Effect of External Irrelevant Distracters on a Visual Search Test in School-Age Children

Abstract

Objective: This study aimed to test the effect of an irrelevant external distracter included in a computer-administered visual search test. Two hypotheses were tested: (a) If the distracter affects performance, attention efficiency will be lowered; (b) if children do not habituate to the distracter, performance will be lower for every item of the test. Method: Distraction was induced changing the screen color unexpectedly several times in each trial—450 children (225 girls and 225 boys) from second to sixth course were tested. This group was compared with a group of 423 children from the same age range who were tested with the same test without distraction. Results: Induced distraction reduced attention efficiency for all ages and for every trial in the treatment group (test with distraction). Speed was lower, but number of errors did not increase. Conclusion: School-age children cope with an irrelevant external distracter by reducing speed, not accuracy.

Keywords

distraction visual search school-age children computerized assessment

“Attention” is a broad term most frequently used to describe the ease by which attention is sustained, the ease by which relevant stimuli are selected while nonrelevant ones are ignored, and the ease in inhibiting automatic answers (Lehman, Naglieri, & Aquilino, 2010; Manly et al., 2001; Posner & Rothbart, 2007). This broadness is reflected in attention not being a unitary brain process, with different neuroanatomical networks carrying out distinct cognitive processes (Posner & Petersen, 1990). One model that fits data for adults and children is a three-factor model supporting separable but correlated attention systems. For children from 6 to 16 years of age, selective attention correlates at .72 with attentional control/switching and at .40 with sustained attention, and attentional control/switching correlates at .60 with sustained attention, at the latent level (Manly et al., 2001).

For the school-age group, visual search and continuous performance tasks (CPTs) are usually the tests used to assess selective and sustained attention (Strauss, Sherman, & Spreen, 2006). Visual search refers to the ease by which similarities and differences between visual images are detected. Targets are usually only slightly different from nontargets that are considered distractive items. CPTs assess the ease to detect a rare target (selective attention) or to inhibit for responding (sustained attention). In the first case, CPT measures inattention or distraction (target appears in 22% of the trials). In the second case, CPT measures response inhibition (target appears in 78% of the trials).

Visual search tasks provide the closest test of attention in a classroom, a street, a market, the playground, and so on, that is, in the real world (Wolfe et al., 2012). This is why in the school-age group, visual search tests are the most frequently used tasks to assess selective attention. In the school-age group, ADHD is the most diagnosed disorder (Willcutt, 2012). ADHD refers to levels of attention and/or activity that do not correspond with chronological age. Visual search tests have been used to differentiate between groups of children rated by teachers as good or poor attenders (Cornish, Wilding, & Hollis, 2008; Wilding, Munir, & Cornish, 2001; Wilding, Pankhania, & Williams, 2007) or between good attenders and those with ADHD (Manly et al., 2001), from attention efficiency indexes. However, practitioners have to check for other nonadaptive facets of attention other than general attention efficiency, when checking for an attention disorder (Diagnostic and Statistical Manual of Mental Disorders [4th ed., text rev.; DSM-IV-TR]; American Psychiatric Association, 2000). One of these aspects is whether a child tends to be distracted by irrelevant stimuli (external and/or internal). The main question that this study raises is how distraction is being measured with visual search tests, and whether it can be improved.

The well-known visual search tests for school-age children are the following (see Table 1 for a summary of their main features). (a) The Visual Search—In this test, children have to locate target pictures, that is, cats, in a random array containing bunnies, houses, and trees. (b) The Visual Attention Test—This is a cancellation test in which the child must mark the targets, that is, two faces, as quickly and accurately as possible. (c) The Number Detection—Here children have to find specified targets, that is, three to six numbers depending on age, and printed in different fonts for older children, while ignoring the other numbers. Children have to work row by row and from left to right (it is a cancellation-type task). (d) The Sky Search—In this test, children have to find a pair of identical spacecraft; it is a serial self-terminating task in which score computations control for differences in motor speed. (e) The Map Mission—This is a timed-out test in which children have 1 min to find small restaurant symbols in a laminated city map, while ignoring other symbols. (f) Visearch, Single Target Search—In this test, children must find “the king monster” by first finding all the hidden monsters, that is, by clicking the vertical black ellipses while ignoring the other colors, shapes, and sizes. (g) The Simple Visual Discrimination of Trees (Divisa)—The child must click on targets, that is, a specific tree in each trial, while ignoring nontargets. This test consists of eight trials, whereas the other tests mentioned above include only one trial.

Table 1.

Visual Search Tests.

