A Comparison of Fragile X Syndrome and Down Syndrome

Home , Down syndrome, Fragile X syndrome

VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION

Signaling Noncomprehension of Language: A Comparison of Fragile X Syndrome and Down Syndrome

Leonard Abbeduto Waisman Center, University of Wisconsin-Madison Melissa M. Murphy Kennedy Krieger Research Institute, Johns Hopkins University Sara T. Kover, Nancy D. Giles, Selma Karadottir, Adrienne Amman, Loredana Bruno, Jee-Seon Kim, Susen Schroeder, Julie A. Anderson, and Kathryn A. Nollin University of Wisconsin-Madison

Abstract Signaling noncomprehension of the spoken messages of others was examined for youth with fragile X or Down syndrome in comparison with each other and nonverbal MA- matched typically developing children. A direction-following task was used in which some of the directions were inadequate. Both syndrome groups signaled noncomprehension less often than did the typically developing children. The ability to signal noncomprehension appropriately was related to a measure of receptive vocabulary and syntax. Preliminary analyses indicated that males with fragile X syndrome signaled noncomprehension less often than did their female peers, even after controlling for differences in nonverbal MA.

For a discourse to be successful, the partici- within-group variability in most domains of lan- pants must fulfill the obligations associated with guage use, including noncomprehension signaling their roles as speaker and listener (Clark, 1996). In (Abbeduto & Rosenberg, 1980; Bedrosian & Prut- the role of listener, a participant must use all avail- ting, 1978). The causes and correlates of such var- able sources of information to construct the iability are poorly understood (Abbeduto & Hes- speaker’s intended meaning. Moreover, the listen- keth, 1997). In this study, we examined the pos- er must signal when comprehension is not possi- sibility that the nature and extent of problems in ble so that the speaker can provide clarification. noncomprehension signaling vary with etiology If the listener fails to signal noncomprehension, by focusing on Down syndrome and fragile X syn- he or she will find it increasingly difficult to con- drome, the two most common genetic causes of struct an accurate representation of the talk and intellectual disabilities (Dykens, Hodapp, & Fin- to make meaningful contributions (Clark & ucane, 2000). We also examined the sources of Schaefer, 1989). Individuals with intellectual dis- between- and within-syndrome differences in non- abilities often fail to signal noncomprehension comprehension signaling. Such data can provide (Abbeduto, Davies, Solesby, & Furman, 1991; Ab- the foundation for language interventions de- beduto, Short-Meyerson, Benson, & Dolish, 1997; signed to meet the unique needs of the individual Abbeduto, Short-Meyerson, Benson, Dolish, & with intellectual disabilities (Dykens et al., 2000; Weissman, 1998; Ezell & Goldstein, 1991; Fujiki Hodapp & Fidler, 1999; Murphy & Abbeduto, & Brinton, 1993). There is, however, considerable 2005).

214 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Signaling noncomprehension requires that 2006). Third, auditory memory is especially im- the listener continuously monitor his or her un- paired (i.e., relative to other facets of memory) in derstanding and formulate linguistic responses Down syndrome (Marcell & Weeks, 1988; Seung that make clear to the speaker what aspects of the & Chapman, 2000), whereas no such asynchrony utterance are problematic and, thereby, the nature has been documented for fragile X syndrome (Dy- of the clarification that the speaker must provide kens et al., 2000). Fourth, maladaptive behavior (Clark, 1996). In turn, this ability presupposes the occurs at relatively low rates in Down syndrome, development of a number of linguistic, cognitive, but is frequent in fragile X syndrome, with the and social–cognitive skills (Abbeduto et al., 1997, latter being troubled by, for example, social anx- 1998; Abbeduto & Short-Meyerson, 2002; Ack- iety (Bregman, Leckman, & Ort, 1988; Mazzocco, erman, 1993; Golinkoff, 1986; van der Meij, Baumgardner, Freund, & Reiss, 1998) and atten- 1988; Robinson & Whittaker, 1985; Tager-Flus- tional difficulties (e.g., Bregman et al., 1988; Cor- berg, 2001; Whitehurst & Sonnenschein, 1985), nish, Sudhalter, & Turk, 2004; Dykens, Hodapp, virtually all of which are delayed or impaired in & Leckman, 1989; Mazzocco, Pennington, & Down syndrome and fragile X syndrome. Hagerman, 1993). Although the relative contri- Most individuals with Down syndrome have butions of language, theory of mind, auditory impairments in numerous cognitive skills, as re- memory, and maladaptive behavior to the typical flected by IQs in the range of mild to moderate development of noncomprehension signaling intellectual disabilities (Chapman & Hesketh, have yet to be determined (Abbeduto et al., 1997), 2000). The range of cognitive impairments is it is reasonable to suppose that phenotypic differ- broader in fragile X syndrome, with virtually all ences in these domains could lead to within- and males and half of females with the full mutation between-syndrome differences in noncomprehen- having a diagnosis of intellectual disabilities (Hag- sion signaling. Moreover, comparisons of Down erman, 1999). Below age-level mastery of the lin- syndrome and fragile X syndrome should be es- guistic system is almost always found in individ- pecially useful in clarifying the contributions of uals with Down syndrome (Chapman, 2003) and these behavioral domains to the development of is characteristic of most males and many females mature noncomprehension signaling. with fragile X syndrome (Murphy & Abbeduto, In light of the importance of noncomprehen- 2003). Performance in the social–cognitive do- sion signaling to successful discourse, it is surpris- main (e.g., in reasoning about mental states) is ing that there are no published studies on the abil- also delayed relative to typically developing age- ity of individuals with Down or fragile X syn- matched peers in Down syndrome (Zelazo, Bur- drome to engage in noncomprehension signaling. ack, Benedetto, & Frye, 1996) and fragile X syn- It is interesting to note, however, that very young drome (Cornish et al., 2005; Garner, Callias, & children with Down syndrome show delays in Turk, 1999; Grant, Apperly, & Oliver, 2007; Lewis learning to respond as speakers to requests for et al., 2006). Thus, the impairments that define clarification from other people (Coggins & Stoel- the behavioral phenotypes of Down and fragile X Gammon, 1982; Scherer & Owings, 1984), al- syndromes are likely to lead to substantial delays though the extent of the delay (e.g., whether ex- in noncomprehension signaling. cessive relative to other domains of language) is Despite the commonalities, there are differ- not clear (Rosenberg & Abbeduto, 1993). In ad- ences in the behavioral phenotypes of Down syn- dition, there is some evidence that the ability to drome and fragile X syndrome. First, the devel- evaluate the fit of a spoken utterance to the dis- opment of the linguistic system, especially its syn- course context is impaired in individuals with tactic aspects, is more delayed in Down syndrome fragile X syndrome, even in adult females with than in fragile X syndrome (Abbeduto et al., typical-range IQs (Simon, Keenan, Pennington, 2003). Second, individuals with Down syndrome Taylor, & Hagerman, 2001). These studies rein- display deficits in theory of mind (i.e., reasoning force the notion that noncomprehension signal- about mental states) that are more severe than ing may pose a special challenge for individuals their deficits in other areas of cognitive function- with Down or fragile X syndrome; however, the ing (Zelazo et al., 1996), whereas individuals with extent of that challenge for either syndrome re- fragile X syndrome reason as accurately about mains to be determined. mental states as do their MA-matched typically In examining noncomprehension signaling, it developing peers (Garner et al., 1999; Lewis et al., is important to recognize that the type of prob-

