. M. Elliott et al.
considered two possible explanations. First, delaying the
presentation of the color patch by 500 msec might provide participants with sufficient time to both process and
suppress activation of the distractor. Alternatively, the
cross-modal Stroop effect might stem from concurrent lexical access of the words presented in the auditory and the
visual modalities. This hypothesis derives from a study
by Schriefers, Meyer, and Levelt (1990), in which they
showed semantic interference in a cross-modal picture
word interference paradigm. Schriefers et al. suggested
that semantic interference stems from competition between the distractor and the picture names, occurring
during the semantically driven stage of lexical access,
called lemma retrieval. Although E. M. Elliott et al. favored the former, suppression account as an explanation
of the time course of the cross-modal Stroop effect, their
data do not provide a basis for ruling out the alternative
concurrent processing explanation.
We propose that examining the development of the
cross-modal Stroop effect will help to resolve the controversies regarding the source and reliability of the interference effect because developmental differences in
selective attention may affect the time course and mag-

nitude of the effect. In developmental studies of the visual Stroop effect, the interference effect has been shown
to decline in magnitude from the age of 7 years to early
adulthood (Comalli, Wapner, & Werner, 1962; Rosinski,
Golinkoff, & Kukish, 1975). Not surprisingly, in children who are not yet skilled readers, the effect does not
occur at all (Ehri, 1976; Schiller, 1966). To study development of selective attention and inhibition in very
young children, researchers have modified the methodology of the Stroop task in a variety of ways. For instance, Gerstadt, Hong, and Diamond (1994) designed a
daynight Stroop task in which 3- to 7-year-olds were instructed to say day when shown a line drawing of
moon and stars and night when shown a line drawing
of the sun. That is, in the experimental condition, the
correct answer was the opposite of what the card depicted. In a control condition, children were instructed to
say day and night in response to specific abstract
geometric patterns. Although even the youngest children
performed well in the control condition, in the experimental condition children younger than 5 years of age
often took very long to respond and made many errors.
Gerstadt et al. concluded that the younger children could
not remember the rules of the game while inhibiting the
matching response (i.e., saying night when shown the
moon and stars). Jerger and her colleagues (Jerger, Martin, & Pirozzolo, 1988; Jerger, Pearson, & Spence, 1999)
modified Green and Barbers (1981, 1983) auditory Stroop
task for use with preschoolers. In Jerger et al. (1988),
3- to 6-year-olds were told to press a Mommy button whenever a male voice was heard and a Daddy button whenever
a female voice was heard. Jerger et al. (1988) found slower
response latencies for conflict trials (e.g., hearing daddy
spoken in a female voice), as compared with either congruent trials (e.g., hearing mommy spoken in a female
voice) or neutral trials (e.g., hearing ice cream spoken in
either voice). Furthermore, the magnitude of the interference effect (i.e., the reaction time [RT] difference between conflict and neutral conditions) decreased significantly between the ages of 3 and 4. Taken together, these
studies have shown development in selective attention,
leading to a decrease in Stroop-like interference, but have
not addressed the specific issue of the independence of
lexical processing in the auditory and the visual modalities. To our knowledge, no previous research has compared the extent to which children and adults are able to
suppress processing of irrelevant spoken words when naming visually presented stimuli.
In the present study, we examined the development of
the cross-modal Stroop effect in 4- to 11-year-old children and adults. Given the inconsistency of reports of the
effect, we adopted the basic procedures of E. M. Elliott
et al. (1998, Experiment 1) to maximize the likelihood of
replicating their results with adults. However, we varied
the timing of the distractor relative to the picture somewhat differently than in E. M. Elliott et al. to provide a
stronger test of the suppression and concurrent processing explanations of the effect. We present two timing

