Figural Fluency: Differential Impairment inpatients with Left Versus Right Frontal Lobe Lesions

Transcription

1 An:hives o/oiltical Neuropsychology. Vol.9. pp.41-5s Primed id!he USA. All right& re&elved. ajs $ Copyright e 1993NationalAcademyof Neuropaycholog)' Figural Fluency: Differential Impairment inpatients with Left Versus Right Frontal Lobe Lesions Ronald M. Ruff,1,2 Charles C. Allen,l Charles E. Farrow,1 Hendrik Niemann,l and ThomasWylie1 1Head Injury Center and 2Departmentof Psychiatry and Division of Neurosurgery, University ofcalifornia, Schoof ofmedicine, San Diego Based on an experimental study, Jones-Gotman and Milner demonstrated that patientswith rightfrontal lobe lesions were impaired on design fluency. We sought a clinicaladaptation anddeveloped a psychometrically soundtechnique for design fluency. The present study explores the validity of the Ruff Figural Fluency Test (RFFT) in discriminating patients with eitherrightfrontalor nonright frontal lobe lesions. In the first of two studies, six subjects with circumscribed focal lesions weregiventhe RFFTandthe Jones-Getman andmilnerfiguralfluency task. In the secondstudy, we identified a larger sample of30 patients withfocal lesions in the rightfrontal, leftfrontal, rightposterior, or left posterior cortex. The results from both studies support the validity of the RFFT as a measure which is sensitive to rightanterior dysfunction. Verbal fluency tests are an integral part of neuropsychological evaluations, because of their descriptive value for assessing word-finding difficulties and their proven sensitivity to left frontal lobe dysfunction (Benton, 1968; Milner, 1964; Perret, 1974; Ramier & Hecaen, 1970). Verbal fluency is measured by asking the patient to produce as many words as possible starting with a specific letter of the alphabet within a set time period (Borkowski. Benton, & Spreen, 1967; Thurstone & Thurstone, 1949). The taxonomy called fluency refers, therefore, to the ability to utilize one or more strategies that maximize the production ofresponses while avoiding response repetition. Address correspondence to: Ronald M. Ruff, PhD, Rehabilitation Center, St. Mary's Hospital and Medical Center, 2235 Hayes Street, San Francisco, CA

2 42 R. M. Ruffet al. Jones-Gotman and Milner (1977) were the first to develop an analogous visuospatial method. Similar to the two-part procedure of Thurstone and Thurstone (1949), they devised a design fluency task in which subjects were first asked to draw different nonverbal designs on a sheet of paper during a 5- min period, and then in a second part, limit the drawings to four parts (lines. circles, etc.) and the time to 4 min. Based on a sample of neurosurgical patients with focal lesions, they demonstrated a double dissociation between verbal and figural fluency and left versus right frontal lobe lesions, respectively. Therefore, their measure of design fluency was one of the few measures sensitive to right anterior dysfunctioning. Jones-Gotman and Milner introduced their task as an experimental measure. Despite its potential for clinical application, no normative data have been made available. Regretfully, their experimental measures do not readily lend themselves to psychometric sophistication for the following reasons. Given the vagueness of the tasks, the drawings vary as a rule in size and complexity between individuals, thus influencing the rate of production. The scoring is cumbersome and presents an unnecessary challenge. Even experienced clinicians all too frequently differ in their opinion as to the nonverbal quality of a drawing. Moreover, it is difficult to judge whether a repeated drawing is a perseveration or not, because no two drawings by hand can ever be exactly the same. Finally, in the second part the drawings are supposed to be restricted to four parts; however, it is frequently impossible within a design to reliably identify the number of parts. Ruff and his colleagues (Ruff. Light, & Evans, 1987) developed a procedure for figural fluency similar to a technique first reported by Regard, Strauss. and Knapp (1982), with the intent of providing the appropriate psychometric features for clinical adaptation. Just as verbal fluency is controlled by the letter with which words should begin. a similar restriction was placed on figural fluency: within configurations of dots, different figures are to be drawn by the subject. The Ruff Figural Fluency Test (RFFT) was normed on 358 normal adults stratified according to gender, age. and education. The test proved to be reliable (test-retest on an N =95: r =.76). Performance depended not on sex, but on age and education. Moreover, figural fluency neither correlates with measures of motor speed nor of verbal fluency. Modest correlations, however, were present with performance IQ. In the literature, significantly inferior RFFf performances have been documented on a clinical sample of 35 head injured patients, for verbal and figural fluency relative to 50 controls (Ruff, Evans, & Marshall. 1986). Additional differentiation was demonstrated between those with severe versus moderate head injuries. Studies with pediatric populations suggest an age-related increase in figural fluency. as well as verbal fluency measures (Evans, Ruff. & Gualtieri. 1985; Vik & Ruff. 1988). Despite the above-mentioned applications of the RFFT. the pivotal question remained: is this measure sensitive to right frontal lobe dysfunction?

