Archives of Clmicd Neuropsychology.

Vol. 11, No. 1, pp. l-9, 1996

Copyright Q 1995 National Academy of Neuropsychology
Printed in the USA. All rights reserved

Pergamon

0887-6177/%

$15.00 + .OO

Analysis of the Money Road-Map Test

Performance in Normal and
Brain-Damaged Subjects
Guy Vingerhoets,
Department

Engelien

of Psychiatry

and

Lannoo,
Neuropsychology,

and Sabien
University

Bauwens
of

Gent

The Money Road-Map

Test (MRMT)
is a paper and pencil assessment of lef-right
discrimination.
Some of the answers require an egocentric
mental rotation in space. In a first experiment
with 63
normal adults. we found that the accuracy
and speed of the left-right
decision process significantly decreased
with an increasing
degree of mental rotation required. A division of the MRMT turns
in three categories
with increasing
mental rotation
was proposed.
In a second study (n = 50),
patrents
with predominantly
parietal
brain lesions performed
significantly
worse than patients
with predominantly
frontal
lesions. The si,@ficant
group difference
in total error score was especially due to the error scores of turns requrring
mental rotation.
The turn type analysis did not
contribute
to the lateralization
of the lesionr

The Standardized Road-Map Test of Direction Sense (Money, Alexander & Walker,
1965), generally referred to as the Money Road-Map Test (MRMT), measures left-right
orientation with and without egocentric mental rotation in space. The MRMT is a paper
and pencil test of left-right discrimination in which the subject traces a dotted pathway
through a city map, and indicates the direction taken at each turn, left or right. Some of
the answers require an egocentric mental rotation, because the dotted pathway follows
an erratic trace both away from and towards the subject, who is not allowed to turn the
map or to make head and body movements to give the correct answer.
In analyzing the performance on the Road-Map Test we should be aware that different
cognitive functions are assessed, namely the left-right discrimination per se and (if
required) the ability to perform an egocentric mental rotation in space to give the correct
left-right response. In examining the MRMT performances of four groups of patients with
localized lesions (right parietal and temporal, left frontal and temporal), Butters, Soeldner,
and Fedio (1972) suggested that the exceptionally high number of errors in the left frontal
patients reflected the tests conceptual demands for making mental spatial rotations. It was

The authors thank the reviewers of this manuscript for their helpful suggestions.
Address correspondence to: Guy Vingerhoets, Department of Psychiatry and Neuropsychology,
University Hospital Gent 4K3, De Pintelaan 185, B-9000 Gent, Belgium.
I

stated, however, that the absence of left parietal and right frontal groups in this study warranted further research to account for the left-right confusion in patients with left hemisphere damage (Lezak, 1983). In contrast to the findings of Butters et al. (1972), research
on the lateralization of mental rotation skill has consistently documented a right hemispheric, and especially parietal, specialization for mental rotation (Corballis & Sergent,
1989; Deutsch, Bourbon, Papanicolaou, & Eisenberg, 1988; Ditunno & Mann, 1990;
Rosier, Schumacher, & Sojka, 1990) whereas the disability of left-right discrimination is
normally associated with left parietal dysfunction (Ratcliff & Newcombe, 1973; Roeltgen,
Sevush, & Heilman, 1983; Strub & Geschwind, 1983).
To account for the fact that left-right discrimination and mental rotation are two different
abilities, several classifications of turn types have been proposed (Flicker, Ferris, Crook,
Reisberg, & Bartus, 1988; Money et al., 1965). Research evidence (Money et al., 1965) and
our own clinical experience with this test, pointed to a differentiation between no rotation
turns (i.e., turns in which no mental rotation is required to state a left-right decision, Figure
lA), half rotation turns (Figure 1B and C), and full rotation turns (Figure 1D). It can also be
argued that the left-right decision in a half rotation turn is more difficult when the half rotation decision follows a full rotation (Figure IC) than when it follows no rotation (Figure 1B).
Research on mental rotation skill repeatedly showed a significant sex difference favoring
males, both under timed test conditions, and conditions in which the effects of time on performance were minimized or eliminated (Gallagher & Johnson, 1992; Kerns & Bembaum,
1991; Resnick, 1993; Thomas & Kail, 1991). It also was argued that gender differences in
mental rotation skill are of limited practical value due to the large variability of the cognitive skill in both sexes (Resnick, 1993).
By assessing the response time and the accuracy
score for each turn of the Road-Map
Test in both male and female normal subjects, we can evaluate several hypotheses suggested
by the literature. First, if the requirement of mental rotation over and above left-right discrimination makes the cognitive task more complex, we can expect lower accuracy scores
A

