Neuropsychology 2014, Vol. 28, No. 6, 894 –904

© 2014 American Psychological Association 0894-4105/14/$12.00 http://dx.doi.org/10.1037/neu0000125

Interhemispheric Collaboration During Digit and Dot Number-Matching in Younger and Older Adults Urvi J. Patel

Brandon K. Barakat

Christopher Newport University

University of California–Los Angeles

Ruben Romero and Daniel Apodaca

Loyola Marymount University

Joseph B. Hellige

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

California State University–Fullerton

Barbara J. Cherry California State University–Fullerton Digit and dot number-matching stimuli were used to replicate findings reported for younger adults by Patel and Hellige (2007) and to explore whether performance would differ for younger versus older participants. Participants were to make numerical matches of digits only, dots only, and digits and dots mixed conditions to determine whether reaction time (RT), percentage error, and efficiency scores that combine latency and accuracy for match trials were better on within- versus across-hemisphere trials. Sixty-six younger and 42 older participants were screened with the Mini-Mental State Examination (MMSE) and the Geriatric Depression Scale. They performed the three experimental conditions and were assessed with Digit Span Forward and Backward subscales from the Wechsler Adult Intelligence Scale-III. Results for younger adults demonstrated a within-hemisphere advantage for the Digits and Mixed conditions and an across-hemisphere advantage for the Dots condition, consistent with previous literature. Older participants showed a stronger within-hemisphere advantage for the Digits condition compared with younger participants and no advantage for within- or across-hemisphere processing for the Mixed condition when RT was considered, but they performed similarly to younger adults when efficiency scores were used and showed a relative across-hemisphere advantage for the Dots condition. Although RT suggests age-related differences in how information is distributed across the hemispheres of the brain, more comprehensive efficiency scores indicate that younger and older adults appear to use similar strategies in the coordination of interhemispheric transfer of information. MMSE scores regardless of age were related to type of task but not to across- versus within-hemisphere performance. Keywords: interhemispheric communication, aging, numerical quantity

is initially projected to the right hemisphere (RH) whereas information presented to the right visual field (RVF) is projected to the left hemisphere (LH). To measure how information is shared across both cerebral hemispheres, participants are instructed to match items that are either projected to the same visual field (i.e., within a single hemisphere) or to separate visual fields (i.e., to different hemispheres). This is accomplished by presenting items to the LVF and RVF on a computer screen while participants focus on a center fixation cross (Hellige, 1993). For a correct decision to be made in the latter case, information must be transferred across the hemispheres. Transfer efficiency can be measured by comparing accuracy and response time on within- versus acrosshemisphere trials. Previous research has shown that in young adults, making a Physical Identity match with letters (e.g., A vs. A) produces a within-hemisphere advantage for reaction time (RT) and percentage accuracy (Banich, 1998; Banich & Belger, 1990). However, making a Name Identity match (e.g., A vs. a), considered more computationally complex, typically produces an acrosshemisphere advantage (Banich, 1998; Banich & Belger, 1990).

The study presented here examined how age differences in cognitive function might influence the efficiency with which information is distributed across the left and right cerebral hemispheres. Visual matching tasks can be used to evaluate hemispheric communication by requiring the individual to rapidly coordinate information transfer from one hemisphere to another. These tasks operate on the anatomical principle that information presented to the left visual field (LVF)

This article was published Online First August 18, 2014. Urvi J. Patel, Department of Psychology, Christopher Newport University; Brandon K. Barakat, Department of Psychology, University of California–Los Angeles; Ruben Romero and Daniel Apodaca, Department of Psychology, California State University–Fullerton; Joseph B. Hellige, Department of Psychology, Loyola Marymount University; Barbara J. Cherry, Department of Psychology and Gerontology Program, California State University–Fullerton. Correspondence concerning this article should be addressed to Urvi J. Patel, who is now at the Gerontology Program, California State UniversityFullerton, Ruby Gerontology Center– 8, 800 N State College Blvd, Fullerton, CA 92831. E-mail: [email protected] 894

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

DIGITS AND DOTS

Studies using two distinctly different formats in mixed conditions, with stimuli such as words and pictures (Koivisto & Revonsuo, 2003), emotional words and faces (Patel, Hellige, Kim, & George, 2007), and digits and dice-like dot patterns (Marks & Hellige, 2003; Patel & Hellige, 2007), generally report more efficient processing within a single hemisphere in young adults. For example, Patel and Hellige (2007) examined how the effect of dividing the processing load across both hemispheres might be influenced by mixing digit and dot formats (e.g., · vs. 1). Participants were instructed to match all digit stimuli, all dot stimuli, or digit and dot mixed stimuli on the basis of numerical quantity. The condition of primary interest was the mixed condition in that presentation of a dot pattern and a digit required that the numerical quantity be determined first to make a comparison; that is, the matching task could not be performed on the basis of physical identity. It was surprising that mixing these formats did not produce an across-hemisphere advantage, although a comparison on the basis of perceptual characteristics would not be sufficient to make a match. In fact, a within-hemisphere advantage was found when participants matched digits and dots on the basis of numerical quantity and small, medium, and large numeric magnitude categories. It may be that the processing of digits and dots involves different cortical activation routes that are sufficiently distinct to allow for parallel processing within the same hemisphere. Behavioral and neuroimaging studies suggest that digit and dot patterns may activate somewhat different cortical pathways although they represent the same numerical quantity (Dehaene, 1992, 1996; Dehaene & Cohen, 1995; Dehaene, DehaeneLambertz, & Cohen, 1998; Dehaene, Dupoux, & Mehler, 1990). Visual field studies report a RVF/LH advantage for processing number words and digits that share letter-like processing because they are both verbal representations of quantity, whereas dot clusters that represent numeric quantity spatially produce either no visual field difference or a LVF/RH advantage (e.g., Adamson & Hellige, 2006; Boles, 1986; Eviatar, 1999; Marks & Hellige, 2003). Functional imaging studies also support the idea that digits and dots may access different routes in that the verbally inclined LH may process digits using their letter-like properties and may involve bilateral activation of the parietal lobe (depending on the demands of the task; e.g., Chochon, Cohen, van de Moortele, & Dehaene, 1999; Eger, Sterzer, Russ, Giraud, & Kleinschmidt, 2003). However, dot patterns either in the form of a familiar dice configuration or random arrangement activate the posterior portion of the right intraparietal sulcus (IPS; e.g., Piazza, Mechelli, Butterworth, & Price, 2002; Sathian et al., 1999; Warrington, 1982; Warrington & James, 1967), which is also activated to a greater degree during a familiar dot pattern compared with digit addition (Venkatraman, Ansari, & Chee, 2005). Despite differences in the overall ability of each hemisphere to process specific numerical representations, research conducted with split-brain patients indicates that both hemispheres can process and compare numerical quantities regardless of format (Colvin, Funnell, & Gazzaniga, 2005). Thus, the literature suggests that the processing of digits versus dots may require nonhomologous regions of a given hemisphere. This may decrease costs for within-hemisphere processing and increase costs for across-hemisphere processing, at least in younger adults. To capitalize on this recent trend of results with different formats that represent the same abstract code but that may be acti-