3 Figural Fluency 43 In the present studies, we tested the hypothesis that patients with circumscribed right frontal lesions would be more impaired on the RFFf as compared to patients with left frontal lobe lesions. When contrasting figural fluency with verbal fluency, a double dissociation was hypothesized. Moreover, in those cases with bifrontal lobe lesions, we predicted that verbal and figural fluency would be significantly compromised. In a first study, we evaluated six patients with clearly confined and well-circumscribed lesions using the RFFf and Jones-Gotman and Milner's task. In a second study we analyzed only the RFFf performance in a larger clinical sample of patients subdivided into four groups with primary lesions disrupting either the posterior or anterior region of either the left or right hemisphere. Subjects STUDY ONE There were six patients with lesions of the frontal lobes identified by one or more computed tomography (CT) scans and a review of the medical records. Janowsky, Shimamura, and Squire (1989) previously described five of these six subjects in their study evaluating the relevance of frontal lobe involvement to human amnesia. Thus, detailed reconstructions from their CT brain scans have been published elsewhere. They later selected the sixth patient, OH, for study and similarly documented that this patient's lesion probably did not exceed the right frontal region. Thus, in Study I we compared two cases with cerebral involvement to either the left or right unilateral frontal lobe and two patients with bilateral lesions, one the result from a brain abcess the other from a cerebral accident (eva). None of the six patients had a significant premorbid history of psychiatric or neurological disorders. Table 1 contains the demographic characteristics, and Table 2 describes the lesion focus. Becauses normative values have already been published (Ruff, 1988; Ruff et al., 1987). no additional control subjects were utilized for either Study 1 or 2. Procedure Ruff Figural Fluency Test (Ruff, 1988). This measure contains five parts (see Figure 1). Each part contains 35 identical stimuli items arranged on a 5 x 7 matrix on white 22 em x 17 cm paper. Each item consists of five dots within a 3 cm square. The dot patterns in Parts I, II, and III are the same with two levels of distractors superimposed in Parts II and III. Parts IV and V have different dot patterns and no distractors. These dot configurations are less symmetrical, with the aim of being even less conducive to verbal strategies (see Figure 1). The rational for creating the different parts is discussed elsewhere (Ruff, 1988).

4 44 R. M. Ruffet ai. TABLEt Demographic Characteristics eases Age Education Work History Current Activities Right Frontal O.H General Laborer Unemployed J.V Retired computer fum Employed as a manager Quality Control Exec., 10 of luxury condominium years Real Estate managment complex. Lives with wife Left Frontal J.D Retired, army nurse for 32 Owns home. Volunteerat years hospital; travels R.L Retired, director of large city Owns home. Cares for school system stockroom invalid wife, reads novels and enjoys gardening Bi-Frontal MD Homemaker, hospital Owns home. Lives with housekeeper for 10 years daughter and grand- Daycare mother children. Enjoys 1V and jigsaw puzzles. Prepares all meals for family G.Y Retired floor covering Owns home. Manages contractor rental property, participates in "gourmet group." Lives with wife TABLE 2 Neurological Features Cases Right Frontal O.H. J.v. Left Frontal J.D. R.L. Bifrontal MD. G.Y. Etiology eva eva 1969 Brain abcess removed 1973 CVA 1985 eva 1985 eva 1976 Lesion Site Small right medical focus in prefrontal region Right dorsolateral prefrontal cortex, with slight involvement of the premotor cortex; some additional involvementof lateral orbital surface of frontal lobe and anterior portion of the insula Primary focus in dorsolateral prefrontal cortex; minimal involvement also of the lateral orbitofrontal cortex Primary focus left dorsolateral prefrontal cortex; slight involvement of lateral orbital surface eva associated with surgery for a left pericauosalartery aneurysm; extensive bilateral lesion ofthe medial frontal cortex including the anterior cingulate cortex of the medial portion of the orbital frontal cortex eva associated with surgery for an anterior communicating artery aneurysm; primary lesion involves left medial frontal cortex

