Lateralized interference in finger tapping: Comparisons of rate and variability measures under speed and consistency tapping instructions

BRAIN

AND

COGNITION

5,

268-279 (1986)

Lateralized Interference in Finger Tapping: Comparisons of Rate and Variability Measures under Speed and Consistency Tapping Instructions DANIEL W. KEE AND KAREN MORRIS California

State University,

Fullerton

KAY BATHURST University

of California,

Los Angeles

AND JOSEPH B. HELLIGE University

of Southern California,

Los Angeles

Forty right-handed college subjects tapped with and without a verbal task under two instructional conditions (tap as quickly as possible vs. tap as consistently as possible) and two levels of verbal production (silent vs. aloud). The tapping task consisted of the alternate tapping of two keys with the index finger of the left vs. right hands, while the verbal task was anagram solution. Three rate and four variability measures of tapping performance were evaluated in the identification of laterahzed interference. The results indicate that reliable lateralized interference, more right-hand than left-hand tapping disruption, was observed only for variability measures under instructions to tap as consistently as possible. Furthermore, only one of these variability measures was sensitive to an increase in lateralized interference produced by verbal production. Because of the limited demonstration of verbal laterality effects with the two-key tapping procedure in this study, This research was funded in part by a small-college Faculty Opportunity Award to Daniel W. Kee from the National Science Foundation and in part by a National Science Foundation Grant (BNS-8217379) to Joseph B. Hellige. Appreciation is expressed to Professor Edward Stearns for his helpful comments concerning aspects of the statistical analyses. The preparation of this manuscript was facilitated by a Faculty Research Grant and a President’s Grant from California State University, Fullerton. Requests for reprints should be addressed to D. W. Kee, Department of Psychology, California State University, Fullerton, CA 92634.

0278-2626186$3.00 Copyright All rights

0 1986 by Academic Press, Inc. of reproduction in any form reserved.

268

LATERALIZED

INTERFERENCE

269

conclusions suggest that the simpler manual task of repetitive tapping of one key should be viewed as the method of choice in future dual-task studies. D 1986 Academic F’ress. Inc.

Finger tapping of the left versus right hand is a manual task condition used frequently in dual-task studies of cerebral asymmetry (see Kinsboume & Hiscock, 1983). In general, right-hand tapping is disrupted to a greater extent than left-hand tapping when a concurrent verbal task is performed. This pattern of lateralized interference in finger tapping is presumed to reflect left-hemisphere specialization for verbal activity because (1) the manual activity of each hand is programmed primarily by the contralateral cerebral hemisphere and (2) two independent tasks should interfere with each other more if they involve the same cerebral hemiphere rather than different cerebral hemispheres (Kinsbourne & Hiscock, 1983). The repetitive tapping of a single key with the index finger is a frequently used finger tapping procedure in dual task studies (see Kinsboume & Hiscock, 1983). With this procedure subjects are instructed to tap as quickly as possible and performance is measured by a rate index. Evidence indicates that this manual task procedure reliably detects both verbal and visuospatial laterality effects. For example, in our own dual-task studies using this single-key procedure, reliable left-hemisphere involvement has been observed for such verbal tasks as paragraph reading (Hellige & Longstreth, 1981; Kee, Hellige, Bathurst, 1982); word memory; rhyme recitation; and anagram solution (Kee, Bathurst, & Hellige, 1983). Righthemisphere involvement has also been observed for subjects’ manual solution of WISC block designs (Hellige & Longstreth, 1981; Kee et al., 1982; Kee, Bathurst, & Hellige, 1984). Furthermore, this single-key procedure is sensitive to variation in the degree of left-hemisphere involvement for verbal tasks due to familial sinistrality (Kee et al., 1983), verbal production (Hellige & Longstreth, 1981; Kee et al., 1982), and gender/handedness differences in pre-school-age children (Kee, Bathurst, & Gottfried, 1985). In contrast to the reliable laterality effects observed with the singlekey procedure, some preliminary evidence reported by Kee et al. (1982) suggests that multikey finger tapping procedures may be less sensitive. For example, Kee et al. (1982) observed reliable left-and right-hemisphere laterality effects with the single-key method for paragraph reading and WISC block design solution, respectively, while such laterality effects were not detected when the manual task required sequential tapping of two or three keys. A review of the extant literature indicates that of the multikey procedures, the alternate tapping of two keys with the index finger is one of the most frequently used (see Kinsbourne & Hiscock, 1983). With this two-key procedure, however, different instructional sets