᭧ American Association on Intellectual and Developmental Disabilities 215 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. lematic message and its context can have dramatic Abbeduto, 2003). Because only a small number of effects on young listeners (Shatz, 1983). In terms females with fragile X syndrome participated in of message type, incompatible messages (i.e., this study, the results are only preliminary in na- those for which there are no reasonable available ture. referents) are highly salient and associated with minimal information-processing constraints and, Method thus, typically developing children as young as 2 or 3 signal noncomprehension of such messages Participants (Lempers & Elrod, 1983; Revelle, Wellman, & The participants were 18 adolescents and Karabenick, 1985). In contrast, it is not until the young adults with fragile X syndrome, 22 adoles- age of 4 and up that typically developing children cents and young adults with Down syndrome, and respond to ambiguous messages (i.e., messages for 17 typically developing 3- to 6-year-olds. Partici- which there are multiple, plausible referents) and pants in the syndrome groups were recruited by other more subtle types of problems (Ackerman, advertisements in local newspapers, mailings to lo- Szymanski, & Silver, 1990; Lempers & Elrod, cal educators and administrators of genetics clin- 1983; Revelle et al., 1985). In terms of context, ics, and notices to families enrolled in a university young typically developing children are less likely research registry. The syndrome groups were also to fully analyze the potential referent array, which recruited nationally through postings on Internet is necessary for identifying instances of noncom- websites and listservs and in the newsletters of na- prehension, the larger that array (Glucksberg, tional organizations focused on developmental Krauss, & Higgins, 1975; Roberts & Patterson, disabilities. Typically developing children were re- 1983). The noncomprehension signaling of ado- cruited locally through university research regis- lescents and adults with intellectual disabilities has tries, community postings, and preschools. A few also been found to be influenced by such vari- of the participants were also included in analyses ables, although studies to date have included only reported in Lewis et al. (2006). samples that were heterogeneous with respect to Sample characteristics are summarized in Ta- etiology and etiology was not a variable of interest ble 1. Participants were selected to achieve a (Abbeduto et al., 1997, 1998; Fujiki & Brinton, group-wise match across the three groups on non- 1993). verbal mental age (MA), which, as described be- In the present study, we had three goals. The low, was measured using three subtests from the first was to determine the extent and nature of the Stanford-Binet Intelligence Scale, 4th edition delay in noncomprehension signaling for individ- (Thorndike, Hagen, & Sattler 1986). Although the uals with Down syndrome or fragile X syndrome. groups did not differ significantly in nonverbal We were interested in documenting the delay in MA, F(2, 54) ϭ 2.03, p ϭ .14, the match achieved this domain of language use relative to typically was less close than suggested by Mervis and Rob- developing children at similar developmental lev- inson (1999) and, thus, nonverbal MA was used els as well as the possibility of syndrome differ- as a covariate in the primary analyses (see Results ences in the extent or nature of the delay. In ad- section). The syndrome groups also did not differ dressing this goal, we examined noncomprehen- in nonverbal IQ on the Stanford-Binet subtests, sion signaling as a function of type of problematic t(38) ϭ .10, p ϭ .92, or chronological age (CA), message and size of the potential referent array. t(38) ϭ .01, p ϭ 1.00. Although the number of The second goal was to determine how the non- males was greatest in the fragile X syndrome comprehension signaling of individuals with group, the groups did not differ significantly in Down or fragile X syndrome is shaped by their gender composition, ␹2(2, N ϭ 57) ϭ 2.44, p ϭ levels of cognition, language, social cognition, .30. and maladaptive behavior. Our third goal was to No participant had more than a mild hearing explore possible differences between males and fe- loss (i.e., a mean pure tone threshold of 30 dB or males with fragile X syndrome in the extent and worse in the better ear across the frequencies of pattern of their noncomprehension signaling. Few 500, 1000, and 2000 Hz) at the time of testing. studies of language have included both males and No participant met criteria for autism (see Lewis females assessed under comparable task condi- et al., 2006, for details). tions; thus, whether they differ quantitatively or Physician or hospital reports providing DNA qualitatively in language is not clear (Murphy & confirmation of the full mutation were available

216 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Table 1. Participant Characteristics by Diagnostic Group Groupa DS (n ϭ 22) FX (n ϭ 18) TD (n ϭ 17) Characteristic Mean SD Mean SD Mean SD CA (in years) 17.59** 3.68 17.58** 3.35 4.54 .58 S-Bb Nonverbal MA (in years) 5.22 1.06 5.03 1.52 4.50 .70 S-Bb Nonverbal IQ 42.32** 6.38 42.56** 8.42 98.00 6.90 OESc age-equivalent (in years) 4.60* 1.78 8.09 4.34 5.62* .89 TACL–3d age-equivalent (in years) 5.62* 1.47 6.83 1.69 5.95 1.06 Proportion correct false belief .29 .28 .47 .35 .50 .39 No. correct digit sequencese 3.14 2.42 3.33 2.03 4.35 1.93 Child Behavior Checklist Total T scoref 55.09 8.29 57.06 9.67 — No. Caucasians 22 — 16 — 14 No. mothers with college degreeg 15 — 13 — 15 — No. males 12 — 13 — 8 — aDS ϭ Down syndrome, FX ϭ fragile X syndrome, TD ϭ typically developing. bBased on administration of the Pattern Analysis, Copying, and Bead Memory subtests of the Stanford-Binet Intelligence Scale, 4th ed. cOral Expression Scale of the Oral and Written Language Scales. Age-equivalents are missing for 1 participant with DS and 2 TD participants. dTest of Auditory Comprehension of Language–3. eBased on the forward recall portion of the Digit Span subtest of the Wechsler Scales of Intelligence for Children, 3rd edition. fChild Behavior Checklist/4–18. gLevel of maternal education was not available for 1 participant with FX. *Significantly different, p Ͻ .05, from FX. **Significantly different, p Ͻ .05, from TD. for all participants with fragile X syndrome. Two verbal IQs in the intellectual disabilities range. males with fragile X syndrome were mosaic. Par- Moreover, the males and females with fragile X ticipants with Down syndrome all had trisomy 21 syndrome overlapped considerably in the range of according to parental report, which was confirmed nonverbal IQs (37 to 57 and 44 to 56, respective- by physician/hospital records of the karyotype re- ly). Thus, the females in this sample should be sults for most of them. Three families had more seen as representing the more affected end of the than one child with fragile X syndrome partici- continuum of females with fragile X syndrome; pate. that is, the half who meet diagnostic criteria for There was no difference across the groups in intellectual disabilities (Hagerman, 1999). racial composition, ␹2(2, N ϭ 57) ϭ 3.91, p ϭ .14, with 52 of the 57 participants self-identifying as White. The groups also did not differ in ma- Measures Providing Putative Predictors of ternal education, ␹2(2, N ϭ 56) ϭ 2.17, p ϭ .34, Noncomprehension Signaling with 43 of the 56 mothers who provided such data The following measures assessed domains im- indicating that they had a 4-year college degree or portant in recognizing and resolving comprehen- higher. sion failures. The measures were administered to Characteristics of the participants with fragile each participant as part of a more comprehensive X syndrome are presented by gender in Table 2. battery. The measures were administered over sev- The males and females differed on nonverbal IQ, eral sessions, typically on the same day as, or with- t(16) ϭ 4.48, p Ͻ .0005, and nonverbal MA, t(16) in a few days of, the noncomprehension signaling ϭ 5.08, p Ͻ .0005, but not age, t(16) ϭ .00, p ϭ task (described below). 1.00. These differences are consistent with previ- Nonverbal cognition. The Bead Memory, Pat- ous findings of greater affectedness in males than tern Analysis, and Copying subtests of the Stan- females (Hagerman, 1999). Note, however, that ford-Binet, 4th edition were administered. These because of our interest in making comparisons be- subtests, which require a minimum of verbal in- tween nonverbal MA-matched syndrome groups, structions and only nonverbal responses, have the females with fragile X syndrome all had non- been used in previous studies to create cognitively

