Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Journal of Speech and Hearing Disorders, Volume 54, 113-124, February 1989 THE COMMUNICATIVE EFFECTIVENESS INDEX: DEVELOPMENT AND PSYCHOMETRIC EVALUATION OF A FUNCTIONAL COMMUNICATION MEASURE FOR ADULT APHASIA JONATHAN L O M A S L A U R APICKARD McMaster University, Hamilton, Ontario, Canada STELLA BESTER HEATHER ELBARD Chedoke-McMaster Hospitals, Hamilton, Ontario, Canada ALAN FINLAYSON Chedoke-McMaster Hospitals and McMaster University, Hamilton, Ontario, Canada CAROLYN ZOGHAIB Chedoke-McMaster Hospitals, Hamilton, Ontario, Canada Groups of aphasic patients and their spouses generated a series of communication situations that they felt were important in their day-to-day life. Using criteria to ensure that the situations were generalizable across people, times, and places, we reduced the number of situations to 36 and constructed an index that allowed the significant others of 11 recovering and 11 stable aphasic individuals to rate their partners' performance in the situations on two occasions 6 weeks apart. These data were then used to evaluate the psychometric properties of the Communicative Effectiveness Index (CETI) as a measure of change in functional communication ability. Further application of a generalization criterion reduced the final index to 16 situations. Results- showed the CETI to be internally consistent and to have acceptable test-retest and interrater reliability. It was valid as a measure of functional communication according to the pattern of correlations found with other measures (Western Aphasia Battery, Speech Questionnaire, and global ratings). Finally, it was responsive to functionally important performance change between testings. Further research with the CETI and its usefulness for clinicians and researchers are discussed. Historically, the focus of aphasia assessment has been on language abilities with general communicative abilities as only a secondary consideration. Furthermore, assessment instruments have been validated with more concern for their ability to discriminate aphasic from nonaphasic performance or one aphasia type from another than for their ability to detect change in the severity of the aphasia over time. The development of an instrument with the intent to measure change does not guarantee that it will actually measure change (Kirshner & Guyatt, 1985; MacKenzie, Charlson, DiGioia, & Kelly, 1986). Despite these historic trends there has been increased interest recently in measuring pragmatic communication skills in aphasia and to do this in the context of withinpatient evaluation of change rather than between-patient discrimination (Behrmann & Penn, 1984; Beukelman, Yorkston, & Lossing, 1984; Holland, 1977; Houghton, Pettit, & Towey, 1982; Milton, Prutting, & Binder, 1983). We were particularly interested in a measure of functional communication that could document the evolution in aphasic individuals of both language and nonlanguage solutions to the communication problems encountered in day-to-day living. The existing instruments developed to measure this area of aphasic performance (Holland, 1980; © 1989, American Speech-Language-Hearing Association Lincoln, 1982; Sarno, 1969) are all unsatisfactory for one or more of the following reasons: They are incomplete in their documentation of psychometric properties (Skenes & McCauley, 1985), correlate poorly with observation of nonverbal communication (Behrmann & Penn, 1984), are not easy for the assessor to either administer or score (Houghton et al., 1982; Swisher, 1979), correlate so well with existing language measures (Holland, 1980) that they are probably not measuring any separate dimension of communication, and have not been shown to be sensitive to functionally important change over time. Therefore, we developed a measure of functional communication for the adult with aphasia that could measure change in performance over time--the Communicative Effectiveness Index (CETI). Based on previous work in the area of disease-specific quality of life measures (Guyatt, Bombardier, & Tugwell, 1986; Kirshner & Guyatt, 1985), we established a set of properties that the instrument should possess. These properties were relevant both to the development of items and the structure of the index, as well as to the evaluation of the index, During development we asked ourselves the following questions: 113 Downloaded from jshd.asha.org on April 26, 2010 0022-4677/89/5401-0113501.00/0 114 Journal of Speech and Hearing Disorders Are both verbal and nonverbal communication assessed? Are patients' values reflected in the instrument? Is performance in daily living assessed? Is the instrument credible to users (face validity)? Is the instrument simple and easy to administer? For the evaluation of the index we answered the following questions: Is there a quantitative measure of change? Is the instrument reliable (i.e., is it reproducible across time, people, and places)? Is the instrument valid (i.e., does it really measure the stated dimension of functional communication)? Is the instrument responsive (i.e., does it measure functionally important change even if this change is small)? These properties or criteria are quite similar to those proposed by the American Psychological Association as standards for test development (APA, 1974) and are used by others in their critical appraisals of language assessment tools (McCauley & Swisher, 1984; Skenes & McCauley, 1985). METHOD lndex Structure In deciding on the structure of the index we wished to ensure that performance (not potential) of communication in daily living was assessed. To do this one must directly observe, or get reports of direct observation of, the aphasic individual in actual communication situations. Alternatively, one might role-play such situations with the aphasic patient. Holland (1980), in her Communicative Abilities in Daily Living, chose the role-playing option; Sarno (1969), in her Functional Communication Profile, assumed that the clinician-assessor would have some direct observation experience of the aphasic patient's performance. However, we wanted our instrument to be simple and easy to administer and therefore rejected the role-playing option. Furthermore, we were not confident that clinician-assessors generally observe the aphasic patient in daily-living situations; thus, they might confuse potential with performance. Therefore, we rejected direct observation by the clinician. We opted for reports from direct observation made by the aphasic individual's significant other (i.e., spouse, relative, neighbor, or friend) who spends enough time with him or her to make accurate judgments of communication performance. Because the purpose of the CETI is to quantitatively assess change in performance over time, the spouses' or significant others' judgments were structured as ratings of the aphasic person's performance in particular situations on a 10-cm visual analogue scale (VAS). To emphasize that we were interested in performance relative to premorbid ability of the individual, not relative to some absolute standard, the anchor on one end of the scale was not at all able and at the other end was as able as before 54 113-124 February 1989 the stroke. We selected the VAS (rather than a Likert-type scale) because we believed it to be particularly sensitive to changes in performance over time (Bond & Lader, 1974). Because the focus of the CETI was on change in performance rather than absolute level of performance, significant others rated the aphasic partner on the communication situations on a series of different occasions. The difference in ratings from one assessment to the next (the change score) was the score of interest. On the basis of recent research