KEE ET AL.

have been used and various measures of manual performance reported (in contrast to the single-key procedure in which subjects are usually told to tap as quickly as possible and performance is consistently indexed by a rate measure). For example, in some studies subjects are told to tap as quickly as possible (McFarland & Ashton, 1978a),tap as consistently as possible (McFarland & Ashton, 1978b), or tap as fast and consistently as possible (McFarland & Ashton, 1978~).Regarding the different measures of tapping performance, an early study by McFarland and Ashton (1978a) suggestedthat rate and variability measuresof tapping performance provide comparable estimates of lateralized interference under concurrent verbal activity conditions. Recent evidence, however, suggests that the two kinds of tapping performance measures may not always provide equivalent estimates of lateralized interference (see Kinsbourne & Hiscock, 1983). For example, McFarland and Ashton (1978~) suggest that variability measures may be more sensitive than rate measures to the kinds of intermittent bursts of interference which can be produced by concurrent verbal activity. Also, Kinsbourne and Hiscock (1983) caution that when the mean and standard deviation of finger tapping performance measures are used as rate and variability indices, respectively, estimates of lateralized interference will have greater redundancy in contrast to the use of the mean and a coefficient of variation. In this latter case, the coefficient of variation adjusts estimates of variability for differences in rate by dividing the standard deviation of finger tapping performance by the mean rate. Thus, differences reported between studies concerning the comparability of rate and variability measures of lateralized interference may be due to the different kinds of variability measures evaluated. In summary, when the single-key finger tapping procedure is used in dual-task studies of left-hemisphere verbal specialization subjects are usually told to tap as quickly as possible and a rate measure of performance is reported. These task conditions have proven very reliable for the measurement of verbal laterality effects. With the two-key procedure, however, different instructional sets are used and various measures of manual performance reported. Thus, with the two-key task it is unclear which combination of instructional set and performance measure is reliably associated with verbal laterality effects and should be considered for use in future dual-task studies. In order to provide this information, an evaluation of the comparability of different rate and variability measures of manual performance under the two instructional sets used most frequently with the two-key procedure (tap as fast as possible vs. tap as consistently as possible) is provided in this study. The measuresof manual performance evaluated were the rate measures of mean taps per second, mean interresponse time between taps (IRT), and the mean tap-to-tap rate (TTR) which is derived from interresponse times. Variability measures included

LATERALIZED

INTERFERENCE

271

the standard deviation of both the interresponse time (SD of IRT) and tap-to-tap (SD of TTR) measures and coefficients of variability for the interresponse time (CV of IRT) and tap-to-tap (CV of TTR) measures. Finally, dual-task studies have indicated that lateralized interference produced by verbal activities can be detected with verbal tasks that do not require verbal production. That is, a verbal laterality effect in dualtask studies can be observed at the “cognitive” level of verbal activity. The addition of verbal production to the verbal activity can increase the extent of lateralized interference observed (Hellige & Longstreth, 1981). In the present study, therefore, verbal production was manipulated in order to determine if the different measures of tapping performance are affected in similar ways by the presence versus absence of verbal production. METHOD Design The design of this study was a 2 x 2 x 2 x 2 x 5 mixed factorial with subject’s sex as the between-subjects’ factors and within-subjects’ factors of instruction (speed vs. consistency); tapping hand (right vs. left); trial (1 vs. 2); and five task conditions (tapping alone, tapping while solving anagrams silently, tapping while solving anagrams aloud, solving anagrams silently, and solving anagrams aloud). The two anagrams alone conditions provided baseline conditions for estimating lateralized interference in verbal performance as well as tapping performance.