᭧ American Association on Intellectual and Developmental Disabilities 217 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Table 2. Characteristics of Participants With Fragile X Syndrome by Gender Males (n ϭ 13) Females (n ϭ 5) Characteristic Mean SD Mean SD CA (in years) 17.58 3.64 17.58 2.79 Nonverbal MAa (in years) 4.31 1.05 6.90 .65 Nonverbal IQa 38.77 6.07 52.40 4.83 aBased on administration of the Pattern Analysis, Copying, and Bead Memory subtests of the Stanford-Binet Intelligence Scale, 4th edition. matched comparisons of the three groups includ- receptive language to noncomprehension signaled in the present study (e.g., Abbeduto et al., ing. 2003; Chapman et al., 1991). Nonverbal MA and Theory of mind. We administered a false belief IQ were computed from the three subtests (see task similar to those commonly used to study the- Table 1). Nonverbal MA was the basis of the ory of mind (Yirmiya, Erel, Shaked, & Solomon- group-wise matching and was used to evaluate the ica-Levi, 1998). In this task, the participant is contribution of nonverbal cognitive ability to asked a series of questions about the beliefs of noncomprehension signaling. various story characters after listening to and Auditory short-term memory. The Digit Span watching a story told and enacted with miniature subtest from the Wechsler Scales of Intelligence figures and props. The test questions assess wheth- for Children, 3rd edition (Wechsler, 1991) was ad- er the participant differentiates between his or her ministered. In this subtest, the participant must own true beliefs and the story characters’ false be- immediately repeat verbatim digit sequences spo- liefs. Some questions require reasoning about a ken by the examiner at the rate of one digit per character’s beliefs about an object’s location (first- second. The number of correctly repeated se- order reasoning), whereas others require reasoning quences (see Table 1) was used to evaluate the about a character’s beliefs about yet another char- contribution of auditory short-term memory to acter’s beliefs about an object’s location (second- noncomprehension signaling. order reasoning). Typically developing children Language. We used the Oral Expression Scale succeed at such tasks near the age of 4, which is of the Oral and Written Language Scales (Carrow- when they recognize that the human mind rep- Woolfolk, 1995) to measure expressive skills. This resents rather than copies the world (Tager-Flus- scale measures numerous dimensions of expres- berg, 2001). The proportion of test questions an- sive language, from vocabulary to syntax to dis- swered correctly (see Table 1) was used to evaluate course-level rules. Each item requires the participant to produce a word, phrase, or sentence in the contribution of theory of mind to noncom- response to a verbal prompt from the examiner. prehension signaling. This task is described in de- The prompts are accompanied by drawings. The tail elsewhere (Lewis et al., 2006). age-equivalent score (see Table 1) was used to eval- Maladaptive behavior. The Child Behavior uate the contribution of expressive language abil- Checklist/4-18 (Achenbach, 1991) was completed ity to noncomprehension signaling. for each participant with fragile X syndrome or Receptive language was assessed using the Down syndrome by his or her parent. This infor- Test for Auditory Comprehension of Language-3 mant-report measure has been widely used for de- TACL-3 (Carrow-Woolfolk, 1999). This test in- cades, including for individuals with developmen- cludes items tapping understanding of vocabulary, tal disabilities. We computed a Total Problems T grammatical morphemes, and syntactic rules and score (see Table 1) as well as T scores for each of relations. Each item requires the participant to the five subscales expected to distinguish the fragile point to the one picture from among several al- X and Down syndrome groups: Withdrawn, Anx- ternatives that correctly conveys the meaning of ious/Depressed, Attention Problems, Thought the word, phrase, or sentence spoken by the ex- Problems, and Social Problems. We used T scores aminer. The total test age-equivalent score (see Ta- to evaluate the contribution of maladaptive be- ble 1) was used to evaluate the contribution of havior to the noncomprehension signaling task.

218 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Noncomprehension Signaling Task tions; however, he or she had to signal noncom- Task overview. Participants, who were tested prehension and, thereby, solicit more information individually, played the role of listener, and a sec- to be sure of making a correct referent choice for ond researcher played the role of speaker. The par- the three types of inadequate directions. In addi- ticipant and speaker sat at a table facing each oth- tion to direction type, we also manipulated the er. The participant had an easel book; each page number of potential referents available per page. contained a colorful scene (e.g., a seascape). Each page included two or four potential refer- Moveable magnetic pieces, each with a colored ents. drawing of an object (e.g., a seashell), were situ- Materials. The participant’s book contained ated at the bottom of the page. A magnetic strip 32 pages (2 practice and 30 experimental items). ϫ in the scene could hold one of the pieces. The Each 28 cm 21.5 cm page depicted a back- speaker also had an easel book containing the ground scene into which a potential referent scenes, but with one of the pieces already printed could be placed. Each potential referent was ϫ on the scene. For each page, the speaker produced drawn on a separate 5 cm 5 cm card. Each card a one-sentence direction indicating which piece had a magnet on its back. A magnet large enough the participant should move into the scene (e.g., to hold a single card was located within each ‘‘Put the seashell on the beach’’). The goal was for scene. The potential referents were arrayed in a the participant to make each page identical to the line at the bottom of the page in a single random order. The scenes and potential referents were speaker’s page. The participant could not see the drawn with a standard clip art computer program. speaker’s page. The pages and cards were laminated and com- The speaker’s direction allowed the intended bined into a single spiral-bound book that stood referent to be unambiguously identified on the up on an easel so that each page could be easily pages of the informative condition (e.g., the po- viewed by the participant but was out of view of tential referents on one page were four crayons of the speaker. different colors, including a red one, and the di- Half of the pages for the experimental items rection was ‘‘Put the red crayon on the box’’). The contained two potential referents and half con- directions were less than fully informative for oth- tained four potential referents. Six two-referent er pages. In some cases, the direction referred to and six four-referent pages were selected at ran- a piece that was not available (i.e., the incompat- dom and assigned to the informative direction ible condition). On one page, for example, the condition. The 18 remaining pages were assigned scene was of a dinner plate and place setting, and at random to the incompatible, ambiguous, and the magnetic pieces were drawings of forks of dif- unfamiliar direction condition, with an equal ferent colors. The speaker said, ‘‘Put the black fork number of pages per condition and an equal num- on the plate,’’ but none of the forks was black. In ber of two- and four-referent pages per condition. other cases, the direction did not contain an ad- The speaker’s directions were scripted for each jective to indicate which piece was the one in- experimental item. Each direction was a single- tended (i.e., the ambiguous condition). On one clause imperative directing the participant to page, for example, the scene was of a painting on move a specific potential referent into the scene. an easel; the magnetic pieces contained drawings Each imperative began with a verb such as put, of paint brushes, each dipped in a different color place,ormove. The basic structure of the impera- of paint. The speaker’s direction was ‘‘Put the tive was Verb ϩ Noun Phrase ϩ Prepositional brush on the painting.’’ Finally, the speaker’s di- Phrase. The directions were six to eight words in rection sometimes contained an adjective whose length. meaning was highly unlikely to be known by the A script of speaker responses to possible par- participant (i.e., the unfamiliar condition). On ticipant verbal responses was also created for each one page, for example, the scene was a sky, and speaker direction. The script specified a pragmat- one magnetic piece depicted a blue hot air bal- ically appropriate speaker response for any possi- loon and another piece depicted a yellow hot air ble signal of noncomprehension from the partic- balloon. The speaker’s direction for this page was ipant (see Table 3). The script also specified the ‘‘Place the azure balloon in the sky.’’ Thus, a par- nature of the speaker’s response should the partic- ticipant could immediately select a referent and ipant respond to the direction by doing some- move it into the scene for the informative direc- thing other than signaling noncomprehension or