NR

HR-A

HR-6

NR

= no

HR A
HR
FR
FIGURE

FR

rotation

half

rotation

following

no rotation

B =

half

rotation

folIowIng

full

full

I. Examples

rotation

rotation
of different

turn

types

in the Money

Road-Map

Test.

and longer response times in left-right decisions requiring an increasing degree of mental
rotation. Second. if a left-right decision in a half rotation turn is more difficult when the half
rotation decision follows a full rotation than when it follows no rotation, we can expect that
a further division in four turn types (Figure 1) will show that accuracy scores and response
times are worse when a right or left approach follows a full rotation movement than when a
right or left approach follows a no rotation movement. Third, based on the sex differences
found in the research on mental rotation skill, we can expect males to perform faster and
more accurately than females in turns where a mental rotation to make a correct left-right
judgement is required.
In a second study, we will compare the error scores for the different turn types of brain
damaged patients with either left frontal, right frontal, left parietal or right parietal lesions.
Because previous research pointed to a parietal involvement in left-right discrimination either
with or without mental rotation, we would hypothesise that patients with parietal lesions
show a higher error score as opposed to patients with frontal lesions. Because an intact left
parietal lobe should be able to make adequate left-right discriminations when no mental rotation is required, we can also expect that patients with right parietal lesions should show a
lower error score on the no-rotation turns as opposed to patients with left parietal lesions.
EXPERIMENT

Method

Subjects. Sixty-three healthy volunteers, consisting of young graduate and undergraduate

university students agreed to take part in the study. The mean age for all subjects was 21.9
(age range 19-29). Twenty-six (41%) were males (mean age 21.3) and 37 (59%) were
females (mean age 22.3).
Each subject was tested individually. The subjects were instructed to follow the
broken line, just as if they were tracing a map. and to indicate aloud whether they turned to the
left or to the right. Their verbal responses were audiotaped. They were free to use their finger
to trace the line, but we allowed no head and body movements. The preliminary practice route
of three turns on the bottom right hand side of the test was given to assure a complete comprehension of the instructions. The subjects were then pointed to the beginning of the standard
test route. The Road-Map sheet was covered and the subjects were instructed to do the test as
fluently, but also as accurately as possible. The test was uncovered, and the subjects were told
to start.

Procedure.

Design. The answer\ were recorded, and the audiotape was analyzed for accuracy and
response time by use of a computerized chronometer. Response time was defined as the
time lapse between two consecutive answer\
The 32 turns of the Money Road-Map standard text route were numbered and classified
by each author separately in one of the three followmg categories: (a) the correct left-right
indication of the turn requires no mental rotation (no rotation, NR), (b) the correct left-right
indication of the turn requires a mental rotation of approximately 90 (half rotation, HR), (c)
the correct left-right indication of the turn requires a mental rotation of more than 90 (but
no more than 180) (full rotation, FR).
The category of the half rotations was turther divided into half rotations following no
mental rotation (HR-A), and half rotations following a full mental rotation (HR-B). A comparison of the separate classifications of e;~ch author showed a complete agreement of the
division in turn types