895

vated by somewhat different cortical routes, the current study was designed to examine the reliability of findings for mixed stimulus formats with young adults and to extend the divided visual-field paradigm with a mixed condition to older adults. As such, one important purpose of the current study was to replicate the findings reported by Patel and Hellige (2007) with digit and dot stimuli in young adults. We predicted a relative within-hemisphere advantage for a simple digit identification condition, a relative acrosshemisphere advantage for a more computationally complex or difficult dot identification condition, and a relative withinhemisphere advantage for a task in which stimuli were mixed (both digits and dots). A second purpose of this study was to determine whether performance in older individuals would show the same patterns. A growing body of literature suggests that neural resources for cognitive processing are less specialized in older compared with younger adults (dedifferentiation) and structurally and functionally show less connectivity between neural regions (see Goh, 2011 for review). Studies have particularly focused on across-hemisphere processing in support of age-related dedifferentiation and loss of connectivity (e.g., Davis, Kragel, Madden, & Cabeza, 2012), but recent research shows that local or within neural regions in addition to more global or across neural regions experience age-related degradation as well (Gee et al., 2011; Geerligs, Renken, Saliasi, Maurits, & Lorist, 2014; Goh, 2011; Schulte, Muller-Oehring, Rohlfing, Pfefferbaum, & Sullivan, 2010). The paradigm used in the study presented here allows for a behavioral assessment of age differences in cognitive performance when the relative advantage for a computational complex task is within hemisphere for younger adults (Patel & Hellige, 2007). Therefore, older compared with younger adults may be more likely to distribute processing across hemisphere if local or withinhemisphere processing is degraded or compromised. Evidence of age-related performance differences in hemispheric communication is seen in behavioral (Cherry, Adamson, Duclos & Hellige, 2005; Cherry et al., 2010; Reuter-Lorenz & Mikels, 2005; Reuter-Lorenz, Stanczak, & Miller, 1999) and imaging studies (Cabeza, Anderson, Locantore, & McIntosh, 2002; Reuter-Lorenz et al., 2000) that use visual matching tasks. Younger and older adults behaviorally demonstrate a relative within-hemisphere processing advantage with simple identification tasks. However, with increased computational complexity, younger adults show an advantage for distributing processing across hemispheres whereas results for older adults have been mixed. Although an acrosshemisphere advantage is still seen in older adults, some studies show this to be stronger in older compared with younger adults (Reuter-Lorenz & Mikels, 2005; Reuter-Lorenz et al., 1999) whereas other studies show this advantage to be weaker (Cherry et al., 2005, 2010). Individual differences in older adults may partially explain these conflicting results. For example, Cherry et al. (2005, 2010) incorporated neuropsychological measures that measured executive function and processing speed and/or an efficiency index (overall RT/accuracy for their identification tasks) in addition to Physical and Name Identity tasks. Cherry et al. (2005) found that for older adults, those who performed better on a Digit Span Backward (DSB) task were those who also showed a reduced across-hemisphere advantage for the Name Identity task. Cherry et al. (2010) also found that a more efficient older adult group and those who performed better on an animal fluency task were those

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

896

PATEL ET AL.

who demonstrated a reduced across-hemisphere advantage for the same task. Bilateral compared with unilateral activation in older compared with younger adults has been found across several cognitive tasks and studies using functional magnetic resonance imaging (fMRI) (Cabeza et al., 2002; Reuter-Lorenz et al., 2000). Two hypotheses have been offered to explain the activation of both hemispheres reported in older adults. First, the shift may represent compensatory mechanisms that recruit homologous regions of the brain to offset decline that may result from age (Reuter-Lorenz et al., 1999). To illustrate, several studies have reported that older adults over-recruit prefrontal cortex regions that are contralateral to those activated in younger adults and that this recruitment may positively affect performance (Davis et al., 2012; Park & ReuterLorenz, 2009). However, note that Davis et al. (2012) report that the structural integrity of the corpus callosum may mediate this functional connectivity between homologous prefrontal regions to a larger degree in older adults than in young adults. The second hypothesis suggests that bilateral processing with age reflects a dedifferentiation of resources by which more regions are now made available for tasks that previously recruited specialized neural mechanisms (Cabeza, 2002; Colcombe, Kramer, Erickson, & Scalf, 2005). Although this could benefit performance in some instances, it could also reflect reduced interhemispheric inhibition via the corpus callosum and thus a failure to confine neural activity to a single hemisphere. From this perspective, additional frontal activation may be associated with detrimental rather than beneficial age-related changes (Buckner & Logan, 2002; Colcombe et al., 2005; Erickson et al., 2007; Kinsbourne, 1980; Logan, Sanders, Snyder, Morris, & Buckner, 2002). Note that the conclusions drawn by these hypotheses are not mutually exclusive and it may be the case that the structural and functional pattern of changes in older adults reflects a combination of disturbances in brain processes and emergence of compensatory mechanisms (Carp, Gmeindl, & Reuter-Lorenz, 2010). This shift in processing with age may be attributed in part to structural and functional age-related changes that result from intrahemispheric cortical and white matter degradation and a decrease in the overall integrity and size of the corpus callosum. Imaging and diffusion tensor imaging (DTI) studies have shown that, as an individual ages, the corpus callosum may become compromised (Abe et al., 2002; Janowsky, Kaye, & Carper, 1996; Sullivan, Pfefferbaum, & Rohling, 2010), and structural decline is associated with functional deficits that lead to poorer cognitive performance by older compared with younger participants (Davis et al., 2009; Sullivan, Rohlfing, & Pfefferbaum, 2010). With age-related declines in cortical areas and fiber tracts that connect within and across hemispheres, clarification of how collaboration occurs across and within the hemispheres may begin to explain how cognition is influenced by aging. For younger adults, Patel and Hellige (2007) claim that there is a relative withinhemisphere advantage for mixed formats because processing of the two formats uses different cortical routes on within-hemisphere trials, reducing competition for limited capacity resources. However, if dedifferentiation occurs within a hemisphere with age, then processing of the two formats will not be as segregated; therefore, a relative within-hemisphere advantage for the mixed condition may not be found. Specifically, cortical decline in older adults might suggest that the separation of processing regions within a

single hemisphere may no longer be available and that resources might have to be sought farther (i.e., across hemispheres) to make a match between the digits and dots. If this is the case, then it would be more likely that information is distributed across the corpus callosum. On the other hand, it may be that corpus callosum decline with age determines that processing will still need to occur within a single hemisphere. These issues are important in understanding how the normal older brain ages and for understanding compensatory and/or dedifferentiation mechanisms that may accompany these changes. Because relative across- versus withinhemisphere processing may also be influenced by individual differences over and beyond age, the study presented here also included the Mini-Mental State Examination (MMSE), the Geriatric Depression Scale (GDS), and Digit Span Forward (DSF) and DSB to explore other factors that might influence how information is distributed across the hemispheres.