᭧ American Association on Intellectual and Developmental Disabilities 219 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Table 3. Types of Noncomprehension Signals and the Speaker’s Responses to Them Examples of responses elicited Type of signal Example from speaker Request for The blue hat? Yes, I meant the blue hat. confirmation This one? (plus holds up card for Actually, I meant the red one. speaker to see) Request for What’s russet mean? Russet is a kind of red. definition What’s tawny? Tawny is another word for orange. What’s a ? (uttered with an in- tonation suggesting that completion by the speaker is expected) Request for specific Which one? Sorry, I meant the yellow duck. information Which fork do you mean? I’m sorry . . . I meant the yellow fork. The red jar . . . russet is another word for red (for unfamiliar directions only). Statement of There is no brown book. Actually, I meant the yellow duck. nonexistence There’s not one like that. Sorry, I meant the blue hat. I can’t find that one. Statement of There are four forks. Actually, I meant the yellow one. existence There are lots of those you know. Sorry, I meant the green letter. Other For example, participant holds up a po- Yes, I meant the blue hat. tential referent to show the examiner Actually, I meant the yellow duck. while looking expectantly. moving a potential referent into the scene. Note sure the participant that there were no prohibi- that nonspecific questions such as ‘‘huh?’’ or tions against talking, an instruction that has been ‘‘what?’’ were treated as requests for repetition of found to be important in ensuring the validity of the original direction rather than signals of non- this task (Abbeduto et al., 1997). comprehension based on previous studies dem- The two practice trials, each of which in- onstrating that such questions most often elicit a volved an informative direction, immediately fol- repetition of the original utterance rather than lowed the reading of the task instructions by the new information from speakers in natural conver- examiner. For the practice items, the participant sation (e.g., Garvey, 1977). and speaker were allowed to compare their pages Two versions of the easel book were created, after each direction with either positive or correc- each comprised of a different random order of the tive feedback provided as necessary. No corrective 30 experimental items. Participants were randomly feedback or opportunity to compare pages was assigned to one or the other order. provided after the practice trials, although non- Procedure. In explaining the task, the examiner contingent general praise (e.g., ‘‘I like how you’re stressed the need to listen carefully and achieve listening’’) was delivered according to a script an exact match with the speaker. The examiner throughout the task. The examiner’s participation also explained that ‘‘you can talk with ࿞࿞࿞ [speaker in the remainder of the task was minimal. name], ask him/her questions, or say anything to The speaker looked at his or her own book him/her. You need to make sure your pictures when producing each direction and maintained match.’’ This latter instruction was designed to as- that focus until the participant had either signaled

220 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. noncomprehension or moved a potential referent informative directions in a 3 (diagnostic group) ϫ into the scene. Eye contact between participant 2 (number of potential referents) ANCOVA. and speaker was thereby avoided so as not to con- Number of potential referents was a within-partic- vey the impression that the speaker was necessar- ipant variable and nonverbal MA was the covari- ily expecting a verbal response from the partici- ate. Only the main effect of diagnostic group was pant. The speaker responded verbally to signals of significant, F(2, 53) ϭ 6.46, p ϭ .003, ␩2 ϭ .20; noncomprehension according to the script. Re- however, even the participants with Down syn- sponses in which the participant simply moved a drome, who were the lowest performing group, potential referent into the scene received no re- did well, selecting the correct referent on a covar- sponse from the speaker. iate-adjusted mean of 5.55 across the two- and The examiner scored the participant’s re- four-referent informative directions. The covari- sponses to the speaker’s directions as they oc- ate-adjusted means for the fragile X syndrome and curred, noting which potential referent was select- typically developing groups were 5.99 and 6.01, ed and whether a signal of noncomprehension respectively. Thus, the participants understood was produced and transcribing any signal pro- the task and could process the linguistic forms duced. The entire session was also audio- and vid- used for the directions, although those with eotaped so that the accuracy of the examiner’s no- Down syndrome were somewhat less capable in tations with regard to the occurrence and tran- this regard. scription of noncomprehension signaling could be checked for accuracy. Scoring noncomprehension signals. Each page of Diagnostic Group Comparisons of the book was scored for the presence or absence Noncomprehension Signaling of a signal of noncomprehension by the partici- The number of trials on which a signal of pant. The types of participant responses that were noncomprehension was produced was analyzed in ϫ ϫ scored as signals of noncomprehension are illus- a 3 (diagnostic group) 4 (direction type) 2 trated in Table 3. We did not distinguish between (number of potential referents) ANCOVA. Direc- responses containing multiple versus single signals tion type and number of potential referents were of noncomprehension. Interrater agreement, cal- within-participant variables, and the covariates culated for 9 participants (3 per diagnostic group), were nonverbal MA and number of correct refer- was found to be 100% for the occurrence of a ent selections on informative directions. We em- signal of noncomprehension. ployed the Greenhouse-Geisser adjustment for all Checks on task materials/manipulations. Addi- significance tests involving the within-participant tional data were collected either from the partici- variables to control for violations of the sphericity pants or from different participants in the devel- assumption. Because there were twice as many in- opmental range of interest in pilot studies to en- formative directions as any other direction type, sure that (a) participants understood the nouns the number of signals produced in response to used in the directions, (b) the meanings of the informative signals was divided by two to ensure adjectives used in the unfamiliar directions were that the scale was constant across conditions. not known by the participants, (c) the shorter Note that covariate-adjusted means are presented length of the ambiguous directions (i.e., in con- throughout. The adjusted means and standard er- trast to the other inadequate directions, they rors for each condition for each diagnostic group lacked an adjective) had no impact on the non- are presented in Table 4. comprehension signaling of the participants, and We found a main effect of diagnostic group, (d) all participants understood the color adjectives F(2, 52) ϭ 6.60, p ϭ .003. Using Fisher’s LSD used and could discriminate among the colors (Levin, Serlin, & Seaman, 1994), we found that used. (Details are available from the first author.) post-hoc comparisons of the diagnostic groups indicated that the typically developing participants Results produced significantly more signals of noncomprehension than did participants in either of the Correct Referent Selections for Informative syndrome groups, who did not differ from each Directions other. The covariate-adjusted mean number of sig- We analyzed the number of correct (i.e., in- nals (elicited by the three directions per each com- tended) referent selections in response to the 12 bination of direction type and number of refer-