Subjects The subjects were 40 (20 males and 20 females) introductory psychology students from California State University, Fullerton. Recent research indicates that right-handed persons with familial dextrality are most consistently associated with latetalized interference implicating left-hemisphere specialization for verbal processing (see Bathurst & Kee, 1983; Kee et al. 1983).Thus, the 40 participating subjects were right-handed from familial dextral backgrounds. The Edinburgh Handedness Inventory (Oldfield, 1971) was used to determine subjects’ degree of manual laterality and to ensure that subjects’ immediate family members were predominantly right-handed. Laterality scores for males (M = 0.78, SD = 0.23) and females (M = 0.82, SD = 0.16) were equivalent on the Edinburgh test, p > .lO. Prior to experimental trials, subjects’ hand preference was verified (Provins & Cunliffe, 1972). Subjects were asked to sign their names and it was the case that all subjects signed their name with the right hand. In addition, subjects were given two tapping trials, one with each hand. Performance was measured in taps per second and all subjects tapped faster with their right hand (M = 4.21, SD = 0.58) than the left hand (M = 3.75, SD = 0.54, p < .05).

Materials and Procedures Each subject participated in two 30-minute experimental sessions on 2 different days of a single week. During each session, subjects were seated at a table upon which were two microswitch keys (Micro Switch BA-2RV2-A2) mounted in a small (15 x 10 x 5.5 cm) metal utility cabinet, which could be positioned at a comfortable distance (approximately 45 to 60 cm) from subjects in front of either their right or left shoulder. The microswitch keys were spaced 6 cm apart (see McFarland & Ashton, 1978~) and a small rubber pad

272

KEE ET AL.

(1.5 cm in diameter) was attached to the top of each key. The keys were connected to an Apple II Plus microcomputer, with a Mountain Hardware Millisecond clock, which recorded subjects’ tapping performance. The following measuresof tapping performance were calculated by the computer: (1) mean taps per second; (2) mean interresponse time (msec) between taps (IRT); (3) mean tap-to-tap rate (‘FIR), which is based on a transformation of each IRT into a taps per second measure [TI’R = lOOO/IRT(msec)]; (4) standard deviation of IRT; and (5) the standard deviation of the TTR. Subjects were told that the purpose of the experiment was to determine how fast or how consistently they could tap under various conditions. During one session subjects were instructed to tap as consistently as possible and to try to maintain the same rate of consistency throughout all trials. Subjects were given an example of a consistent rate of tapping which consisted of an untimed practice trial during which subjects were instructed to tap along with a ticking sound produced by the microcomputer at a rate of 380 msec as suggested by McFarland and Ashton (1978b). Following this, subjects were given two timed (15 set) practice trials during which they were asked to maintain the same tapping rate. During the other session, subjects were instructed to tap as fast as possible and to try to maintain the same rate of speed throughout the trials. Subjects were given one untimed and two timed practice trials prior to the start of the test trials. In addition, subjects were given one practice anagram trial during each session to ensure that they understood the instructions concerning anagram solution. The order in which subjects participated in the two experimental sessions was counterbalanced. All trials were 15 set and timed by the microcomputer, which produced a buzz that started and stopped each trial. Subjects were given two trials in each of the task conditions and were instructed to look down at the table prior to the start of each trial, then to look up when they heard the buzz and begin tapping and/or perform the verbal task. Solving anagrams, a concurrent task used in a previous study (Kee et al., 1983) constituted the verbal task. In the tapping alone condition, subjects were told to tap continuously during the trial and were instructed to look at a blank piece of paper, positioned in front of them on a typing stand. This was done to prevent subjects’ visual guidance of their finger tapping (see Lomas, 1980). In the concurrent verbal/tapping condition, subjects were asked to perform the tapping task as in the alone condition while simultaneously performing the verbal task. The verbal stimuli were placed in front of the subjects prior to the start of each trial. In the verbal alone condition, subjects were asked to solve anagrams continuously for 15 sec. In both the concurrent verbal/tapping and the verbal alone conditions, the verbal task was performed both silently and aloud. In the silent conditions, subjects were asked to solve the anagrams without saying anything aloud. In the aloud conditions, subjects were asked to describe each step in the anagram solution aloud. Immediately following each trial in the concurrent verbal tapping and the verbal alone conditions, subjects were asked to write down from memory the word solutions for anagrams they had just solved. Within each of the instructional conditions, one-half of the subjects performed the three tapping task conditions first with their right hand, and one-half performed the tapping conditions first with their left hand. The two anagram conditions were randomly intermixed within each set of three tapping conditions performed by each hand. The order in which the five task conditions were performed was counterbalanced by a Latin square. The verbal stimuli consisted of 32 lists of single-solution anagrams. The anagrams were arranged vertically and typed on 13 x 20-cm index cards with an Orator 61-10 type size. Each stimulus list consisted of four anagrams which had a criterion of low to medium imagery (range = 1.90 to 5.74 on a 7-point scale, M = 4.10, SD = 0.98) and medium to high familiarity (range = 2.02 to 6.55 on a 7-point scale, M = 3.85, SD = 1.01) (Gilhooly & Hay, 1977). Low-imagery words were selected because evidence suggests that high-imagery words may implicate the non-language-dominant hemisphere in language processing (Seamon & Gazzaniga, 1973). High-familiarity words were selected in order to ensure that subjects would recognize the words when the letters were unscrambled.