5 FiguralFluency 45 TheRuffFigural Auency Test (Parts I-V) ~ I ]I TIt nz: V FIGURE 1. Configuration or test stimuli in parts I-V. We presented sample items for each part with the following instruction: "In front of you are 3 squares. each containing 5 dots. Note that the arrangement of the five dots is always the same. I want you to connect two or more dots by always using straight lines. The purpose of the test is for you to make as many figures (or designs) as possible. but each design has to be different in some way from all the others." Following completion of the sample, we gave feedback ifidentical designs were made. Subsequently, we presented the test page with the following instructions: "On this page please draw as many different figures (or designs) as possible. Start in the upper left square and work from left to right (examiner points out order). Just connect at least two dots with a straight line and remember, work as quickly as possible and make every figure different. Get ready, go." Design Fluency Measure (Jones-Gotman & Milner, 1977). There were two parts to this test. The first condition required the participant to invent drawings that do not represent actual or identifiable objects without scribblings. We made white sheets of paper available to the subject and set the time limit at 5 min. In the second condition, the participants were instructed that the drawings must consist of only four components. such as lines or circles, and examples were discussed. The time was limited to 4 min. Two independent judges, blind to the nature and aims of the study, scored the test using the procedures outlined by Jones-Gottman and Milner (1977). The number of unique designs. repetitions, and errors make up the scoring, and any inconsistencies were resolved by the first author. Neuropsychological Variables. In a first session, the Controlled Oral Word Association Test by Benton and Hamsher (1976, revised 1978), the Wisconsin Card Sorting Test (Heaton, 1981), and the WAIS-R (Wechsler. 1981) were administered. In a subsequent session, we administered the Design Fluency Measure, followed by the RFFT. Design Fluency Measure RESULTS The combined results of the two conditions of the Jones-Gotman and Milner design fluency task fell in the expected direction for right frontal

6 46 R. M. Ruffet al. patients, that is, both cases only generated 12 unique designs (see Table 3). The unique design production of our right frontal cases was lower than the mean number of designs generated by the right frontal cases reported by Jones-Oatman and Milner (1977). There was an unexpected disparity between the two left frontal cases, because J.D.'s performance was clearly above and R.L.'s was slightly below the level of the two right frontal lobe cases, In the bifrontal cases, the results were also unexpected because both performances fell above those of the right frontal patients. RFFT The two patients with right frontal lobe cases generated approximately 40% fewer unique figures than those with left frontal lobe lesions (see Table 3). The patients with bifrontallesions scored at roughly comparable levels for generating unique figures as the right frontal lobe patients. However, the TABLE 3 Results of Neuropsy.:hological Testing on the Left Frontal, Bifrontal, and Right Frontal Lesion Patients Left Frontal Bifrontal Right Frontal Neuropsychological Test R.L. 1.D. M.D. G.Y. O.H. 1.V. Jones-Gorman and Milner Design Ruency Task: Novel Responses 5-min Novel Responses 4-min Novel Responses Total Ruff Figural Fluency Test: Adjusted Total Unique Figures * 34* 26* 27* Percentile Rank * 2* <1* <1* Controlled Oral Word Association Test: Adjusted Total * Percentile Rank * Wisconsin Card Sorting Test Categories Achieved 3 5 0* }'I' 3* 4 % Perseverative Errors '" 42* 48* 21 WAIS-R: Full Scale IQ Block Design (scaled score) 9 9 5** 5** 5** 6"'* *lmpaired as defined by published norms (i.e., ~ 2nd percentile rank or :?: 2 SD from Mean). **lmpaired as defined by a scaled score value of ~ 6.

