Acta Psychologica 155 (2015) 1–7

Contents lists available at ScienceDirect

Acta Psychologica
journal homepage: www.elsevier.com/ locate/actpsy

Evaluation of the attentional capacities and working memory of early and

a r t i c l e i n f o a b s t r a c t

Article history: Although attentional processes and working memory seem to be significantly involved in the daily activities
Received 18 March 2014 (particularly during navigating) of persons who are blind and who use these abilities to compensate for their
Received in revised form 12 November 2014 lack of vision, few studies have investigated these mechanisms in this population. The aim of this study is to
Accepted 22 November 2014
evaluate the selective, sustained and divided attention, attentional inhibition and switching and working
Available online 12 December 2014
memory of blind persons. Early blind, late blind and sighted participants completed neuropsychological tests
PsycINFO classification:
that were designed or adapted to be achievable in the absence of vision. The results revealed that the early
3299 Vision & Hearing & Sensory Disorders blind participants outperformed the sighted ones in selective, sustained and divided attention and working
memory tests, and the late blind participants outperformed the sighted participants in selective, sustained and
Keywords: divided attention. However, no differences were found between the blind groups and the sighted group in the
Visually impaired persons attentional inhibition and switching tests. Furthermore, no differences were found between the early and late
Early blindness blind participants in this set of tests. These results suggest that early and late blind persons can compensate for
Late blindness the lack of vision by an enhancement of the attentional and working memory capacities.
Attentional processes
© 2014 Elsevier B.V. All rights reserved.
Working memory
Neuropsychological assessment

1. Introduction guidance of locomotion) and that basic sensorial functions are not
improved.
Because vision is the most useful sense for processing spatial infor- Thus, several authors have proposed that blindness might result in
mation (Thinus-Blanc & Gaunet, 1997), the navigation of persons the enhancement of higher-level cognitive processes, such as attentional
who are visually impaired requires that they manage a large amount processes, and have suggested that navigation without vision appears to
of information through available sensorial modalities, mobility aids involve the mobilization of selective attention and attentional inhibition
and mnesic information. Moreover, because navigation is a dynamic (of irrelevant information), sustained attention, divided attention, atten-
activity, temporal constraints require the regular updating of processed tional switching and working memory (Espinosa, Ungar, Ochaita, Blades,
information. & Spencer, 1998; Geruschat & Turano, 2007; Kujala, Lehtokoski, Alho,
There is a long-standing debate regarding the possibility that blind Kekoni, & Näätänen, 1997; Occelli, Spence, & Zampini, 2013). Some of
persons compensate for their lack of vision through improvements these attentional processes have previously been studied in blind
in tactile and auditory sensitivities. Although some results are in agree- individuals.
ment with this perspective (Gougoux et al., 2004; Wan, Wood, Reutens, For example, Collignon, Renier, Bruyer, Tranduy, and Veraart (2006)
& Wilson, 2010), there is strong evidence that blindness does not lead to and Collignon and De Volder (2009) found that early blind persons
improvements in basic sensorial acuity (Lewald, 2002; Pascual-Leone & answer more rapidly (but not more accurately) than sighted persons
Torres, 1993) but rather leads to more efficient processing of available in Go–NoGo tasks composed of auditory and tactile stimuli. In a more
information (Grant, Thiagarajah, & Sathian, 2000; Hatwell, 2003). In recent study, Lerens and Renier (2014) found that early blind partici-
agreement with this latter assumption, Hugdahl et al. (2004) reported pants are faster than sighted ones in the detection of auditory targets
that the perceptual improvements observed in blind persons occur coming from frontal and/or peripheral locations and that the improved
only for perceptual functions that involve higher-level processes (such performance of blind persons is greater when there are multiple sound
as speech discrimination and the detection of echoes as signals for the sources. Thus, the results obtained in these studies suggest an enhance-
ment of spatial selective attention in early blind persons. Additionally, to
measure the inhibition process, Collignon et al. (2006) and Collignon and
⁎ Corresponding author at: IFSTTAR-LESCOT, 25 Avenue François Mitterrand — Case 24,
De Volder (2009) compared the rates of false alarms (i.e., when partici-
69675 Bron Cedex, France. Tel.: +33 4 72 14 24 65. pants respond to a distractor) of blind and sighted participants and
E-mail address: [email protected] (C. Pigeon). found no differences between the two groups. Furthermore, Collignon

http://dx.doi.org/10.1016/j.actpsy.2014.11.010
0001-6918/© 2014 Elsevier B.V. All rights reserved.
2 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7