Method Participants Participants (N ⫽ 108) consisted of younger adults and older adults. All were native English speakers and reported normal to corrected-to-normal vision in both eyes. Participants with any history of concussion or brain injury were excluded. Ninety-six participants were right-handed, three were left-handed, and six were ambidextrous on the basis of self-report. Older participants were 42 volunteers (20 males, 22 females) from a university chapter of the Osher Lifelong Learning Institute (OLLI), a chartered program that provides learning opportunities to adults aged 55 years and older. Older adults ranged in age from 62 to 88 years (M ⫽ 74, SD ⫽ 6.4). Table 1 provides additional information about the demographics of older adults and their performance on neuropsychological measures. Participants were not compensated for their participation. The younger adult group consisted of 66 (28 males, 38 females) university student volunteers. Student participants received extra credit toward various courses in psychology. Younger adult participants ranged in age from 18 to 23 years (M ⫽ 19.5, SD ⫽ 1.6). See Table 1 for additional details about the demographics of younger participants and their performance on neuropsychological measures. Table 1 Means, SD, and Ranges for Age, Education, MMSE, GDS, DSF Score, and DSB Score for Younger and Older Participants Younger (n ⫽ 66) 38 F; 28 M

a

Age (years) Education (years)a MMSEb GDSa DSF DSB

Older (n ⫽ 42) 22 F; 20 M

M (SD)

Range

M (SD)

Range

19.5 (1.6) 13.8 (1.6) 29.5 (0.9) 5.3 (3.3) 10.7 (2.2) 6.6 (2.01)

18–23 12–18 26–30 0–13 6–16 4–12

73.8 (6.4) 16.7 (2.4) 29.0 (1.2) 2.4 (2.2) 10.6 (2.3) 7.6 (2.5)

62–88 12–25 25–30 0–8 7–15 4–13

Note. F ⫽ female, M ⫽ male. a Younger and older significantly different, p ⬍ .001. older significantly different, p ⬍ .05.

b

Younger and

DIGITS AND DOTS

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Apparatus and Stimuli Stimuli for the digits and dots experiment were presented on two different computers: a MacIntosh IIci computer with an Apple Color RGB monitor (model #M1297) and a power Macintosh G3 with an Apple Multiple Scan720 (model #M4552). Monitor presentation screens were matched to approximate a 12-in. diagonal. Stimulus presentation was controlled and RT and error scores were collected using a MacProbe software package from Aristometrics, Inc. (Hunt, 1993). An adjustable chin rest was mounted in front of the computer monitor to ensure a viewing distance of 60 cm from the monitor screen. The height of the chin rest and chair were adjusted for each participant according to their size to allow for an optimal viewing angle and comfort. All stimuli were presented as black images on a white background. Stimuli consisted of two different types of number representation; digits from 1 to 6 or a pattern of one to six dots on a die figure (see Figure 1 for examples). Digit stimuli were presented in 24-point Elite font. In each trial, stimuli were presented in groups of three and were organized in the form of an upside-down triangle (see Figure 1). A central fixation cross was located at approximately 1° of the visual angle horizontally and vertically. Two stimuli were presented above the point of eye fixation, one in each visual field, and the third stimulus appeared below the point of eye fixation in either the LVF or the RVF. When projected on the

897

viewing screen, single digits subtended approximately 0.4° of the visual angle horizontally and 0.8° of the visual angle vertically. The edge of the upper digits closest to the center of a fixation cross was displaced approximately 1.2° of the visual angle above the fixation cross and 3.0° of the visual angle horizontally, one to the LVF and one to the RVF. The edge of the lower digit closest to the center of the fixation cross was displaced approximately 1.2° of the visual angle below the fixation cross and 1.5° of the visual angle horizontally, to either the LVF or RVF. Dot stimuli were generated using Canvas 3.5 graphics software and consisted of outline squares containing one to six dots. When projected on the viewing screen, each dot pattern subtended approximately 1.5° of the visual angle horizontally and 1.5° of the visual angle vertically. The corner of the upper dot patterns closest to the center of a fixation cross was displaced approximately 1.2° of the visual angle above the fixation cross and 2.7° of the visual angle horizontally, one to the LVF and one to the RVF. The corner of the lower dot pattern closest to the center of the fixation cross was displaced approximately 1.2° of the visual angle below and 1.5° of the visual angle horizontally, to either the LVF or RVF. Stimulus displays consisted of a digits-only condition (three digits arranged in the form of an upside triangle), a dots-only condition (three figures arranged in the form of an upside triangle with dice-like patterns), and a mixed condition that presented digit stimuli and dot stimuli. The mixed condition displays consisted of two different formats. Half of the participants received mixed format displays that comprised two digits above the fixation point and a dot pattern below the fixation point. The other half of participants received mixed format displays with two dot patterns above the fixation point and a digit below the fixation point.

Neuropsychological Measures

Figure 1. Examples of stimulus displays: Across-Hemisphere match trials (top row), Within-Hemisphere match trials (middle row), and nonmatch trials (bottom row). LVF and RVF indicate the visual field to which the bottom letter is presented. Mixed 1 refers to the Mixed format condition in which the upper items were digits and the lower item was a dot pattern and Mixed 2 refers to the Mixed format condition in which the format of the upper and lower items were reversed.

MMSE. The MMSE is a brief, 30-item questionnaire that provides an estimate of cognitive impairment by tapping into domains such as orientation, concentration, and memory (Folstein, Folstein, & McHugh, 1975). It can be administered within 5–10 min and requires individuals to respond to simple questions and problems. A correct answer to each question is awarded 1 point for a possible total of 30 points; a score ⱖ25 points indicates intact functioning whereas a score below this cutoff may indicate varying degrees of cognitive impairment with low scores suggesting the presence of dementia. GDS. The GDS is a 30-item self-report questionnaire designed to identify depressive symptoms in older adults (Yesavage et al., 1982). It is used to assess mood over the past week and requires individuals to respond to questions in a yes/no format. One point is awarded to each answer that is consistent with a scoring grid for a possible cumulative score of 30 points; a score ⱕ9 indicates normal mood whereas a score above this cutoff indicates varying degrees of depressive symptoms with high scores suggesting more severity. DSF and DSB. DSF and DSB subtests from the Weschler Adult Intelligence Scale (3rd edition) were used to assess working memory/executive function (Wechsler, 1997). DSF requires participants to repeat a string of digits in the same order they hear them. Participants are presented the same digit length for two trials and if their response is correct for at least one of these trials, then the digit length increases by one until

898

PATEL ET AL.

the participant gets two trials of the same digit length incorrect (up to nine digits). DSB requires participants to report the digit series in the reverse order that they hear them. As in DSF, the digit series increases in length until participants miss two spans of the same length.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Procedure Upon the participant’s arrival to the laboratory, researchers tested the participant’s vision via the Near Vision Test Card and recorded demographic information via a questionnaire prompt to obtain information about the participant’s gender, education, age, and handedness. After gathering demographic information, the researcher administered the MMSE to the participant. Participants were then administered the first of the three stimulus format conditions (digits only, dots only, and mixed format) of the computer experiment. The order of stimulus format conditions was counterbalanced. The computer experiment of the study presented here used an adapted version of the Patel and Hellige (2007) paradigm. Each stimulus format condition began with instructions and examples of the stimulus configurations followed by a block of 24 practice trials. During the practice trials, feedback generated through MacProbe was provided if an incorrect response was given (i.e., making a response on a nonmatch trial or failing to respond on a match trial). After each practice trial, the participants were allowed to ask the researcher any questions about the test before proceeding to the testing trials. Participants were instructed that no feedback would be given for the testing trials. With the exception of the use of three rather than four blocks to limit possible fatigue, the visual matching tasks were the same as those used by Patel and Hellige (2007). A preliminary analysis with the Patel and Hellige (2007) data comparing three versus four blocks across all conditions provided evidence that using a shortened version would still show the same pattern of effects (rs ranged from .97 to .99, n ⫽ 48, all ps ⬍ .001 for Stimulus Condition by Across/Within correlations). As such, each stimulus condition consisted of three testing blocks of 64 trials each and a short break after each block for a total of 192 testing trials per condition. Of the 64 trials in each block, there were 32 match trials (bottom stimulus matched one of the top two stimuli) and 32 nonmatch trials (bottom stimulus did not match either of the top two stimuli). For the match and nonmatch trials, the bottom stimulus was presented 16 times to the LVF/RH and 16 times to the RVF/LH. For each stimulus condition, participants were instructed to place the index fingers of both hands on the keyboard space bar in front of them and to press the space bar with both index fingers as quickly as possible if the numeric quantity represented by the bottom stimulus matched either of the numeric quantities represented by the top two stimuli. Participants were told to refrain from responding if there was not a match. Participants were also told to look directly at the center of the fixation cross when it appeared at the beginning of each trial and to maintain that eye position until after the stimuli had disappeared from the screen. Each trial began with a black fixation cross on a white screen that remained on the screen until the offset of the stimulus display. A beep sounded 1,050 ms after the presentation of the fixation cross, and after 1,050 ms a three-item stimulus display was presented for 195 ms. The next trial was immediately initiated by the participant’s re-