᭧ American Association on Intellectual and Developmental Disabilities 221 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Table 4. Frequency of Noncomprehension Signaling: Covariate-Adjusted Means and Standard Errors (SEs) Groupa DS (n ϭ 22) FX (n ϭ 18) TD (n ϭ 17) Condition Mean SE Mean SE Mean SE Informative directions Two-referent arrays .04 .05 .00 .05 .16 .05 Four-referent arrays .06 .06 .06 .06 .15 .06 Incompatible directions Two-referent arrays .93 .31 1.59 .32 2.23 .34 Four-referent arrays .99 .31 1.60 .32 2.32 .34 Ambiguous directions Two-referent arrays .54 .31 1.00 .32 2.00 .34 Four-referent arrays .60 .29 .78 .30 2.23 .32 Unfamiliar directions Two-referent arrays .56 .27 .70 .28 1.77 .30 Four-referent arrays .45 .28 .74 .29 2.22 .31 aDS ϭ Down syndrome, FX ϭ fragile X syndrome, TD ϭ typically developing. ents) was .52, .81, and 1.64 for the Down syn- typically developing children on both the unfa- drome, fragile X syndrome, and typically devel- miliar and the ambiguous directions. In the case oping groups, respectively. The effect of of incompatible directions, the marginally signif- diagnostic group was moderate to large, ␩2 ϭ .20, icant finding reflected the fact that the partici- according to the guidelines proposed by Cohen, pants with Down syndrome produced fewer sig- Cohen, West, and Aiken (2003). nals of noncomprehension on the incompatible We also found a significant Diagnostic Group directions than did the typically developing par- ϫ Direction Type interaction, F(4.1, 106.7) ϭ ticipants. None of the comparisons between the 4.24, p ϭ .003, with an ␩2 of .14, indicating a two syndrome groups were significant. small to moderate effect size. Simple effects tests The interaction of diagnostic group and num- ϭ were conducted to examine differences among di- ber of referents was also significant, F(2, 53) ϭ ␩2 ϭ agnostic groups separately for each type of direc- 3.46, p .04, .12. Simple effects tests were tion using the Holm sequential procedure to pre- conducted to examine differences among diag- vent inflation of Type I error (Holm, 1979; Levin, nostic groups separately for the two-referent and four-referent condition using the Holm sequential Serlin, & Seaman, 1994). This procedure required procedure. This procedure required the largest F the largest F for diagnostic group to reach an al- ␣ for diagnostic group to reach an alpha of .025 for pha of .013 for significance (i.e., /4); the next significance (i.e., ␣/2) and the smallest to reach an largest F, an alpha of .017 (␣/3); the next largest, ␣ alpha of .05. The effect of diagnostic group was an alpha of .025 ( /2); and the smallest, an alpha significant for both the four-referent, F(2, 52) ϭ of .05. The effect of diagnostic group was signif- 7.88, p ϭ .001, and two-referent, F(2, 52) ϭ 5.24, ϭ icant for unfamiliar directions, F(2, 52) 7.62, p p ϭ .008, conditions. Post-hoc comparisons (using ϭ ϭ .001, and ambiguous directions, F(2, 52) Fisher’s LSD technique) indicated that the partic- ϭ 6.39, p .003, but just failed to reach significance ipants with fragile X syndrome and those with for incompatible directions, F(2, 52) ϭ 3.93, p ϭ Down syndrome produced fewer signals of non- .026. The effect of diagnostic group was not sig- comprehension than did the typically developing nificant for the informative directions. Post-hoc children on both the two- and four-referent con- comparisons (using Fisher’s LSD technique) in- ditions, although the magnitude of the difference dicated that the participants with fragile X syn- between groups was greater on the four- than the drome and those with Down syndrome produced two-referent arrays. None of the comparisons be- fewer signals of noncomprehension than did the tween the two syndrome groups were significant.

222 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Comparisons of Males and Females With noncomprehension signaling and the various pre- Fragile X Syndrome dictors. The dependent measure was created by As noted previously, the participants with first computing the proportion of inadequate (i.e., fragile X syndrome were virtually perfect in se- incompatible, ambiguous, and unfamiliar) direc- lecting the correct referent in response to infor- tions that elicited signals of noncomprehension mative directions, with a covariate-adjusted mean from a participant and then subtracting the pro- of 5.99 across the two- and four-referent arrays. portion of informative directions that elicited Thus, we did not perform any statistical analyses such signals from the participant, thereby con- involving gender and this dependent measure. trolling for any indiscriminate signaling. This pro- The number of signals of noncomprehension portional dependent variable was subjected to an produced by the participants with fragile X syn- arcsine transformation prior to analysis. Because drome was analyzed in a 2 (gender) ϫ 4 (direction the number of predictors was large relative to the type) ϫ 2 (number of potential referents) AN- sample size, we adopted a conservative approach COVA. Direction type and number of potential to model building, limiting the number of predic- referents were within-participant variables, and tors while still testing whether the relationships nonverbal MA was the covariate. (Number of cor- among variables differed across diagnostic groups rect referent selections on informative directions (see Abbeduto et al., 2006, for a similar approach). was not included as a covariate because there was The regression analysis proceeded in three no variability on this measure.) Again, the number steps. In Step 1, we entered two dummy variables of signals produced for informative directions was to represent the three diagnostic groups (Cohen divided by 2. et al., 2003). The first dummy variable indexed We found that males with fragile X syndrome whether the participant had Down syndrome and signaled noncomprehension less often than did the second, whether the participant had fragile X ϭ ϭ females with the syndrome, F(1, 15) 3.46, p syndrome. A significant coefficient for a dummy ␩2 ϭ .04 (one-tailed), .19 (i.e., a moderate effect variable indicated that the indexed group and the size). The (covariate-adjusted) mean frequencies typically developing group differed on the depen- for males and females were .51 and 1.72, respec- dent variable. In Step 2, nonverbal MA, TACL-3 tively. No other main or interaction effects were age-equivalent score, Oral Expression Scale age- significant. equivalent score, proportion correct on the false We also re-ran the Diagnostic Group ϫ Di- ϫ belief task, and number of correctly recalled se- rection Type Number of Potential Referents quences on the digit span task were entered si- ANCOVA described in the previous section, ex- multaneously. We also entered gender at Step 2 cluding females with fragile X syndrome, but in- because of the gender differences in the frequency cluding the females with Down syndrome or typ- of noncomprehension signaling among the fragile ical development because there was no reason to X syndrome participants described in the preced- expect gender differences in the latter two groups. Our aim was to determine whether the diagnostic ing section. In Step 3, we entered interactions be- group differences in noncomprehension signaling tween the predictors that were significant at Step and the interactions of diagnostic group with di- 2 and the dummy variables indexing the diagnos- rection type and with number of potential refer- tic groups. Each interaction was represented by ents had somehow been distorted by the inclusion the product of the dummy variable and predictor of the females with fragile X syndrome in that variable. A significant coefficient for an interac- analysis. Although the means of the fragile X and tion term indicated that the relationship between Down syndrome groups were even more similar the significant Step 2 predictor and the dependent to each other in the re-analysis than in the pri- variable differed between the diagnostic group in- mary analysis that included females with fragile X dexed by the dummy variable and the typically syndrome, the pattern of statistical significance developing group. This approach to testing inter- was unchanged. (The results of this re-analysis are actions reflected the assumption that the interac- available from the first author.) tions would be ordinal (i.e., the relationships would vary in strength but not direction across Examination of Putative Predictors of groups), which means that significant interactions Noncomprehension Signaling were likely to be associated with significant main Multiple regression was used to examine the effects. We expected the predictors to be related relationship between a measure of ‘‘appropriate’’ positively (or not at all) to the dependent variable