Test	Age range	Test measuring	Scores	Measuring distraction	Developmental pattern
The Visual Search test from the NEPSY (Klenberg, Korkman, & Lahti-Nuuttila, 2001). Paper and pencil test	3 to 12 years	Visual search (random array)	Time to completion, omission errors, commission errors	No	Increasing from 3 to 7, plateau; increasing from 8 to 10. Plateau from 10 to 12
The Visual Attention test from the NEPSY (Klenberg et al., 2001). Paper and pencil test	5 to 12 years	Visual scanning of linear arrays of pictures. Cancellation task	Time to completion, omission errors, commission errors	No	Increasing from 5 to 10 and plateau to 12
Number Detection test from the Cognitive Assessment System (Naglieri, 1997). Paper and pencil test	5 to 17 years, 11 months	Visual search, cancellation task	Efficiency Index (hits minus commissions)/time taken	No	Performance improves from 5 to 7 steadily and from 8 to 15
The Sky Search from the Test of Everyday Attention for Children (Manly et al., 2001). Paper and pencil test	6 to 16 years	Visual search, cancellation task, 20 targets among 108 distractors	Time per target (final score adjusted for motor speed)	No. It can be deduced for longer times	Continuous decreasing time per target from 6 to 16
The Map Mission also from Test of Everyday Attention for Children (Manly et al., 2001). Paper and pencil test	6 to 16 years	Time out visual search, cancellation task with fixed time (1 min)	Number of targets correctly marked.	No	Continuous increasing number of correct targets from 6 to 16
Visearch, single target search (Wilding, Munir, & Cornish, 2001). Computerized assessment	4 to 12 years	Visual search: 25 targets randomly located among 100 shapes in all	False alarms (commission errors), time/hit, distance/hit	No	—
The Simple Visual Discrimination of Trees (Divisa; Santacreu, Shih, & Quiroga, 2011). Computerized assessment	6 to 12 years	Visual Search (random array). Eight trials/screens	Global Attention Index, commission errors, omission errors, Task Organization Index, Distraction-Hastiness Index	No	Global Attention Index steadily increases from 6 to 12 while commission and omission errors decrease. Task Organization Index and Distraction-Hastiness Index do not show differences between age groups but within groups

Scores from visual search tests reflect attention efficiency either as time per target (Sky Search, Visearch), as accuracy (Map Mission), or as accuracy to total time (Visual Search, Visual Attention, Number Detection, Divisa). In all cases, higher scores reflect higher selective attention. Some of these tests provide other scores that add information to that obtained by the global index. Complementary scores can refer to accuracy, speed, and variability, but only Divisa provides scores for organization and for distraction-hastiness.

Distraction means that attention is not focused but is divided between the object to be attended and the source of distraction. Distractions can come from the personal world (fantasy and emotions) and the external environment where the person is functioning (Santostefano, 1985/1995). Distraction is usually measured through tasks that rely on interference-control paradigms, as in the Flankers Task and the Stroop Test (Foster & Lavie, 2008; Stins, Polderman, Boomsma, & de Geus, 2005). Interference-control paradigms measure what some researchers have called inner distraction (Santostefano, 1988), because an inner item produces the interference (similar to what happens in the real life when some thoughts or images distract you while driving). By contrast, outer distraction is when another person, a noise, or something else disturbs you while performing a task (e.g., outer distracters are advertisements on the road while driving). However, distractions coming from the external environment are not measured by the common interference-control paradigms.

In the visual search tasks, distracters are nontarget stimuli, but to find targets you need to compare them with nontargets, and so nontargets form part of the task and cannot be ignored. A genuine distracter should be nonrelevant for the task at hand, as in the Fruit Distraction Test (FDT) by Santostefano (1988). In this test, the child must read the colors of fruits (10 rows of 5 columns) while ignoring the small black and white drawings that surround the colored fruits. FDT measures the ability to focus attention on relevant stimuli while nonrelevant ones are encountered. Thus, visual search tests assess distraction in an indirect manner. The measurement of selective attention usually relies on variations in the type and amount of nontarget stimuli included (Wolfe & Horowitz, 2004). Nontarget stimuli are considered the distracter items. Nevertheless, targets and nontargets constitute the task’s stimuli, and similarities and differences between them, as well as their amount, allow defining and measuring task complexity and difficulty (Wolfe et al., 2012). But, in a strict sense, they do not measure the distraction or the influence of an irrelevant (distracting) stimulus.

This study was designed to incorporate a genuine distracter in a visual search test. The test selected was Divisa (Santacreu, Shih, & Quiroga, 2011), because it is a computerized test; it includes separate scores for attention efficiency, task organization, and distraction-hastiness; and its psychometric properties are quite good (see “Method” section). To induce distraction, screen color changed several times in each trial in Divisa-D (Divisa with Distraction). Screen color is a nonrelevant piece of information for the task of searching trees. The presentation of a distracting stimulus, which is not part of the information to be attended, influences task performance because it generates perceptual overload (Lavie, 1995, 2005). Using a response-competition paradigm, Lavie and Cox (1997) showed that when perceptual load is high, distracters do not interfere with performance (measured as response time for incongruent stimuli minus response time for congruent stimuli, and so response times were almost equal for both types of stimuli). When perceptual load was low, however, the response time for incongruent stimuli minus response time for congruent stimuli was higher.

The same effect has been confirmed with children. Using the same task as Huang-Pollock, Carr, and Nigg (2002), Lavie and Cox (1997) confirmed that for children from 7 to 12 years of age, as perceptual load increases, the effects from incompatible distracters decrease. This interference cannot be explained by perceptual grouping. Couperus (2011) also obtained changes in processing as a function of perceptual load from 7 to 19 years of age and more. Younger children needed more exposure time to maintain a similar level of accuracy on high perceptual load trials than did adult participants. Thus, the selective attention of children develops following different trajectories for automatic and effortful attentional processes, as postulated by Lavie (1995) and as shown by studies with children (Couperus, 2011; Huang-Pollock et al., 2002). The reduced effect of the distracter stimulus on tasks with high perceptual load has been shown with different paradigms (response-competition, implicit learning about spatial configuration of irrelevant stimuli, positive priming, and negative priming; see Lavie, 2005, for a review) with an exception: when using distracters with social meaning in a name categorization task (Lavie, Ro, & Russell, 2003). Perceptual load should be low to maintain its distraction feature; otherwise (high perceptual load), distracters are not processed.