7 Figural Fluency 47 fewest unique figures were generated by the two patients with right frontal lesions. According to the normative data, the patients with right frontal and bifrontal lesions fell into the impaired range. whereas the two left frontal patients fell into the borderline and low average ranges. Neuropsychological Variables The verbal fluency measure did not allow a double dissociation with the RFFr or Design Fluency Measure. As shown in Table 3, the performance of one patient (G.Y.) with bifrontal involvement fell within the impaired range. Neither of the left frontal patients displayed an impaired performance. However, a borderline deviation was noted in the case of R.L. On the Wisconsin Card Sorting Test, the performances of bifrontal patients and one right frontal patient (O.H.) fell within the impaired range on the percentage of perseverative errors and categories achieved. A comparison of the Block Design subtest of the WAIS-R demonstrates that the right frontal and bifrontal lobe patients were at comparable levels while the patients with left frontal lobe lesions performed at a higher or average range (see Table 3). Similarly, the Full IQ scores for the two left frontal lobe patients fell above the scores ofthe other four patients. DISCUSSION The data confirmed the hypothesized sensitivity of the RFFf in both cases with right frontal lesions. The two right frontal cases produced the fewest number of unique designs. Moreover, as hypothesized, both of the bifrontal and right frontal cases fell within the impaired range when compared to age and educationally matched norms. The findings also suggest concurrent validity with the Design Fluency Measure by Jones-Getman and Milner. However, among the six patients unexpected variation occurred. We expected the bifrontallobe patients to perform below or at the level of the right frontal lobe patients, but this was not the case for one bifrontal patient (G.Y.) on the design fluency task. One of the left frontal lobe patients (R.L.) also deviated in the unexpected direction on the 5-min. but not the 4-min section of the Design Fluency Task. Although the results ofthe RFFf fall consistently in the expected direction, the small number of patients limits the conclusions that can be drawn. Especially because the IQ scores were substantially higher for the two left frontal lobe patients, it is difficult to determine what role this may play. Therefore, we introduced a second study with a larger sample size. Moreover, comparisons needed to include not only frontal lesion groups, but also patients with left or right posterior lesions.

8 48 R. M. Ruffet al. STUDY TWO Subjects We reviewed archival evaluations from the past 8 years at the University of California Medical Center at San Diego to establish our sample (N =30). Table 4 presents the demographic and neurological characteristics of the four groups. Most of the lesions in these patients were trauma related (N = 27). Because brain trauma cases are generally not well suited for lesion studies, rigid selection criteria were introduced. Only patients with one or more CTscan reports were considered. According to CT summary statements, we initially eliminated out of hundreds of clinical cases all patients with focal and diffuse damage. This resulted in a pool of 37 cases with unilateral focal damage. These cases were further submitted to two reviewers who sorted the summary statements into one of five categories; left frontal, right frontal, right posterior, left posterior, or lesions involving anterior and posterior regions. Only those patients who were sorted into the same focal lesion group by both reviewers were retained, and 10 cases were eliminated in this fashion. All brain trauma patients were assessed as outpatients at least 6 months post trau- TABLE 4 Demographic and Neurological Characteristics of Lesion Groups Right Frontal Left Frontal R. Posterior L. Posterior Gender Male Female Age M SD Education M SD Injury Type Closed-head Cardiovascular Accident Seizure Tumor CT Scan Subdural Hematoma 5 1 Intracranial Hemorrhage Contusion Infarct Neurosurgery

9 Figural Fluency 49 rna, while being oriented and not in a state of posttraumatic amnesia; (this was assessed with the Galveston Orientation and Amnesia Test and all patients achieved a score above 75). None of the patients had a premorbid history of hospitalization for a psychiatric disorder or substance abuse. All patients were right-handed. and none suffered from hemiparalyses. However. motor slowing or hemiparesis was evident on the Finger Tapping Test for two patients with left and three with right frontal lobe damage. These five outliers were able to complete the task but achieved significantly inferior tapping rates for the contralateral hand. Note that on other standardized measures, these patients were able to adequately draw and write with their dominant hand. Procedure Because the subjects were selected from archival evaluations, examiners administered the RFFT (but not the Design Fluency Measure) as part of a comprehensive neuropsychological examination, namely the San Diego Neuropsychological Test battery; for a description of this test battery and its construct validity, see Baser and Ruff (1987). RFFl' RESULTS Table 5 lists the means and standard deviations for total unique figures generated by different lesion groups. Accordingly. a one-way analysis of variance (ANOVA) calculated for total unique figures adjusted for age and education revealed a significant main effect for the lesion groups [F(3, 26) =4.32, P =.013). Three planned comparisons contrasted the right frontal with each of the other focal lesion groups. We adjusted the alpha level to p <.017 according to a Bonferroni correction. Results of t-tests strongly supported our hypothesis, because the patients with right frontal pathology generated significantly fewer designs than those with left frontal [t(14) = 3.04, p =.004], left posterior [t(14) = 3.21, P =.003]. or right posterior lesions [t(12) = 2.63, P =.011). Moreover, the left and right posterior groups performed virtually at the same level in terms oftotal figures. The left frontal sample produced fewer unique designs than the posterior groups, but the difference was not statistically significant; (p >.05). Figural versus Verbal FLuency Jones-Gorman and Milner (1977) reported a double dissociation between verbal and design fluency and left versus right frontal lobe lesions. To exam-