showing that the size of the change score is unaffected but the variance around it is reduced when raters are allowed to see their ratings from previous occasions (Guyatt, Berman, Townsend, & Taylor, 1985), we decided to adopt this strategy for the administration of the CETI. Raters therefore saw on the VAS where they had rated the performance of the aphasic person on their previous assessments for each communication situation (see Figure 1). Index Content To ensure that the communication situations that would be rated were representative of patient values and daily-living activities, we elicited situations from aphasic individuals themselves. The procedures employed, and the reliability or validity of the various judgments made by experts during the process, have been described in more detail elsewhere (Lomas, Pickard, & Mohide, 1987). What follows is a brief description and summary of the results. Under the auspices of the local Stroke Recovery Association, we assembled for a 4-hr meeting a group of 9 I THE COMMUNICATIVE EFFECTIVENESS INDEX I Fiaum~ 1. Schematic of the use of the visual analogue scale for rating at initial and repeat testing. Downloaded from jshd.asha.org on April 26, 2010 LOMAS ET AL.: The C o m m u n i c a t i v e E f f e c t i v e n e s s I n d e x stroke survivors who had been or still were aphasic; 7 of them were accompanied by their spouse or significant other. Using the nominal group process technique of Delbecq, Van de Ven, and Gustafson (1975), we asked for their responses to the question " I n which situations does a stroke survivor have to be able to get his meaning across and to understand what someone else means?" The nominal group process allows each participant 10 min of individual consideration of the question before responses are elicited from members in turn--the aim is to provide each participant with an equal opportunity to respond. Responses are recorded in view of the entire group, and the process continues until no m e m b e r has any further responses to offer. As far as possible the exact wording of the respondents is recorded. Following this the group then judges redundancies between responses and rewords those that can be combined or that it feels need improvement, thus arriving at a final list. The process was repeated with a second group of 5 aphasic stroke survivors and their spouses who were recruited from past patients of the speech pathology service at Chedoke-McMaster Hospitals; this group provided a comparability check on the representativeness of the situations generated by the first group. Both groups had individuals with long-standing aphasia and a variety of severities, ranging from mild residual anomia to persisting global aphasia. In the case of the few individuals in the latter category it was, obviously, the spouse or significant other who volunteered the wording for responses. To perform the comparability check, an expert panel of eight (five speech-language pathologists and three neuropsychologists) independently applied explicit rules developed by us (see Appendix A) to each of the group's lists after noncommunication situations had been removed. This was in order to (a) assign the communication situations to one of the four categories of basic need, life skill, social need, or health threat, thus comparing the relative proportions in each category for each list; and (b) use the first patient group's list as the "index set" and match comparable situations from the second patient group, thus dividing the latter into matched and unmatched situations. The levels of agreement on categorization among the eight raters are presented in Table 1. The large number of TABLE 1. Levels of agreement between the eight raters on categorization decisions. %of ~of No. of raters assigning item to category Patient Group 1 items Patient Group 2 items All 8 agreed on category 7 of 8 agreed on category 6 of 8 agreed on category 5 of 8 agreed on category 4 of 8 agreed on category Less than 4 agreed on category 21.2 27.3 27.3 21.2 3.0 0 10.5 28.9 31.6 18.5 10.5 0 115 raters and possible categorizations precluded the use of conventional measures of interrater reliability. However, we could use the percentage of item categorization decisions falling into each of the possible levels of agreement. In no case did fewer than four of the eight panelists agree on a category assignment, and for the majority of the assignments (70%-75%), six or more of the panelists were in agreement. The results of the comparability check are presented in Table 2. After exclusion of noncommunication situations, the number of situations generated by each group was roughly comparable (33 and 38). The percentage in each of the communication categories was not significantly different between the two groups [×2(3) = 1.65, p = .65], Finally, 52% of the communication situations from the second patient group were completely matched to situations generated by the first patient group. We were therefore satisfied that the patient groups were not generating unique and nongeneralizable communication situations. The final pool of situations for initial evaluation of the C E T I was arrived at in the following manner. 1. An initial pool was constructed from all communication situations generated by the first patient group plus all the unmatched situations from the second patient group. 2. The expert panel met as a group and judged each situation, retaining only those that they felt satisfied three criteria: not redundant with another situation on the list, assessed performance rather than potential, and relevant to both institutional and community environments. The last of these criteria was important to ensure that the C E T I would be applicable for aphasic persons living in both community and institutional settings. There were 51 situations in the initial pool of finalized items--those from the first patient group plus unmatched situations from the second patient group. After application of the three criteria by the expert panel, 36 situations remained. These 36 situations, worded as closely as possible to the actual wording used by the item-generaTABLE 2. Comparability of the communication situations generated by two patient groups. Patient Group 1 Number of situations generated Number of noncommunication situations generated Number of communication situations generated Percentage in each communication category: Basic Need Health Threat Life Skill Social Need Percentage of situations matched to Patient Group 1 Downloaded from jshd.asha.org on April 26, 2010 Patient Group 2 40 44 7 6 33 38 16% 3% 37% 44% 20% 9% 36% 35% - 52% 116 Journal of Speech and Hearing Disorders tion groups, were structured as our initial CETI. It was then administered to the significant others of two groups of aphasic persons to evaluate its reliability, validity, and sensitivity to change. 