sponse via the computer keyboard or after 2,100 ms without a response. After the first stimulus condition, participants were given the GDS to complete. Before continuing to the second stimulus condition, participants were offered a short break after completing the GDS. Participants then continued with the second stimulus condition, which was followed by the Digit Span test. Participants were administered both trials of each item for the DSF and DSB subscales. For each subscale, if participants received a score of zero on both trials of any one item, the test was discontinued and the participant’s score up until that point was recorded. After completion of the Digit Span test, participants were administered the last stimulus condition and were subsequently debriefed after the completion of all conditions and testing materials.

Data Analysis The mean of median RT and percentage error (PE) were used as dependent variables in a series of mixed analyses of variance (ANOVAs) with Age, MMSE group (30; ⬍30), Stimulus Format (Digits only, Dots only, Digits and Dots mixed) and Across/Within Hemisphere as independent variables. Correlations were also conducted to determine whether stimulus format or relative across/within processing on each of the three computer tasks was associated with the MMSE or GDS. An ␣ level of .05 was used to determine significance for ANOVAs whereas an ␣ level of .01 was used for correlations because of the number of comparisons.

Results The analyses reported below include all 108 participants because initial analyses with right-handed participants only (n ⫽ 98) yielded the same pattern of results. (On the basis of self-report, 98 participants were right-handed, 4 were left-handed, and 6 were ambidextrous.) Recall that the Mixed condition consisted of two versions: one with digits on top and dots on the bottom and a second with dots on top and digits on the bottom. Preliminary analyses revealed no differences between these two conditions; therefore, they were collapsed into one “Mixed” condition. In addition, data were collapsed across the visual field for the bottom letter because initial results were not significantly different for visual field (although for RT, there was a trend for a LVF/RH advantage, consistent with findings reported by Banich [2003]; Eviatar, Hellige, & Zaidel [1997]; and Patel & Hellige [2007]). Because MMSE and GDS were significantly different for younger versus older adults, correlations were conducted between these measures and within-subject variables to determine whether these variables should be included in subsequent analyses. MMSE but not GDS was included as a grouping variable (score of 30 or ⬍30) in final analyses. (GDS was not significantly correlated with any outcome variables). Results for Digits, Dots, and Mixed conditions are reported first, with RT, percentage of errors, and an inverse efficiency score (to combine RT and PE in the same dependent variable: RT/% accuracy). For the sake of clarity, omnibus analyses are reported first, with the main effect of Age and the highest-order interactions described.

DIGITS AND DOTS

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Digits, Dots, and Mixed Tasks We first considered whether RT and PE yielded the same pattern of results by correlating all values from the 12 conditions of Stimulus Format (3), Across/Within Hemisphere (2), and WithinVisual Field (2). The slight positive correlation between RT and PE suggests that there was no differential speed–accuracy tradeoff across the experimental conditions, r ⫽ .09, n ⫽ 12, p ⫽ .78. RT. Using the mean of median RT as the dependent variable, an Age (young, old) by MMSE group (30; ⬍30) by Stimulus Format (digits, dots, mixed) by Across/Within Hemisphere mixed repeated measures ANOVA was conducted. We found a main effect of Age, with younger compared with older participants responding faster, F(1, 104) ⫽ 31.10, MSe ⫽ 1,192,748.56, p ⬍ .001, ␩p2 ⫽ .23 (Younger M ⫽ 678 ms; Older M ⫽ 769 ms). Main effects of Stimulus Format and Across/Within Hemisphere were significant. For Stimulus Format, RT was shortest for Digits (641 ms), longer for Dots (725 ms), and longest for the Mixed condition (804 ms), F(2, 104) ⫽ 351.63, MSe ⫽ 1,260,919.91, p ⬍ .001, ␩p2 ⫽ .77. Pairwise comparisons demonstrated that all conditions were significantly different from each other, ps ⬍ .001. For the Across/Within Hemisphere main effect, there was a relative within-hemisphere advantage, with overall shorter RT for withinhemisphere trials (720 ms) compared with across-hemisphere trials (726 ms), F(1, 104) ⫽ 4.52, MSe ⫽ 5,137.25, p ⫽ .036, ␩p2 ⫽ .042. The highest-order interaction was Age by Stimulus Format by Across/Within Hemisphere, F(2, 208) ⫽ 6.55, MSe ⫽ 6,258.08, p ⫽ .002, ␩p2 ⫽ .059, and the Stimulus Format by Across/Within Hemisphere interaction was also significant, F(2, 208) ⫽ 17.44, MSe ⫽ 16,449.42, p ⬍ .001, ␩p2 ⫽ .14 (see Figure 2 for details about mean differences). Consistent with Patel and Hellige (2007), for younger participants, pairwise comparisons indicated a relative within-hemisphere advantage for Digits, an across-hemisphere advantage for Dots, and a within-hemisphere advantage for the Mixed condition, ps ⬍ .05. However, for older participants, pairwise comparisons demonstrated that the across-hemisphere trial

Figure 2. Across minus within-hemisphere RT differences as a function of stimulus format (Digits, Dots, Mixed) for younger and older participants. Negative scores indicate a relative across-hemisphere advantage and positive scores reflect a relative within-hemisphere advantage.

899

Table 2 RT (in ms) and PE as a Function of Stimulus Format for Within- Versus Across-Hemisphere Conditions in Younger and Older Adults Younger (n ⫽ 66)

Digits RT Across Within Across-within Dots RT Across Within Across-within Mixed RT Across Within Across-within Digits PE Across Within Across-within Dots PE Across Within Across-within Mixed PE Across Within Across-within

Older (n ⫽ 42)

Mean

SE

Mean

SE

603.60 591.77 ⫹12

10.78 9.56 —

702.11 667.18 ⫹35

12.47 11.06 —

681.01 693.90 ⫺13

13.07 13.33 —

754.98 769.25 ⫺14

15.13 15.43 —

757.37 737.84 ⫹20

12.72 12.39 —

858.41 861.57 ⫺3

14.73 14.34 —

01.47 01.18 ⫹00.3

00.82 01.17 —

09.66 09.75 ⫺00.1

00.95 01.35 —

08.65 10.04 ⫺01.4

01.46 01.55 —

18.60 22.70 ⫺04.1

01.70 01.79 —

03.80 03.61 ⫹00.2

01.22 00.99 —

12.50 11.42 ⫹01.1

01.42 01.14 —

Note. Negative across/within hemisphere mean difference scores indicate a relative across-hemisphere advantage and positive scores reflect a relative within-hemisphere advantage.