10 50 R. M. Ruffet al. TABLE5 Descriptive Statistics for the Ruff Figural Fluency Test and for the Controlled Oral Word Test by Lesion Groups R. Frontal L. Frontal R. Posterior L. Posterior RFFT UniqueFigures M SD Error Ratio M OOJI SD COW Adjusted Score M SD % Impaired Note. RFFTerror ratio = perseverative errors/unique designs. inc the impact of lesion site on verbal fluency in this sample, we calculated a one-way ANOVA using adjusted scores (by education, sex, and age) on the Controlled Oral Word Association Test. The analysis yielded no main effect for lesion groups [F(3, 26) = 1.31, P =.29]. This finding is consistent with the results of Study One. As shown in Table 5, however, the number of subjects classified as impaired according to published norms (Benton & Hamsher, 1978) was higher in the left frontal group than in the other lesion groups. Neuropsychological Variables Table 6 presents selected neuropsychological functioning for each patient group across several domains. To evaluate motor and visuomotor skills, oneway ANOVAs were calculated on scores for the Finger Tapping Test and Grooved Pegboard Test. Scores on both tests were calculated for dominant hand, nondominant hand, and the difference between dominant and nondorninant hand performances. While the scores were consistent with expected patterns given lesion focus (see Table 6), significant main effects were noted only on the Grooved Pegboard Test for the time difference scores [F(3, 26) =3.45, p =.03] and for the time used with the nondorninant hand [F(3, 26) =7.22, p =.001]. Post hoc comparisons using a Tukey test with the alpha level set at.05 revealed that right frontal patients displayed a greater nondominant minus dominant hand difference in time on the Grooved Pegboard Test than did the left frontal patients; right frontal patients required significantly more time with their nondorninant hand relative to left frontal and posterior lesion groups. No other group comparisons were significant.

11 Figural Fluency 51 TABLE 6 Neuropsychological Characteristics of Lesion Groups Within Functional Areas Lesion Group LF RF LP RP Measures Within Functional Areas M SD M SD M SD M SD Motor Skills Finger Tapping Dominant Nondominant Dominant- Nondominant Grooved Pegboard Dominant Nondominant Dominant- Nondominant Spatial Skills Block Design Scaled Score Rey Figure: Correct Elements Attention and Short-term Memory Digit Span Forward Letter Span Letter Span wi delay Block Span Block Span wi delay Digit Symbol Scaled Score Intelligence Test Verbal IQ Performance IQ Full Scale IQ Vocabulary Spatioconstructional deficits were assessed using the Block Design subtest of the WAIS-R and the copy of the Rey Complex Figure Test (RCFT). For the copy of the RCFT, the number of elements that were placed incorrectly or distorted were subtracted from the total number of elements which were present and correctly placed (referred to as correct elements in Table 6). Separate 2 x 2 (caudality vs. laterality) ANOVAs were calculated using the Block Design scale score and the RCFT correct elements score. Analysis of the Block Design subtest revealed significantly higher scores for the left hemisphere versus right hemisphere lesion groups [F(I, 26) =6.54, p =.017]. The main effect for anterior versus posterior lesion groups and the interaction effect were not significant. On the copy ofthe RCFT, no significant group differences were noted for either the anterior versus posterior nor left versus right hemisphere comparisons.