54 113-124 February 1989 TABLE 3. Description of patient samples. Characteristic Number Age The Patient Sample Sex The C E T I was administered on two occasions, 6 weeks apart, to 11 patients recovering from recent onset of aphasia and 11 stable aphasic patients. Inclusion criteria for both groups were: spoken and written knowledge of English, a spouse or significant other was available who had contact with the aphasic individual at least three times a week, a score of less than 93.8 on the Western Aphasia Battery [this score is the cutoff value for aphasic vs. nonaphasic individuals (Kertesz, 1979)]. In addition, for the recovering group, initial test administrations had to be possible within 6--10 weeks postonset. For the stable group, the initial administrations had to be 15 months or more postonset, the attending clinician had to judge the patient as stable, and patients could not be receiving formal language treatment. The attending clinicians used informal assessments to judge whether the patients were stable. The aphasic patients involved in the initial generation of communication situations were excluded from this portion of the study. The patients were recruited from two regional referral centers from admission log books (recovering group) or from consecutive speech pathology department files going back from 15 months prior to the study start date (stable group). The recovering group were all receiving language treatment from the regional referral centers. Table 3 describes the two patient groups. The slightly greater age of the recovering sample (65.4 years vs. 57 years) represents a recent shift in the demographics of stroke patients referred to the centers. The majority of patients in both patient groups had suffered a cerebral vascular accident (CVA). In the recovering group, 10 patients were suffering from their first unilateral infarct and all but one of these were in the left hemisphere; the other patient had a recent left hemisphere infarct after a previous right hemisphere lesion. At initial testing the types of aphasia, according to t h e classification system of Kertesz (Kertesz & Poole, 1974), were: 4 global, 5 Broca, 1 Wernicke, and 1 transcortical sensory. In the stable group, 9 patients had unilateral left hemisphere lesions with no previous infarcts, and 2 patients had previous infarcts prior to the latest left hemisphere lesion--one in the left and one in the right hemisphere. The types of aphasia were: 7 Broca, 1 conduction, and 3 anomic. Only one potential subject for the study was excluded because of the unavailability of a significant other meeting the criterion of at least three contacts per week. Eight of the significant others in the recovering group were spouses, 2 were daughters, and 1 was a sister. All 11 were spouses for the stable group. The number of weeks postonset for initial assessment fell within our desired ranges of 6--10 weeks for the Etiology Recovering M SD M F CVAa Aneurysm Unknown Time of initial testing (weeks postonset) Time between initial and repeat testing (weeks) Severity (WAB Aphasia Quotient) b Initial test Repeat test Description of significant others M SD M SD Range M SD M SD Spouse Child Sibling aCVA = cerebral vascular accident, Battery (Kertesz, 1982). 11 65.4 8.3 8 3 11 0 0 7.2 2.4 7.1 1.3 5.7-9.7 28.1 21.3 33.7 33.7 8 2 1 Stable 11 57.0 14.2 10 1 8 2 1 163 61 8.2 1.1 6.0-9.5 60.0 26.4 59.0 29.8 11 0 0 bWAB = Western Aphasia recovering group and greater than 65 weeks for the stable group. Second assessments were completed at slightly longer than our planned 6-week interval but were not significantly different between the recovering (7.1 weeks) and stable (8.2 weeks) groups, t(20) = 1.1, n.s. The relative severity of the aphasia in the two groups, as measured by the Aphasia Quotient (AQ) of the Western Aphasia Battery, was as expected; at initial assessment the value of the recovering group's AQ was about half that of the stable group. Furthermore, the AQ did show significant improvement in the recovering group between first (M = 28.1) and second (M = 33.7) testing [t(10) = 3.16, p = .01], supporting the fact that they were recovering; the stable group's AQ showed little change (Ms = 60, 59). Finally, the existence of a period of spontaneous recovery in aphasic patients (Culton, 1969; Lomas & Kertesz, 1978) enabled us to use the recent onset recovering group to assess the C E T I ' s ability to detect functionally important change. I f the index was responsive to change, then there would be a significant improvement in scores from initial to second testing in recovering patients. The stable group enabled us to assess the test-retest reliability of the CETI. I f scores on the index were reproducible over time, there would be a strong relationship between the scores at initial and second testing in stable patients. Data on both groups combined contributed to assessments of the internal reliability and the validity of the index. Test Administration Tests were administered by a trained research assistant Downloaded from jshd.asha.org on April 26, 2010 LOMAS ET either at the centers or the aphasic patients' homes. In addition to the CETI, a number of other assessments were conducted. The Western Aphasia Battery (Kertesz, 1982) was administered to obtain a score on a traditional "language" assessment tool. The Speech Questionnaire (Lincoln, 1982) was used as an alternate measure of functional communication that, although poorly documented as to its validity and significantly focused on functional language, was easy and simple to administer. This assessment was completed by the significant other in the presence of the research assistant. The significant others were also asked to complete a single global rating scale that asked them to assess the partner's "overall language and communication skill during the previous week" (possible score of 1--extremely poor to 7--excellent). Finally, 11 aphasic patients had available at least two eligible significant others. We had them both independently complete the C E T I to provide some information on the interrater reliability of the index. The C E T I requires a brief training period with the significant other to familiarize him/her with the use of the VAS. In addition, raters were explicitly instructed to think not only of verbal communication but also of all other forms of communication and understanding when rating the performance of the aphasic individual in the specific communication situations. It was stressed that the assessment was of the individual's overall ability to get his/her meaning across or understand someone else's meaning in daily-living situations, using any communi- cation means at their disposal. The raters were also told to feel free not to rate a particular communication situation if they felt that it was not relevant or did not apply to their partner's situation. This instruction was included to facilitate the final item reduction by revealing which communication situations were considered generally applicable across the variety of settings and people to whom the C E T I would eventually be administered. Perhaps most important, it also ensured that raters provided assessments only on items describing situations in which they had actually observed their partner's performance; raters were explicitly instructed to rate actual performance and not to infer potential performance for unobserved situations. The rating for each situation was converted into a score by laying a template marked with l-ram divisions over the 10-cm VAS and reading off a value between i and 100. The total C E T I score was converted to a 100-point maximum by dividing the sum of the individual situation ratings by the total number of situations. A high score indicated good performance and a low score, poor performance. AL.: The Communicative Effectiveness Index 117 generalizability of each item was revealed by the number of significant others who actually chose to rate their partner on it (i.e., the number who had actually observed performance of the behavior described in an item). To guarantee that the final items were as generalizable as possible, we adopted the stringent criterion that we would include only those situations for which 21 of the possible 22 significant others gave a rating at both initial and repeat testings. Application of the above criterion excluded 20 items to leave a final set of 16 communication situations on the C E T I (see Appendix B). Eight of the 16 items were rated by all 22 significant others at both initial and repeat testing, and 8 items by 21 of the 22 at both initial and repeat testing. All remaining reliability