᭧ American Association on Intellectual and Developmental Disabilities 223 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. and, thus, main effects were evaluated with one- ile X syndrome signaled noncomprehension on tailed tests. We used two-tailed tests to evaluate only 30% of the inadequate directions they heard all interactions. This analytical strategy allowed us compared to 70% for the typically developing to assess potential sources of both within- and be- comparison children. Moreover, the advantage of tween-group differences in appropriate noncom- the typically developing children over the two prehension signaling. syndrome groups in noncomprehension signaling The Step 1 model indicated that the two was evident for both the two- and four-referent dummy variables indexing the group contrasts arrays. At the same time, both syndrome groups were significant, F(2, 51) ϭ 6.78, p ϭ .002. The were virtually perfect in selecting the intended ref- addition of the six predictors (main effects) at Step erent for informative directions, suggesting that 2 was associated with a significant change in the their low frequencies of noncomprehension sig- R2, F(6, 45) ϭ 2.47, p ϭ .04; however, the only naling did not arise solely from limitations in their predictor of the six that yielded a significant beta linguistic knowledge or ability to process the spe- at Step 2 was the TACL age-equivalent score, ␤ cific language forms used in the noncomprehen- ϭ .47, t ϭ 1.93, p ϭ .03 (one-tailed). The betas sion signaling task. This claim is further supported for the dummy variables representing the two by the fact that the typically developing children group contrasts remained significant at Step 2. At did not differ from either syndrome group in their Step 3, we entered the two interaction terms for scores on the TACL-3, which is a standardized group and TACL age-equivalent score. The addi- test of receptive language. Thus, the findings sug- tion of the interaction terms did not lead to a gest that youth with fragile X syndrome or Down significant change in the R2, although the Step 3 syndrome are (a) poor at monitoring their com- model yielded an adjusted R2 of .28, F(10, 43) ϭ prehension, thereby failing to recognize when 3.03, p ϭ .005. The only predictor associated with they do not understand; and/or (b) unable to cre- a significant beta at Step 3 was the TACL age- ate and execute a plan for soliciting corrective in- equivalent score, ␤ϭ.64, t ϭ 1.81, p ϭ .04 (one- formation from the speaker, thereby allowing de- tailed). Neither group contrast was significant in tected problems to go uncorrected. It may also be the Step 3 model. that these youth view the noncomprehension sig- We also conducted regression analyses ex- nal as a sign of their failure rather than the failure amining the relationship between Child Behavior of the message and thus refrain from signaling non- Checklist scores and the arcsine-transformed mea- comprehension so that they, like the adults with sure of appropriate noncomprehension signaling. mild intellectual disabilities studied by Edgerton These analyses involved only the two syndrome (1993), can assume a ‘‘cloak of competence.’’ In groups and were conducted in the same way as any event, given the likely negative consequences the regression described previously, with one anal- for everyday comprehension in a host of settings, ysis including the Child Behavior Checklist Total from school to informal conversations with peers score as predictor and the other analysis including and care providers, it is important that noncom- the five subscale scores as predictors. Scores on prehension signaling be a target of intervention. the Child Behavior Checklist did not contribute Unfortunately, there have been few attempts to to prediction in either analysis. develop effective interventions in this area (see Ezell & Goldstein, 1991, for an exception). We also found that differences in noncom- Discussion prehension signaling between the typically developing children and the syndrome groups were less Diagnostic Group Differences and Similarities pronounced for incompatible directions than for in Noncomprehension Signaling ambiguous or unfamiliar directions. Of the three Our first goal was to determine the extent and types of inadequate directions, the problem ex- nature of the delay in noncomprehension signal- emplified in the incompatible directions is the ing for individuals with Down syndrome or fragile most salient and least difficult to resolve and, X syndrome. We found that both syndrome thus, is the first to be dealt with successfully by groups were less likely to signal noncomprehen- young typically developing children (e.g., Lem- sion of inadequate directions than were the MA- pers & Elrod, 1983; Revelle et al., 1985). Thus, matched typically developing children. In fact, on although noncomprehension signaling is quite de- average individuals with Down syndrome or frag- layed in fragile X and Down syndrome, it is not

224 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. qualitatively different compared to typically de- only scores on the TACL-3, which was designed veloping children. to measure the ability to understand a range of The two syndrome groups did not differ sig- lexical items and grammatical elements, patterns, nificantly in their rates of noncomprehension sig- and rules, made a unique contribution to the de- naling in any comparison. The only marginal dif- pendent measure, with higher TACL-3 scores be- ference occurred for incompatible directions, on ing related to more frequent noncomprehension which only the Down syndrome participants signaling. Moreover, the relation between the tended to differ from the typically developing par- TACL-3 and noncomprehension signaling was ticipants. Despite the generally similar levels of similar across the two syndrome groups. performance exhibited by the two syndrome It is not surprising that knowledge of word groups, it is unlikely that their low rates of non- meanings and syntactic structure would have an comprehension signaling are simply a manifesta- impact on noncomprehension signaling; for ex- tion of their having intellectual disabilities. This ample, the ability to parse the syntactic structure conclusion is suggested by the fact that several of a sentence is likely to facilitate recognition of researchers using a variety of methodologies have the source of noncomprehension and creation of not found differences in noncomprehension sig- a plan for soliciting precisely the sort of linguistic naling rates between youth with mental retarda- information needed to resolve the problem. Var- tion who are heterogeneous with respect to etiol- iation in noncomprehension signaling, however, ogy and MA-matched typically developing chil- cannot be reduced simply to differences in recep- dren (Abbeduto et al., 1997; Abbeduto et al., tive lexical and syntactic ability. The syntax and 1998; Ezell & Goldstein, 1991; Fujiki & Brinton, vocabulary of the directions in the noncompre- 1993; Rueda & Chan, 1980; however, see Abbe- hension signaling task were quite simple and cho- duto et al., 1991, for an exception). Additional sen so that they would be within the competence studies in which investigators include a nonspe- of the participants. In fact, all three groups dis- cific intellectual disabilities comparison in addi- played a high level of skill in processing infor- tion to the two syndrome groups (Dykens et al., mative directions, which were syntactically iden- 2000), however, are needed to verify this claim. If tical to the inadequate directions. Moreover, the differential impairments in noncomprehension participants with fragile X syndrome were poor at signaling are demonstrated for individuals with signaling noncomprehension despite TACL-3 fragile X syndrome or Down syndrome, on the scores that exceeded those of the participants with one hand, and individuals with other forms of Down syndrome and that were not significantly intellectual disabilities, on the other hand, it different from those of the typically developing would be important to determine the causes of children. Thus, noncomprehension signaling re- those differences, with the possibilities including quires skills that extend beyond linguistic knowl- not only the genetic and psychological character- edge. The implication of this conclusion is that istics of the individuals, but their environmental traditional language intervention, which is fo- histories as well. From a clinical perspective, such cused largely on teaching new language targets a finding would suggest the need to develop dif- (Brady & Warren, 2003), is likely to have rather ferent types of language interventions, or at least minimal impacts on comprehension monitoring interventions with different targets, for individuals and the other skills needed to engage in effective with these two syndromes compared to individu- noncomprehension signaling. Future researchers als with intellectual disabilities of other origins. should focus on evaluating more broadly con- ceived language interventions. Contrary to expectations, individual differ- Putative Predictors of Noncomprehension ences in nonverbal cognition, social cognition, au- Signaling ditory memory, and maladaptive behavior did not Our second goal was to determine how non- make unique contributions to noncomprehension comprehension signaling is shaped by an individ- signaling. This may reflect the small sample size ual’s levels of nonverbal cognition, language, the- or limitations of the measures; however, we have ory of mind, auditory memory, and maladaptive also failed to find contributions from these do- behavior and whether there are differences be- mains to other aspects of social communication tween the syndromes and MA-matched typically in these syndromes or in typically developing chil- developing children in this regard. We found that dren (Abbeduto et al., 2006), suggesting that, at