CROSS-MODAL STROOP
conditions. The first is an exact replication of the 0-msec
(simultaneous) condition of E. M. Elliott et al. (1998), at
which they found a reliable cross-modal Stroop effect in
adults. However, because of developmentaldifferences in
processing speed (Kail, 1991, 1992), for the 2500-msec
condition, we lengthened the interval between presentation of the distractor and the color patch to provide children with sufficient time to fully process the word prior
to seeing the patch. That is, we increased the stimulus
asynchrony (SA) to rule out the possibility of concurrent
lexical access of the distractor and the color name even
in our youngest participants. In our experiment, the color
patch appeared 500 msec after the offset of the word, as
opposed to 500 msec after the onset, as in E. M. Elliott
et al. This manipulation of SA allows us to evaluate the
competing accounts of the cross-modal effect proposed
by E. M. Elliott et al.
If the concurrent lexical processing account is correct,
we would expect the interference effect to occur only
with simultaneous presentation of the color patch and
the distractor, irrespective of the age of the participant.
However, if the suppression account is correct, we would
expect to observe the interference effect at both timing
conditions in young children, owing to their requiring
more time to process and inhibit the distracting words.
Furthermore, owing to a developmental progression in
the ability to selectively attend to color naming while inhibiting irrelevant auditory distractors, we would predict
that the cross-modal Stroop effect would be robust in
children and become progressively weaker with age.
That is, we would suspect that the relatively small interference effect observed in previous studies with adults is
the end point of a trajectory in the development of selective attention.
METHOD
Participants
Fifteen 4- to 5-year-olds (6 boys, 9 girls; mean age, 4 years
11 months; range, 4 years 1 month to 5 years 11 months), fifteen 6to 7-year-olds (7 boys, 8 girls; mean age, 6 years 11 months; range,
6 years 2 months to 7 years 8 months), fifteen 9- to 11-year-olds (9
boys, 6 girls; mean age, 10 years 6 months; range, 9 years 1 month
to 11 years 11 months), and 30 undergraduates (14 men, 16 women;
mean age, 22 years; range, 1734 years) took part in the study. All
the participants had corrected-to-normal vision and normal hearing
and were prescreened to ensure no color perception def iciencies.
The sample was ethnically diverse and predominantly of middleclass background. The children were recruited using flyers posted
at the College of Staten Island of the City University of New York.
They were tested in a psychology laboratory and received stickers
and t-shirts for their participation. The undergraduates were recruited from introductory psychology classes and received extra
credit for their efforts. In the adult group, we tested twice the number of participants than in the child groups, owing to the small effect sizes obtained in previous studies (Cowan & Barron, 1987;
E. M. Elliott et al., 1998).
Apparatus and Stimuli
The experiment was conducted using a Macintosh Power PC
computer equipped with PsyScope experimental software (Cohen,
MacWhinney, Flatt, & Provost, 1993) and SoundEdit 16. The visual

361

stimuli consisted of six color patches, measuring 3 3 3 in., surrounded by a black background. The colors were red, white, blue,
green, yellow, and purple, as in E. M. Elliott et al. (1998). Spoken
words corresponding to the six colors, along with six noncolor
words (dry, single, real, short, fewer, and best), were presented over
headphones in a naturally spoken female voice. The noncolor words
were matched to the color words in terms of word frequency and
syllable structure (Carroll, Davies, & Richman, 1971; Elliott et al.,
1998). The duration of the color words ranged from 315 to
512 msec, with a mean of 411 msec, and the duration of the noncolor
words ranged from 308 to 496 msec, with a mean of 410 msec. The
PsyScope software controlled the presentation of the spoken words
and color patches. Each stimulus picture was paired with every
stimulus word except its own (e.g., the red color patch was never
paired with the word red ). That is, in each block of trials, items
were randomly selected with replacement from a list combining
each color patch with each color word (except for the matching
color word) and each noncolor word. A microphone was set up in
front of the participant, with a voice key triggering a button box
timer that recorded response latencies for color naming. RTs for
correct responses were measured from the onset of the color patch
to the triggering of the voice key by the participants response.
Procedure
Each participant was tested individually in a quiet room. The participants were seated in front of the computer monitor, with a microphone adjusted to be approximately 1 in. away from their mouth. Six
color patches (red, blue, white, green, purple, and yellow), printed
on white paper from a color printer, were shown to the participant
as the experimenter read the color names aloud. The participant was
then asked to repeat the color words to ensure color recognition.
The participants then were told that they were going to play a computer game and were asked to wear a set of headphones. The participants were told that the object of the game was to say what you
see. The experimenter said, In this game, color patches are going
to appear on the screen and your job is to name the colors as quickly
as possible. Sometimes, you are going to hear words over the headphones, but try your best to ignore the word and say only what you
see. The words may confuse you sometimes because they may be
different from what you see, but you should concentrate and say
what you see. The participants were told to name the color of each
color patch as soon as it appeared, and not to talk otherwise. After
36 trials, a smiley face appeared to indicate that the block of trials
was finished. At the end of each block, the participants were told
that they did a great job, and the children received stickers for their
efforts.
Each trial began with a white fixation cross, which remained on
the screen for 2,500 msec. In the 0-msec condition, the timing of the
onset of the auditory distractor was simultaneous with presentation
of the color patch and coincided with the disappearance of the fixation cross. In the 2500-msec condition, the timing of the onset of
the auditory distractor was simultaneous with the disappearance of
the fixation cross, and the color patch appeared 500 msec after the
offset of the word. In the silent condition, the color patch replaced
the fixation cross after 2,500 msec, and no auditory distractor occurred. The participants response activated a voice key that
recorded the RT for the trial and triggered the disappearance of the
color patch. After 200 msec, the fixation cross for the next trial appeared. The experimenter kept a log of the participants responses
as a record of errors and lost trials.
Each participant received five blocks of trials. The first block
consisted of 18 trials of the silent condition. This allowed the participants to practice naming the color patches as they appeared on
the computer monitor and allowed the experimenter to adjust the
sensitivity of the voice key. If the participants made any errors during this block (i.e., misnaming a color patch), they were corrected.
The second block comprised 18 practice trials with auditory distractors to allow the participant to get used to hearing the words in