and validity evaluations were carried out on these 16 items. For the analyses that follow, the 8 missing values were imputed using a program that took the patient's scores on all the other items and weighted it by all other patients' scores on that particular item. DISTRIBUTION OF SCORES Figure 2 displays the distribution of total scores on the C E T I for both groups at the initial assessment. Table 4 provides the actual scores of the individuals in both groups for the initial and repeat assessments. The mean C E T I scores for the group reflected the greater severity of the recovering group (44.8) compared to the stable group (68.0). This shows that, although the C E T I was not developed for use on the absolute scores (the focus is on the change in score over time), the absolute scores may have some validity as indicators of the severity of an aphasic individual's problems. Of course, on an individual patient basis, some of the recov- RECOVERING SAMPLE (N=11) # OF PAT I ENTS MEAN I/ : I, 44.8 S,D, = 12,7 STABLE SAMPLE (N=11) # OF PATIENTS RESULTS FINAL ITEM REDUCTION 0-10 '11-20'21-30 "31-40 MEAN = 68.0 The 36 communication situations were initially selected to be generalizable across people and settings on the basis of the expert panel's judgments. The true '81-9 J 91-1 S.D, = 16,8 CETI TOTAL SCORE FIGURE 2. Distribution of total scores on the Communicative Effectiveness Index at initial testing. Downloaded from jshd.asha.org on April 26, 2010 118 Journal o f Speech and Hearing Disorders 54 TABLE 4. Scores on the Communicative Effectiveness Index (CETI) (max. = 100) and change scores on the Western Aphasia Battery (WAB; Kertesz, 1982) (max. = 10O) and the Speech Questionnaire (SQ; Lincoln, 1982) (max. = 18). Patient Recovering group 1 2 3 4 5 6 7 8 9 10 11 M SD Stable group 1 2 3 4 5 6 7 8 9 10 11 M SD Initial test score CETI Repeat test score Change score CETI CETI WAB SQ 57 47 31 46 64 48 24 53 47 25 48 65 65 49 66 72 54 28 59 62 44 51 8 18 18 20 8 6 4 6 15 19 3 44.8 12.7 55.9 12.6 11.4 6.6 57 62 75 51 79 73 77 98 84 41 52 64 63 75 51 80 74 88 89 85 40 40 68.0 16.8 68.1 18.0 7 1 0 0 1 1 11 -9 1 -1 -12 0 6.3 8 13 -3 21 4 17 0 9 0 8 2 7.2 7.6 10 4 -14 -2 2 7 5 5 -9 -5 -4 -0.1 7.4 4 -2 0 2 0 2 4 1 5 3 5 2.2 2.3 1 2 0 -4 -1 0 0 0 0 -1 1 -0.2 1.5 e r i n g group still had absolute scores h i g h e r than the stable group. T h e scores approximate a normal distribution with evid e n c e that significant others were using all or most of the VAS w h e n doing their ratings with enough room left at the top and bottom of the scale for both deterioration and i m p r o v e m e n t in performance to b e d e t e c t e d (MacKenzie & Charlson, 1986). Because the i n t e n t of the C E T I is not to d i s c r i m i n a t e b e t w e e n a p h a s i a types, no a t t e m p t was m a d e to relate these total scores to the nature of the aphasic patient's deficits. T a b l e 4 also p r o v i d e s the change scores from the AQ of the W e s t e r n A p h a s i a Battery (WAB) and from the S p e e c h Questionnaire. 1 It is i n t e r e s t i n g to note that in the recove r i n g group, with the exception o f 2 patients (6 a n d 8), the C E T I reveals t h e same or larger i m p r o v e m e n t s in perform a n c e than the WAB. Also, in the stable group, with the F e b r u a r y 1989 e x c e p t i o n of 3 patients (7, 8, a n d 11), the C E T I shows s m a l l e r changes in p e r f o r m a n c e than the WAB for this clinically d e t e r m i n e d stable p o p u l a t i o n who are not rec e i v i n g s p e e c h therapy. I n d e e d , 7 of the 11 patients have change scores of zero or one on the C E T I , w h e r e a s none have such small c h a n g e scores on t h e WAB. T h e greater t e n d e n c y for the WAB to show changes in performance, in this stable p o p u l a t i o n , m a y reflect its r e l i a n c e on a s s e s s m e n t of p e r f o r m a n c e on a specific occasion (when patients m a y or m a y not b e "'at t h e i r best"); as o p p o s e d to the C E T I ' s r e q u e s t for ratings of p e r f o r m a n c e at the general t i m e i m m e d i a t e l y p r e c e d i n g the a s s e s s m e n t session (thus a v e r a g i n g the " g o o d " and " b a d " days). RELIABILITY Internal Reliability I f the items w i t h i n the a s s e s s m e n t tool are all measuring the same d i m e n s i o n (in this case, c o m m u n i c a t i o n effectiveness), t h e r e s h o u l d b e a strong r e l a t i o n s h i p both a m o n g t h e m and b e t w e e n each one a n d the total score (i.e., the tool s h o u l d b e i n t e r n a l l y consistent). O f course, this r e l a t i o n s h i p s h o u l d not b e perfect; o t h e r w i s e t h e r e w o u l d b e no justification for h a v i n g m o r e than one item in the index. Two m e a s u r e s of internal c o n s i s t e n c y w e r e u n d e r t a k e n on the scores of b o t h groups c o m b i n e d (N = 22). T h e s e results, and the results of the o t h e r r e l i a b i l i t y assessments, are p r e s e n t e d in T a b l e 5. First, w e c a l c u l a t e d the correlations b e t w e e n the score o b t a i n e d on each i n d i v i d ual item a n d the total score (item-total correlations). Second, w e c a l c u l a t e d the C r o n b a c h a l p h a statistic (Cronbach, 1951), w h i c h c o n c e p t u a l l y c o m p u t e s the composite correlation from all the correlations of split-half comparisons of all the item scores. T h e overall t r e n d in t h e item-total correlations is toward high values (11 of the 16 items c o r r e l a t e d at .50 or greater with the total), i n d i c a t i n g that scores on the TABLE 5. Reliability assessments of the Communicative Effectiveness Index. 1. Internal reliability (calculated on initial scores of both patient samples, N = 22) a) Item-total correlation #items >.75 4 .5-.75 7 .25-.50 3 <.2,5 2 b) Cronbach's alpha: .90 Total 16 2. Test-retest reliability (calculated on stable group only, (n = 1Although Lincoln (1982) advises against pooling the scores on the Speaking and Understanding scales of the Speech Questionnaire, we have done so in this report of change scores for simplicity of presentation. When used in the validity analysis later in the paper, the scores on the two scales are kept separate. 113-124 i1) Intraclass correlation = .94 95% confidence limits = .87-.99 3. Interrater reliability (calculated on change scores of eligible from both patient samples, n = 1I) Intraclass correlation = .73 95% confidence limits = .62-.81 Downloaded from jshd.asha.org on April 26, 2010 LOMAS ET AL.: The Communicative Effectiveness Index individual items are strongly related to, but not totally coincident with, the total score. The Cronbach alpha of .90 is high and supports the good internal reliability of the index. 119 correlation is acceptable, espeeially given the fairly subjective nature of the rating judgments that were required of the significant others, and suggests that the change scores obtained from one rater are generalizable to others. Test-Retest Reliability VALIDITY Just as it is important to demonstrate that an instrument can detect change when real change in performance has occurred (see Validity section), it is equally important to be able to show that it does not detect any change over time when performance has stabilized (i.e., test-retest reliability). We calculated test-retest reliability using the stable group of patients, which, given our inclusion criteria for the group, should not show any change in mean performance on the C E T I between the two testings (see Table 5). We did indeed find this to be the case with a mean score at first testing of 68.0 (SD = 16.8) and at second testing of 68.1 (SD = 18.1). The standard error of this mean difference of 0.1 between first and second testing was estimated to be 5.87.