mean was significantly different from the within-hemisphere trial mean only for the Digits condition, p ⬍ .001. Taken together, the Age by Stimulus Format by Across/Within Hemisphere interaction was driven by a stronger within-hemisphere advantage for older than younger adults on the Digits condition, and the lack of a within-hemisphere advantage for older compared with younger adults on the Mixed condition, ps ⬍ .05. Table 2 displays means (SE) and across/within hemisphere trial mean difference scores for younger and older adults with positive difference values demonstrating a within-hemisphere advantage and negative values demonstrating an across-hemisphere advantage. PE. Using mean PE as the dependent variable, an Age (young, old) by MMSE group (30; ⬍30) by Stimulus Format (digits, dots, mixed) by Across/Within Hemisphere mixed repeated measures ANOVA was conducted. We found a main effect of Age, with younger compared with older participants producing fewer errors, F(1, 104) ⫽ 41.55, MSe ⫽ 01.24, p ⬍ .001, ␩p2 ⫽ .29 (Younger M ⫽ 04.8%; Older M ⫽ 14.1%). A main effect of Stimulus Format was found such that PE was smallest for Digits (5.5%), largest for Dots (15.0%), and in between for the Mixed condition (7.8%), F(2, 208) ⫽ 48.82, MSe ⫽ 0.47, p ⬍ .001, ␩p2 ⫽ .41. Pairwise comparisons demonstrated that all conditions were significantly different from each other, ps ⬍ .001. Unlike the RT data in which the poorest performance was for the mixed condition, for PE, the poorest performance was for Dots. This result is consistent with findings reported by Patel and Hellige (2007). However, note that

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

900

PATEL ET AL.

younger participants made few errors in general, and error rates in the easiest conditions demonstrate a floor effect. The highest-order significant interaction was Age by MMSE group by Stimulus Format, F(2, 208) ⫽ 4.00, MSe ⫽ 0.034, p ⫽ .029, ␩p2 ⫽ .037 (Greenhouse-Geisser), such that MMSE scores did not affect accuracy performance for younger adults for any stimulus condition, but they did for older adults for the more difficult and/or complex stimuli Dots condition (30 score M ⫽ 16.7%; ⬍30 score M ⫽ 24.6%), p ⫽ .015, such that PE was higher for those older adults who scored more poorly on the dementia screening. Note that the Age by Stimulus Format by Across/Within Hemisphere interaction was not significant, F(2, 208) ⫽ 1.58, MSe ⫽ 0.004, p ⫽ .209, ␩p2 ⫽ .015, but the Stimulus Format by Across/ Within Hemisphere interaction was significant, F(2, 208) ⫽ 6.29, MSe ⫽ 0.016, p ⬍ .01, ␩p2 ⫽ .057 (see Figure 3 for details about mean differences). To test a priori hypotheses, pairwise comparisons were examined and demonstrated that there was no advantage for within- or across-hemisphere processing for stimulus format for younger participants and for older participants, and an acrosshemisphere advantage was found only for the Dots condition, p ⬍ .01. Table 2 reports means (SE) and across/within hemisphere mean difference scores for young versus older participants for all conditions. Inverse efficiency score. To obtain a single measure of processing efficiency, RT and PE were combined into a single index, the inverse efficiency score. Using the equation originally outlined by Townsend and Ashby (1983), inverse efficiency scores were calculated by dividing RT by accuracy (1 – error rate) for across/ within hemisphere trials across all stimulus formats for each participant. Efficiency scores are expressed in milliseconds and lower scores indicate more efficient performance. Using inverse efficiency scores as the dependent variable, an Age (young, old) by MMSE group by Stimulus Format (digits, dots, mixed) by Across/Within Hemisphere mixed ANOVA was

conducted. We detected a main effect of Age, with younger compared with older participants responding more efficiently, F(1, 104) ⫽ 38.56, MSe ⫽ 6,703,042.29, p ⬍ .001, ␩p2 ⫽ .27 (Younger M ⫽ 719 ms; Older M ⫽ 935 ms). The MMSE group main effect was also significant with an advantage for participants who obtained 30 on the MMSE over participants who scored ⬍30, F(1, 104) ⫽ 4.803, MSe ⫽ 834,942.95, p ⫽ .031, ␩p2 ⫽ .044 (30 score M ⫽ 789 ms; ⬍30 score M ⫽ 865 ms). For Stimulus Format, performance was best for Digits (695 ms), followed by the Mixed (891 ms) and Dots condition (895 ms), F(2, 208) ⫽ 108.75, MSe ⫽ 2,482,587.73, p ⬍ .001, ␩p2 ⫽ .51. Pairwise comparisons demonstrated that Digits was significantly different from Dots and Mixed, ps ⬍ .001. The highest-order interaction was Age by MMSE group by Stimulus Format, F(2, 208) ⫽ 4.47, MSe ⫽ 102,075.88, p ⫽ .013, ␩p2 ⫽ .041. MMSE scores did not affect performance for younger adults for any stimulus condition, but they did for older adults for the more difficult and/or complex Dots condition (30 score M ⫽ 910 ms; ⬍30 score M ⫽ 1,115 ms) and Mixed condition (30 score M ⫽ 935 ms; ⬍30 score M ⫽ 1,084 ms), ps ⬍ .05, with older adults who scored more poorly on the dementia screening demonstrating less efficient performance. Note that the Age by Stimulus Format by Across/Within Hemisphere interaction was not significant, F(2, 208) ⫽ 1.82, MSe ⫽ 14,243.74, p ⫽ .166, ␩p2 ⫽ .017, but consistent with our RT findings and those of Patel and Hellige (2007), the Stimulus Format by Across/Within Hemisphere interaction was significant, F(2, 208) ⫽ 10.38, MSe ⫽ 81,520.032, p ⬍ .001, ␩p2 ⫽ .091 (see Figure 4 for details about mean differences). To test a priori hypotheses, pairwise comparisons for younger adults indicated no significant difference between conditions, all ps ⬎ .05, but the pattern of means was in the expected direction, with a relative within-hemisphere advantage for Digits, an acrosshemisphere advantage for Dots, and a within-hemisphere advantage for the Mixed condition. For older participants, pairwise comparisons demonstrated that the across-hemisphere trial mean was significantly different from the within-hemisphere trial mean only for the Dots condition, p ⬍ .05. Note that the direction of results for older adults is consistent with the pattern found for younger adults. Table 3 reports means (SE) and across/within hemisphere mean difference scores for young versus older participants for all conditions.

Correlations Separate correlations (all participants, younger adults, older adults) were also conducted to determine whether stimulus format or relative across/within processing on each of the three computer tasks was associated with DSF and DSB, GDS, and MMSE scores. No significant correlations were found.

Discussion

Figure 3. Across minus within-hemisphere PE differences as a function of stimulus format (Digits, Dots, Mixed) for younger and older participants. Negative scores indicate a relative across-hemisphere advantage and positive scores reflect a relative within-hemisphere advantage. Some error bars do not show because they are smaller than the circle symbols.