We found no significant difference between both sexes on the variable of age, and on the
performance of a Dutch version of the National Adult Reading Test (NART), that offers an
estimate of verbal intelligence.
time. A three (turn type: NR/HR/FR) by two (sex) ANOVA shows a significant
difference in response time for the three turn types [F(2, 114) = 99.91, p < .OOl], with gradually longer response times for an increasing degree of mental rotation required.
Newman-Keuls
post hoc analysis comparing means indicates significant differences
between FR and both NR and HR response times. There is also a significant main effect of
sex [F( 1, 57) = 5.38, p < ,051, with consistent longer response times for females. We found
no significant interaction effect between sex and turn type.
A four (turn type: NRkIR-A/HR-B/FR)
by two (sex) ANOVA shows similar significant
main effects in response time for sex [F( I, 57) = 5.70, p < .05] and turn type [F(3, 171) =
52.54, p < .OOl] (Figure 2). Newman-Keuls post hoc analysis comparing means indicates
significant differences between FR and both NR and HR-A only. Again, there is no significant interaction effect between sex and turn type.

Response

Accuracy data. A three (turn type: NR/HR/FR) by two (sex) ANOVA shows significant differences in the percentage of correct answers for the three turn types [F(2, 118) = 30.73,
p < .OOl], with a gradually decreasing accuracy score for an increasing degree of required
mental rotation. Newman-Keuls post hoc analysis comparing means indicates significant
differences between each turn type. There is a significant main effect of sex [F(l, 59) =
4.07, p -C.05] with a higher female error score. We found no significant interaction effect
between sex and turn type.
Again, we found similar main effects for a four (turn type: NR/HR-A/HR-B/FR) by two
(sex) ANOVA, indicating a significant effect of turn type [F(3, 177) = 18.70, p < .OOl] and
3

IO0

rer

r,

2.5

.
*

95

.
.
I.5

.
.

A males
q females
FIGURE

2. Response

times and accuracy

scores

for different

turn

types.

a significant effect of sex [F( 1, 59) = 5.63, p = .05] on the percentage of correct answers
(Figure 2). Newman-Keuls post hoc analysis comparing means reveals significant differences between FR and all the other turn types, and between HR-B and NR. There is no
interaction effect between sex and turn type.
Discussion
Our first hypothesis, namely an increasing response time and a decreasing accuracy score
with an incremental degree of mental rotation required, was confirmed for both dependent
variables in both a three and a four type division of turns. Post hoc analysis however,
showed no significant differences between HR-A and HR-B results for both response time
and accuracy score in the four type division of turns. The hypothesis that half rotation turns
following a full rotation are more difficult than half rotation turns following no rotation was
not statistically confirmed.
Marked differences between turn types were found for the three-turn scoring system.
This was most notable for the accuracy scores for which significant differences between all
three turn types were found. Because clinical interpretation of the Money Road-Map Test is
usually based on the error score only (and not on response time), this result confirms the
value of a classification in three turn types.
Based on the mental rotation research, we also hypothesized a significant sex difference in
left-right discrimination requiring mental rotation favoring males. This hypothesis was confirmed for both response time and accuracy score.