In this study, a low perceptual load distracter was included. The main goal was to know how, and how much, performance is influenced by the distracter. To induce distraction, the screen color changed several times in each trial. The irrelevant stimuli, screen color, competed with targets for perceptual processing. Two hypotheses were tested: (a) If the distracter affects performance, attention efficiency will be lowered at the mean level for all the age groups; (b) if distracter is pervasive (children do not habituate to it), performance will be lowered in every item; otherwise, only the first items will be affected. Prior to testing the hypotheses, the assumed low perceptual load of the distracter as well as the reliability indexes of the test’s scores and their developmental sensitivity were checked.

Method

Participants

In this study, 450 children, from second to sixth primary school course (7-12 years old), were assessed. The whole group was divided into five course groups: second grade (n = 92, M age = 7.20, SD = 0.54), third grade (n = 84, M age = 8.12, SD = 0.33), fourth grade (n = 84, M age = 9.17, SD = 0.41), fifth grade (n = 94, M age = 10.16, SD = 0.39), and sixth grade (n = 96, M age = 11.22, SD = 0.41). They were attending state schools in the surroundings of a big city in Spain. There were 225 girls and 225 boys. The total number of assessed children was 497. Participants were excluded when teachers referred severe learning problems, even if not having been diagnosed (n = 14)—if having a Learning Disorder (LD) or an ADHD diagnosis (n = 11) or if children were following a special curriculum due to special needs (n = 15, usually immigrants who are not fluent in Spanish). These 47 children were also tested to avoid them feeling excluded from the group, but their scores were not registered.

The sample from this study did not differ either by sex, χ²(1, n = 874) = .019, p = .893, or by mean age in each course, F(4, 864) = 0.395, p = .812, from the first study on Divisa (Quiroga, Santacreu, Montoro, Martínez-Molina, & Shih, 2011). Thus, the performance from the two samples will be compared.

Parental consent was obtained for the students’ participation prior to assessment, and the study was carried out in accordance with the Personal Data Protection Law and following School Board policies on the use of research data from each of the two public schools in which the tests were administered. Data on socioeconomic status (SES) were available at the school level. Two schools participated in this study from a wider group of five schools that participated in a wider study (Santacreu et al., 2011). Both schools showed predominantly average SES, meaning medium to high qualification of parents and annual incomes from 20,000 to 30,000 Euros (http://www.ine.es/daco/daco42/sociales11/sociales.htm).

Procedure

Testing was conducted in the schools during the academic schedule. Each school has a computer room that was used to test 30 children each time. Before testing, teachers certified that all children had enough experience using a computer and a mouse.

Instructions for the task were projected on a big screen. After the instructions were read and the task was explained, if necessary, the children were helped individually. Three experimental psychologists were present in the assessment of each group.

Instruments

Divisa is a visual search task administered and scored online. A demo version can be found at http://www.uam.es/proyectosinv/psimasd/ (click on “Current Projects” and go to the first task left upper corner). This demo includes instructions for adults. When Divisa was elaborated, however, instructions for children were developed. Norms were obtained from a big nonclinical sample of 1,442 children from second to sixth grade (Santacreu et al., 2011).

In Divisa, the child should click on the trees that match the sample (see Figure 1). The target tree differs from trial to trial. The test consists of eight trials (items or screens). Each trial consists of a 12 × 12 matrix, where 14 trees identical to the target (10%), plus 30 trees different from the target, are located. Resting cells are empty. The first trial (screen) performance is not included when computing scores, because it is intended to introduce the task to the child. Going from one item to the next, the child must press the “following” button when he or she considers that all trees have been selected. The time limit for each trial is 60 s. When a tree identical to the target tree is selected, it disappears; otherwise nothing occurs. Divisa can be administered in Spanish or English (http://web.teaediciones.com/DIVISA–Trees-Simple-Visual-Discrimination-Test.aspx—to download the Manual of the Test in English).

Figure 1.

Divisa test: Instructions screen and example of a trial.

From the performance, nine scores are obtained (see Table 2 for a wide description): Global Attention Index (GAI), commission errors, omission errors, Task Organization Index (TOI), mean response time for commission errors, mean response time for hits, Distraction-Hastiness Index (DHI), interval between presses, and liability between trials. Mean reaction time for commission errors and mean reaction time for hits are computed separately to know what is the reason for a commission error (Conners, Epstein, Angold, & Klaric, 2003). DHI is computed as the mean response time for commission errors minus mean response time for hits. If the mean reaction time for commission errors is high (the child shows delayed answers when failing), the DHI is positive and the meaning of the score is inattention. If the mean reaction time for commission errors is low (participant shows a quick answer when failing), the DHI is negative and the meaning of the score is precipitation or hastiness.

Table 2.

Divisa Scores Description.

Divisa indexes	Computations	Concept being measured	Theoretical description
Global Attention Index	Number of correct responses minus number of commission errors, divided by duration. Computed for each item and for the test	Speed at which a correct execution is performed	Attention efficiency measure (Lehman, Naglieri, & Aquilino, 2010)
Commission errors	Number of clicks on nontarget trees	Lack of attention	Attention deficit measure due to lack of inhibition of automate responses (Klimkeit, Sheppard, Lee, & Bradshaw, 2004)
Omission errors	Total number of target trees not clicked	Sustained attention deficit independent from response time	Sustained attention measure (Barkley, 1991)
Task Organization Index	Summing one point each time target tree clicked is in the same row or columns than the previous one	Organization level when performing the task	Consciousness (Hernández, Sánchez-Balmisa, Madrid, & Santacreu, 2003)
RTE	Mean time for commission errors (nontarget trees and clicks on the screen)	Inattention when high and hastiness when low	Inattention (Conners, Epstein, Angold, & Klaric, 2003)
RTH	Mean time for hits	Slowness when high and quickness when low	Speed of visual search
Distraction-hastiness	RTE–RTH	Attention deficit due to distraction or hastiness	Slow performance when failing due to distraction versus quick performance when failing due to hastiness (impulsivity)
Interval between presses	Mean time between a press and the next one. Trial duration divided by number of presses	Distracted—inner oriented when high	Performance variability, distraction when high
Liability between trials	Standard deviation among trials for the Global Attention Index	Distracted—inconsistent when high	Performance variability, loosing task goal when high

Note. RTE = response time for commission errors; RTH = response time for hits.