The first objective of the study presented here was to confirm the pattern of results found in a prior study in younger adults (Patel & Hellige, 2007); that is, evidence of a relative within-hemisphere advantage for a simple Digits identification condition, a relative across-hemisphere advantage for a more computationally complex or more difficult Dots identification condition, and a relative within-hemisphere advantage for a Mixed condition that required

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

DIGITS AND DOTS

Figure 4. Across- minus within-hemisphere inverse efficiency differences as a function of stimulus format (Digits, Dots, Mixed) for younger and older participants. Negative scores indicate a relative acrosshemisphere advantage and positive scores reflect a relative withinhemisphere advantage. Some error bars do not show because they are smaller than the circle symbols.

digit and dot identification. Results for younger adults in the current study clearly replicated the findings reported by Patel and Hellige (2007). Younger adults demonstrated a within-hemisphere advantage for the simple Digits condition, which is consistent with prior literature (Banich & Belger, 1990). Younger participants demonstrated an across-hemisphere advantage for the more complicated Dots condition. Worth noting is that several studies suggest that although the Dots condition can be accomplished based on the physical characteristics of stimuli, the task may be more computationally complex than a simple perceptual match because participants may convert dot stimuli to their corresponding magnitude. To illustrate, in Patel and Hellige (2007) and our study, the Dots only condition yielded a greater percentage of errors than the other two stimulus conditions and by approximately the same amount. The results for the digits and dots Mixed condition are consistent with performance of the same condition reported by Patel and Hellige (2007) in that despite the increase in computational complexity, participants demonstrated a within-hemisphere advantage. Recall that Patel and Hellige (2007) suggest that a withinhemisphere advantage for this last condition may be due to different cortical activation routes within the same hemisphere, where processing modules/regions are far enough apart to limit interference. The second objective of this study was to observe whether this same pattern of results would be found with older adults. Similar to younger adults, older adults demonstrated a within-hemisphere advantage for the simple Digits condition, consistent with prior literature (Cherry et al., 2010; Reuter-Lorenz et al., 1999). Also consistent was a difference between younger and older adults such that older adults showed a significantly stronger withinhemisphere advantage than younger participants for the Digits condition. This effect has now been reported across several studies using either number or letter stimuli (Cherry et al., 2005, 2010;

901

Reuter-Lorenz & Mikels, 2005; Reuter-Lorenz et al., 1999; the study presented here). This suggests that for younger and older adults, there are enough processing resources within a single hemisphere to effectively perform a task that requires simple digit (or letter) identification. The stronger within-hemisphere advantage for older adults also suggests that the costs of distributing the processing load of such a task across hemispheres may be higher for older than younger adults, consistent with studies that show that across-hemisphere processing in older but not younger adults is moderated by the integrity of white matter in the corpus callosum (Davis et al., 2012). The Dots condition demonstrated an across-hemisphere advantage in older adults that was similar to that of younger participants, although RT was significantly longer and PE was greater for older than younger adults (see Tables 2 and 3). Performance of younger and older adults in the Dots condition suggests that this task may involve more processing than a task that requires a simple perceptual match. Although the Dots condition produced a greater percentage of errors than the Digits and Mixed condition for both groups, the percentage of errors was very low for the younger compared with older adults, which suggests that the condition may be more difficult but not necessarily more computationally complex. It is interesting to note that even without magnitude conversion, there is evidence that processing dots compared with digits lengthens RT (see Nuerk et al., 2005; Vorberg & Blankenberger, 1993). In the study presented here, regardless of whether this was due to task difficulty and/or computationally complexity, the Dots condition demonstrated a relative across-hemisphere advantage regardless of age. Although younger adults showed a relative within-hemisphere advantage for the Mixed condition, this effect was attenuated in older adults when considering RT (see Table 2 and Figure 2); that is, older adults showed neither a within- nor across-hemisphere advantage. However, with RT and PE combined into efficiency scores, this difference disappeared such that no difference was found between age groups (see Table 3 and Figure 3). Although

Table 3 Inverse Efficiency Score (in ms) as a Function of Stimulus Format for Within- Versus Across-Hemisphere Conditions in Younger and Older Adults Younger (n ⫽ 66)

Digits Across Within Across-within Dots Across Within Across-within Mixed Across Within Across-within

Older (n ⫽ 42)

Mean

SE

Mean

SE

613.51 599.88 ⫹14

16.96 24.84 —

791.75 775.52 ⫹16

19.63 28.75 —

757.20 780.27 ⫺23

32.70 33.49 —

970.86 1053.83 ⫺83

37.85 38.76 —

792.17 768.87 ⫹23

28.12 24.33 —

1019.66 999.38 ⫹21

32.55 28.16 —

Note. Negative across/within hemisphere mean difference scores indicate a relative across-hemisphere advantage and positive scores reflect a relative within-hemisphere advantage.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

902

PATEL ET AL.

combining RT and accuracy into one metric is conceptually appealing, one limitation is its possible effect of increasing variability (see Bruyer & Brysbaert, 2011). However, note that for younger adults (Tables 2 and 3), standard errors are very similar regardless of whether RT, PE, or efficiency scores are used. Furthermore, for older adults, consistent with the trend found for RT and PE, the efficiency score yielded larger standard errors compared with errors displayed by younger individuals. Although the cortical route hypothesis is consistent with the idea that local or within-hemisphere processing regions may decline with age, evidence for this is indirect at best. Although studies on neural processing demonstrate local age-related dedifferentiation, decreased structural connectivity, reduced inhibition in regional networks in older adults, and less ability to disengage the default network in older compared with younger adults (Gee et al., 2011; Geerligs et al., 2014; Goh, 2011; Reuter-Lorenz & Cappell, 2008; Schulte et al., 2010), these have not been linked to specific behavioral tasks such as the one explored in the study presented here. Therefore, alternative explanations for within-hemisphere processing on a computationally complex task should be considered. Braun and colleagues (2011) explored methodological manipulations (practice, shape of stimuli, orientation of stimuli) that could possibly influence within- versus across-hemisphere advantages in younger adults, suggesting that not only the corpus callosum but (also) limitations and integrity of the commissures may influence within- versus across-hemisphere processing and may need to be considered in future studies. The pattern of within- versus across-hemisphere processing in older adults may be the result of a combination of gray and white matter changes with age, specific task demands, and individual differences. Use of the digit and dot stimuli in the study presented here suggests that specific task demands should consider task difficulty and computational complexity, especially when older adults are included as participants. The study presented here demonstrates a promising paradigm to tease apart some of these aging issues within hemisphere if used in conjunction with functional and structural assessments of gray and white matter integrity. For example, Davis et al. (2012) used a word-matching task and measured fMRI (blood oxygen level-dependent [BOLD] activity) and DTI at the same time. They found that better performance on the matching task was related to more bilateral involvement in older adults, but that this relationship was influenced by the integrity of the corpus callosum. That is, older participants with greater white matter connectivity relied more on interhemispheric pathways to coordinate activity between hemispheres. Individual differences measured in the study presented here included DSF and DSB and scores on the MMSE. No associations were found between span measures and across- versus withinhemisphere processing (unlike Cherry et al., 2005 but consistent with Cherry et al., 2010). MMSE scores were sensitive to the Dot and Mixed conditions for older adults, with better performance on both tasks for those with higher MMSE scores, but MMSE scores did not fall within the potential dementia range and were not associated with across- versus within-hemisphere processing. Limitations of the study presented here included the use of two different computers and monitors; however, the size of display and rate of presentation were matched and a comparison of performance on both computers with younger adults revealed no significant differences on the basis of the computer. Younger and older

adults were tested in separate buildings and rooms, but in locations they were each familiar with and conditions within each room were the same. It should be noted that our older participants were high-functioning, independent-living individuals, which limits our ability to generalize to all older adults. Moreover, younger participants were selected from a student sample; therefore, inferences are limited to the college population. Finally, a future study that uses a more complex matching task (e.g., five-item display) to prevent a ceiling effect in error performance in younger participants might allow for a more meaningful way of parsing any age-related performance differences in hemispheric communication. In conclusion, the study presented here replicated prior findings (Patel & Hellige, 2007) such that young adults produced a relative within-hemisphere advantage for Digits and Mixed stimuli and an across-hemisphere advantage for Dots. One hypothesis is that even when stimuli are computationally complex, if there are different cortical activation routes far enough apart within a single hemisphere, then a within-hemisphere advantage can be found. Results for older adults were similar in the study presented here, but whether or not older adults are more likely to distribute processing for stimuli considered computationally complex across hemispheres appears to be related to multiple factors. For example, integrity of the corpus callosum and individual differences appear to modulate hemispheric communication. (Although measures for individual differences in the study presented here were not associated with interhemispheric processing, past studies using similar measures were [Cherry et al., 2005; Cherry et al., 2010]). The format of stimuli to be matched has recently also shown to differentially affect performance during such tasks, in which various stimulus formats may use different and parallel processing pathways. Future research could expand on issues related to age-related within-hemisphere processing to clarify these findings by using other types of mixed stimuli (e.g., words and pictures; emotional words and faces) and/or by increasing the number of items to match. Moreover, studies such as the one by Davis et al. (2012) could also investigate age-related neural and behavioral withinhemisphere processing within the same experiment.