et al. (2006), Collignon and De Volder (2009) and Kujala et al. (1997) (Hatwell, 2003). Late blind persons have better abilities to understand
found that early blind individuals have divided attention abilities that spatial arrangements as a whole than do early blind persons because
are improved compared to those of sighted persons; these individuals late blind persons have experienced the simultaneous character of vi-
are faster at detecting targets among auditory and tactile stimuli. Thus, sion (Thinus-Blanc & Gaunet, 1997). Thus, the spatial processing re-
blind persons are familiar with the simultaneous use of auditory and quired for navigation involves a greater cognitive effort (particularly
tactile systems and are less impaired by the cost of sharing attention. in terms of attentional and working memory levels) for early blind
Additional studies have been conducted regarding the working than for late blind persons. This difference can result in greater attention
memory of blind persons, and these studies have primarily focused and working memory capacities in early than in late blind persons.
on the tactile modality. For example, Bliss, Kujala, and Hämäläinen Nonetheless, although some authors have studied these two sub-
(2004) compared the performance of early and late blind and sighted populations separately (Bliss et al., 2004; Ruggiero & Iachini, 2010),
participants in N-back tasks. The blind participants outperformed the the majority of the studies on attention have focused only on early
sighted ones in a tactile task involving raised letters. The performance blind persons (Collignon & De Volder, 2009; Collignon et al., 2006;
of the sighted participants in a visual task was better than that of the Kujala et al., 1997; Lerens & Renier, 2014).
blind participants in the task with the raised letters but did not differ The first aim of the present study was to design a set of cognitive
from the performance of the blind participants in a task with Braille tests that assesses selective attention, sustained attention, divided
characters. Thus, the performance of blind persons and sighted persons attention, inhibition, attentional switching and working memory and
in an N-back task seems to be related to the participants' familiarity that is accessible to persons with visual impairments. The second aim
with the material used. Using a haptic adaptation of the Corsi Block- was to compare the performance of early and late blind persons with
Tapping task, Ruggiero and Iachini (2010) found that the tactile forward the performance of sighted people in this set of tests. According to the
span of late blind participants was higher than that of sighted partici- theory that blind persons overcome their lack of vision via improve-
pants, and the performance of early blind participants was intermediate. ments in higher-level cognitive processes, we propose that the perfor-
The tactile backward spans of the three groups were similar. However, mance of the blind participants should be superior to the performance
the differences in the performance of the groups can be explained by of the sighted participants and that the performance of the early blind
different familiarity with the tactile modality for spatial perception. participants should be superior to the performance of the late blind
Thus, these two studies emphasize the need to use identical materials participants.
and to control the familiarity level between the groups being tested.
To account for tactile processing in blind persons, Cohen, Voss, Lepore,
2. Methods
and Scherzer (2010) and Cohen, Scherzer, Viau, Voss, and Lepore
(2011) proposed the addition of a tactile subsystem to the working
2.1. Participants
memory model of Baddeley and Hitch (1974). Using different concur-
rent tasks during the realization of a tactile working memory task,
Fourteen early blind participants (10 females and 4 males) aged
Cohen et al. (2011) concluded that the tactile subsystem of the working
27–52 (mean age = 35 years, SD = 8.6), ten late blind participants
memory of blind persons has a spatial nature. The working memory
(5 females and 5 males) aged 28–51 (mean age = 39 years, SD = 7.4)
capacities of blind children have also been assessed. Although some
and 24 sighted participants (15 females and 9 males) aged 22–50
authors have shown that blind children's performance is superior to
(mean age = 31 years, SD = 8.2) took part in the study. The blind partic-
that of sighted children in working memory and short-term memory
ipants had no residual eyesight or were able to see bright light and had
tasks (Withagen, Kappers, Vervloed, Knoors, & Verhoeven, 2013),
no other neurological or sensory-motor impairments. Only three early
other authors have found that blind children outperform sighted
blind participants were not blind from birth; these participants became
children in short-term memory tasks, but not in working memory
blind at three, five and six years old. The late blind participants became
tasks (Swanson & Luxenberg, 2009). Moreover, Hull and Mason
blind between the ages of 14 and 46 years (mean age of onset = 15.9
(1995) demonstrated that blind children achieve greater forward and
years, SD = 9.5). The blindness of the participants was due to various
backward spans than do sighted children. Although it is difficult to
etiologies. The early and late blind participants were able to move inde-
draw conclusions from some studies of the working memory of blind
pendently outside of their homes with a white cane (18) or a guide dog
persons because of the use of different tasks for sighted and blind partic-
(6). Six subtests from the verbal portion of the Wechsler Adult Intelli-
ipants, the overall trend of the relevant studies tends to demonstrate an
gence Scale — Revised (WAIS-R, Wechsler, 1955; 1981) were used to
enhancement of working memory in blind persons.
determine that the intellectual functioning of the three groups was sim-
Regarding the sustained attention and attentional switching capaci-
ilar (F(2, 45) = 0.595, p = 0.556). Although this verbal portion of the
ties of blind persons, no studies have been conducted. However, Kujala
test assesses only verbal IQ, this was the only portion of the test that
et al. (1997) suggested that blind persons would have an attentional
is available to evaluate the intellectual functioning of visually impaired
switching mechanism that is more efficient than that of sighted persons
persons. According to Price, Mount, and Coles (1987), this verbal part is
due to the more extensive practice with the available perceptual modal-
appropriate for assessing persons who are blind and provides a verbal
ities, particularly during pedestrian traveling.
IQ that correlates with the IQ acquired with the complete scale at a
Despite the important involvement of attention in navigation by
level of 0.95. The simple reaction times of the participants were also
blind persons and studies of some of the attentional capacities of blind
compared. Using a simple reaction time test, we demonstrated that
persons, a global consideration of all of the attentional processes of
the three groups of participants exhibited equivalent sensory-motor ra-
this population has not been performed, possibly due to the difficulty
pidity (F(2, 45) = 3.628, p = 0.644). The sighted participants were
of assessing attentional processes without the visual modality because
blindfolded during the testing. The subjects provided written informed
attention has traditionally been considered through the lens of vision.
consent prior to participation. This experiment was approved by the
Clinical professionals note that the number of neuropsychological
ethics committee of the IFSTTAR.
tests accessible to persons with visual impairments is limited and that
there are no appropriate norms for this population (Hill-Briggs, Dial,
Morere, & Joyce, 2007). 2.2. Procedures and materials
In addition, most of the studies discussed did not distinguish
between the two main sub-groups of the population of persons who The participants were seated in a quiet room during the experiment.
are blind, early and late blind people. Indeed, blindness has different The duration of the session varied from 90 to 120 min. The different
effects on cognitive processes depending on the age at which it occurs attentional components of the participants were assessed with tests
 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7 3