LATERALIZED

INTERFERENCE

273

Across subjects, each stimulus list was used an equal number of times in each alone and concurrent verbal condition. Each of the four anagrams could be solved using one of the following keys ranging from easy to hard: 23145, 54123, 32145, and 51432 (see Rees & Israel, 1935). For example, the letters in the word logic using a 23145 key, would be rearranged into oglic. Each list of four anagrams was presented in order from easy solution to hard solution in order to prevent the subject from becoming discouraged from participating in the verbal task.

RESULTS

Analysis of Manual Performance The following rate measures of tapping performance were selected as dependent variables for analysis: mean taps per second; mean interresponse time (msec) between taps (IRT); and mean tap-to-tap rate (TTR) which was based on the transformation of subject’s IRTs into corresponding taps per second equivalents [‘ITR = lOOO/IRT(msec)].Variability measures of tapping performance selected for analysis included standard deviation of interresponse time (SD of IRT); standard deviation of tap-to-tap rate (SD of TTR); coefficient of variability for IRT (CV of IRT) which is the SD of IRT divided by mean IRT; and coefficient of variability for TTR (CV of TTR) which is the SD of TTR divided by the mean TTR. Percentage reduction in tapping performance was calculated for each of the seven dependent measures. Percentage reduction scores were derived using the following formula: [TA - TC/TA] x 100, where TA is tapping performance in the tapping alone condition and TC is tapping performance during concurrent verbal performance. A preliminary analysis on the raw scores and the percentage reduction scores yielded the same conclusions about lateralized interference.’ Therefore, only the results from the percentage reduction analyses are reported. For the rate measures of mean taps and mean TTR, a positive reduction score indicates interference (i.e., tapping rate decreased) and a negative score indicates facilitation (i.e., tapping rate increased). Alternately, for the percentage reduction scores for mean IRT, a negative score indicates interference and a positive score indicates facilitation. For the four variability measures-SD of IRT, SD of TTR, CV of IRT, and CV of TTR-a negative score indicates interference (i.e., tapping variability increased) and a positive score indicates facilitation (i.e., tapping variability decreased). Lateralized interference in manual performance was evaluated in four separate multivariate analyses which provided the following comparisons: rate measures nested within speed instructions; rate measures nested within consistency instructions; variability measures nested within speed I Raw score manual performance data is available from the authors. We would also be happy to provide the data file for subjects’ performance on all measures for secondary analyses.

274

KEE ET AL.

instructions; and variability measures nested within consistency instructions. The independent variables for each of these four analyses were sex (male vs. female); order (speed instructions first vs. consistency instructions first); trial (1 vs. 2); hand (right vs. left); and condition (silent vs. aloud). Between-subject factors were sex and order, and withinsubject factors were trial, hand, and condition. Prior to performing the multivariate analyses, the data were screened for outliers in accord with procedures suggested by Tabachnick and Fidel1 (1983). An outlier was defined by a standard score greater than + 3.00 or less than - 3.00. It was found that only 1.18% of the data points fell outside this standard score range. These outliers were “trimmed” using procedures suggested by Tabachnick and Fidel1 to percentage reduction scores equivalent to a standard score value of + 3.00 or - 3.00. A summary of the multivariate statistics for hand, condition, and Hand x Condition effects are presented in Table 1. As can be seen, only variability measures of tapping performance under consistency instructions are associated with significant effects involving hand which implicate lateralized interference in manual performance. Assessment of the univariate F ratios for the hand effect on the four variability measures revealed that only the SD of TTR, F(l, 36) = 4.63, p = .04, and CV of TTR, F(1, 36) = 3.98, p = .05, yielded significant hand main effects indicating greater variability in right-hand tapping than left-hand tapping (M SD of TTR: right hand = -58.61 (SD 73.62), left hand = -40.64 (SD 64.44); M CV of TTR: right hand = - 54.46 (SD 69.48) left hand = -37.42 (SD 65.29)). Although the same pattern of lateralized interference was observed on the remaining variability measures (M SD of IRT: right hand = -79.27 (SD = 118.72), left hand = -55.65 (SD = 121.78); M CV TABLE 1 SUMMARY