References Abe, O., Aoki, S., Hayashi, N., Yamada, H., Kunimatsu, A., Mori, H., . . . Ohtomo, K. (2002). Normal aging in the central nervous system: Quantitative MR diffusion-tensor analysis. Neurobiology of Aging, 23, 433– 441. doi:10.1016/S0197-4580(01)00318-9 Adamson, M. M., & Hellige, J. B. (2006). Hemispheric differences for identification of words and nonwords in Urdu-English bilinguals. Neuropsychology, 20, 232–248. doi:10.1037/0894-4105.20.2.232 Banich, M. T. (1998). The missing link: The role of interhemispheric interaction in attentional processing. Brain and Cognition, 36, 128 –157. doi:10.1006/brcg.1997.0950 Banich, M. T. (2003). Interacting hemispheres: A means of modulating attention. In E. Zaidel and M. Iacoboni (Eds.), The parallel brain: The cognitive neuroscience of the corpus callosum (pp. 267–270). Cambridge, MA: MIT Press. Banich, M. T., & Belger, A. (1990). Interhemispheric interaction: How do the hemispheres divide and conquer a task? Cortex, 26, 77–94. doi:10.1016/ S0010-9452(13)80076-7 Boles, D. B. (1986). Hemispheric differences in the judgment of number. Neuropsychologia, 24, 511–519. doi:10.1016/0028-3932(86)90095-3

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

DIGITS AND DOTS Braun, C. M. J., Achim, A., Gauvin, G., Desjardins, S., Roberge, C., & Suffren, S. (2011). New variations of intrahemispheric processing indexed by the Dimond paradigm. The American Journal of Psychology, 124, 163–176. doi:10.5406/amerjpsyc.124.2.0163 Bruyer, R., & Brysbaert, M. (2011). Combining speed and accuracy in cognitive psychology: Is the inverse efficiency scores (IES) a better dependent variable than the mean reaction time (RT) and the percentage of errors (PE)? Psychologica Belgica, 51, 5–13. doi:10.5334/pb-51-1-5 Buckner, R. L., & Logan, J. (2002). Frontal contributions to episodic memory encoding in the young and elderly. In E. A. Parker, E. L. Wilding, & T. Bussey (Eds.), The cognitive neuroscience of memory encoding and retrieval. Philadelphia, PA: Psychology Press. Cabeza, R. (2002). Hemispheric asymmetry reduction in older adults: The HAROLD model. Psychology and Aging, 17, 85–100. doi:10.1037/08827974.17.1.85 Cabeza, R., Anderson, M. D., Locantore, J. K., & McIntosh, A. R. (2002). Aging gracefully: Compensatory brain activity in high-performing older adults. NeuroImage, 17, 1394 –1402. doi:10.1006/nimg.2002.1280 Carp, J., Gmeindl, L., & Reuter-Lorenz, P. A. (2010). Age differences in the neural representation of working memory revealed by multi-voxel pattern analysis. Frontiers in Human Neuroscience, 4. doi:10.3389/ fnhum.2010.00217 Cherry, B. J., Adamson, M., Duclos, A., & Hellige, J. B. (2005). Aging and individual variation in interhemispheric collaboration and hemispheric asymmetry. Aging, Neuropsychology, and Cognition, 12, 316 –339. Cherry, B. J., Yamashiro, M., Anderson, E., Barrett, C., Adamson, M. M., & Hellige, J. B. (2010). Exploring interhemispheric collaboration in older compared to younger adults. Brain and Cognition, 72, 218 –227. doi:10.1016/j.bandc.2009.09.003 Chochon, F., Cohen, L., Van de Moortele, P. F., & Dehaene, S. (1999). Differential contributions of the left and right inferior parietal lobules to number processing. Journal of Cognitive Neuroscience, 11, 617– 630. doi:10.1162/089892999563689 Colcombe, S. J., Kramer, A. F., Erickson, K. I., & Scalf, P. (2005). The implications of cortical recruitment and brain morphology for individual differences in inhibitory function in aging humans. Psychology and Aging, 20, 363–375. doi:10.1037/0882-7974.20.3.363 Colvin, M. K., Funnell, M. G., & Gazzaniga, M. S. (2005). Numerical processing in the two hemispheres: Studies of a split-brain patient. Brain and Cognition, 57, 43–52. doi:10.1016/j.bandc.2004.08.019 Davis, S. W., Dennis, N. A., Buchler, N. G., White, L. E., Madden, D. J., & Cabeza, R. (2009). Assessing the effects of age on long white matter tracts using diffusion tensor tractography. NeuroImage, 46, 530 –541. doi:10.1016/j.neuroimage.2009.01.068 Davis, S. W., Kragel, J. E., Madden, D. J., & Cabeza, R. (2012). The architecture of cross-hemispheric communication in the aging brain: Linking behavior to functional and structural connectivity. Cerebral Cortex, 22, 232–242. doi:10.1093/cercor/bhr123 Dehaene, S. (1992). Varieties of numerical abilities. Cognition, 44, 1– 42. doi:10.1016/0010-0277(92)90049-N Dehaene, S. (1996). The organization of brain activations in number comparison: Event-related potentials and the additive-factors method. Journal of Cognitive Neuroscience, 8, 47– 68. doi:10.1162/jocn.1996.8 .1.47 Dehaene, S., & Cohen, L. (1995). Towards an anatomical and functional model of number processing. Mathematical Cognition, 1, 83–120. Dehaene, S., Dehaene-Lambertz, G., & Cohen, L. (1998). Abstract representations of numbers in the animal and human brain. Trends in Neurosciences, 21, 355–361. doi:10.1016/S0166-2236(98)01263-6 Dehaene, S., Dupoux, E., & Mehler, J. (1990). Is numerical comparison digital? Analogical and symbolic effects in two-digit number comparison. Journal of Experimental Psychology: Human Perception and Performance, 16, 626 – 641. doi:10.1037/0096-1523.16.3.626