EXPERIMENT

Method
Subjects. Fifty brain-damaged patients with localized cerebral lesions, who had been
referred to our department for neuropsychological testing during the period from 1983 to
1993, were selected for the study. Only the protocols of the patients with either a left
frontal, right frontal, left parietal, or a right parietal lesion were selected. A prerequisite for
selection was that the reported lesions were documented by reliable neurological and neuroradiological data (computerized tomography, CT), and that the detected lesions were
restricted to the cerebral lobe involved. However, because of the inherent limits of the CT
neuroimaging procedures (Damasio & Damasio, 1989), the existence of micro-structural or
nonmorphological brain damage outside the specified region can not be excluded. In some
of our cases, and especially in the cases with head trauma, the existence of more widespread
brain damage is even likely, although no objective evidence for such damage was found.
These considerations necessitate a cautious interpretation of the results.
All patients were alert and medically stable at the time of the examination. Of the 13
patients with left frontal lesions, 7 suffered contusion following head trauma, 4 suffered
intracerebral hematoma, 1 suffered a posttraumatic extradural hematoma and a subdural
hygroma for which neurosurgery was performed, and 1 was operated on for a brain tumor. Of
the 12 right frontal lesion patients, 5 suffered contusion following head trauma, 3 suffered an
intracerebral hematoma, 3 had a brain tumor (2 were operated and 1 was inoperable), and 1
patient suffered from epileptic seizures with a clear right frontal focus (as was illustrated by
EEG and a right frontal hypodense spot on MRI). Of the 12 left parietal lesion patients, 5 suffered contusion following head trauma, I was operated on for a meningeal abscess, 3 suffered
intracerebral hematoma, and 3 suffered from stroke. Of the 13 right parietal lesion patients, 2

suffered contusion following head trauma . 2 suffered intracerebral hematoma, 4 were operated
on for removal of brain tumors, I suffered from an encephalographically localized irritative
focus after an epidural hematoma, and 4 patients suffered from stroke.
All 50 patients were individually assessed with the Money Road-Map Test as a
part of a neuropsychological investigation for which they were referred by their physicians.
The Money Road-Map Test was given as a paper and pencil test and they were instructed to
follow the broken line, just as if they were tracing a map, and to indicate by writing down an
L for left or an R for right whether they turned to the left or to the right. Turning the paper or
head and body movements were not allowed. The preliminary practice route of three turns
was used to assure a complete comprehension of instructions. The patients were then pointed
to the beginning of the standard test route. The performance was not timed, and the patients
were told to do the test as accurately as possible.

Procedure.

The 32 turns of the Money Road-Map Test were classified in three turn types, as
described earlier. The number of errors for each category of turns was used as the dependent variable.

Design.

Table 1 shows the subject variables for each group. We found no significant differences
in gender distribution, age, and years of education between the four groups. This was also
true when the total frontal and total parietal groups were compared.
When we compared the total frontal and total parietal groups (regardless of the side of
the lesion), we found a significant difference for the total number of errors on the RoadMap Test [t(48) = -2.69, p = .Ol]. The frontal group (n = 25) averaged I.8 errors, whereas
the parietal group (n = 25) averaged 5.2 errors. If we looked at the different turn types separately, we found no significant difference between both groups for the no rotation turns, but
we found significant differences for the half rotation turns [t(48) = -2.39, p < ,051 and the
full rotation turns [t(48) = -2.58, p = ,011.
In addition, we also evaluated if group differences in error score could be found when the
side of the lesions was taken into account. The average number of errors on the total test
and for each category of turn type for each type of lesion are listed in Table 2.
A one-way analysis of variance showed a trend towards differences in total error score
between the four groups, but this trend did not reach statistical significance [F(3, 46) = 2.61,
p < .06]. Analysis of the different turn types separately only showed a significant difference
for the half rotation turn errors between the four groups [F(3, 47) = 3.51, p < .05].
Newman-Keuls post hoc analyses comparing means revealed significant differences between
the left parietal and both the left and right frontal groups and the right parietal group. A comparison of the error scores of males (n = 37i and females (n = l3), regardless of lesion,

Subject

Variables

TAH1.E I
for Each Lesion

Group:

Mean

Age
1s 5 ( I X.0)
76.X (12.7)
3x.x (I 5.2)
4l.h(lh.h)

(SD)
Educatm
10.x
10.2
10.9
11.8

(2.5)
(1.6)
(2.8)
(2.1)

Average

Number

TABLE
2
on the Total Test and for Each Turn
Lesion Group:
Mean (SD)

Total Error
Score

Lesion
Left
Right
Leti
Right

of Errors

frontal
frontal
parietal
parietal

2.2
I .4
6.0
4.5

(2.5)
(2.9)
(7.7)
(3.41

Type According

Half-Rotation
Errors
0.2
0.2
I.1
0.2

(0.41
(0.6)
(I.91
(0 4,

0.5
0.3
2.5
0.9

(I .O)
(0.7)
(3.2)
(1.2)

to

Full-Rotation
Errors
1.5
0.9
2.4
3.4

(1.5)
(I 3)
(3.1)
(2.7)

showed no significant difference for the total error score and for the error scores of the different turn types as a function of sex of subject.