~ The intraclass correlation (rho) was .94, with a confidence interval of .87 to .99. The individual patient scores shown in Table 4 indicate that this test-retest reliability was a group effect, with the scores of some aphasic patients in the stable group changing by up to 12 points. It is not clear whether the changes in these patients' scores reflect measurement error or the variability in clinicians' judgments about what constituted a "stable patient." Clarification of the meaning of these individual patient score changes will have to await further study of the properties of the CETI. Either way, we were most interested in the group effect, which clearly demonstrated test-retest reliability. Interrater Reliability In addition to being reproducible across time, the ratings made of one individual's performance should also be reproducible across different raters at the same point in time (interrater reliability). With two significant others available for 11 of our 22 patients (6 recovering and 5 stable), we were able to calculate the interrater reliability of the CETI. Because the focus of the instrument is on the change in score from one assessment to the next, we calculated the intraclass correlation coefficient (rho) on the change score produced by the 11 pairs of raters (see Table 5). For these 11 patients the mean change score of the first raters was 6.4 (SD = 6.5) and of the second raters 8.3 (SD = 9.4); the intraclass correlation, adjusted to represent the population of potential raters (Berk, 1979), was .73 with 95% confidence limits of .62 to .81. This 2This estimate of the standard error is based on the estimated variance of the difference score, derived from the pooled variance of the scores on first and second testing, corrected for test reliability using the formula SE = SDV~T~-~ Face Validity Face validity pertains to the degree to which an instrument appears to be measuring what it was designed for. As such it relates significantly to our criterion of "credible to users." The extensive involvement of the aphasic patients themselves in generating the items used on the C E T I should ensure that the index has high face validity. Furthermore, the item reduction process involved clinicians, the aphasic individuals, and their significant others in culling from an initial pool of 51 items the 16 that were most representative of communicative effectiveness in the aphasic patient. Thus, clinicians familiar with the deficits encountered in aphasia should find the items listed in Appendix B credible as measures of functional communication. Construct Validity The construct validity of an instrument is a measure of the extent to which it actually measures what it was designed for. This is assessed by comparing the scores on the instrument to scores obtained from other measures of the same individual's performance. We had three concurrent measures available for comparison with the CETI. First, we used the Western Aphasia Battery (WAB), a traditional measure of language ability that has b e e n validated as being able to reliably discriminate among aphasic types and between individuals who do or do not have aphasia (Shewan & Kertesz, 1980). However, although the WAB has been used, it has not been evaluated as a measure of change in aphasic performance over time. We used the Aphasia Quotient (AQ) as the WAB score of interest (maximum score = 100). The second measure was the Speech Questionnaire (SQ), a brief self-administered instrument designed as a measure of functional language ability in aphasia, which has b e e n shown to be internally reliable using Guttman scaling and to have good test-retest reliability (Lincoln, 1982) but has not been assessed for its validity against other measures or for its ability to measure change over time. We used the scores from the Speaking (maximum = 13) and Understanding (maximum = 5) subscales separately because Lincoln reports that they cannot be combined. Finally, we had the single global ratings done by each of the significant others to evaluate the overall ability of their partner in both language and communication (maximum score = 7). Performance was rated as excellent, good, fairly good, fair, not too good, poor, or extremely poor. Downloaded from jshd.asha.org on April 26, 2010 120 Journal of Speech and Hearing Disorders 54 113-124 Each of these measures was obtained at both the initial testing and the repeat testing. Commensurate with our a priori hypothesis that functional communication is a separate but overlapping dimension to language, we made a series of predictions about the relationship between the scores on the C E T I and scores on these other measures. We predicted that the strongest associations would be seen between the C E T I and the global ratings by the significant others; although because these global ratings included more assessment of language ability than the C E T I (communication and language were assessed by the global ratings), we did not expect the correlations to be extremely high. Shewan and Kertesz (1980) claim that the WAB has some components that measure functional communication, but the focus of the WAB on language assessment led us to expect only a moderate but nevertheless significant correlation with the CETI. Finally, the correlation of both the Speaking and Understanding subseales of the SQ with the C E T I was not expected to be very different from that between the C E T I and the WAB, despite the supposed focus of the SQ on functional language ability. This a priori hypothesis was based on scrutiny of the items in the SQ; they exclusively concerned language issues (e.g., "Does he/she say phrases spontaneously?") rather than communication performance in different daily-living situations. The Spearman rank correlation coefficients between the C E T I and the three measures for initial and repeat testings were calculated on all 22 of the aphasic individuals from both groups (except for some of the global ratings where ratings for fewer than 22 patients were available) (Table 6). We did indeed find that the highest correlations were with the global ratings by significant others at initial and repeat testing. The WAB had the next highest correlations and the SQ similar but slightly lower correlations. All the correlations between the C E T I and the other measurements at initial and repeat testing were significant at at least the .05 level. Our belief that the SQ, despite its stated objective, was measuring language rather than communication found some support from the fact that it correlated highly with the WAB: .89 at initial testing and .87 at repeat testing. However, correlations involving the SQ should be treated TABLE 6. Correlations between the Communicative Effective- ness Index and three other measures of aphasic performance. Global ratings by significant others Western Aphasia Battery b Speech Questionnaire ° Speaking Understanding N Initial test Repeat test a 22 .79** .61"* .62** .52** 22 22 .46* .47* .43* .56* ~The number of raters at initial testing was 19 and at repeat testing was 21. bKertesz, 1982. °Lincoln, 1982. *p < .05. **p < .01. February 1989 cautiously insofar as the other instruments combine measurement of speaking and understanding, whereas the SQ measures them separately. The pattern of correlations found between the C E T I and these other measures suggests that it possibly measures a separate but overlapping dimension to the traditional language assessment instrument. Sensitivity to Change in Performance Because the principal objective of the C E T I is to assess changes in performance over time (i.e., evaluate progress), it was particularly important to