OF MULTIVARIATE

F RATIOS FOR THE HAND, CONDITION

CONDITION,

Consistency instructions Source

df

x

Speed instructions F

Hand Condition Hand x Condition

Variability dependent measures 4,33 3.89* 4,33 29.74* 4,33 5.88*

Hand Condition Hand x Condition

3,34 3,34 3,34

*p < .05.

AND HAND

EFFECTS

d!

F

4,33 4,33 4,33


Rate dependent measures
1.80 18.59*
LATERALIZED

275

INTERFERENCE

of IRT: right hand = -82.60 (SD = 114.63), left hand = -60.01 (SD = 113.25)), the effects were not statistically reliable, p’s > .20. Significant multivariate F ratios were observed in all analyses for the condition effect, indicating an increase in disruption of tapping performance when verbal production was required. However, only one significant multivariate F for the Hand x Condition interaction was detected. This significant effect was found in the analyses of variability measures under consistency instructions. Univariate analyses of variance indicated a highly significant interaction on the dependent variable of SD of TTR, F(l, 36) = 9.57, p = .004, and a marginally significant effect for CV of TTR, F(l, 36) = 3.28, p = .08. For the other two dependent variables, the univariate F’s for the Hand x Condition interaction were less than 1. Table 2 presents the means for the SD of TTR and CV of TTR relevant to the Hand x Condition interaction. Descriptively it appears that the magnitude of lateralized interference was larger under the verbal production condition. Tests of simple effects indicate reliable hand differences in the aloud condition for SD of TTR, t(78) = 2.67, p < .05, but not in the silent condition, p > . 10. Finally, analyses indicated that the factors of Subject’s Sex, Order, and Trial did not alter conclusions about the patterns of lateralized interference observed. Thus, these factors are not treated in this discussion. In summary, reliable detection of lateralized interference in tapping performance was observed only for variability measures under the instructional condition to tap as consistently as possible. Univariate analyses suggest that of the four variability measures, only the SD and CV of TTR proved to be reliable measures of hand differences in manual performance under concurrent verbal task conditions. Furthermore, only the SD of TTR was sensitive to detecting the increase in lateralized interference produced by verbal production. TABLE 2 MEANS (SD)

FOR THE HAND

x CONDITION INTERACTION FOR THE DEPENDENT SD OF TTR AND cv OF TTR

SD of TTR

Silent Aloud M

OF

CV of TTR

Hand Condition

VARIABLES

Hand

Left

Right

- 16.30 (52.06) -64.97 (66.69) -40.64

-21.89 (48.12) -95.33 (76.57) - 58.61

M

- 19.10 -80.15

Left

Right

- 12.91 (50.48) -61.92 (69.39) - 37.42

- 22.44 (44.38) - 86.47 (75.31) - 54.46

M

- 17.68 - 74.20

276

KEE ET AL.

Zntercorrelational

Analysis

Assessment of the intercorrelations among the different dependent variables provides additional evidence concerning their comparability as indices of tapping performance. Table 3 summarizes these intercorrelations within instructional condition. These results suggest that rate measures are highly correlated with each other, speed instructions: r = .91 to - 99; consistency instructions: r’s = .93, while variability measures are moderately to highly correlated with each other, speed instructions: r = .71 to .94; consistency instructions: r = .79 to .97. In contrast, the correlational analyses indicate that rate and variability measures are not TABLE 3 INTERCORRELATIONSAMONG RATE AND VARIABILITY MEASURESFOR THE Two INSTRUCTIONAL CONDITIONS

Mean taps

Mean IRT

Mean TTR Consistency

Mean taps Mean IRT Mean TTR SD of IRT SD of TTR cv of IRT cv of TTR

SD of IRT

SD of TTR

cv of IRT

cv of TTR

instructions

1.00 -.98

1.00

.98

- .98

1.oo

- .43

.44

- .33

1.oo

- .04

.05

.04

.79

1.00

- .23

.24

-.12

.97

.85

1.oo

- .33

.34

- .25

.87

.95

.86

1.00

Speed instructions Mean taps Mean IRT Mean TTR SD of IRT SD of TTR cv of IRT cv of TTR