362

HANAUER AND BROOKS

41,516, p , .001], accounting for 10% of the variance.
Overall RTs for naming colors decreased markedly with
age. RTs tended to be faster for noncolor word trials than
for color word trials and were faster when the distractor
occurred in advance of the color patch (2500-msec SA)
rather than simultaneously with the color patch (0-msec
SA). The main effects were qualified by significant twoway interactions of age with distractor type [F(3,71) 5
13.10, MSe 5 28,871, p , .001], accounting for 2% of
the variance, age with SA [F(3,71) 5 27.43, MS e 5
41,516, p , .001], accounting for 4% of the variance,
and distractor type with SA [F(1,71) 5 11.80, MSe 5 16,
613, p , .001], accounting for less than 1% of the variance. The three-way interaction of age, distractor type,
and SA was not significant.
To test for the existence of a cross-modal Stroop effect
at each age, planned comparisons were conducted, comparing RTs for color word and noncolor word trials at
each SA condition. For the 2500 SA condition, the 4- to
5-year-olds named color patches paired with color words
more slowly than color patches paired with noncolor
words [F(1,14) 5 9.45, MS e 5 57,335, p , .01], accounting for 2% of the total variance. In contrast, for the
other age groups, RTs for color word and noncolor word
trials were statistically equivalent when the distractor
was presented 500 msec before the color patch. Thus, for
all groups except the youngest children, delaying presentation of the color patch by 500 msec provided sufficient time for activation of the distractor to be fully suppressed. The lack of a cross-modal Stroop effect in adults
with the distractor presented in advance of the color
patch replicated the findings of Elliott et al. (1998). For
the 0-msec SA condition, the RT difference for color
word and noncolor word trials was highly significant for
the 4- to 5-year-olds [F(1,14) 5 20.38, MS e 5 57,335,
p , .001], accounting for 5% of the variance, and for the
6- to 7-year-olds [F(1,14) 5 33.40, MSe 5 11,098, p ,
.001], accounting for 5% of the variance, and was reliable for the 9- to 11-year-olds [F(1,14) 5 5.13, MSe 5
12,982, p , .05], accounting for 2% of the variance, and
for the adults [F(1,29) 5 6.09, MS e 5 1,369, p , .05],
accounting for less than 1% of the variance. Taken together, the planned comparisons show that the crossmodal Stroop effect is much more robust in children than

the headphones. If the participants made any errors in this block

(e.g., repeating the auditory stimulus instead of naming the color
patch), they were corrected. After the two blocks of practice trials,
three blocks of 36 test trials (108 total trials) were presented with
each block, consisting of 12 silent trials, 12 color word trials (6 at
each SA), and 12 noncolor word trials (6 at each SA), occurring in
a random order. In the test blocks, the participants were not corrected for any errors. If, at any point during the experiment, a child
became distracted or started a conversation, the experimenter
paused the experiment until the child calmed down and trials could
be resumed.

RESULTS
Reaction Times
Since the distributions of RTs were positively skewed,
especially for the youngest children, who often took very
long to respond, we used median RTs for correct trials as
the dependent variable. Table 1 presents median RTs for
each age group as a function of distractor type and SA.
Overall, color naming was fastest for the 2500-msec
condition, intermediate for the silent condition, and
slowest for the 0-msec condition.This ordering of RTs as
a function of distractor condition exactly replicated Elliott et al. (1998) and was likely due to the fact that the
timing of the color patch, relative to both the fixation
point and the distractor, varied randomly across trials
within each block. Thus, because it was impossible to
predict for any trial exactly when the color patch would
appear, when the distractor preceded the color patch
(2500-msec SA), it alerted the participant to prepare a
response. In the following statistical analyses, we will
examine RTs for trials involving auditory distractors and
will not consider the silent condition further.
RTs were analyzed in a mixed design analysis of variance (ANOVA) with SA (2500 or 0 msec) and distractor
type (color word or noncolor word) as within-subjects
factors and age (4- to 5-year-olds, 6- to 7-year-olds, 9- to
11-year-olds, or adults) as a between-subjects factor.1
The dependent variable was the median RT for naming
the color patches. The ANOVA showed main effects of
age [F(3,71) 5 48.98, MSe 5 265,525, p , .001], accounting for 50% of the total variance, distractor type
[F(1,71) 5 40.30, MSe 5 28,871, p , .001], accounting
for 2% of the variance, and SA [F(1,71) 5 184.18, MSe 5

Table 1
Median Reaction Times (in Milliseconds, With Standard Errors) as a Function of Auditory
Stimulus Condition (Stimulus Asynchrony [SA] and Distractor Type) and Participant Age
Auditory Stimulus Condition
2500-msec SA
Silent

Color Word

0-msec SA

Noncolor Word

Color Word

Noncolor Word

Age

Median

SE

Median

SE

Median

SE

Median

SE

Median

SE

45 years
67 years
911 years
Adult

1,481
1,096
,770
,597

132
65
33
22

1,361
,962
,680
,557

162
49
25
13

1,092
,913
,675
,571

78
41
22
15

2,015
1,495
,970
,630

197
98
69
29

1,620
1,273
,875
,606

123
62
49
25

NoteN 515 for each child group; N 5 30 for adults.