that were either standardized or designed for the study because no test 2.2.1.4. Sustained attention test. This task is similar to the previous
was available for visually impaired persons. task, but 360 stimuli were presented, and the task lasted 9 min. The in-
structions and measures were the same. To examine variations in
2.2.1. Tests designed for the study attentional level over time, the measures were analyzed in three parts
In a pilot study (unpublished) with four early blind persons and ten (i.e., the first part, the second part and the last part of the task).
sighted persons, the subtests 2, 3 and 5 of a French version of the Test of
Everyday Attention (TEA) adapted by Allain, Kefi, and Gall (2002) were
used to assess the attentional capacities of the participants. The ceiling 2.2.1.5. Divided attention test. This task was composed of two subtasks
effect obtained revealed that the TEA subtests used are too easy for that were performed simultaneously. The participant was asked to per-
persons without attentional disorders. Then, attentional tests were form a task that was identical to the selective attention test and at the
specifically designed for the present study. same time was required to count the occurrences of one digit (presented
All of the tasks described below were programmed with the eight times) aloud. For the first task, hits, false alarms, omissions and
software SuperLab Pro (version 2.0.4). The programming of tests, the reaction times were recorded. For the second task, the score (number
presentation of the stimuli and the recording of the responses and of digits counted) was recorded.
reaction times were performed on a Dell Latitude D620 laptop.
For the N-back task and the selective, sustained and divided atten-
2.2.2. Standardized test
tion tasks, the participant was instructed to respond by pressing the
space bar on the keyboard with the index finger of his dominant hand
2.2.2.1. Digit span (WAIS-R, Wechsler, 1955; 1981). This test is composed
as rapidly as possible. Button presses within 200 milliseconds (ms) of
of two subtests called the Digit Forward and Digit Backward tests. In
the target stimulus onset and before the next stimulus onset were clas-
both subtests, sequences of digits are presented and gradually length-
sified as hits, and presses at any other time were scored as false alarms.
ened by single digits until the participant fails two consecutive
sequences. In the Digit Forward task, the participant is asked to repeat
2.2.1.1. N-back test. The N-back test is used to test working memory,
the sequence in the same order of presentation. In the Digit Backward
particularly the processes of updating (Kane, Conwal, Miura, & Colflesh,
task, the participant is asked to repeat the sequence in reverse order.
2007). Consonants were presented one at a time. The interstimuli
The dependent variables are the forward and backward spans (i.e., the
interval was 1500 ms. There were ten targets among 60 stimuli in each
maximal numbers of items that could be repeated without errors) and
condition. There were four conditions (0, 1, 2 and 3-back) that differed
the standard age score (obtained by adding the scores in the two condi-
in the level of difficulty involved. In the 0-back condition, the participant
tions and converting the obtained score based on the participant's age).
had to respond every time a target consonant was presented; in the
1-back condition, he had to respond when the presented consonant
was the same as the previously presented consonant; in the 2-back 3. Results
condition, a response was required when the consonant was the same
as the one preceding the previous one; and in the 3-back condition, a The data (scores and/or reaction times) from the tests were statisti-
response was required when the consonant was the same as the one cally analyzed with one-way analyses of variance (ANOVAs) to compare
presented three consonants previously. The numbers of responses and the performances of the three groups (i.e., the early blind, late blind and
reaction times were recorded. sighted groups). For some of the data from the divided attention test
and digit span test, repeated-measures ANOVAs using a design with 3
2.2.1.2. Plus–minus task. This task is used to test attention switching groups (between subject factor: early blind, late blind and sighted
(Miyake et al., 2000). In the original version (Spector & Biederman, groups) × 2 conditions (within-subject factors) were used. For the
1976), three lists of 30 two-digit numbers are presented on a sheet of N-back test, a repeated measures ANOVA with a design of 3 groups
paper. On the first list, the participant is instructed to add three to (between subject factor: early blind, late blind and sighted groups) ×
each number and write the answers. On the second list, he is asked to 4 conditions (within-subject factors: 0-back, 1-back, 2-back and 3-
subtract three from each number and write the answers. On the third back conditions) was used. If the variances were not homogeneous
list, he is required to alternate between adding three to and subtracting (as determined with Mauchly's test), the Greenhouse–Geisser correc-
three from the numbers and writing the answers. The participant is tion was used. Based on significant F values, the Fisher LSD test was
asked to complete each list as quickly and accurately as possible. The used for the post hoc analyses. The level of significance for all statistics
cost of switching between the two operations is calculated as the differ- was p b 0.05. The means and standard deviations for each of the perfor-
ence between the time required to complete the third list and the aver- mance measures for each of the groups are shown in Table 1.
age of the times required to complete the first two lists. We adapted this
task to the auditory modality and used the same numbers. For each list,
the numbers were audibly presented by the computer. When a number 3.1. Selective attention test
was presented, the participant performed the appropriate operation
and then provided his answer verbally. Then, the experimenter pressed A one-way ANOVA of the scores revealed a significant effect of group
the space bar, and the next number was presented. At the same time, (F(2, 45) = 4.274, p = 0.020). Post hoc analyses demonstrated that the
the experimenter noted the errors and recorded the time required to early blind participants scored significantly higher than did the control
complete each list with a stopwatch. The cost of switching was calculated participants (p = 0.011), the late blind participants scored significantly
in the same manner used for the original version, and the result served as higher than did the control participants (p = 0.045), and there was no
a dependent variable. significant difference between the early and late blind participants
(p = 0.782). A one-way ANOVA on the reaction times revealed a signif-
2.2.1.3. Selective attention test. Consonants (targets) and one-digit icant effect of group (F(2, 45) = 7.106, p = 0.002). Post hoc analyses
numbers (distractors) were audibly presented in a random order with demonstrated that the early blind participants (p = 0.002) and late
interstimuli intervals of 1000, 1500 and 2000 ms (randomized). The blind participants (p = 0.010) exhibited significantly shorter reaction
participant was asked to press the space bar each time he heard a con- times than did the control participants, but there was no significant
sonant. One hundred twenty stimuli were presented; 1/3 of the stimuli difference between the early and late blind participants (p = 0.778;
were targets, and the task lasted 3 min. Hits, false alarms, omissions and Fig. 1.). A one-way ANOVA of the false alarms did not reveal a significant
reaction times were recorded. effect of group (F(2, 45) = 1.715, p = 0.118).
4 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7

Table 1
Means and standard deviations for each of the performance measures of the groups.