12 52 R. M. Ruffetal. Immediate retention and short-term recall was assessed according to the Digit Span (WAIS-R; forward), letter, and block span with and without a 20-s interference delay. Psychomotor speed was evaluated according to the Digit Symbol subtest (WAIS-R). The groups did not significantly differ on these attention and short-term memory tasks, although the right frontal group achieved lower overall performances. Verbal IQ and Performance IQ were also roughly equivalent between groups, although the mean Verbal IQ for right posterior patients was 11.5 points higher than the right frontal patients. However, the scaled scores on the Vocabulary subtest of the WAIS-R, were virtually the same for all groups. Accuracy ofrfft Classifications Based on the stratified normative sample of the RFFf (Ruff, 1988; Ruff et al., 1987), which indirectly served as a control group, we calculated T-scores and percentile ranks. We defined an impairment as a score falling two or more standard deviations below the normative mean, stratified according to age and education. We evaluated this impairment index for the patients of both Studies One and Two. To establish diagnostic accuracy when discriminating right frontal from nonright frontal subjects, we calculated the true clinical hit rate using the normative values (i.e., the true classifications of clinical and normal subjects added together and divided by the total N). Accordingly, 85% of the sample was accurately classified. The sensitivity (i.e., the true clinical classifications divided by the true clinical and false normal classifications) was also adequate at 60%. Finally, specificity (i.e., the true normal classifications divided by the combined normal and false clinical classifications) reached 96 percent. GENERAL DISCUSSION The right frontal lobe represents nearly a fourth of the cerebral cortex and has in the past been more or less elusive to localization based on neuropsychological measures. It was for this reason that the Jones-Gotman and Milner Design Fluency Measure received so much attention in the field of neuropsychology. However. its clinical application is restricted not only by a lack of norms but also due to inherent difficulties in achieving psychometric sophistication. Both present studies documented the RFFT as being sensitive to right frontal lobe lesions. When compared to the Design Fluency measure, the RFFT has the advantage of being psychometrically sound and adequately normed. Verbal fluency was not found to be statistically sensitive to left frontal lobe pathology, and therefore no duplication was achieved of the double dissociation reported by Jones-Gotman and Milner (1977). The literature reports mixed findings. and other authors have also encountered difficulties in dupli-

13 Figural Fluency 53 eating a tight association between left frontal lobe pathology and an impairment in verbal fluency (Pendleton & Heaton, 1982). The neuropathology of the subjects in Study One was carefully delineated according to neuroradiological evidence suggesting that the lesions caused by the CVAs were confined to the anterior portions, or more specifically, prefrontal regions of the cortex. However. some question can arise. despite clear neuroimaging, as to whether vascular deficiencies affected other cerebral regions. Similar questions arise when studying patients with lesions caused by trauma. All too frequently. focal damage is accompanied by diffuse damage. Thus, we dropped those cases with diffuse damage or multifocal lesions demonstrated on CT scans. However, in the literature diffuse axonal loss has been noted even in the absence of positive CT scans (e.g., Ruff, Cullum & Luerssen. 1989). Therefore, our selection in Study Two may be criticized as less precise than patient accrual, which excludes trauma cases. We also acknowledge that our selection procedure of eliminating diffuse cerebral damage as detected by CT-scan may be flawed. However, it should be noted that both of these factors decreased the likelihood of obtaining significant results by increasing Type II Error. Moreover, the neuropsychological performance of our sample proved evidence for the validity of our method in that the patterns obtained were generally consistent with expected performance given lesion site. One apparent inconsistency was demonstrated by our patient's performance on constructional tasks. becasue the patients with posterior and anterior lesions reached comparable levels. However, posterior damage alone is frequently not sufficient to produce constructional apraxia, which is more prevalent among patients with post-rolandic damage involving multilobe lesions to either the parietal-temporal, parietal-occipital, or parietal-temporal-occipital areas, particularly in the right hemisphere (Benson & Barton, 1970; Piercy, Hecean, & de Aguriaguerra, 1960). A review ofct-scan reports in our posterior lesion groups revealed that only two left posterior patients and one right posterior patient suffered multiple lesions involving the areas listed above. This may account for the absence of significant constructional deficits according to the mean performance ofour posterior lesion groups. The combined results of Studies One and Two, which were conducted on substantially different samples in terms ofage and etiology ofneuropathology, yields strength to confirmation of our hypotheses. Although a cross-validation ofthe results on another focal lesion sample is needed, the present results suggest that the selected cut-off values for impairment for unique designs on the RFFT are quite accurate in discriminating patients with right frontal versus nonright frontal pathology. From a cognitive standpoint, however, the question remains as to exactly what functions are being assessed by the RFFT. At one level the RFFT requires the drawing of different figures. which depends on a fine motor coordination and spatioconstructional skills. The present results were not accounted for by group differences in motor speed, dexterity or spatioconstructional