CROSS-MODAL STROOP
in adults and occurs at a wider range of SA conditions in
very young children than in older children and adults.
Finally, we examined proportional RT differences
when a color word distractor occurred, in comparison
with a noncolor word distractor. These proportions were
computed by dividing RT differences between color and
noncolor distractor conditions by RTs for the noncolor
distractor condition. For the 2500-msec SA condition,
the proportional differences were 24.6% for the 4- to 5year-olds, 5.4% for the 6- to 7-year-olds, 0% for the 9to 11-year-olds, and 2.5% (in the opposite direction) for
the adults. For the 0-msec SA condition, the proportional
differences were 24.4% for the 4- to 5-year-olds, 17.4%
for the 6- to 7-year-olds, 10.9% for the 9- to 11-yearolds, and 4.0% for the adults. Thus, we see further evidence of the decline in the effect size of the interference
effect with age.
Errors
Table 2 presents error rates for each age group as a
function of distractor type and SA. Trials were counted
as errors when the participant misnamed the color patch
(e.g., seeing the color green, but saying yellow). If a
participant stuttered (e.g., by saying ah prior to naming the color), failed to make a response within 7 sec, or
did not speak loudly enough for the voice key to activate,
or if the voice key was triggered prematurely (e.g., by
heavy breathing in the direction of the microphone or
kicking or hitting the desk), the trial was considered lost.
Of the 108 total test trials, the 4- to 5-year-olds averaged
5.7 lost trials, the 6- to 7-year-olds averaged 1.9 lost trials, the 9- to 11-year-olds averaged 1.8 lost trials, and the
adults averaged 1.4 lost trials. Error rates were computed
as proportions of completed trials, taking into consideration the number of lost trials for each condition.
Errors were analyzed in a mixed design ANOVA with
within-subjects factors of SA (2500 or 0 msec) and distractor type (color word or noncolor word) and a betweensubjects factor of age (4- to 5-year-olds, 6- to 7-yearolds, 9- to 11-year-olds, or adults). Error proportions
were arcsine transformed prior to statistical analysis, as
is recommended for proportional data (Cohen & Cohen,
1983). Overall, the main effect of age was significant
[F(3,71) 5 11.69, MSe 5 0.10, p , .001], and accounted
for 16% of the total variance. Although error rates were

relatively low at all ages, there was a significant decline

in errors with age. The only other significant effect in
the ANOVA was a two-way interaction of distractor type
and SA [F(1,71) 5 5.24, MSe 5 0.05, p , .05], accounting for 1% of the variance. When the distractor occurred
simultaneously with the color patch (0-msec SA), the
error rate for color word trials was significantly higher
than that for noncolor word trials [F(1,71) 5 4.91,
MSe 5 0.05, p , .05]. At the 2500-msec SA, error rates
for color word and noncolor word trials were statistically
equivalent. Planned comparisons of error rates for color
word and noncolor word trials were conducted for each
age group separately but revealed no significant differences, owing to low error rates and low statistical power.
DISCUSSIO N
In this study, we examined the development of selective attention, using the cross-modal Stroop task, testing
children 411 years of age as well as adults. Prior studies of selective attention (e.g., Dempster, 1992; Lane &
Pearson, 1982; Zelazo, Carter, Reznick, & Frye, 1997)
have demonstrated that the ability to tune out distracting
information and suppress inappropriate action develops
throughout childhood and adolescence. These studies
have not specifically addressed the issue of whether one
can ignore irrelevant spoken language while selectively
attending to a visual stimulus. The cross-modal Stroop
task proved to be an ideal measure for this purpose, in
that it enabled us to test individuals of a wide range of
ages, using identical procedures. We discovered that interference from hearing of irrelevant color words while
one is naming colors decreased markedly with age. Thus,
although noisy environments have been shown to disrupt
adult work performance (Banbury & Berry, 1998), our
results suggest that children may be even more drastically affected, for instance, by irrelevant spoken language
from television or adult conversation.
It is of interest that, in our pilot work with eight 3year-olds, we found that the majority of children at this
age could not perform the cross-modal Stroop task at all.
In the practice condition, when asked to name the color
patches in the absence of auditory distractors, the 3-yearolds performed accurately. However, when the distractors were presented along with the color patches, all but

Table 2
Mean Percentages of Errors (with Standard Errors) as a Function of Auditory Stimulus
Condition (Stimulus Asynchrony [SA] and Distractor Type) and Participant Age
Auditory Stimulus Condition
2500-msec SA
Silent

Color Word

363

0-msec SA

Noncolor Word

Color Word

Noncolor Word

Age

SE

SE

SE

SE

SE

45 years
67 years
911 years
Adult

2.7
2.9
1.5
1.5

1.0
0.8
0.8
0.5

5.1
1.1
0.9
0.7

1.9
0.7
0.6
0.4

4.0
3.3
1.0
0.3

1.4
1.0
0.6
0.2

8.2
4.3
1.9
1.0

2.5
1.4
1.1
0.5

6.5
1.9
0.8
0.7

2.7
1.2
0.5
0.4

NoteN 5 15 for each child group; N 5 30 for adults.