᭧ American Association on Intellectual and Developmental Disabilities 225 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. least in the developmental range of our partici- Differences in Noncomprehension Signaling by pants (approximately 3 to 7 years), nonverbal cog- Males and Females With Fragile X Syndrome nition, social cognition, auditory memory, and Our third goal in the study was to explore maladaptive behavior do not constrain acquisition possible gender differences in noncomprehension or use of the skills underlying noncomprehension signaling among those with fragile X syndrome. signaling. All four domains, however, have been Despite including only a small number of females found to impact other aspects of language at other with this syndrome, we found that they were more points in development for individuals with devel- likely to signal noncomprehension than were their opmental disabilities as well as for typically de- male counterparts. This difference emerged developing children (e.g., Belser & Sudhalter, 1995; spite statistical equation of males and females on McDuffie, Chapman, & Abbeduto, in press; Tag- nonverbal MA through ANCOVA and despite er-Flusberg, 2001). Together, such findings under- the fact that we included only those females score the importance of expanding language as- whose nonverbal IQs placed them among the sessment to include domains such as auditory lower half of females with fragile X syndrome (i.e., memory while tailoring the focus of the assess- those meeting criteria for an intellectual disabili- ment of nonlinguistic domains to the child’s de- ty). Moreover, even the males with fragile X syn- velopmental level and the profile of language drome were virtually perfect in selecting the in- problems. In terms of typical development, the tended referent for informative directions. Thus, findings suggest that claims about the pervasive the difference in noncomprehension signaling be- role played by theory of mind and auditory mem- tween males and females was not due simply to ory in communication will need to be more con- the within-syndrome variations in cognitive and strained and nuanced, recognizing that the role of linguistic ability that are inherently correlated skills in these domains will vary across develop- with gender. This is consistent with recent find- mental periods and different facets of communi- ings concerning various aspects of expressive lan- cation. guage, including the verbal perseveration that is We note, however, that we evaluated non- so characteristic of individuals with fragile X syn- comprehension signaling under ‘‘ideal’’ condi- drome (Murphy & Abbeduto, 2007). Such find- tions; that is, testing occurred in a quiet room ings suggest that differences in the behavioral phe- with a skilled adult partner, with few distractions, notypes of males and females with fragile X syn- and without other people involved or available to drome are not simply quantitative in nature, re- participate in the interaction. Of course, this is a flecting differences in severity; instead, different very different context than is true of most of the profiles of relative strengths and weaknesses may daily experience of a youth with Down syndrome characterize males and females with the syndrome or fragile X syndrome (e.g., in school). The do- as well. Indeed, the rate of noncomprehension sig- mains of nonverbal cognition, social cognition, naling in females with fragile X syndrome ap- auditory memory, and maladaptive behavior proached that observed for the typically devel- might make more important contributions to oping matches, suggesting that below-MA rates of noncomprehension signaling in these everyday noncomprehension signaling characterize largely contexts than was observed in this study. Future males with the syndrome. It will be an important researchers, therefore, should examine noncom- task for future researchers to more completely prehension signaling and its determinants in a characterize the language phenotypes of males range of contexts, including laboratory-based tasks and females, which will require studies using tasks in which the demands on nonverbal cognition, and measures that allow direct comparison of per- social cognition, auditory memory, and the con- formance across the two (Murphy & Abbeduto, straints on the occurrence of maladaptive behav- 2003). ior are systematically manipulated. Moreover, we included only very broad summary measures of nonverbal cognition and language ability and Conclusion measures of rather narrow ‘‘slices’’ of theory of mind and auditory memory. Measurement of a We must acknowledge three limitations of the different set of skills from each of these domains study. First, the sample size, especially with regard might lead to different conclusions and should be to the comparison of males and females with frag- evaluated. ile X syndrome, was small; thus, there is a need

226 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al. to replicate the findings and to do so with larger retardation. American Journal on Mental Retar- samples. Second, although we examined noncom- dation, 95, 551–562. prehension signaling in relation to other dimen- Abbeduto, L., Murphy, M. M., Cawthon, S. W., sions of behavioral functioning, we assessed only Richmond, E. K., Weissman, M. D., Kara- narrowly within some domains of interest and dottir, S., & O’Brien, A. (2003). Receptive lan- with only a single method of measurement per guage skills of adolescents and young adults domain. Moreover, other aspects of the behavior- with Down or fragile X syndrome. American al phenotypes of the two syndromes distinguish Journal on Mental Retardation, 108, 149–160. them and may impact the signaling of noncom- Abbeduto, L., Murphy, M. M., Richmond, E. K., prehension (e.g., visual search). Third, we includ- Amman, A., Beth, P., Weissman, M. D., Kim, ed only individuals who did not also have an au- J.-S., Cawthon, S. W., & Karadottir, S. (2006). tism diagnosis, which means that we omitted as Collaboration in referential communication: many as one fourth of the population of persons Comparison of youth with Down syndrome with fragile X syndrome (Demark, Feldman, & or fragile X syndrome. American Journal on Holden, 2003). There is evidence that these two Mental Retardation, 111, 170–183. subgroups differ in both their degree of impair- Abbeduto, L., & Rosenberg, S. (1980). The com- ment, with the co-morbid subgroup being lower municative competence of mildly retarded functioning on average, and in their profiles of adults. Applied Psycholinguistics, 1, 405–426. behavioral strengths and weaknesses (Hepburn, Abbeduto, L., & Short-Meyerson, K. (2002). Lin- Hayes, Hagerman, & Rogers, 2004; Lewis et al., guistic influences on social interaction. In H. 2006). It will be interesting to determine whether Goldstein, L. Kaczmarek, & K. M. English variations in noncomprehension signaling also (Eds.), Promoting social communication in chil- distinguish between individuals with fragile X syn- dren and youth with developmental disabilities drome with and without a co-morbid diagnosis of (pp. 27–54). Baltimore: Brookes. autism. Abbeduto, L., Short-Meyerson, K., Benson, G., & Despite these limitations, the present study Dolish, J. (1997). Signaling of noncompre- has yielded new, clinically important data about hension by children and adolescents with the serious challenges facing youth with fragile X mental retardation. Effects of problem type syndrome or Down syndrome as listeners. Indeed, and speaker identity. Journal of Speech, Lan- the failure to signal noncomprehension can have guage, and Hearing Research, 40, 20–32. pervasive negative effects in an interaction, as mis- Abbeduto, L., Short-Meyerson, K., Benson, G., understandings that are not resolved are sure to Dolish, J., & Weissman, M. (1998). Under- be compounded as the interaction progresses. standing referential expressions: Use of com- Moreover, when coupled with, as displayed in this mon ground by children and adolescents with study, a tendency to select a referent despite inadequacies in the messages heard, listeners with mental retardation. Journal of Speech, Language, these syndromes are likely to ‘‘guess’’ wrong and and Hearing Research, 41, 348–362. say or do things that are inappropriate to the Achenbach, T. M. (1991). Child Behavior Check- speaker’s intent, leading to further breakdowns in list/4–18. Burlington: University of Vermont. the interaction. This study has also demonstrated Ackerman, B. P. (1993). Children’s understanding the advantage of including both another syn- of the speaker’s meaning in referential com- drome group and a typically developing compar- munication. Journal of Experimental Child Psy- ison group when evaluating behavioral aspects of chology, 55, 56–86. syndrome phenotypes and of comparing males Ackerman, B. P., Szymanski, J., & Silver, D. and females with fragile X syndrome directly using (1990). Children’s use of the common ground the same measures. in interpreting ambiguous referential utteranc- es. Developmental Psychology, 26, 234–245. Bedrosian, J. L., & Prutting, C. A. (1978). Com- References municative performance of mentally retarded Abbeduto, L., Davies, B., Solesby, S., & Furman, adults in four conversational settings. Journal L. (1991). Identifying the referents of spoken of Speech and Hearing Research, 21, 79–95. messages: The use of context and clarification Belser, R. C., & Sudhalter, V. (1995). Arousal dif- requests by children with and without mental ficulties in males with fragile X syndrome: A