1.00 -99

1.00

.91

- .91

1.oo

- .50

.52

-.26

1.00

- .23

.21

.ll

.71

1.00

- .40

.38

-.lO

.94

.84

1.00

- .43

.42

-.15

.76

.90

.84

1.00

LATERALIZED

INTERFERENCE

277

strongly related to each other, speed instructions: r = - .lO to 52; consistency instructions: r = .04 to .44. Analysis

of Verbal Performance

A 2 (instruction: speed vs. consistency) x 2 (hand: right vs. left) x 2 (condition: silent vs. aloud) x 2 (task: solving anagrams while tapping vs. solving anagrams alone) x 2 (trial: 1 vs. 2) x 2 (sex: male vs. female) repeated measures analysis of variance was conducted on the number of anagrams solved in order to evaluate verbal performance. The condition effect, F(1, 38) = 5.58, p = .02, was significant, indicating that fewer anagrams were solved under the aloud condition (M = 1.41, SD = 1.21) than the silent condition (M = 1.55, SD = 1.21). The hand effect was not significant, F < 1, indicating that number of anagrams solved was the same for the right-hand (M = 1.48, SD = 1.23) and lefthand (M = 1.48, SD = 1.19) trials. Furthermore, the absence of a Hand x Task Condition interaction, F < 1, indicates a general absence of lateralized interference in verbal performance. A final significant effect was observed for the four-way interaction of Instruction x Hand x Task x Subject’s Sex, F(1, 38) = 6.63, p = .Ol. Comparisons between the left versus right hand, within the eight conditions defined by the factorial combination of Instruction x Task x Sex, did not reveal any laterality differences in verbal performance (p’s > .05). The absence of lateralized effects on verbal performance observed in this study is consistent with previous findings and is not inconsistent with the hypothesis that when two tasks are competing for the resources of a single hemisphere, there will be interference in the easier of the two tasks, in this case, the tapping (Kinsbourne & Hiscock, 1983). In addition, the equivalent verbal performance observed for the right hand and left hand under concurrent trials allows for an unambiguous interpretation of lateralized interference in tapping performance as reflecting hemisphere-specific verbal processing. DISCUSSION

The present study was designed to identify instructional set and performance measure combinations associated with reliable lateralized interference in dual-task performance. The expected pattern of results, greater interference in right-hand tapping than left-hand tapping produced by the verbal task, was obtained only under consistency instructions with the variability measures of SD of TTR and CV of TTR. Evidence of lateralized interference was not provided by rate measures under either instructional condition or variability measureswithin the speed instructional condition. Furthermore, only the variability measure of SD of TTR in conjunction with consistency instruction was sensitive to the increased lateralized interference produced by verbal production. These findings

KEE ET AL.

suggest that when the alternate tapping of two keys is used as the manual task, lateralized interference is most reliably detected using variability measures under instructions to tap as consistently as possible. The absence of significant lateralized interference with rate measures under speed instructions, rate measures under consistency instructions, and variability measures under speed instructions is surprising given previous research (McFarland & Ashton, 1978a, 197%). A major methodological difference between this study and the previous work of McFarland and Ashton in which reliable lateralized interference effects were reported with speed instruction is the amount of time subjects were allowed to practice tapping prior to the test trials. McFarland and Ashton allowed subjects practice periods of 15-25 min, whereas in the present study, subjects were only allowed one short (approximately lo-20 s) untimed practice trial and two timed (15 s) practice trials with each hand. It may be that the manual task of alternate finger tapping of two keys with one finger is a fairly complex task which requires a longer practice period in order for it to be performed adequately under speed instructions. Therefore, significant interference in the concurrent task conditions may have been masked. In this connection, it is important to note that when speed instructions are used with the simpler manual task of repetitive tapping of one key with the index finger, reliable lateralized interference in tapping rate is consistently observed in the absence of lengthy practice (e.g., Hellige & Longstreth, 1981; Kee et. al.; 1983). The present results suggest that when the alternate tapping of two keys with the index finger is selected as the manual task condition, a combination of consistency tapping instruction and variability measures (SD of TTR or CV or TTR) of manual performance should be used. Alternately, the repetitive tapping of a single key with the index finger may be viewed as the method of choice for dual-task studies of cerebral laterality. Recall that the single-key procedure is consistently associated with reliable laterality effects with a simple rate measure of performance in the study of both verbal/language and visuospatial activity (e.g., Hellige & Longstreth, 1981; Kee, 1984; Kee et al, 1982; 1983; 1984). Furthermore, this simpler manual task is probably less affected by such factors as spatial accuracy and force modulation in tapping and does not require lengthy preexperimental practice. In summary, lateralized interference implicating left-hemisphere specialization for verbal processing was detected only with the variability measures of tapping performance. These findings indicate that rate and variability measures do not always provide comparable estimates of cerebral lateralization in dual-task studies. This interference was found only under instructions to tap as consistently as possible. Thus, if the alternate finger tapping of two keys is selected for use in dual task studies, lateralized interference will be best detected using a combination of variability mea-