364

HANAUER AND BROOKS

2 of the children consistently repeated the words they

had heard, even though they knew that the rules were to
say what you see. Only these 2 children were able to
complete a single block of 36 test trials and produced
only 28% and 40% correct responses for this block. Attempts to train the 3-year-olds to ignore the spoken
words while naming the color patches led to child frustration and forced us to abandon testing younger children. Clearly, the auditory modality was highly salient
and nearly impossible for our youngest participants to
ignore. In comparison with the 3-year-olds, none of the
4- to 5-year-olds produced errors in over 20% of the trials with distractors, and only 3 children had error rates
over 10%.
Our discovery of a developmental trajectory in performance of the cross-modal Stroop task helps to clarify
the literature on the effect. With only adults being tested
in previous cross-modal Stroop experiments, some researchers have found reliable interference (Cowan, 1989;
Cowan & Barron, 1987; E. M. Elliott & Cowan, 2001;
E. M. Elliott et al., 1998), whereas others have not (Miles
& Jones, 1989; Miles et al., 1989). What is most pronounced in our data is the finding that the cross-modal
Stroop effect becomes progressively weaker with development and is a relatively small effect in adults. The fact
that a robust cross-modal Stroop effect appears when
children are tested runs counter to Miles and his colleagues assertion that interference from irrelevant spoken
words occurs only in tasks involving short-term memory.
E. M. Elliott et al. (1998) considered two explanations
of the cross-modal Stroop effect. They argued that semantic interference might occur in a temporary speech
buffer unless there is sufficient time for the spoken word
to be suppressed. Consistent with this hypothesis was
their observation that the cross-modal Stroop effect fully
disappeared in adults if the auditory distractor was presented 500 msec in advance of the color patch. E. M. Elliott et al., however, could not rule out an alternative hypothesis that the cross-modal Stroop effect stems from
concurrent lexical processing in the auditory and the visual modalities (Schriefers et al., 1990). According to
the concurrent processing account, the interference effect occurs when there is temporal overlap in accessing
the lemmas for the spoken word and name of the color
patch and stems from competition among lemmas during
lexical access.
To test these two accounts, we varied the timing of
presentation of the auditory distractor, relative to presentation of the color patch. We hypothesized that the
cross-modal Stroop effect would be seen over a wider
range of SA in children than in adults, owing to differences in information-processing speed (Kail, 1991,
1992), coupled with less efficient suppression mechanisms (Bjorklund & Harnishfeger, 1990; Livesey &
Morgan, 1991; Tipper, Bourque, Anderson, & Brehaut,
1989). In one condition, the color patch occurred simultaneously with the distractor. With simultaneous presentation, both suppression and concurrent processing ac-

counts would predict a cross-modal Stroop effect to

occur, replicating E. M. Elliott et al. (1998). In the other
condition, the color patch was delayed and appeared
500 msec after the offset of the distractor. We suspected
that a 2500-msec stimulus offset interval would be sufficient to prevent temporal overlap in lexical access for
the auditory distractor and the color patch. Thus, the occurrence of a cross-modal Stroop effect, even in the
youngest children, would be inconsistent with the concurrent processing account. Our results favored the suppression account: Our youngest participants showed a
large interference effect at both SAs, whereas older participants showed interference only with simultaneous
presentation of the distractor and the color patch. Thus,
delaying presentation of the color patch by 500 msec
provided older children and adults, but not younger children, with sufficient time to fully suppress the irrelevant
stimulus.
One possible objection to this interpretation of our results is that, given the relative inefficiency of word retrieval processes in children, a 2500-msec SA might not
be sufficiently long to prevent simultaneous lexical activation of the distractor and the color patch names in very
young children. Under normal listening conditions, lexical access occurs rapidly, even in young children. This
has been shown in recent on-line studies of the development of word recognition, indicating that children as
young as 24 months of age access the meanings of familiar spoken words prior to their offsets (Fernald, Pinto,
Swingley, Weinberg, & McRoberts, 1998; Swingley,
Pinto, & Fernald, 1999). Word recognition based on partial input has been observed in 24-month-olds even when
the words in the response set share initial consonants, as
in doll and doggie (Swingley et al., 1999). In gating studies examining word recognition for the larger vocabularies characteristic of school-aged children, L. L. Elliott,
Hammer, and Evan (1987) and Walley, Michela, and
Wood (1995) demonstrated that 5- to 7-year-olds readily
identify words on the basis of partial input. For example,
Walley et al. estimated that kindergartners (mean age,
6 years) isolated target words after hearing only 66% of
the sound pattern (i.e., for words with an average length
of 617 msec, the words were identified at 406 msec). Although informative, these research f indings unfortunately do not indicate how long lexical activation lingers
in the developing system. Thus, it is impossible for us to
guarantee that in the cross-modal Stroop task, where participants could use the distractor as a cue to anticipate
the onset of the color patch, there would be no temporal
overlap in lexical activation of the target and the distractor names.
The suppression account of the cross-modal Stroop effect is supported by related findings in the developmental literature on the phenomenon of negative priming.
Negative priming occurs when a participant must respond
to a stimulus that was to be ignored on the preceding trial.
Suppose, for example, in the original Stroop task, that on
Trial 1 a participant saw the word blue printed in green