demonstrate that it was sensitive to and able to measure functionally important change. It was for this reason that we selected patients for a "recovering group"; spontaneous recovery assured us of a patient sample in which functionally important change would likely occur. I f the C E T I failed to show significantly improved scores for this group between the initial and repeat testing, then it would clearly not be sensitive to or able to measure this functionally important change. Therefore, using only the 11 aphasic patients from the recovering group, we did a repeated measures analysis of variance on their initial and repeat test scores on the CETI. The mean total score at initial testing was 44.8 (SD = 12.7) and at repeat testing was 55.9 (SD = 12.6). The difference between these mean scores was significant, F(1, 10) = 32.4, p < .002. No such significant difference was found for the stable patients, where the mean total score at initial testing (68.0) was not significantly different from that at repeat testing (68.1). The mean change score for the recovering group was 11.4, with a standard deviation of 6.6 and a range of 3-19. This change, however, was not significantly correlated with the group's change in AQ scores on the Western Aphasia Battery; the Spearman rank correlation coefficient was only .36 (df = 11, p > .D. DISCUSSION This study used aphasic patients to generate items for a functional communication measure and patients' significant others to rate performance on these items. On the basis of psychometric evaluation of 11 recovering and 11 stable aphasic patients, the resulting 16-item Communicative Effectiveness Index appears to be reliable and valid. Both the inclusion/exclusion criteria and the scores on traditional language measures indicate that one group was indeed recovering function, whereas the other was stable. Although some individuals in each group may not have shown recovering or stable scores, the two groups were selected to have the potential to recover or to have a high probability of being stable. Obviously, this did not guarantee that all group members would recover or be stable, but it did result in group means that clearly represent recovering and stable samples. The stable pa- Downloaded from jshd.asha.org on April 26, 2010 LOMAS ET AL.: The Communicative Effectiveness Index tients were used to assess test-retest reliability, and the recovering patients contributed to the assessment of the CETI's sensitivity to change in performance. The remaining aspects of instrument evaluation combined both groups of patients for analysis. The evaluations of reliability produced clearly acceptable values for use of the CETI in group studies and adequate values for use with individual patients. However, further studies would be useful to clarify the CETI's reliability for use with the latter. The most contentious question is whether the results support the validity of the CETI as a measure of functional communication. There are two aspects to this question. First, there is the question of whether significant others are unbiased reporters of aphasic patient performance. Second, there is the issue of whether the relationship of scores on the CETI and scores on the other measures supports the hypothesis of a dimension of functional communication that is separate from but overlapping with the language dimension. Spouses (or significant others) have been shown to have a good level of agreement with aphasic patients on the presence or absence of particular communication problems, as well as on the severity of the problem (Shewan & Cameron, 1984). Of particular interest for the CETI is that when there is disagreement between the spouse and the aphasic patient on the precise level of severity, spouses are equally likely to underestimate (31% of problems) as to overestimate (27%) severity as judged by the patients themselves (Shewan & Cameron, 1984). In contrast, some studies have shown that, relative to clinicians (not relative to the aphasic patient), spouses overestimate performance (e.g, Helmiek, Watamori, & Palmer, 1976). However, because we were looking for validation of spouse assessments as proxies for the aphasic patients, we did not consider as relevant those studies that contrasted spouse and clinician assessments. Furthermore, the validity of clinicians' assessments of as subjective an area as functional communication in the aphasic patient is potentially problematic because of their limited opportunity to view the patient when he or she is actively engaged in functional "real-life" behaviors. Indeed, we have shown elsewhere (Lomas et al., 1987) that the types of communication situations that are considered important for functional behavior in aphasia are significantly different for clinicians and aphasic patients. This finding also reinforced our decision to go directly to the patients and their partners for item generation for the CETI. It also calls into question the widespread use of clinicians' views as the basis for constructing items for any "quality of life" assessment instrument. Thus, for the purpose of rating performance, we believe that spouses are reasonable proxy respondents for the aphasic patient, with no apparent bias toward overestimating ability. Furthermore, for the CETI it would not even matter if spouses did overestimate ability because of the focus of the index on the change score. As long as ability was overestimated an equivalent amount at each testing, then the validity of the change score would remain intact. 121 Assessment of construct validity presents a particular problem for the development of instruments in new areas because of the absence of an alternative measure of the dimension under consideration with which to compare the new measure. For aphasia assessment tools that focus on discriminating among types of aphasia and relating this to localization of the damaged areas of the brain (e.g., the Western Aphasia Battery or the Boston Diagnostic Aphasia Examination) such a "gold standard" for comparison (and, therefore, validation) is available from CT scans or other "hard" evidence of the actual location of damage. However, what can be considered the "gold standard" of functional communication against which the CETI can be validated? One of the reasons to develop such a measure is the belief that existing aphasia assessment instruments do not, or do not fully, assess functional communication and would not, therefore, correlate particularly highly with the CETI. We adopted the approach of specifying, a priori, the relationship we would expect to see between the CETI and three other measures (WAB, Speech Questionnaire, and global ratings) if the CETI was measuring a separate dimension of functional communication (Cronbach & Meehl, 1955). The actual correlations we obtained were consistent with these specifications. Further evidence for the recovery of functional communication separately from the recovery of language ability came from the observation that the recovering group's change scores on the CETI did not correlate significantly with their change scores on the WAB. This result is not inconsistent with our hypothesis of a separate dimension of functional communication, with a separate recovery course from language, and that this dimension is being measured by the CETI but not by traditional instruments. A similar argument has been made by Sarno in discussing the relationship between the Functional Communication Profile and the Neurosensory Centre Comprehensive Examination for Aphasia (Sarno, Sarno, & Levita, 1971). It might be argued that validation should involve correlations of the CETI scores with direct observations of actual communication behavior. This approach was used by Holland in validation