Reliability indexes (Cronbach’s α) range from .77 for omission errors to .95 for the GAI (Quiroga et al., 2011). Divisa’s GAI correlates at .59 with the total number of symbols processed and at .57 with the concentration performance, both scores from the d2 test by Brickenkamp (1962/2009). Predictive power falls between 70.2% and 96.5% of success when comparing control and ADHD children, depending on chronological age (higher predictive power for older children).

In this study, the screen color of Divisa changed 7 times in each trial (Divisa-D), at a random occurrence ranging from 5 to 7 s, always unexpectedly. The colors were chosen to be different from any part of the search stimuli.

Results

Reliability indexes (Cronbach’s α) for Divisa-D scores were computed. The obtained results are very similar to those obtained for the original version: GAI, α = .95; commission errors, α = .93; omission errors, α = .85; interval between presses, α = .92; and mean response time for hits, α = .86.

To test the developmental sensitivity of Divisa-D, a MANOVA was performed to check whether the scores from Divisa-D differentiate between the age groups as the scores of Divisa do, that is to say, if Divisa and Divisa-D are comparable in terms of developmental sensitivity. Comparable would mean that the interaction between course and version is not statistically significant and so two independent samples of more than 400 children will show the same developmental trends in the main variables. The main course effect was statistically significant, F_{Pillai Test}(36, 3,072) = 13.16, p < .001, $η_{p}^{2}$ = .13, but the interaction Course × Divisa version was not statistically significant, F_{Pillai Test}(36, 3,072) = 0.951, p = .553, $η_{p}^{2}$ = .01, as expected. Thus, Divisa-D is as sensitive as Divisa to developmental changes. Table 3 includes descriptive (M and SD) and comparative statistics (F and $η_{p}^{2}$ ) for course and for each score. The interesting point to underline is that the higher effect sizes ( $η_{p}^{2}$ ) for the comparison between courses appear in the GAI and in the response time for hits, and are equal in magnitude (d = .39), showing clearly that developmental increase in the GAI is due to the increase in speed. To support this conclusion, the next effect size in magnitude is for the interval between presses. Figure 2 shows the developmental pattern for Divisa-D main scores. This figure displays age patterns, expressed, for each score, in standard scores (M = 0, SD = 1) computed for the whole sample (N = 450). These data show an increasing pattern with age for the GAI (sharp slope), two decreasing patterns for commission and omission errors (slight slope), and two parallel patterns for TOI and DHI (qualitative differences, not quantitative).

Table 3.

Mean and Standard Deviation for Each Course Group and Test (Divisa and Divisa-D).

	Second (n = 66)		Third (n = 112)		Fourth (n = 78)		Fifth (n = 67)		Sixth (n = 101)						F _{Course × Divisa version}
Divisa	M	SD	M	SD	M	SD	M	SD	M	SD	F _Course	$η_{p Course}^{2}$	F _{Divisa version}	$η_{p Divisa Version}^{2}$	F _{Course × Divisa version}
Global Attention Index	0.45	0.12	0.54	0.12	0.58	0.14	0.65	0.14	0.77	0.17	139.62***	.39	28.21***	.03	0.88
Commission errors	2.55	1.87	1.61	4.52	1.45	4.62	1.36	2.54	0.71	1.87	5.74***	.03	2.84		1.22
Omission errors	3.76	3.17	3.18	3.34	2.41	3.20	1.30	1.52	1.68	1.71	16.17***	.07	20.59***	.02	0.37
Task Organization Index	17.59	3.45	18.05	3.50	17.29	3.18	17.82	3.53	17.97	4.01	1.38		2.37		0.33
Distraction-Hastiness Index	0.39	1.13	0.36	1.07	0.29	1.10	0.23	0.70	0.21	0.76	0.29		0.14		0.28
Interval between presses	2.08	0.43	1.77	0.37	1.70	0.43	1.50	0.34	1.29	0.30	74.73***	.26	31.56***	.04	1.52
Liability between trials	0.11	0.04	0.12	0.05	0.13	0.06	0.14	0.06	0.14	0.05	15.93***	.07	4.21*	.01	1.20
RTE^a	2.07	1.16	1.75	1.14	1.57	1.19	1.37	0.72	1.19	0.78	13.19***	.06	8.71**	.01	0.27
Response time for correct responses	1.69	0.35	1.40	0.27	1.28	0.28	1.14	0.23	1.00	0.20	135.24***	.39	29.66***	.03	1.31
	Second (n = 92)		Third (n = 86)		Fourth (n = 85)		Fifth (n = 95)		Sixth (n = 96)
Divisa-D	M	SD	M	SD	M	SD	M	SD	M	SD
Global Attention Index	0.39	0.12	0.46	0.13	0.54	0.14	0.62	0.13	0.75	0.18
Commission errors	4.22	9.64	3.31	9.58	1.40	5.83	1.07	3.66	0.93	1.93
Omission errors	5.16	5.26	3.88	3.78	3.71	4.32	2.37	3.41	2.51	2.52
Task Organization Index	17.07	3.33	17.30	4.03	17.05	3.12	17.40	4.13	18.01	3.53
Distraction-Hastiness Index	0.29	1.43	0.20	0.84	0.32	1.30	0.30	1.34	0.22	1.21
Interval between presses	2.32	0.53	2.05	0.62	1.97	1.02	1.58	0.35	1.39	0.33
Liability between trials	0.10	0.04	0.10	0.05	0.13	0.07	0.13	0.05	0.15	0.06
RTE^b	2.34	1.49	1.85	0.83	1.92	1.34	1.61	1.26	1.46	1.17
Response time for correct responses	1.82	0.42	1.60	0.44	1.40	0.38	1.20	0.23	1.07	0.22