CROSS-MODAL STROOP
and on Trial 2, he or she saw the word red printed in blue.
Identifying blue as the print color on Trial 2 would take
longer than it would if it had not been the ignored stimulus in the previous trial. It has been argued (e.g., Driver &
Baylis, 1993; Neill, 1977; Neill & Westberry, 1987; Tipper
et al., 1989) that the magnitude of the negative priming effect is a direct reflection of the efficiency of suppression
mechanisms and is negatively correlated with susceptibility to distraction. Tipper et al. proposed that the developmental differences seen in studies of selective attention
(e.g., Doyle, 1973; Gerstadt et al., 1994; Jerger et al.,
1988; Zuckier & Hagan, 1978) are due to childrens underutilization of the suppression mechanism. In support
of this idea, Tipper et al. observed that 7- to 8-year-olds
showed greater Stroop interference than did adults and
that adults showed greater negative priming than did 7- to
8-year-olds. Driver and Baylis found similar tradeoffs in
adults between the magnitude of the interference effect and
the size of the negative priming effect in a cross-modal
Stroop-like paradigm. It is noteworthy in this regard that
Gernsbacher and her colleagues (e.g., Gernsbacher &
Faust, 1991; Gernsbacher, Varner, & Faust, 1990) have
observed that the efficiency of suppression mechanisms
varies considerably among adults and is a strong predictor of individual differences in general comprehension
skill (Gernsbacher, 1997). For example, less skilled comprehenders have been shown to less efficiently suppress
inappropriate meanings of homophones (e.g., reject CALM
after reading He had a lot of patients), as well as information across pictorial and printed word modalities (i.e.,
ignore words superimposed on pictures or pictures superimposed on words). These findings led us to predict
that less skilled comprehenders would experience greater
interference in the cross-modal Stroop task, at a wider
range of SAs, than would more skilled comprehenders.
It is also important to note that recent studies (Brooks
& MacWhinney, 2000; Brooks, MacWhinney, & Evans,
1995), using the cross-modal pictureword interference
paradigm, have demonstrated, in both children and adults,
that related words presented in the auditory modality can
prime (i.e., facilitate), as well as inhibit, picture-naming
latencies. Brooks et al. examined associative priming at
a range of SAs, and found that picture-naming latencies
for pictures paired with associated words (e.g., hearing cat
when shown a picture of a dog) were significantly faster
than latencies for pictures paired with unrelated words.
This effect was robust in children and adults, even with
the auditory stimulus occurring 550 msec in advance of
the picture. Thus, participants did not suppress activation
of auditory stimuli that were potentially relevant to the
task at hand. Together with the present study, Brooks
et al.s results indicate that suppression is a flexible,
nonobligatory process that may be invoked under conditions in which it is beneficial, with developmental
change occurring in the efficiency and flexibility of the
underlying mechanisms.
In addition to development in suppression, other executive factors likely contributed to the failures of selective attention observed in the cross-modal Stroop task.