Mean (SD)

Test Variable Early blind Late blind Control

Selective attention test Score 38.8 (1.3) 38.5 (1.2) 36.6 (3.2)
False alarm number 0.6 (0.8) 1.3 (1.1) 0.7 (0.9)
Reaction time (ms) 662.4 (55.4) 669.5 (55.3) 731.2 (65.3)
Sustained attention test Score 117.1 (4.3) 116.4 (3.6) 109.5 (10.6)
Reaction time (ms) 686.4 (62.2) 696.9 (66.1) 755.3 (72.4)
Divided attention test First task score 35.4 (3.9) 34.9 (3.5) 30.6 (5.7)
First task reaction time (ms) 740.2 (62.5) 740.5 (62.3) 846.5 (110.4)
Second task score 6.6 (1.2) 6.5 (1.0) 7.4 (1.0)
Auditory N-back test 0-back score 10.0 (0.0) 9.9 (0.3) 9.9 (0.2)
1-back score 10.0 (0.0) 9.6 (0.7) 9.9 (0.2)
2-back score 9.0 (1.6) 9.3 (1.6) 8.2 (1.8)
3-back score 6.1 (1.6) 5.4 (2.0) 5.3 (1.4)
0-back reaction time (ms) 520.1 (69.5) 539.2 (87.3) 591.0 (137.7)
1-back reaction time (ms) 558.8 (50.1) 607.0 (91.6) 642.3 (101.0)
2-back reaction time (ms) 681.9 (91.5) 740.1 (140.8) 838.0 (132.7)
3-back reaction time (ms) 824.0 (114.1) 965.1 (285.9) 1010.1 (135.6)
Digit span Forward span 7.1 (1.9) 6.6 (1.8) 6.9 (1.2)
Backward span 5.9 (1.5) 5.5 (0.9) 5.0 (1.2)
Standard age score 15.0 (3.8) 13.1 (2.7) 12.9 (2.9)
Auditory plus–minus task Costs of switching 9.3 (11.2) 7.3 (9.9)a 8.1 (11.7)
a
One late blind participant did not perform the plus–minus task adequately; thus, for this task, there were nine participants in late blind group.

3.2. Sustained attention test the scores revealed an effect of group (F(2, 141) = 4.930, p = 0.012),
an interaction effect between part and group (F(4, 141) = 2.761,
A one-way ANOVA of the scores revealed a significant effect of group p = 0.032) and no significant effect of part (F(2, 141) = 0.962, p =
(F(2, 45) = 4.930, p = 0.012). Post hoc analyses demonstrated that the 0.386). A two-way ANOVA on the reaction times revealed an effect of
early blind participants (p = 0.007) and the late blind participants group (F(2, 141) = 5.091, p = 0.010), no significant effect of part
(p = 0.029) scored significantly higher than did the control partici- (F(2, 141) = 0.933, p = 0.397) and no interaction effect between part
pants, but there was no significant difference between the early and and group (F(4, 141) = 1.295, p = 0.282; Fig. 1.).
late blind participants (p = 0.826). A one-way ANOVA on the reaction
times revealed a significant effect of group (F(2, 45) = 5.422, p = 3.3. Divided attention test
0.008). Post hoc analyses revealed that the early blind participants
(p = 0.001) and late blind participants (p = 0.003) exhibited signifi- A one-way ANOVA on the scores for the first task revealed a signifi-
cantly shorter reaction times than did the control participants, but cant effect of group (F(2, 45) = 5.301, p = 0.009). Post hoc analyses
there was no significant difference between the early and late blind demonstrated that the early blind participants (p = 0.006) and the
participants (p = 0.994). The performances of the participants were late blind participants (p = 0.024) were significantly faster than the
decomposed and analyzed in three parts to determine whether there control participants, but there was no significant difference between
were variations in attentional level over time. A two-way ANOVA on the early and late blind participants (p = 0.821). A one-way ANOVA

*
9000
** *** **
*
**
8000 **
7000
E y bllind
Earl
Reaction times (ms)
( )

6000
L e Bliind
Late
5000
C ntrol
Con
4000

3000

2000

1000

0
Seelecttive attenntion test Suustaiined
d atteentioon teest Divided atteentio
on teest

Fig. 1. Mean reaction times (with standard deviations of the means) exhibited by the three groups in the selective, sustained and divided attention tests. *p ≤ 0.05; **p ≤ 0.01;***p ≤ 0.001.
 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7 5