LATERALIZED

INTERFERENCE

279

sures and consistency instructions. However, a better choice for future studies would be the one-key tapping procedure. REFERENCES Bathurst, K., & Kee, D. W. 1983, August. Verbal and visuospatial cerebral asymmetry in left-handers. Paper presented at the 91st annual convention of the American Psychological Association, Anaheim, CA. Gilhooly, K. J., & Hay, D. 1977. Imagery, concreteness, age of acquisition, familiarity, and meaningfulness values for 205 five-letter words having single-solution anagrams. Behavioral Research Methods and Instrumentation, 9, 12-17. Hellige, J. B., & Longstreth, L. E. 1981. Effects of concurrent hemisphere-specific activity 19, 395-405. on unimanual tapping rate. Neuropsychologia, Kee, D. W. 1984. Comments on Hughes and Sussman’s time-sharing study of cerebral laterality in language-disordered and normal children. Brain and Language, 22, 197203.

Kee, D. W., Bathurst, K., & Hellige, J. B. 1983. Lateralized interference of repetitive finger tapping: Influence of familial handedness, cognitive load, and verbal production. Neuropsychologia, 21, 617-624. Kee, D. W., Bathurst, K., & Hellige, J. B. 1984. Lateralized interference in finger tapping: Assessment of block design activity. Neuropsychologia, 22, 197-203. Kee, D. W., Bathurst, K., & Gottfried, A. W. 1985, April. Sexual dimorphism during early childhood in left-hemisphere language specialization. Paper presented at the biennial meeting of the Society for Research in Child Development, Toronto, Canada. Kee, D. W., Hellige, J. B., & Bathurst, K. 1982, November. Cerebral lateralization: Assessment of concurrent and manual task activities. Paper presented at the annual meeting of the Psychonomic Society, Minneaplois, MN. Kinsboume, M., & Hiscock, M. 1983. Asymmetries of dual-task performance. In J. Hellige (Ed.), Cerebral hemisphere asymmetry: Methods, theory and applications (pp. 225334). New York: Praeger Press. Lomas, J. 1980. Competition within the left hemisphere between speaking and unimanual task performance without visual guidance. Neuropsychologia, 14, 141-149. McFarland, K., & Ashton, R. 1978a. The influence of brain lateralization of function on a manual skill. Cortex, 14, 102-111. McFarland, K., & Ashton, R. 1978b. The influence of concurrent task difficulty on manual performance. Neuropsychologia, 16, 735-741. McFarland, K., & Ashton, R. 1978~. The lateralized effects of concurrent cognitive and motor performance. Perception and Psychophysics, 23, 344-349. Oldfield, R. C. 1971.The assessment and analysis of handedness: The Edinburgh Inventory. Neuropsychologia, 9, 97-113. Provins, K. A., & Cunliffe, P. 1972. The reliability of some motor performance tests of handedness. Neuropsychologia, 10, 199-206. Rees, H. J., & Israel, H. E. 1935. An investigation of the establishment and operation of mental sets. Psychological Monograph: General and Applied. 46, l-26. Seamon, J. G., & Gazzaaniga, M. S. 1973. Coding strategies and cerebral laterality effects. Cognitive

Psychology,

5, 249-256.

Tabachnick, B. G., & Fidell, L. S. 1983. Using multivariate & Row.

statistics.

New York: Harper