365

Executive skills essential for purposeful goal-directed

activity, such as planning and monitoring action, selfregulation, impulse control, holding goals in active
memory, maintaining proper response sets, and making
use of feedback (Anderson, 1998; Gnys & Willis, 1991),
are thought to be mediated by the frontal cortex and the
prefrontal cortex, which are the slowest maturing areas
in the brain (Anderson, 1998; Bell & Fox, 1992; Chugani,
Mazziotta, & Phelps, 1993). Researchers (e.g., Diamond
& Taylor, 1996; Gnys & Willis, 1991; Levin et al., 1991;
Passler, Isaac, & Hynd, 1985; Zelazo et al., 1997) have
used such tasks as the Tower of London, the Wisconsin
Card Sorting Task, the gono-go task, and delayed alternation to study developmental change in semantic associations, concept formation, mental flexibility, planning,
and problem solving. We would argue that development
in executive functioning greatly affects childrens ability
to play experimental games properly and most likely influenced performance on the cross-modal Stroop task.
REFERENCES
Anderson, V. (1998). Assessing executive function in children: Biological, psychological, and developmental considerations. Neuropsychological Rehabilitation, 8, 319-349.
Baddeley, A. (1986). Working memory. Oxford: Oxford University
Press.
Baddeley, A. (1997). Human memory: Theory and practice. Boston:
Allyn & Bacon.
Banbury, S., & Berry, D. C. (1998). Disruption of office-related tasks by
speech and office noise. British Journal of Psychology, 89, 499-517.
Bedi, G., Halperin, J. M., & Sharma, V. (1994). Investigation of
modality specific distractibility in children. International Journal of
Neuroscience, 74, 79-85.
Bell, M. A., & Fox, N. A. (1992). The relations between frontal brain
electrical activity and cognitive development during infancy. Child
Development, 63, 1142-1163.
Bjorklund, D. F., & Harnishfeger, K. K. (1990). The resources construct in cognitive development: Diverse sources of evidence and a
theory of inefficient inhibition. Developmental Review, 10, 48-71.
Brooks, P. J., & MacWhinney, B. (2000). Phonological priming in
childrens picture naming. Journal of Child Language, 27, 335-366.
Brooks, P. J., MacWhinney, B., & Evans, J. (1995, March). Phonological and semantic priming in childrens picture naming. Poster
presented at the Biennial Meeting of the Society for Research in
Child Development, Indianapolis, IN.
Carroll, J. B., Davies, P., & Richman, B. (1971). The American Heritage word frequency book. Boston: Houghton Mifflin.
Chugani, H., Mazziotta, J. C., & Phelps, M. E. (1993). Positron
emission tomography study of human brain functional development.
In M. H. Johnson (Ed.), Brain development and cognition: A reader
(pp. 125-143). Malden, MA: Blackwell.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.
Cohen, J., MacWhinney, B., Flatt, M., & Provost, J. (1993).
PsyScope: An interactive graphic system for designing and controlling experiments in the psychology laboratory using Macintosh computers. Behavioral Research Methods, Instruments, & Computers,
25, 257-271.
Comalli, P., Wapner, S., & Werner, H. (1962). Interference effects of
Stroop color word test in childhood, adulthood, and aging. Journal of
Genetic Psychology, 100, 47-53.
Cowan, N. (1989). The reality of cross-modal Stroop effects. Perception & Psychophysics, 45, 87-88.
Cowan, N., & Barron, A. (1987). Cross-modal, auditoryvisual
Stroop interference and possible implications for speech memory.
Perception & Psychophysics, 41, 393-401.

366

HANAUER AND BROOKS

Cowan, N., & Saults, J. S. (1995). Memory for speech. In H. Winitz

(Ed.), Human communication and its disorders (Vol. 4, pp. 81-170).
Timonium, MD: York.
Dempster, F. N. (1992). The rise and fall of the inhibitory mechanisms:
Toward a unified theory of cognitive development and aging. Developmental Review, 12, 45-75.
Diamond, A., & Taylor, C. (1996). Development of an aspect of executive control: Development of the abilities to remember what I said
and to Do as I say, not as I do. Developmental Psychobiology, 29,
315-334.
Doyle, A. (1973). Listening to distraction: A developmental study of
selective attention. Journal of Experimental Child Psychology, 15,
100-115.
Driver, J., & Baylis, G. C. (1993). Cross-modal negative priming and
interference in selective attention. Bulletin of the Psychonomic Society, 31, 45-48.
Ehri, L. C. (1976). Do words really interfere in naming pictures? Child
Development, 47, 502-505.
Elliott, E. M., & Cowan, N. (2001). Habituation to auditory distractors in a cross-modal, colorword interference task. Journal of Experimental Psychology: Learning, Memory, & Cognition, 27, 654667.
Elliott, E. M., Cowan, N., & Valle-Inclan, F. (1998). The nature of
cross-modal colorword interference effects. Perception & Psychophysics, 60, 761-767.
Elliott, L. L., Hammer, M. A., & Evan, K. E. (1987). Perception of
gated, highly familiar spoken monosyllabic nouns by children,
teenagers, and older adults. Perception & Psychophysics, 42, 150157.
Fernald, A., Pinto, J. P., Swingley, D., Weinberg,A., & McRoberts,
G. W. (1998). Rapid gains in speed of verbal processing by infants in
the 2nd year. Psychological Science, 9, 228-231.
Gernsbacher, M. A. (1997). Two decades of structure building. Discourse Processes, 23, 265-304.
Gernsbacher, M. A., & Faust, M. E. (1991). The mechanism of suppression: A component of general comprehension skill. Journal of
Experimental Psychology: Learning, Memory, & Cognition, 17, 245262.
Gernsbacher, M. A., Varner, K. R., & Faust, M. E. (1990). Investigating differences in general comprehension skill. Journal of Experimental Psychology: Learning, Memory, & Cognition, 16, 430-445.
Gerstadt, C. L., Hong, Y. J., & Diamond, A. (1994). The relationship
between cognition and action: Performance of children 3 1/27 years
old on a Stroop like daynight test. Cognition, 53, 129-153.
Gnys, J. A., & Willis, W. G. (1991). Validation of executive function
tasks with young children. Developmental Neuropsychology, 7, 487501.
Green, E. J., & Barber, P. J. (1981). An auditory Stroop effect with
judgments of speaker gender. Perception & Psychophysics, 30, 459466.
Green, E. J., & Barber, P. J. (1983). Interference effects in an auditory
Stroop task: Congruence and correspondence. Acta Psychologica,
53, 183-194.
Hays, W. L. (1988). Statistics (4th ed.). Fort Worth, TX: Holt, Rinehart
& Winston.
Jerger, S., Martin, R. C., & Pirozzolo, F. J. (1988). A developmental study of the auditory Stroop effect. Brain & Language, 35, 86104.
Jerger, S., Pearson, D. A., & Spence, M. J. (1999). Developmental
course of auditory processing interactions: Garner interference and
Simon interference. Journal of Experimental Child Psychology, 74,
44-67.
Kail, R. (1991). Processing time declines exponentially during childhood and adolescence. Developmental Psychology, 27, 259-266.