of the reaction times in the first task revealed a significant effect of achieved significantly shorter reaction times than did the control partic-
group (F(2, 45) = 8.325, p = 0.001). Post hoc analyses demonstrated ipants (p = 0.000). However, there were no significant differences be-
that the early blind participants (p = 0.009) and the late blind partici- tween the early and late blind participants (p = 0.094) or between
pants (p = 0.034) exhibited significantly shorter reaction times than the late blind and control participants (p = 0.094). No interaction effect
did the control participants, but there was no significant difference be- was found between group and condition (F(6, 189) = 1.756, p =
tween the early and late blind participants (p = 0.810). A one-way 0.141). The group differences in reaction times were studied using sep-
ANOVA on the scores for the second task did not reveal a significant arate one-way ANOVAs for each condition. For the 0-back condition,
effect of group (F(2, 45) = 1.172, p = 0.319). To determine the cost this analysis did not reveal a significant effect of group (F(2, 45) =
induced by the second task on the performance of the first task, the 1.969, p = 0.151). The analysis revealed a significant effect of group in
scores and reaction times for the first task were compared to those of the 1-back condition (F(2, 45) = 4.053, p = 0.024). Post hoc analyses
the selective attention test. A two-way ANOVA on the scores revealed demonstrated that the reaction times of the late blind participants did
a significant effect of condition (F(1, 93) = 59,680, p = 0.000). Post not differ from those of the early blind participants (p = 0.189) or
hoc analyses demonstrated that scores in the selective attention condi- from those of the control participants (p = 0.289), but the early blind
tion were higher than the scores in the divided condition (p = 0.000). A participants achieved shorter reaction times than did the control partic-
significant effect of group was found (F(2, 93) = 5.765, p = 0.006). Post ipants (p = 0.007). The analysis revealed a significant effect of group in
hoc analyses demonstrated that the early blind participants (p = 0.004) the 2-back condition (F(2, 45) = 7.968, p = 0.001). Post hoc analyses
and the late blind participants (p = 0.019) scored significantly higher demonstrated that the reaction times of the early blind participants
than the control participants, but there was no significant difference be- did not differ from those of the late blind participants (p = 0.260),
tween the early and late blind participants (p = 0.792). The analysis did but the early blind participants achieved shorter reaction times than
not reveal a significant interaction effect between group and condition did the control participants (p = 0.000), and the late blind participants
(F(2, 93) = 2.741, p = 0.075). A two-way ANOVA on the reaction achieved shorter reaction times than did the control participants
times revealed a significant effect of condition (F(1, 93) = 68.446, p = (p = 0.033). The analyses revealed a significant effect of group in the
0.000). Post hoc analyses demonstrated that the reaction times in the 3-back condition (F(2, 45) = 5.296, p = 0.009). Post hoc analyses dem-
selective attention condition were shorter than the reaction times in onstrated that the reaction times of the late blind participants did not
the divided condition (p = 0.000). A significant effect of the group differ from those of the early blind participants (p = 0.062) or from
was observed (F(2, 93) = 9.438, p = 0.000). Post hoc analyses demon- those of the control participants (p = 0.433); however, the early blind
strated that the early blind participants (p = 0.000) and the late blind participants achieved shorter reaction times than did the control partic-
participants (p = 0.002) exhibited significantly shorter reaction times ipants (p = 0.002).
than did the control participants, but there was no significant difference
between the early and late blind participants (p = 0.896). The analysis 3.5. Digit span
did not reveal a significant interaction effect between group and condi-
tion (F(2, 93) = 2.073, p = 0.138; Fig. 1). A two-way ANOVA on the scores revealed a significant effect of
condition (F(1, 93) = 48.760, p = 0.000). Post hoc analyses demon-
3.4. N-back test strated that the forward span scores were higher than the backward
span scores (p = 0.000). No group effect (F(2, 93) = 1.299, p =
A two-way ANOVA on the scores revealed a significant effect of con- 0.283) or interaction effect between group and condition was
dition (F(3, 189) = 144.918, p = 0.000). Post hoc analyses demonstrated found (F(2, 93) = 1.867, p = 0.427). The group differences in scores
that scores in the 0-back condition did not differ significantly from the were also studied using separate one-way ANOVAs for the two con-
scores in the 1-back condition (p = 0.139) but were significantly higher ditions. These analyses did not reveal a significant effect in the for-
than those in the 2- and 3-back conditions (p = 0.000). The scores in the ward span (F(2, 45) = 0.360, p = 0.700) or backward span tests
1-back condition were higher than those in the 2- and 3-back conditions (F(2, 45) = 1.977, p = 0.150).
(p = 0.000), and the scores in the 2-back condition were higher than Standard scores were calculated by adding the scores of the two
those in the 3-back condition (p = 0.000). The analysis did not reveal a conditions and converting the obtained score based on the participant's
significant effect of the group (F(2, 189) = 1.395, p = 0.258) or an inter- age. A one-way ANOVA on the standard age scores did not reveal a
action effect between group and condition (F(6, 189) = 1.512, p = significant effect of group (F(2, 45) = 2.775, p = 0.077).
0.203). The group differences in the scores were also examined using
separate one-way ANOVAs for each condition. For the 0-back condition, 3.6. Plus–minus task
the analysis did not reveal a significant effect of group (F(2, 45) = 0.706,
p = 0.499). The analysis revealed a significant effect of group in the A one-way ANOVA on the costs of switching did not reveal a signif-
1-back condition (F(2, 45) = 4.706, p = 0.016). Post hoc analyses dem- icant effect of group (F(2, 45) = 0.096, p = 0.909).
onstrated that the scores of the late blind participants were lower than
those of the early blind participants (p = 0.008) and those of the control 4. Discussion
participants (p = 0.008), but the scores of the early blind participants
did not differ from those of the control participants (p = 0.721). The The aims of this study were to design a set of cognitive tests that
analysis did not reveal significant differences between the three groups assessed selective attention, sustained attention, divided attention, inhi-
in the 2-back condition (F(2, 45) = 1.864, p = 0.167) or the 3-back con- bition, attentional switching and working memory and that was accessi-
dition (F(2, 45) = 1.085, p = 0.345). A two-way ANOVA on the reaction ble to persons with visual impairment and to compare the performances
times revealed a significant effect of condition (F(3, 189) = 119.918, of early blind, late blind and sighted people on this set of tests.
p = 0.000). Post hoc analyses demonstrated that reaction times in the First, each process will be discussed separately, then a more general
0-back condition were significantly shorter than those in the 1-back discussion will address the absence of differences observed between the
condition (p = 0.001), the reaction times in the 1-back condition two blind groups.
were significantly shorter than those in the 2-back condition (p =
0.000), and the reaction times in the 2-back condition were significantly 4.1. Enhanced attentional and working memory processes in blindness
shorter than those in the 3-back condition (p = 0.000). Moreover, the
analysis revealed a significant effect of group (F(2, 189) = 8.158, p = In the selective attention test, the early and late blind participants
0.001). Post hoc analyses demonstrated that the early blind participants exhibited higher scores and shorter reaction times than did the sighted
6 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7