Note. Divisa-D = Divisa with Distraction; RTE = response time for commission errors.

n = 359.

n = 424.

p < .05. **p < .01. ***p < .001.

Figure 2.

Developmental patterns for Divisa-D scores.

To test the induced distraction level (low perceptual load), the GAIs for two different age groups (7-9 vs. 10-12) were compared. A variance analysis with two independent variables (Divisa test and age group) was performed. The results show no statistically significant interaction between age group and type of Divisa test, F(1, 870) = 0.634, p = .426, on the GAI. Thus, the distraction feature influences younger and older children equally, as expected. This result supports the assumption made that changing the color of the screen induces a low perceptual load.

To test the first hypothesis, a variance analysis was performed with type of Divisa as the independent variable. The dependent variable was the GAI. The obtained results show the GAI for Divisa-D to be lower than that of Divisa, as expected, M_Divisa = 0.608, SD = 0.18; M_Divisa-D = 0.552, SD = 0.18; F(1, 872) = 20.675, p < .001, d = .31, while variability remains similar, F_Levene(1, 872) = 0.82, p = .775. Thus, the distraction feature affects task efficiency with a medium effect size while maintaining variability between age groups. Three more ANOVAs were performed to test whether the distraction feature affects the three components of the GAI in the same or different way. Results show that the number of commission errors is similar in Divisa and Divisa-D, M_Divisa = 1.47, SD = 4.12; M_Divisa-D = 2.16, SD = 6.89; F_{Brown–Forsythe}(1, 872) = 3.29, p = .074. However, the number of hits and the duration differ between the two tests—hits: M_Divisa = 95.5, SD = 2.85; M_Divisa-D = 94.3, SD = 6.02; F_{Brown–Forsythe}(1, 872) = 15.03, p < .001, d = .32; duration: M_Divisa-D = 186, SD = 62.89; M_Divisa = 167, SD = 46.16; F_{Brown–Forsythe}(1, 872) = 24.42, p < .001, d = .33. These results show that the distraction feature lowers the speed of performance (lower number of hits and higher duration) without increasing inaccuracy (same number of commission errors).

For the second hypothesis, a MANOVA was performed to test whether the GAI was equally influenced through the trials or only at the beginning of the task when the distraction (changing of screen color) appeared unexpectedly for the first time. The MANOVA results show a Divisa version main effect, F_{Pillai Test}(7, 866)= 7.56, p < .001, $η_{p}^{2}$ = .06, meaning that for all the trials, Divisa-D differs from Divisa. Table 4 includes means and standard deviations for each trial (from 2 to 8) for the two Divisa versions as well as the univariate comparative statistic F. The data in Table 4 show that the GAI in each trial is lower for Divisa-D, confirming the influence of the distraction through the complete task and not only at the beginning—that is, it is not possible to avoid the influence of the distraction feature from the beginning to the end of the task.

Table 4.

Means, Standard Deviations, and Comparative Statistics for the Global Attention Index in Each Trial (Divisa and Divisa-D).

	Divisa		Divisa-D			$η_{p}^{2}$
	M	SD	M	SD	F(1, 872)	$η_{p}^{2}$
2	0.51	0.19	0.47	0.18	7.93**	.01
3	0.69	0.21	0.60	0.22	38.35***	.04
4	0.58	0.19	0.53	0.20	12.66***	.01
5	0.71	0.22	0.66	0.24	10.80**	.01
6	0.61	0.19	0.55	0.22	14.91***	.02
7	0.57	0.18	0.54	0.20	7.38**	.01
8	0.73	0.23	0.66	0.23	20.68***	.02

Note. Divisa-D = Divisa with Distraction.

p < .01. ***p < .001.

Discussion

This study was designed to include an irrelevant external distracter in a visual search test already published, that is, Divisa (Santacreu et al., 2011), and to check its influence comparing the performance of two big groups (sample size over 400)—one that took the Divisa and the other that completed the Divisa-D (with distraction). To our knowledge, no previous study has tested the influence of an irrelevant external stimulus on a visual search test with school-age children. Moreover, no other visual search test includes separate scores for task organization and distraction-hastiness, as Divisa does. The distraction feature included was a change in the screen color where stimuli were located. This feature is supposed to be a low perceptual load, following the proposal of Lavie (1995, 2005).

The results of this study show that Divisa-D preserves and even improves the reliability of its scores, with Cronbach’s alphas in the range of .85 to .95, which are high values (Cronbach, 1951; Schmitt, 1996). This is an important point and deserves to be underlined, because the distraction feature introduced in Divisa-D does not alter the order of individual differences tested through trials, that is, all trials differentiate equally between participants (which is the meaning of a high alpha) as Divisa does.