903

Eger, E., Sterzer, P., Russ, M. O., Giraud, A., & Kleinschmidt, A. (2003). A supramodal number representation in human intraparietal cortex. Neuron, 37, 719 –726. doi:10.1016/S0896-6273(03)00036-9 Erickson, K. I., Colcombe, S. J., Wadhwa, R., Bherer, L., Peterson, M. S., Scalf, P. E., . . . Kramer, A. F. (2007). Training-induced plasticity in older adults: Effects of training on hemispheric asymmetry. Neurobiology of Aging, 28, 272–283. doi:10.1016/j.neurobiolaging.2005.12.012 Eviatar, Z. (1999). Cross-language tests of hemispheric strategies in reading nonwords. Neuropsychology, 13, 498 –515. doi:10.1037/0894-4105 .13.4.498 Eviatar, Z., Hellige, J. B., & Zaidel, E. (1997). Individual differences in lateralization: Effects of gender and handedness. Neuropsychology, 11, 562–576. doi:10.1037/0894-4105.11.4.562 Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189 –198. doi:10.1016/00223956(75)90026-6 Gee, D. G., Biswat, B. B., Kelly, C., Stark, D. E., Margulies, D. S., Shehzed, Z., . . . Milham, M. P. (2011). Low frequency fluctuations reveal integrated and segregated processing among the cerebral hemispheres. NeuroImage, 54, 517–527. doi:10.1016/j.neuroimage.2010.05 .073 Geerligs, L., Renken, R. J., Saliasi, E., Maurits, N. M., & Lorist, M. M. (2014). A brain-wide study of age-related changes in functional connectivity. Cerebral Cortex. doi:10.1093/cercor/bhu012 Goh, J. O. S. (2011). Functional differentiation and altered connectivity in older adults: Neural accounts of cognitive aging. Aging and Disease, 2, 30 – 48. Hellige, J. B. (1993). Hemispheric asymmetry: What’s right and what’s left. Cambridge: Harvard University Press. Hunt, S. (1993). Macprobe software manual. Castro Valley, CA: Aristometrics. Janowsky, J. S., Kaye, J. A., & Carper, R. A. (1996). Atrophy of the corpus callosum in Alzheimer’s disease versus healthy aging. Journal of the American Geriatrics Society, 44, 798 – 803. Kinsbourne, M. (1980). Attentional dysfunction and the elderly: Theoretical models and research perspectives. In L. W. Poon, J. L. Fozard, L. S. Cermak, D. Arenberg, & L. W. Thompson (Eds.), New directions in memory and aging: Proceedings of the George A Talland Memorial Conference (pp. 113–129). Hillsdale, NJ: Erlbaum. Koivisto, M., & Revonsuo, A. (2003). Interhemispheric categorization of pictures and words. Brain and Cognition, 52, 181–191. doi:10.1016/ S0278-2626(03)00054-X Logan, J. M., Sanders, A. L., Snyder, A. Z., Morris, J. C., & Buckner, R. L. (2002). Under-recruitment and nonselective recruitment: Dissociable neural mechanisms associated with aging. Neuron, 33, 827– 840. doi: 10.1016/S0896-6273(02)00612-8 Marks, N. L., & Hellige, J. B. (2003). Interhemispheric interaction in bilateral redundancy gain: Effects of stimulus format. Neuropsychology, 17, 578 –593. doi:10.1037/0894-4105.17.4.578 Nuerk, H.-C., Wood, G., & Willmes, K. (2005). The universal SNARC effect: The association between number magnitude and space is amodal. Experimental Psychology, 52, 187–194. doi:10.1027/1618-3169.52.3.187 Park, D. C., & Reuter-Lorenz, P. (2009). The adaptive brain: Aging and neurocognitive scaffolding. Annual Review of Psychology, 60, 173–196. doi:10.1146/annurev.psych.59.103006.093656 Patel, U. J., & Hellige, J. B. (2007). Benefits of interhemispheric collaboration can be eliminated by mixing stimulus formats that involve different cortical access routes. Brain and Cognition, 63, 145–158. doi:10.1016/j.bandc.2006.10.007 Patel, U. J., Hellige, J. B., Kim, J. G., & George, P. (2007). Interhemispheric collaboration for matching emotions signified by words and faces. Presented at the Psychonomic Society, Long Beach, CA.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

904

PATEL ET AL.

Piazza, M., Mechelli, A., Butterworth, B., & Price, C. J. (2002). Are subitizing and counting implemented as separate or functionally overlapping processes? NeuroImage, 15, 435– 446. doi:10.1006/nimg.2001 .0980 Reuter-Lorenz, P. A., & Cappell, K. (2008). Neurocognitive aging and the compensation hypothesis. Current Directions in Psychological Science, 18, 177–182. Reuter-Lorenz, P. A., Jonides, J., Smith, E. S., Hartley, A., Miller, A., Marshuetz, C., & Koeppe, R. A. (2000). Age differences in the frontal lateralization of verbal and spatial working memory revealed by PET. Journal of Cognitive Neuroscience, 12, 174–187. doi:10.1162/089892900561814 Reuter-Lorenz, P. A., & Mikels, J. A. (2005). A split-brain model of Alzheimer’s disease? Behavioral evidence for comparable intra and interhemispheric decline. Neuropsychologia, 43, 1307–1317. doi:10.1016/j.neuropsychologia.2004.12 .007 Reuter-Lorenz, P. A., Stanczak, L., & Miller, A. C. (1999). Neural recruitment and cognitive aging: Two hemispheres are better than one, especially as you age. Psychological Science, 10, 494 –500. doi:10.1111/1467-9280.00195 Sathian, K., Simon, T. J., Peterson, S., Patel, G., Hoffman, J. M., & Grafton, S. T. (1999). Neural evidence linking visual object enumeration and attention. Journal of Cognitive Neuroscience, 11, 36–51. doi:10.1162/089892999563238 Schulte, T., Muller-Oehring, E. M., Rohlfing, T., Pfefferbaum, A., & Sullivan, E. V. (2010). White matter fiber degradation attenuates hemispheric asymmetry when integrating visuomotor information. The Journal of Neuroscience, 30, 12168 –12178. doi:10.1523/JNEUROSCI.2160-10.2010 Sullivan, E. V., Pfefferbaum, A., & Rohling, T. (2010). Longitudinal study of callosal microstructure in the normal adult aging brain using quantitative DTI fiber tracking. Developmental Neuropsychology, 35, 233– 256. doi:10.1080/87565641003689556

Sullivan, E. V., Rohlfing, T., & Pfefferbaum, A. (2010). Quantitative fiber tracking of lateral and interhemispheric white matter systems in normal aging: Relations to timed performance. Neurobiology of Aging, 31, 464 – 481. doi:10.1016/j.neurobiolaging.2008.04.007 Townsend, J. T., & Ashby, F. G. (1983). The stochastic modeling of elementary psychological processes. New York, NY: Cambridge University Press. Venkatraman, V., Ansari, D., & Chee, M. W. L. (2005). Neural correlates of symbolic and non-symbolic arithmetic. Neuropsychologia, 43, 744 – 753. doi:10.1016/j.neuropsychologia.2004.08.005 Vorberg, D., & Blankenberger, S. (1993). Mental representation of numbers. Sprache & Kognition, 12, 98–114. Warrington, E. K. (1982). The fractionation of arithmetical skills: A single case study. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 34, 31–51. doi:10.1080/14640748208400856 Warrington, E. K., & James, M. (1967). Tachistoscopic number estimation in patients with unilateral cerebral lesions. Journal of Neurology, Neurosurgery & Psychiatry, 30, 468 – 474. doi:10.1136/jnnp.30.5.468 Wechsler, D. (1997). Wechsler Adult Intelligence Scale-Revised. New York, NY: Psychological Corporation. Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M., & Leirer, V. O. (1982). Development and validation of a geriatric depression screening scale. Journal of Psychiatric Research, 17, 37– 49. doi:10.1016/0022-3956(82)90033-4

Received August 21, 2013 Revision received May 16, 2014 Accepted June 12, 2014 䡲