participants, and the scores and reaction times of the early blind and late faster than the sighted participants only in the 2-back condition. How-
blind participants were equivalent. There were no differences in the ever, there were no significant differences in the reaction times between
number of false alarms between the three groups. These results are con- the early and late blind participants in any of the four conditions. The
sistent with those of Collignon et al. (2006) and Collignon and De Volder lack of differences between groups in the 0-back condition is not
(2009), who found that early blind persons are faster (but achieve surprising because this condition does not involve working memory;
equivalent scores) than sighted persons in a selective attention test. it only requires the detection of a single stimulus. Therefore, this condi-
However, our results are more robust because we found a significant tion can be considered a control condition. Condition effects on the
difference between the scores of the two groups. As in the studies of scores (excluding the 0- and 1-back tasks in which the scores were
Collignon et al. (2006) and Collignon and De Volder (2009), we did similar) and reaction times were found. These findings indicate
not observe any difference between the three groups in the number of that the performances (in terms of both scores and rapidity) of the
false alarms, which suggests that the inhibition capacities of the three participants decreased when the difficulty level of the condition was
populations were equivalent. However, the test used (as with those increased. The N-back test used by Bliss et al. (2004) did not allow the
used by Collignon et al., 2006 and Collignon & De Volder, 2009) does authors to conclude that blind persons exhibited better performance
not specifically assess inhibition; rather, the number of false alarms of than sighted persons because the participants were not compared
the selective attention test was used to measure inhibition. It will be with the same stimuli and did not have the same familiarity with the
necessary to develop a specific test to assess the inhibition processes stimuli. The N-back test of the present study allowed for such a compar-
of blind persons. Although the results are comparable, the test used ison. However, no difference was found between the early and late blind
in the present research is somewhat different from those used by persons. Regarding the digit span test, there were no differences
Collignon et al. (2006) and Collignon and De Volder (2009). In between the three groups in either the forward or backward span. In
Collignon's studies, the participants judged the spatial localization of their study, Rokem and Ahissar (2009) found that early blind partici-
stimuli, whereas in the present study, the participants judged the cate- pants exhibit greater forward spans than do sighted participants, but
gory (digit or letter) of stimuli. The present study therefore showed that the backward spans of these two groups are equivalent. It is important
the superiority of blind persons in selective attention exceeds the spatial to note the presence of a ceiling effect only for the blind participants
domain. in the present study (this effect was observed in 6 early and 2 late
In the sustained attention test, the early blind participants exhibited blind participants in the forward span test and 3 early blind participants
significantly higher scores and shorter reaction times than did the sight- in the backward span). Thus, the spans obtained here do not actually
ed participants. The late blind participants exhibited shorter reaction account for the mean number of items that the blind participants were
times but did not exhibit higher scores than the sighted participants. able to retrieve from working memory. This number of items might
The early blind participants and the late blind participants exhibited have been higher if a longer span test had been used, and differences be-
equivalent scores and reaction times. As explained in the Results tween groups might have been found. Because no ceiling effect was men-
section, to study the variation in attentional level over time, the perfor- tioned in the study of Rokem and Ahissar (2009), it is difficult to compare
mances were decomposed and analyzed in three parts. A significant their results to those of the present study. Thus, the ceiling effect that was
interaction effect between group and part was found, but no part effect present only for the blind participants suggests (but does not confirm)
was found, which suggests that the mean performance of all of the that blind persons exhibit improved spans, and the results of the N-back
participants did not significantly fluctuate with time. The present test confirm the superiority of the working memory of blind persons.
study seems to be the first to investigate the sustained attention of Because the three groups did not exhibit significant differences in
blind persons and to show that the blind persons (early and late) reaction times in the simple reaction time test or the 0-back condition,
were more efficient than the sighted persons in maintaining attention the faster reaction of the blind participants in the attentional tests and
over 10 min. other conditions of the N-back test did not depend on faster stimuli
In the divided attention test, the early blind and late blind partici- detections or faster motor responses. Moreover, because the faster reac-
pants exhibited significantly higher scores and shorter reaction times tion times of the blind participants were not accompanied by lower
than did the sighted participants, and the early and late blind partici- scores, the rapidity seems to be unrelated to a speed–accuracy trade-
pants exhibited equivalent scores and reaction times in the first task. In off. Hence, the faster reaction times suggest attentional enhancements
the second task, the performances of the three groups were equivalent. in the blind participants, as found in previous studies (Collignon & De
The scores and reaction times obtained in the first task were compared Volder, 2009; Collignon et al., 2006; Kujala et al., 1997).
with those obtained in the selective attention task. An effect of the con- Taken together, our results demonstrate that blind persons have bet-
dition was found; the task was easier when performed alone than ter attentional capacities than do sighted persons, particularly in terms
when performed with a secondary task. A group effect was also found; of selective, sustained, and divided attention and working memory
the early and late blind participants exhibited better performance than capacities, in the absence of enhanced simple reaction times or superior
did the sighted participants, but no interaction effect between group intellectual functioning. At the cortical level, the attentional modula-
and condition was found. This study confirms the results of Kujala et al. tions of blind persons are also observable. Activations of the visual
(1997), Collignon et al. (2006) and Collignon and De Volder (2009), cortex in early blind persons (but not in sighted ones) during auditory,
who found that early blind participants are faster in divided attention tactile and bimodal attention tasks have been reported in an fMRI study
tasks than sighted participants but additionally shows that late blind (Weaver & Stevens, 2007). Weaver and Stevens (2007) reported that
persons are more efficient than sighted persons in divided attention. the additional occipital activity observed in blind persons during atten-
An auditory version of the plus–minus task was used to assess atten- tional processing might explain the enhanced attentional abilities of this
tional switching, and no significant between-group differences in the population, including those observed in the present study.
costs of switching were found. This study seems to be the first to evaluate Navigation without vision, which involves the processing and
attentional switching in blind persons, but it failed to identify enhanced manipulation of a large amount of information, can lead to the
performance among blind persons. over-development of attentional mechanisms and working memory,
The N-back and digit span tests were used to assess working mem- as is the case for other cognitive strategies, such as numerical skills.
ory. In the N-back test, there were significant differences in the scores Indeed, superior numerical abilities have been found in blind persons
only in the 2-back condition, in which the late blind participants exhib- (Castronovo & Delvenne, 2013; Castronovo & Seron, 2007), and these
ited lower scores than did the early blind and sighted participants. findings can be explained by the fact that navigating without vision in-
The early blind participants were faster than the sighted participants volves numerical processing, such as quantitative judgments (e.g., time
in the 1, 2, and 3-back conditions, and the late blind participants were and distance estimations and numbers of footsteps between two
 C. Pigeon, C. Marin-Lamellet / Acta Psychologica 155 (2015) 1–7 7