of the Communicative Abilities in Daily Living (Holland, 1980). However, such a validation strategy for the CETI would have been redundant and, if the observations were rated or scored by clinicians, potentially misleading. It would have been redundant because the structure of the index was such that spouses were already rating the functional communication ability of their aphasic partners based on observations of their actual communication behavior. Our acceptable levels of interrater reliability attest to the stability of this rating. If the ratings of actual communication behavior were performed by clinicians, then they would be, to some extent, subjective, and we have no reason to believe that they would be any more valid as a measure of a particular patient's functional communication ability than ratings done by the spouse. In fact, we have reason to believe that they would be less valid compared to a spouse-there is only limited exposure of the clinician to the Downloaded from jshd.asha.org on April 26, 2010 122 Journal of Speech and Hearing Disorders particular communication foibles of a particular patient. The existing evidence comparing clinician and spouse judgments demonstrated that they do not coincide (Helmick et al., 1976; Lomas et al., 1987). Therefore, in the area of functional communication measurement, it would not appear to be a reasonable premise that ability can be more objectively defined by expert clinicians than by individuals such as spouses who have significantly greater exposure to the aphasic patients' behavior. Obviously, the small number of patients involved in the evaluation of the C E T I means that further validation and reliability assessments should be undertaken. Future work can profitably explore the relationship between changes on the C E T I and other language and communication measures; this will assist in specifying more clearly the degree to which functional communication is separate from other aspects of aphasic patients' difficulties. The higher scores of the stable group compared to the recovering group suggest that, although not designed for this purpose, the C E T I could be useful in measuring absolute severity of, and not just change over time in, functional communication deficits. Studies in other patient groups would clarify this potential use of the index. Finally, studies on larger numbers may clarify the reliability of the C E T I for use with individual patients. Future work with the C E T I would also help to answer some important questions that were not addressed by the current evaluation. For instance, the high proportion of men in both our patient groups (18 out of 22) raises the question of whether the index is truly generalizable across the sexes. With more women would different results have been obtained or would different communication situations have been included as items? However, it should be r e m e m b e r e d that more males than females generally become aphasic (McGlone, 1980). Nevertheless, the items on the C E T I appear, and were constructed to be, generalizable across the sexes (as well as across places and time). It is interesting to note that of the 16 final items on the index, 8 were from the category of Social Need, 3 from Life Skill, 4 from Basic Need, and only 1 from the Health Threat category (see Appendix A). The major value of this instrument is as a measure of change in communicative effectiveness, thus enabling clinicians to evaluate individual patients' progress in therapy and researchers to evaluate recovery of communication ability in group studies of the effectiveness of therapy. The C E T I does not purport to be a comprehensive aphasia assessment instrument; it does not substitute for the traditional language assessments. It has not been designed (nor has it been validated) as a tool to discriminate among aphasic types or between those with aphasia and those without aphasia, although it does appear to potentially reflect levels of severity in an aphasic population. It should not be used to predict responsiveness to therapy unless future work is able to validate it for such a use. In its present form and at the present stage of evaluation, clinicians and researchers should feel confident using the change scores obtained from one testing to the next as valid and reliable measures of an aphasic 54 113--124 February 1989 patient's progress at functionally communicating in everyday life situations. The critical value for the C E T I ' s change scores, above which individual aphasic patients can be d e e m e d to have made a clinically important improvement, can be estimated by two methods. If one assumes that the combined treatment effect and spontaneous recovery of the recovering group is equivalent to clinically important improvement, then the difference between the mean initial and repeat scores of this group (11.4) represents the critical value--scores at or above this level signify improvement. Alternatively, one may consider that the critical value is the change score that would exclude all patients in the stable group from having exhibited clinically important improvement [i.e., the largest change score in the stable group (12.0)]. Both methods yield a change score of approximately 12 over a 6--8 week period as the initial estimate of the critical value. Of course, if some of the recovering patients were not in fact recovering, or if some of the stable patients were not in fact stable, this will be an overestimate of the critical value. Corroboration or revision of this initial estimate will have to await further development work with the CETI. One final characteristic of the C E T I worth mentioning is its reliance on assessments made by significant others. This not only gives, in a structured fashion, first-hand evidence on communication performance to the clinician, but it also involves an often motivated and concerned spouse in assessment of progress and orients him or her to the full repertoire of potential communication behaviors available to his or her partner. Given that the clinician can only be in contact with the aphasic individual for a few hours a week, increased involvement of those who live with or spend large amounts of time with him/her may lead to improved results in therapy outcomes. The Communicative Effectiveness Index is an attempt to make available an instrument that will be able to measure some of these more functional changes that, though occurring during the aphasic patient's recovery process, have so far eluded measurement. ACKNOWLEDGMENTS We wish to thank the following for their assistance during the study: the Hamilton-Wentworth Stroke Recovery Association; Allan Kroll and the staffof Niagara Region Rehabilitation Center; and Drs. Geoffrey Norman, Gordon Guyatt, Cindy Shewan, and Judith Trotter. This study was supported by grants from the Ontario Ministry of Health (#01265) and Bauer Funds from the Health Sciences Gerontological Research Group, McMaster University. The first author is supported as a career scientist by the Ontario Ministry of Health (#0135). REFERENCES AMEmCAN PSYCHOLOGICALASSOCIATION.