Kail, R. (1992). Processing speed, speech rate, and memory. Developmental Psychology, 28, 899-904.
Lane, D. M., & Pearson, D. A. (1982). The development of selective
attention. Merrill-Palmer Quarterly, 28, 317-337.
Levin, H. S., Culhane, K. A., Hartmann, J., Evankovich, K., Mattson, A. J., Harward, H., Ringholz, G., Ewing-Cobbs, L., &
Fletcher, J. M. (1991). Developmental changes in performance on
tests of purported frontal lobe functioning. Developmental Neuropsychologist, 7, 377-395.
Livesey, D. J., & Morgan, G. A. (1991). The development of response
inhibition in 4- and 5-year-old children. Australian Journal of Psychology, 43, 133-137.
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative review. Psychological Bulletin, 109, 163-203.
Miles, C., & Jones, D. M. (1989). The fallacy of the cross-modal
Stroop effect: A rejoinder to Cowan (1989). Perception & Psychophysics, 45, 85-86.
Miles, C., Madden, C., & Jones, D. M. (1989). Cross-modal, auditory
visual Stroop interference: A reply to Cowan and Barron. Perception
& Psychophysics, 45, 77-81.
Neill, W. T. (1977). Inhibitory and facilitatory processes in selective
attention. Journal of Experimental Psychology: Human Perception
& Performance, 3, 444-450.
Neill, W. T., & Westberry, R. L. (1987). Selective attention and the
suppression of cognitive noise. Journal of Experimental Psychology:
Learning, Memory, & Cognition, 13, 327-334.
Passler, M. A., Isaac, W., & Hynd, G. W. (1985). Neuropsychological development of behavior attributed to frontal lobe functioning in
children. Developmental Neuropsychology, 1, 349-370.
Rosinski, R. R., Golinkoff, R. M., & Kukish, K. S. (1975). Automatic
semantic processing in a pictureword interference task. Child Development, 46, 247-253.
Schiller, P. H. (1966). Developmental study of colorword interference. Journal of Experimental Psychology, 72, 105-108.
Schriefers, H., Meyer, A. S., & Levelt, W. J. (1990). Exploring the
time course of lexical access in language production: Pictureword
interference studies. Journal of Memory & Language, 29, 86-102.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions.
Journal of Experimental Psychology, 18, 643-662.
Swingley, D., Pinto, J. P., & Fernald, A. (1999). Continuous processing in word recognition at 24 months. Cognition, 71, 75-108.
Tipper, S. P., Bourque, T. A., Anderson, S. H., & Brehaut, J. C.
(1989). Mechanisms of attention: A developmental study. Journal of
Experimental Child Psychology, 48, 353-378.
Walley, A. C., Michela, V. J., & Wood, D. R. (1995). The gating paradigm: Effects of presentation format on spoken word recognition by
children and adults. Perception & Psychophysics, 57, 343-351.
Zelazo, P. D., Carter, A., Reznick, J. S., & Frye, D. (1997). Early
development of executive function: A problem-solving framework.
Review of General Psychology, 1, 198-226.
Zuckier, H., & Hagan, J. W. (1978). The development of selective attention under distracting conditions. Child Development, 49, 870873.
NOTE
1. For all analyses, we computed effect sizes by using eta-squared,
which indicates the proportion of variance accounted for by the effect
(Hays, 1988). These effect sizes are provided as percentages of the total
variance.
(Manuscript received August 27, 2001;
revision accepted for publication July 30, 2002.)