landmarks). The cognitive mechanisms of blind persons, particularly Baddeley, A.D., & Hitch, G. (1974). Working memory. Psychology of Learning and
Motivation, 8, 47–89.
their attentional processes, should be studied during navigation activi- Bliss, I., Kujala, T., & Hämäläinen, H. (2004). Comparison of blind and sighted participants'
ties to better understand the involvement of attentional capacities in performance in a letter recognition working memory task. Cognitive Brain Research,
compensation for a lack of vision during navigation, as has been done 18, 273–277.
Castronovo, J., & Delvenne, J. -F. (2013). Superior numerical abilities following early visual
with semi-sighted persons with glaucoma (Geruschat & Turano, 2007). deprivation. Cortex, 49(5), 1435–1440.
Castronovo, J., & Seron, X. (2007). Numerical estimation in blind subjects: Evidence of the
4.2. Similar performances in early and late blindness impact of blindness and its following experience. Journal of Experimental Psychology:
Human Perception and Performance, 33, 1089–1106.
Cohen, H., Scherzer, P., Viau, R., Voss, P., & Lepore, F. (2011). Working memory for braille
In the present research, no differences were found between early is shaped by experience. Communicative & Integrative Biology, 4, 227–229.
and late blind persons (with the exception of the 1-back condition Cohen, H., Voss, P., Lepore, F., & Scherzer, P. (2010). The nature of working memory for
Braille. PLoS ONE, 5.
scores). Similar results have been reported by Bliss et al. (2004). The
Collignon, O., & De Volder, A.G. (2009). Further evidence that congenitally blind partici-
performance of late blind persons was nearly always intermediate to pants react faster to auditory and tactile spatial targets. Canadian Journal of
those of early blind and sighted participants, and more significant differ- Experimental Psychology, 63, 287–293.
ences were found between the early blind and sighted participants than Collignon, O., Renier, L., Bruyer, R., Tranduy, D., & Veraart, C. (2006). Improved selective
and divided spatial attention in early blind subjects. Brain Research, 1075, 175–182.
between the late blind and sighted participants. Although these findings Espinosa, M.A., Ungar, S., Ochaita, E., Blades, M., & Spencer, C. (1998). Comparing methods
suggest that early blind participants potentially have superior capacities for introducing blind and visually impaired people to unfamiliar urban environments.
than do late blind participants, the differences between the two popula- Journal of Environmental Psychology, 18, 277–287.
Geruschat, D.R., & Turano, K.A. (2007). Estimating the amount of mental effort required
tions were not sufficiently large to be statistically significant. This lack of for independent mobility: Persons with glaucoma. Investigative Ophthalmology &
significance does not seem to be the result of the lack of sensitivity of the Visual Science, 48, 3988–3994.
tests because score differences were found between the sighted and Gougoux, F., Lepore, F., Lassonde, M., Voss, P., Zatorre, R.J., & Belin, P. (2004). Pitch
discrimination in the early blind. Nature, 430(6997), 309.
blind participants in addition to the reaction time differences that Grant, A.C., Thiagarajah, M.C., & Sathian, K. (2000). Tactile perception in blind Braille
have typically been found in other studies (Collignon & De Volder, readers: A psychophysical study of acuity and hyperacuity using gratings and dot
2009; Collignon et al., 2006; Kujala et al., 1997). These results suggest patterns. Perception & Psychophysics, 62(2), 301–312.
Hatwell, Y. (2003). Psychologie cognitive de la cécité précoce. (Paris: Dunod).
that the sensitivities of the tests designed for the present study were Hill-Briggs, F., Dial, J.G., Morere, D.A., & Joyce, A. (2007). Neuropsychological assessment
good. of persons with physical disability, visual impairment or blindness, and hearing
The absence of differences may be due to the duration of blindness impairment or deafness. Archives of Clinical Neuropsychology, 22, 389–404.
Hugdahl, K., Ek, M., Takio, F., Rintee, T., Tuomainen, J., Haarala, C., et al. (2004). Blind
among the late blind participants included in this study. These partici-
individuals show enhanced perceptual and attentional sensitivity for identification
pants had been without vision for an average of 15.9 years (SD = 9.5 of speech sounds. Cognitive Brain Research, 19, 28–32.
years). It is possible that the practice of daily activities without vision Hull, T., & Mason, H. (1995). Performance of blind children on digit-span tests. Journal of
during this relatively long period of time led to enhancements of atten- Visual Impairment and Blindness, 89, 166.