14 54 R. M. RuffetaL skills. On yet a different level, this test also measures fluid and flexible thinking and the ability to create novel responses without repetitions similar to the verbal fluency task. Most likely, frontal lobe pathology disrupts the ability to switch between various strategies to maximize production of responses while at the same time avoiding response repetition. Our data suggests that the right frontal lobe effect cannot be attributed to group-specific limitations in attention or short-term memory. but rather suggest a limitation in fluid and flexible thinking. On a more behavioral level. it will be of interest to determine the ecological validity of RFFf. In other words. the question will need to be answered as to how a deficiency on the RFFf corresponds with problems in everyday life functioning. Acknowledgment: We are grateful to Larry Squire, Joyce Zouzounis, and their colleagues for sharing their patients with us and collaborating in the data collection of Study One. We are also indebted to Larry Marshall for providing consistent encouragement. Finally, the presented research was, in part, supported by a grant from Learning Services Corporation. REFERENCES Baser, C. A., & Ruff, R. M. (1987). Construct validity of the San Diego Neuropsychological Battery. Archives ofclinical Neuropsychology, 2, Benson, D., & Barton, M. (1970). Disturbances in constructional ability. Cortex, 6, Benton, A. L. (1968). Differential behavioral effects in frontal lobe disease. Neuropsychologia, 6, Benton, A. L., & Hamsher, K., des (1976 revised 1978). Multilingual Aphasia Examination. Iowa City: University of Iowa. Borkowski, J. G., Benton, A. L., & Spreen, O. (1967). Word fluency and brain damage. Neuropsychologia, S, Evans, R., Ruff, R. M., & Gualtieri, C. (1985). Verbal fluency and figural fluency in bright children. Perceptual and MotorSkills, 61, Heaton, R. K. (1981). A Manual/or the Wisconsin CardSorting Test. Odessa, FL: Psychological Assessment Resources, Inc. Janowsky, J. S., Shimamura, A. P., & Squire, L. R. (1989). Memory and metamemory: Comparisons between patients with frontal lobe lesions and amnesic patients. Psychology, 17(1),3-11. Jones-Gotman, M., & Milner, B. (1977). Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia, 15, Milner, B. (1964). Some effects of frontal lobectomy in man. In J. M. Warren & K. Akert (Eds.), The frontal granular cortexand behavior (pp ). New York: McGraw-Hm. Pendleton, M. G., Heaton, R. K., Lehman, R. A., & Hulihan, D. (1982). Diagnostic utility of the Thurstone Word Fluency Test in Neuropsychological Evaluations. Journal 0/ Clinical Neuropsychology,4, Perret, E. (1974). The left frontal lobe of man and the suppression of babitual responses in verbal categorical behavior. Neuropsychologia, 12, Piercy, M., Hecean, H., & de Ajuriaguerra, J. (1960). Constructional apraxia associated with unilateral cerebral lesions -left and right sided cases compared. Brain, 83, Ramier, A. M., & Hecaen, H. (1970). Role respectif des attelntes frontales et de Ja lateralisation lesionnelle dans Ies deficits de la "fluence verbale." Revue Neurologique (Paris), 123,

15 Figural Fluency 55 Regard, M., Strauss, E., & Knapp, P. (1982). Children's production on verbal and non-verbal fluency tasks. Perceptual and Motor Skills, 55, Ruff, R. M. (1988). Ruff Figural Fluency Test Administration Manual. San Diego: Neuropsychological Resources. Ruff, R M., Culhnn, C. M., & Luerssen, T. G. (1989). Brain imaging and neuropsychological outcome in traumatic head injury. In: E. D. Bigler, R. A. Yeo, & E. Turkheimer (Eds.), Neuropsychologicaljunction and brain imaging (Chap. 6) (pp ). New York: Plenum Press. Ruff, R. M., Evans, R., & Marsball, L. (1986). Impaired verbal and figural fluency after head injury. Archives o/clinicalneuropsychology, 1, Ruff, R. M., Light, R., Evans, R (1987). The Ruff Figural Fluency Test: A nonnative study with adults. Developmental Neuropsychology, 3, Thurstone, L. L., & Thurstone. T. (1949). Examiner Manual/or the SRA Primary Mental Abilities lest. Chicago: Science Research Associates. Vile, P., & Ruff, R. M. (1988). Children's Figural Fluency Performance: Development of strategy use. Developmental Neuropsychology. 4, Wechsler, D. (1981). Wechsler Adult Intelligence Scale-Revised (Manual). The Psychological Corporation. New York: Harcourt, Brace, Jovanovich Publishers.