As for the influence of distraction incorporated into Divisa, the results show, first, that performance efficiency is slightly reduced, the size of this effect being medium to low (d = .31), following the standards of Cohen (1988). Moreover, and importantly, the groups’ variability remains similar. When the components of the GAI are compared, commission errors did not differ between the two groups but differed on hits and duration. Thus, the inclusion of an external irrelevant distracter lowered performance of all age groups without increasing inaccuracy. Again, the size effects are medium to low (.32 for hits and .33 for duration). Because Divisa has a time limit of 60 s per trial, the lower number of hits in Divisa-D is due to the need to use more time to find the targets in each trial. This conclusion is reinforced because the interval between presses increases in Divisa-D in all age groups. Results from this study show that distraction influences speed but not accuracy, which means that normal school-age children cope with the distraction induced by Divisa-D by reducing speed and so maintaining their accuracy level. Due to the transsectional nature of this study, obtained effects are at the mean level but, as should be pointed later, individual computations (Divisa-D and Divisa) would complete this picture in terms of the increase of time that distraction produces in each child.

Second, the external distracter used in Divisa-D affects performance across all trials. Participants could not avoid being distracted through the task even when realizing that the screen color changes from time to time. There is no habituation effect. As in the real life, distracters cannot be anticipated and so can neither be avoided nor modulated. This fact underlines the ecological validity of the test.

The distraction feature included in Divisa-D was supposed to be a low perceptual load. If the distraction feature produces a low perceptual loading for all ages, the two age groups will be affected in their performance. Otherwise, if the distraction feature included in Divisa-D produces a high perceptual load, young children would not have been distracted and their performance would have been almost the same between Divisa and Divisa-D, as can be deduced from Maylor and Lavie (1998). The obtained data show that the included distraction influences the performance of younger (7-9 years of age) as well as older children (10-12 years of age). Thus, the introduced perceptual loading does not increase task difficulty until the point where it becomes necessary to use all attentional resources to perform it (as should be the case for the youngest, it should result in having a high perceptual load). These results cannot be compared with previous studies, because this is the first study that includes an irrelevant external distracter on a visual search test. However, the studies from Lavie and colleagues have shown increasing response times when the task includes an irrelevant distracter (Foster & Lavie, 2008; Lavie & Cox, 1997).

The confirmation of the low perceptual load of the induced distraction supports the fact that Divisa-D maintains the developmental sensitivity that Divisa has. The developmental patterns depicted in Figure 2 are almost identical to the ones obtained for Divisa (Santacreu et al., 2011). Neither interaction (between course and Divisa version) was statistically significant, meaning that the developmental patterns remain similar between both tests. The inclusion of an irrelevant distracter has not altered the properties of the test and so the psychometric goodness of Divisa can be extended to Divisa-D. Moreover, the principal effect of course shows that children increase GAI with age ( $η_{p Course}^{2}$ = .39), reduce the response time for hits ( $η_{p Course}^{2}$ = .39), and reduce the interval between presses ( $η_{p Course}^{2}$ = .26). These results clearly indicate that development means reducing the time needed to complete the task with the same accuracy. Couperus (2011) reached a similar conclusion about differences in latencies as a function of age; children needed more time to maintain accuracy when exposure time was reduced. Moreover, Wetzel, Widmann, and Schröger (2009) concluded that the response time of children 7 and 8 years of age increased when a distracter was present. This study reinforces and extends this conclusion also to 10 to 12 years old. Although different tasks were used, all studies point to the same conclusion: Development from 7 to 12 years of age in the performance of visual search tasks means increasing speed (reducing response time), but when confronted to irrelevant distracters, all age groups increase the mean of response time while maintaining the differences among them.

Besides the GAI, Divisa includes two interesting scores, the TOI and the DHI, which deserve some comments. Performance from the groups tested did not differ, either in the TOI or in the DHI. Thus, the induced distraction does not alter each child’s searching order. A higher TOI score means a more ordered search (through rows and columns from top to down and left to right). No previous research has studied the search sequence in visual search tasks. The most similar variable is conscientiousness style in adults assessed through objective tasks (Hernández, Lozano, Shih, & Santacreu, 2009; Hernández, Sánchez-Balmisa, Madrid, & Santacreu, 2003). The results from these studies show that the search sequence is a qualitative trait; actually norms from Divisa also show this point (Santacreu et al., 2011).

As for the DHI, the induced distraction does not lead to a higher distraction mean level of the group. It should be kept in mind that the DHI is computed as the response time for error minus the response time for hits. A positive mean (as the one obtained for all the age groups tested in this study and in the former with Divisa) indicates that the response time for errors is higher than response time for hits. As the obtained results show, response time for error and response time for hits increased, showing again that an irrelevant external stimulus lowers the performance speed in a visual search test as Divisa.

All the obtained results point to the conclusion that children would lower speed performance, but it could be the case that the change in colors had increased the perceptual difficulty of the target stimuli. However, whether this was the case, commission errors would also have been increased. Children in this study had a lower number of hits and spent more time to collect them, but the number of commission errors they made did not increase. Thus, the slowed response seems to be the result of induced distraction not of perceptual factors. Nevertheless, future studies should specifically be designed to test this possibility. Moreover, some children could have been more distracted than others because they took the test in groups. Testing in this study was supervised by three psychologists to warrant that children completed the task silently and not disturbing other children. However, this possibility cannot be completely refused.