Kane, M.J., Conwal, A.R.A., Miura, T.K., & Colflesh, G.J.H. (2007). Working memory,
tion and working memory that were nearly as important as those attention control, and the n-back task: A question of construct validity. Journal of
observed in the early blind persons. Further investigations with larger Experimental Psychology: Learning, Memory, and Cognition, 33, 615–622.
samples are required to determine the effect of the age of blindness Kujala, T., Lehtokoski, A., Alho, K., Kekoni, J., & Näätänen, R. (1997). Faster reaction times
in the blind than sighted during bimodal divided attention. Acta Psychologica, 96,
onset on improvements in attentional and working memory abilities. 75–82.
Lerens, E., & Renier, L. (2014). Does visual experience influence the spatial distribution of
4.3. Conclusions auditory attention? Acta Psychologica, 146, 58–62.
Lewald, J. (2002). Opposing effects of head position on sound localization in blind and
sighted human subjects. European Journal of Neuroscience, 15(7), 1219–1224.
The present study seems to be the first attempt to provide an overall Miyake, A., Friedman, N.P., Emerson, M.J., Witzki, A.H., Howerter, A., & Wager, T.D. (2000).
assessment of the attentional functioning of blind persons, although The unity and diversity of executive functions and their contributions to complex
“frontal lobe” tasks: A latent variable analysis. Cognitive Psychology, 41, 49–100.
several other studies have focused on specific cognitive mechanisms. Occelli, V., Spence, C., & Zampini, M. (2013). Auditory, tactile, and audiotactile information
Indeed, our results indicate that blindness (since birth or appearing in processing following visual deprivation. Psychological Bulletin, 139, 189–212.
adulthood) seems to lead to information processes and manipulations Pascual-Leone, A., & Torres, F. (1993). Plasticity of the sensorimotor cortex representation
of the reading finger in Braille readers. Brain, 116(1), 39–52.
that are fast and efficient.
Price, J. R., Mount, G. G., & Coles, E. A. (1987). Evaluating the visually impaired: neuropsy-
The set of tests developed for this study is easy to use and might be chological technique. Journal of Visual Impairment and Blindness, 48, 20–30.
helpful to allow clinicians to appreciate the cognitive functioning of Rokem, A., & Ahissar, M. (2009). Interaction of cognitive and auditory abilities in congen-
people with visual impairments. Considering the accelerated aging of itally blind individuals. Neuropsychologia, 47, 843–848.
Ruggiero, G., & Iachini, T. (2010). The role of vision in the Corsi block-tapping task:
the population, the increasing numbers of persons with age-related Evidence from blind and sighted people. Neuropsychology, 24, 674–679.
visual diseases (such as age-related macular degeneration) and the Spector, A., & Biederman, I. (1976). Mental set and mental shift revisited. The American
declines in attentional and working memory functions that occur with Journal of Psychology, 89(4), 669.
Swanson, H.L., & Luxenberg, D. (2009). Short-term memory and working memory in
normal and pathological aging, it is important to have tools that can children with blindness: Support for a domain general or domain specific system?
evaluate these processes in the absence of the visual modality. Child Neuropsychology, 15, 280–294.
Thinus-Blanc, C., & Gaunet, F. (1997). Representation of space in blind persons: Vision as a
spatial sense? Psychological Bulletin, 121(1), 20–42.
Acknowledgments Wan, C.Y., Wood, A.G., Reutens, D.C., & Wilson, S.J. (2010). Early but not late-blindness
leads to enhanced auditory perception. Neuropsychologia, 48, 344–348.
We gratefully thank all the volunteers and the Fédération des Weaver, K.E., & Stevens, A.A. (2007). Attention and sensory interactions within the
occipital cortex in the early blind: An fMRI study. Journal of Cognitive Neuroscience,
Aveugles et Handicapés Visuels de France, Association Valentin
19, 315–330.
Haüy de Lyon, Association Valentin Haüy de Marseille and Associa- Wechsler, D. (1955). Manual for the Wechsler Adult Intelligence Scale.
tion Valentin Haüy de Grenoble for their collaboration. Wechsler, D. (1981). Wechsler Adult Intelligence Scale Revised. New-York: Psychological
Corporation.
Withagen, A., Kappers, A.M.L., Vervloed, M.P.J., Knoors, H., & Verhoeven, L. (2013). Short
References term memory and working memory in blind versus sighted children. Research in
Developmental Disabilities, 34, 2161–2172.
Allain, P., Forgeau, M., Kefi, M. Z., Etcharry-Bouyx, F., & Le Gall, D. (2002). Evaluation des
troubles attentionnels chez des traumatisés crâniens sévères: intérêt d’une adapta-
tion francophone du Test of Everyday Attention. Revue de Neuropsychologie, 12(3),
401–435.