(1974). Standards for educational and psychological tests. Washington, DC: Author. BEHRMANN, M., ~ PENN, C. (1984). Non-verbal communication of aphasic patients. British Journal of Disorders of Communi- Downloaded from jshd.asha.org on April 26, 2010 LOMAS ET AL.: cation, 19, 155-168. BERK, R. A. (1979). Generalizability of behavioral observations: A clarification of interobserver agreement and interobserver reliability. American Journal of Mental Deficiency, 83, 460472. BEUKELMAN, D. R., YORKSTON, K. M., & LOSSlNG, C. A. (1984). Functional communication assessment of adults with neurogenie disorders. In A.S. Halperin & M.J. Fuhrer (Eds.), Functional assessment in rehabilitation (pp. 101-115). Baltimore: Paul H. Brookes. BOND, A., & LADER, M. (1974). The use of analogue scales in rating subjective feelings. British Journal of Medical Psychology, 47, 211-218. CRONBACH, L. J. (1951). Coefficient alpha in the internal structure of tests. Psychometrika, 16, 297-334. CRONBACH,L. J., & MEEHL, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281-300. CULTON,G. L. (1969). Spontaneous recovery from aphasia.Journal of Speech and Hearing Research, 12, 825-832. DELBECQ, A.L., VAN DE VEN, A. H., & GUSTAFSON, D . H . (1975). Group techniques for program planning--A guide to nominal group and delphi processes. Boston: Glenview, Scott, Foresman, and Co. GUYATT, G. H., BERMAN, L. B., TOWNSEND, M., & TAYLOR, D.W. (1985). Should study subjects see their previous responses? Journal of Chronic Diseases, 38, 1003-1007. GUYATT, G. H., BOMBARDIER, C., & TUGWELL, P. (1986). Measuring disease-specific quality of life in clinical trials. Canadian Medical Association Journal, 134, 889-895. HELMICK, J.W., WATAMORI, T. S., & PALMER, J.M. (1976). Spouses' understanding of the communication disabilities of aphasic patients.Journal of Speech and Hearing Disorders, 41, 238-243. HOLLAND, A. L. (1977). Some practical considerations in aphasia rehabilitation. In M. Sullivan & M. Kommers (Eds.), Rationale for adult aphasia therapy (pp. 345-349). Lincoln: University of Nebraska Medical Centre. HOLLAND, A. L. (1980). Communicative abilities in daily living: A test of functional communication for aphasic adults. Baltimore: University Park Press. HOUGHTON, P. M., PETTIT, J. M., & TOWEY, M. P. (1982). Measuring communication competence in global aphasia. In Proceedings of the Clinical Aphasiology Conference (pp. 28-37). Minneapolis: BRK Publishers. KERTESZ, A. (1979). Aphasia and associated disorders: Taxonomy, localization and recovery. New York: Grune and Stratton. KERTESZ,A. (1982). Western Aphasia Battery. New York: Grune and Stratton. KERTESZ, A., & POOLE, E. (1974). The Aphasia Quotient: The taxonomic measurement of aphasic disability. Canadian Journal of Neurological Sciences, 1, 1-17. KIRSHNER, B., & GUYATT, G. (1985). A methodological framework for assessing health indices.Journal of Chronic Diseases, 8, 27-36. LINCOLN, N. B. (1982). The Speech Questionnaire: An assessment of functional language ability. International Rehabilitation Medicine, 4, 114-117. LOMAS,J., ~ KERTESZ,A. (1978). Patterns of spontaneous recovery in aphasic groups: A study of adult stroke patients. Brain and Language, 5, 388-401. LOMAS, J., PIGKARD, L., & MOHIDE, A. (1987). Patient versus clinician item generation for quality of life measures: The case of language disabled adults. Medical Care, 25, 764-769. MACKENZIE, R. C., & CHARLSON, M. E. (1986). Standards for the use of ordinal scales in clinical trials. British Medical Journal, 292, 40-43. MACKENZIE,R. C., CHARLSON,M. E., DIGIOlA, D., & KELLY,K. (1986). Can the Sickness Impact Profile measure change? An example of the scale assessment. Journal of Chronic Diseases, 39, 429-438. McCAULEY, R., & SWISHER, L. (1984). Psychometric review of language and articulation tests for preschool children. Journal The Communicative Effectiveness Index 123 of Speech and Hearing Disorders, 49, 34-42. MCGLONE, J. (1980). Sex differences in human brain asymmetry: A critical survey. Behavioural and Brain Sciences, 3, 215-226. MILTON, S. B., PRUTTING, C. A., & BINDER, G. M. (1983). Appraisal of communicative competence in head injured adults. In R. H. Brookshire (Ed.), Clinical aphasiology: Conference proceedings (pp. 114-123). Minneapolis: BRK Publishers. SAmqO, M. T. (1969). Functional communication profile. New York: New York University Medical Centre. SARNO, M.T., SARNO, E., & LEVITA, E, (1971). Evaluating language improvement after completed stroke. Archives of Physical Medicine and Rehabilitation, 52, 73-78. SHEWAN, C. M., & CAMERON, H. (1984). Communication and related problems as perceived by aphasic individuals and their spouses. Journal of Communication Disorders, 17, 175-187. SHEWAN, C. M., & KERTESZ, A. (1980). Reliability and validity characteristics of the Western Aphasia Battery (WAR). Journal of Speech and Hearing Disorders, 45, 308-324. SKENES, L. L., & McCAULEY, R. J. (1985). Psychometric review of nine aphasia tests. Journal of Communication Disorders, 18, 461-474. SWISHER, L. (1979). Functional Communication Profile (FCP) (Review). In F. L. Darley (Ed.), Evaluation of appraisal techniques in speech and language pathology (pp. 17-20). Redding, MA: Addison Wesley. Received September 30, 1986 Accepted January 14, 1988 Requests for reprints should be sent to Jonathan Lomas, Clinical Epidemiology and Biostatistics, McMaster University, 1200 Main Street West, Hamilton, Ontario L8N 3Z5 Canada. APPENDIX A The following definitions were provided to the expert panel: Communication Categories 1. Basic Need--Communication is required to meet basic needs (e.g., toileting, eating, grooming, positioning). 2. Health Threat--Physical well-being or health is dependent on effective communication (e.g., calling for help after falling, giving or receiving information about one's medical condition). 3. Life Skill--Giving or receiving information that is necessary to accomplish everyday living (e.g., shopping, home maintenance, use of telephone, understanding traffic symbols). 4. Social Need--Communication that is primarily social in nature (i.e., communication with others as an end in and of itself such as dinner table conversation, playing cards, writing a letter to a friend). Matching Situations should be considered matched if the inclusion of both of the situations in a communication assessment scale would provide no extra information beyond the inclusion of only one of the situations. APPENDIX B F i n a l 16 Items of the C o m m u n i c a t i v e Effectiveness Index (CETI) Please R a t e 's ability a t . . . 1, Getting somebody's attention. 2, Getting involved in group conversations that are about him/ her. Downloaded from jshd.asha.org on April 26, 2010 124 Journal of Speech and Hearing Disorders 3. Giving yes and no answers appropriately. 4. Communicating his/her emotions. 5. Indicating that he/she understands what is being said to him/ her. 6. Having coffee-time visits and conversations with friends and neighbors (around the bedside or at home). 7. Having a one-to-one conversation with you. 8. Saying the name of someone whose face is in front of him/ her. 9. Communicating physical problems such as aches and pains. 54 113-124 F e b r u a r y 1989 10. Having a spontaneous conversation (i.e., starting the conversation and/or changing the subject). i i . Responding to or communicating anything (including yes or no) without words. 12. Starting a conversation with people who are not close family. 13. Understanding writing. 14. Being part of a conversation when it is fast and there are a number of people involved. 15. Participating in a conversation with strangers, 16. Describing or discussing something in depth. Downloaded from jshd.asha.org on April 26, 2010 The Communicative Effectiveness Index: Development and Psychometric Evaluation of a Functional Communication Measure for Adult Aphasia Jonathan Lomas, Laura Pickard, Stella Bester, Heather Elbard, Alan Finlayson, and Carolyn Zoghaib J Speech Hear Disord 1989;54;113-124 This article has been cited by 13 article(s) which you can access for free at: http://jshd.asha.org/cgi/content/abstract/54/1/113#otherarticles This information is current as of April 26, 2010 This article, along with updated information and services, is located on the World Wide Web at: http://jshd.asha.org/cgi/content/abstract/54/1/113 Downloaded from jshd.asha.org on April 26, 2010