So should we use Divisa, Divisa-D, or both? The obtained results show that both tests have very good reliability indexes and developmental sensitivity. Both tests could be used with a child or group of children to compute within-individual differences (for the five main scores, i.e., GAI with distraction–GAI without distraction) that should be compared with the age group mean. This was the procedure Santostefano (1988) followed with the FDT. Santostefano showed that distracters influence all children but not in the same amount. Thus, the interesting point is to assess how much distraction each child experiences when compared with his or her age group, when coping with the same distracter. To use both tests conjointly gives extra information from within-individual differences scores, very valuable for practitioners: the specific amount of distraction a child experiences when confronted to irrelevant distracters.

This study let some questions to be answered in future researches. The effect of other types, and/or content, of irrelevant external stimuli remains to be studied, such as moving objects in the periphery. In this sense, Bartram, Ware, and Calvert (2003) showed that, at least in adults, moving cues (moticons) are easier and quicker to detect. Moticons could be promising distracters to be included in visual search tests to represent those distractions that persist but move, while performing a task.

It also remains to be studied how distracters influence performance in clinical groups (as, for example, ADHD) in a visual search test. One possibility is that ADHD children would increase commission errors but maintain speed performance as other researchers have found with other tasks (Manly et al., 2001; Mason, Humphreys, & Kent, 2003). Another interesting question that remains to be studied is whether distraction influences task order performance in ADHD children. As we have shown, distraction does not alter task order performance in a nonclinical group.

In summary, in this study, we have started differentiating between distractions coming from the personal world (fantasy and emotions) and distractions coming from the external environment as was formerly proposed by Santostefano (1985/1995). From this distinction, we have stated that interference-control paradigms are assessing inner but not outer distractions. To assess outer distractions (distracters coming from the environment where the person is functioning), we have proposed to include irrelevant external distracters to visual search tests. Using Divisa (Santacreu et al., 2011), this study has shown how an irrelevant external distracter influences performance. Children from 7 to 12 years of age cope with an irrelevant distracter by reducing performance speed, but not accuracy. Moreover, response time for errors and response time for hits are higher in the five age groups tested. Interestingly, the sequence of searching was not altered by the induced distraction, showing that an ordered pattern of search could rely more on temper characteristics in school-age children. Summing up, the inclusion of an irrelevant distracter in visual search tests improves the assessment of individual differences in selective attention in school-age children.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

M. A. Quiroga, PhD, is a university professor of differential psychology at the Universidad Complutense de Madrid; fellow of the International Society for the Study of Individual Differences, and present president of the Asociación Iberoamericana para la Investigación de las Diferencias Individuales (Iberoamerican Society for Research on Individual Differences). Her main research interests have focused on cognitive styles and on intelligence. Now she is developing new computerized cognitive measures for children and adults.

J. Santacreu, PhD, is a university professor of personality and psychological assessment at the Universidad Autónoma de Madrid; fellow of the Iberoamerican Society for Research on Individual Differences. His main research interests include the development of WEB objective tests in cognitive and personality areas: dynamic spatial ability, attention, and risk-taking behavior.

C. López-Cavada received his MA degree in clinical psychology at Universidad Complutense de Madrid in 2011. He works in a Juvenile Correctional, while completing a master’s degree in forensic sciences.

E. Capote is a child psychologist. She received her MA in research methodology at Universidad Camilo José Cela. Presently, she works in a primary school where she assesses and provides treatment for children with learning disabilities and ADHD.

D. Morillo is a PhD student in statistics and behavioral sciences at Universidad Autónoma de Madrid. He received his MA in cognitive psychology from Universidad Complutense de Madrid in 2011, having previously received an MA in telecommunication engineering in 2007. His research interests include psychological assessment and software development as a means for psychological assessment and intervention.

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

Barkley

R. A.

(1991). The ecological validity of laboratory and analogue assessment methods of ADHD symptoms. Journal of Abnormal Psychology, 19, 149-178.

Bartram

Ware

Calvert

(2003). Moticons: Detection, distraction and task (special issue on Notifications and Interruptions). International Journal of Human-Computer Studies, 58, 515-545.

Brickenkamp

(2009). d2, Test de Atención 3ª Edición [d2, Test of Attention, 3rd ed.]. Madrid, Spain: TEA Ediciones. (Original work published 1962)

Cohen

(1988). Statistical power analysis for the behavioral sciences. Hillsdale, MI: LEA.

Conners

C. K.

Epstein

J. N.

Angold

Klaric

(2003). Continuous performance test performance in a normative epidemiological sample. Journal of Abnormal Child Psychology, 31, 555-562.

Cornish

K. M.

Wilding

J. M.

Hollis

(2008). Visual search performance in children rated as good or poor attenders: The differential impact of DAT1 genotype, IQ, and chronological age. Neuropsychology, 22, 217-225.

Couperus

J. W.

(2011). Perceptual load influences selective attention across development. Developmental Psychology, 47, 1431-1439.

Cronbach

(1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-334.

10.

Foster

Lavie

(2008). Failures to ignore entirely irrelevant distractors: The role of load. Journal of Experimental Psychology: Applied, 14, 73-83.

11.

Hernández

J. M.

Lozano

J. H.

Shih

P. C.

Santacreu

(2009). Validez convergente de dos pruebas de evaluación de la minuciosidad [Convergent validity of two tests of conscientiousness]. Psicothema, 21, 133-140.

12.

Hernández

J. M.

Sánchez-Balmisa

Madrid

Santacreu

(2003). La evaluación objetiva de la minuciosidad: Diseño de una prueba conductual [Objective assessment of conscientiousness: Design of a new behavioral task]. Análisis y modificación de conducta, 29, 455-477.