The results confirm the hypothesis that patients with parietal lesions make more errors on
the Money Road-Map Test than do patients with frontal lesions. Further analysis of this
finding by using the proposed division in three turn types shows that the difference is predominantly made by the errors for half rotation turns and the full rotation turns. Errors predominate in turns where the performance requires mental rotation over and above normal
left-right discrimination.
When the frontal and parietal groups are further divided by left or right hemispheric
localisation of the lesions, the differences in total error score are no longer significant, and
our second hypothesis, namely a lower no-return error score in patients with right parietal as
opposed to patients with left parietal lesions was not statistically confirmed.
Our findings do not support the suggestion of Butters et al. (1972). in that the high error
score of their left frontal group reflected the tests conceptual demands for making mental spatial rotations. Our left frontal group showed a relatively low error score as opposed to both
parietal groups. Moreover, the frontal groups distinguished themselves from the parietal
groups by significantly lower error scores in turns where mental rotation was involved. The
results of our study show evidence for a parietal involvement of mental rotation skill,
although a right hemispheric predominance for this skill could not be confirmed. It must be
stressed however, that these results were obtained in a selected sample of patients, and that
several remarks have to be made to allow for a critical interpretation of the results. First, all
patients suffered from brain lesions that were macrostructurally confined to either left frontal,
right frontal, left parietal or right parietal lobe, as evidenced by CT scans. Given the limits of
computerized tomography, microstructural or nonmorphological brain damage outside the
specified region cannot be excluded. Second, we have no data on the intellectual abilities of
the patients, or the intellectual comparability of the groups. However, the level of education as
expressed in years of schooling is highest in the poorly performing parietal groups, although
statistical analysis shows no significant difference in educational level between the groups.
In view of these remarks, we conclude that the Money Road-Map Test can be a useful
instrument in the differentiation of predominantly frontal versus predominantly parietal
involvement in a restricted number of cases. The proposed division in turn types is not essential
for this differentiation, but it can offer additional support for a possible parietal involvement if
the high number of errors on turns requiring a mental spatial rotation is in contrast with a relatively normal performance in left-right discrimination on turns requiring no mental rotation.
Finally, we also agree with the comment of Resnick ( 1993) that gender differences in
mental rotation skill are of limited practical value, except for the extremes, due to the large

G. Vingerhoets.

E. lmtnoo.

and

S. Bauwens

variability of this cognitive skill in both sexes. A comparison of male and female performance in our sample, regardless of the lesion, found no significant differences in error score
as a function of sex.

CONCLUSION
Presenting the Money Road-Map Test performance as a unitary measure would disregard
the fact that at least two different cognitive processes, namely left-right discrimination and
mental spatial rotation, are involved. In a first experiment we showed that the accuracy and
the speed of the left-right decision process differs significantly with the degree of mental spatial rotation required. A division of the MRMT-turns in three categories (absent, moderate,
and high degree of mental rotation) was proposed. That mental rotation was, indeed, fundamentally involved was suggested by identical gender differences for turns requiring mental
rotation as could be predicted on the basis of research on mental rotation skill with other,
more classic, mental rotation tasks.
In a second study, the clinical relevance of this proposal was evaluated in a number of
patients with neuroradiologically confirmed focal brain lesions. The Money Road-Map
Test performance of these subjects showed that the patients with predominantly parietal
lesions performed significantly worse than the patients with predominantly
frontal
lesions. A more detailed analysis with the division of turn types showed that the significant group difference in total error score was predominantly due to the error scores of
turns requiring mental spatial rotation. The question remains whether the qualifying factor
is the mental spatial rotation task itself, or if it is the superposition of mental rotation on a
task of left-right discrimination. More research is needed to corroborate and clarify these
findings. The turn types analysis did not otfer a diagnostic contribution to the lateralization of the lesions.

REFERENCES
Butters.

N.. Soeldner.

C., & brdio,

P. (1972~.

Comparlwn

of parietal

and frontal

lobe spatial

deficits

in man:

Extrapersonal
v%. personal (egocentric)
space. trrc~eptual
artd Motor Skiffs. 34, 27-34.
Corballis,
M. C., & Sergent. J. (1989). Hemispherw
specialisation
for mental rotation. Cortex, 25, 15-25.
Damasio, H., & Damasio, R. D. (1989). Lesion anulws
in neuropsychology.
New York: Oxford University
Deutsch,
G., Bourbon,
W. T., Papanicolaou,
A. C.. & Eisenberg,
H. M. (1988).
activation
of regional cerebral blood flow. Neurop.sw.holo~ia.
26. 445452.

Visuospatial

Dltunno,
P. L., & Mann, V. A. (1990). Right hemisphere
hpecialisatwn
for mental
damaged subjects. Conex, 26, 177-l 88.
Flicker, C., Ferris, S. H., Crook, T.. Reisberg, B., (L Hartuh. K. T (I 9X8). Equivalent
mal aging and Alzheimers
Gallagher,
S. A., & Johnson,
adolescents.
G&d
Child
Kerns.

K. A., & Berenbaum.

Sex dlftcrcnce\

rotation

in normals

spatial-rotation

via

and brain

deficits

in nor-

Erprrimenrrrl
Neuropsychology,
10, 387-399.
Ilmits on performance
of mental rotation by gifted

disease. Journal of C(rnr~x~/ md

E. S. (19Y2). The elfect of time
Quarterly,
36. 1% 22
S. A. (1991)

Press.

tasks compared

In spatial

ability

in children.

Behavior

21,

Genetics,

383-396.
Lezak. M. D. (1983). Neuropsycholopxf
Money.
I.. Alexander,
D., & Walker,
Baltimore:
Johns Hopkins
Press.
Ratcliff,
G., & Newcombe,
F. (1973).
bral lesions. Journul
Resnick,
S. M. (1993).

(2.s.w~wwnt (2nd ed ). New York: Oxford University

H. T ( IYhSi
UWU~J/. A stundardizrd
road-map
Spatial orienratwn

of Neurology,
Neurosurgr~,
Sex differences
in mental

71-79.
Roeltgen, I). P, Sevush, S.. & Hellman.
NeuroL,,qy.
40, 4h- 47

I man

Effects

of left, right,

und I\vchuu~.
36, 448454.
rocarwn~:
An effect of time Ilmits?

K. M. ( 19X 7) lurc

Gerstmnnns

syndrome

from

Press.
test of direction

and bilateral
Brain
a focal

and

sense.

posterior
Cognition,

lesion.

Archives

cere-

21,
of

Rosier.

F., Schumacher,

ing mental
Strub.

rotation

G., & Sojka.

and mental

R. L.. & Geschwind,

New

neuropsychology.
Thomas,

H., & Kail,

Intelligence,

What
German

N. (1983). Localisation
York: Academic
Press.

R. (1991).

15. 11-32.

B. ( 1990).

arithmetic.

Sex differences

the bram

Journal

reveals

when

of Psychology,

in Gerstmann
m speed of mental

syndrome.
rotation

it thinks:

Event-related

potentials

dur-

Localisafion

in

14, 185-203.
In A. Kertesz
and the X-linked

(Ed.),

genetic

hypothesis.