13.

Huang-Pollock

C. L.

Carr

T. H.

Nigg

J. T.

(2002). Development of selective attention: Perceptual load influences early versus late attentional selection in children and adults. Developmental Psychology, 38, 363-375.

14.

Klenberg

Korkman

Lahti-Nuuttila

(2001). Differential development of attention and executive functions in 3- to 12-year-old Finnish children. Developmental Neuropsychology, 20, 407-428.

15.

Klimkeit

E. I.

Sheppard

D. M.

Lee

Bradshaw

J. L.

(2004). Bimanual coordination deficits in attention deficit/hyperactivity disorder (ADHD). Journal of Clinical and Experimental Neuropsychology, 26, 999-1010.

16.

Lavie

(1995). Perceptual load as a necessary condition for selective attention. Journal of Experimental Psychology: Human Perception and Performance, 21, 451-468.

17.

Lavie

(2005). Distracted and confused? Selective attention under load. Trends in Cognitive Science, 9, 75-82.

18.

Lavie

Cox

(1997). On the efficiency of visual selective attention: Efficient visual search leads to inefficient distractor rejection. Psychological Science, 8, 395-398.

19.

Lavie

Russell

(2003). The role of perceptual load in processing distractor faces. Psychological Science, 14, 510-515.

20.

Lehman

E. B.

Naglieri

J. A.

Aquilino

S. A.

(2010). A national study on the development of visual attention using the cognitive assessment system. Journal of Attention Disorders, 14, 15-24.

21.

Manly

Anderson

Nimmo-Smith

Turner

Watson

Robertson

I. H.

(2001). The differential assessment of children’s attention: The Test of Everyday Attention for Children (TEA-Ch), normative sample and ADHD performance. Journal of Child Psychology, Psychiatry and Allied Disciplines, 42, 1065-1081.

22.

Mason

D. J.

Humphreys

G. W.

Kent

L. S.

(2003). Exploring selective attention in ADHD: Visual search through space and time. Journal of Child Psychology and Psychiatry, 44, 1158-1176.

23.

Maylor

E. A.

Lavie

(1998). The influence of perceptual load on age differences in selective attention. Psychology and Aging, 13, 563-573.

24.

Naglieri

J. A.

(1997). Planning, attention, simultaneous, and successive theory and the cognitive assessment system: A new theory-based measure of intelligence. In Flanagan

D. P.

Genshaft

J. L.

Harrison

P. L.

(Eds.), Contemporary intellectual assessment. Theories, tests and issues (pp. 247-267). New York, NY: Guilford.

25.

Posner

M. I.

Petersen

S. E.

(1990). The attention system of human brain. Annual Review of Neuroscience, 13, 25-42.

26.

Posner

M. I.

Rothbart

M. K.

(2007). Research on attention networks as a model for the integration of psychological science. Annual Review of Psychology, 58, 1-23.

27.

Quiroga

M. A.

Santacreu

Montoro

Martínez-Molina

Shih

P. C.

(2011). Computerized assessment of attention for children from 7 to 11 years old: DiViSA-UAM and TACI-UAM. Clínica y Salud, 22(1), 3-20.

28.

Santacreu

Shih

P. C

Quiroga

M. A.

(2011). Divisa, trees simple visual discrimination test. Manual. Madrid, Spain: Tea Ediciones. Retrieved from http://web.teaediciones.com/DIVISA–;Trees-Simple-Visual-Discrimination-Test.aspx

29.

Santostefano

(1988). The cognitive control battery. Los Angeles, CA: Western Psychological Services.

30.

Santostefano

(1995). Integrative psychotherapy for children and adolescents with ADHD. Northvale, NJ: Jason Aronson. (Original work published 1985)

31.

Schmitt

(1996). Uses and abuses of coefficient alpha. Psychological assessment, 8, 350-353.

32.

Stins

J. F.

Polderman

J. C. (Tinca)

Boomsma

de Geus

E. J. C.

(2005). Response interference and working memory in 12-year-old children. Child Neuropsychology, 11, 191-201.

33.

Strauss

Sherman

E. M. S.

Spreen

(2006). A compendium of neuropsychological tests. Administration, norms and commentary (3rd ed.). New York, NY: Oxford University Press.

34.

Wetzel

Widmann

Schröger

(2009). The cognitive control of distraction by novelty in children aged 7-8 and adults. Psychophysiology, 46, 607-616.

35.

Wilding

Munir

Cornish

(2001). The nature of attentional differences between groups of children differentiated by teacher ratings of attention and hyperactivity. British Journal of Psychology, 92, 357-371.

36.

Wilding

Pankhania

Williams

(2007). Effects of speed and accuracy instructions on performance in a visual search task by children with good or poor attention. British Journal of Psychology, 98, 127-139.

37.

Willcutt

E. G.

(2012). The prevalence of DSM-IV attention-deficit/hyperactivity disorder: A meta-analytic review. Neurotherapeutics, 9, 490-499.

38.

Wolfe

J. M.

Horowitz

T. S.

(2004). What attributes guide the deployment of visual attention and how do they do it? Nature Review Neuroscience, 5, 1-7.

39.

Wolfe

J. M.

Kluender

K. R.

Levi

D. M.

Bartoshuk

L. M.

Herz

R. S.

Klatzky

R. L.

Merfeld

D. M.

(2012). Attention and scene perception. In Wolfe

J. M.

. (Eds.), Sensation and perception (3rd ed., pp. 187-218). Sunderland, MA: Sinauer Associates.