᭧ American Association on Intellectual and Developmental Disabilities 227 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

preliminary report. Developmental Brain Dys- Demark, J. L., Feldman, M. A., & Holden, J. J. A. function, 8, 270–279. (2003). Behavioral relationship between au- Brady, N. C., & Warren, S. F. (2003). Language tism and fragile X syndrome. American Journal interventions for children with mental retar- of Medical Genetics, 108, 314–326. dation. In L. Abbeduto (Ed.), International re- Dykens, E. M., Hodapp, R. M., & Finucane, B. view of research in mental retardation (Vol. 27, M. (2000). Genetics and mental retardation syn- pp. 231–253). New York: Academic Press. dromes: A new look at behavior and interventions. Bregman, J. D., Leckman, J. F., & Ort, S. I. (1988). Baltimore: Brookes. Fragile X syndrome: Genetic predisposition to Dykens, E. M., Hodapp, R. M., & Leckman, J. F. psychopathology. Journal of Autism and De- (1989). Adaptive and maladaptive functioning velopmental Disorders, 18, 343–354. of institutionalized and noninstitutionalized Carrow-Woolfolk, E. (1995). Oral and Written fragile X males. Journal of the American Acad- Language Scales. Allen, TX: DLM Teaching emy of Child and Adolescent Psychiatry, 26, 427– Resources. 430. Carrow-Woolfolk, E. (1999). Test for Auditory Edgerton, R. B. (1993). The cloak of competence: Stig- Comprehension of Language–3. Allen, TX: ma in the lives of the mentally retarded (rev. and DLM Teaching Resources. updated ed.). Berkeley: University of Califor- Chapman, R. (2003). Language and communica- nia Press. tion in individuals with DS. In L. Abbeduto Ezell, H. K., & Goldstein, H. (1991). Observation- (Ed.), International review of research in mental al learning of comprehension monitoring retardation (Vol. 27, pp. 1–34). New York: Ac- skills in children who exhibit mental retarda- ademic Press. tion. Journal of Speech and Hearing Research, 34, Chapman, R. S., & Hesketh, L. J. (2000). The be- 141–154. havioral phenotype of individuals with Down Fujiki, M., & Brinton, B. (1993). Comprehension syndrome. Mental Retardation and Developmen- monitoring skills of adults with mental retar- tal Disabilities Research Reviews, 6, 84–95. dation. Research in Developmental Disabilities, Chapman, R., Schwartz, S., & Kay-Raining Bird, 14, 409–421. E. (1991). Language skills of children and ad- Garner, C., Callias, M., & Turk, J. (1999). Exec- olescents with DS: I. Comprehension. Journal utive function and theory of mind perfor- of Speech, Language, and Hearing Research, 34, mance of boys with fragile-X syndrome. Jour- 1106–1120. nal of Intellectual Disability Research, 43, 466– Clark, H. H. (1996). Using language. New York: 474. Cambridge University Press. Garvey, C. (1977). The contingent query: A de- Clark, H. H., & Schaefer, E. F. (1989). Contrib- pendent act in conversation. In M. Lewis & uting to discourse. Cognitive Science, 13, 259– L. A. Rosenblum (Eds.), Interaction, conversa- 294. tion, and the development of language (pp. 63– Coggins, T. E., & Stoel-Gammon, C. (1982). Clar- 93). New York: Wiley. iﬁcation strategies used by four Down’s syn- Golinkoff, R. M. (1986). ‘‘I beg your pardon’’: drome children for maintaining normal con- The preverbal negotiation of failed messages. versational interaction. Education and Training Journal of Child Language, 13, 455–476. of the Mentally Retarded, 17, 65–67. Grant, C. M., Apperly, I., & Oliver, C. (2007). Is Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. theory of mind understanding impaired in (2003). Applied multiple regression/correlation for males with fragile X syndrome? Journal of Ab- the behavioral sciences. Mahwah, NJ: Erlbaum. normal Child Psychology, 35, 18–28. Cornish, K., Burack, J. A., Rahman, A., Munir, F., Hagerman, R. J. (1999). Neurodevelopmental disor- Russo, N., & Grant, C. (2005). Theory of ders. Oxford: Oxford University Press. mind deﬁcits in children with fragile X syn- Hodapp, R. M. & Fidler, D. J. (1999). Special ed- drome. Journal of Intellectual Disability Research, ucation and genetics: Connections for the 49, 372–378. 21st century. Journal of Special Education, 33, Cornish, K., Sudhalter, V., & Turk, J. (2004). At- 130–137. tention and language in fragile X. Mental Re- Holm, S. (1979). A simple sequentially rejective tardation and Developmental Disabilities Research multiple test procedure. Scandinavian Journal Reviews, 10, 11–16. of Statistics, 6, 65–70.

228 ᭧ American Association on Intellectual and Developmental Disabilities VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

Lempers, J. D., & Elrod, M. M. (1983). Children’s dation syndromes: The role of joint attention. appraisal of different sources of referential Infants and Young Children, 18, 47–59. communicative inadequacies. Child Develop- Murphy, M. M., & Abbeduto, L. (2007). Gender ment, 54, 509–515. differences in repetitive language in fragile X Levin, J. R., Serlin, R. C., & Seaman, M. A. syndrome. Journal of Intellectual Disability Re- (1994). A controlled, powerful multiple-com- search, 57, 387–390. parison strategy for several situations. Psycho- Philofsky, A, Hepburn, S. L., Hayes, A., Hager- logical Bulletin, 115, 153–159. man, R., & Rogers, S. J. (2004). Linguistic and Lewis, P., Abbeduto, L., Murphy, M. M., Giles, cognitive functioning and autism symptoms N., Richmond, E. K., Bruno, L., & Schroeder, in young children with fragile X syndrome. S. (2006). Cognitive, language, and social-cog- American Journal on Mental Retardation, 109, nitive skills of individuals with fragile X syn- 208–218. drome with and without autism. Journal of In- Revelle, G. L., Wellman, H. M., & Karabenick, J. tellectual Disability Research, 50, 532–545. D. (1985). Comprehension monitoring in pre- Marcell, M. M., & Weeks, S. L. (1988). Short-term school children. Child Development, 56, 654– memory difficulties and Down’s syndrome. 663. Journal of Mental Deficiency Research, 32, 153– Robinson, E. J., & Whittaker, S. J. (1985). Learn- 162. ing about verbal referential communication in Mazzocco, M. M., Baumgardner, T., Freund, L. the early school years. In K. Durkin (Ed.), S., & Reiss, A. L. (1998). Social functioning Language development during the school years among girls with fragile X or Turner syn- (pp. 155–171). London: Croom Helm. drome and their sisters. Journal of Autism and Rosenberg, S., & Abbeduto, L. (1993). Language Developmental Disorders, 28, 509–517. and communication in mental retardation: Devel- Mazzocco, M. M. M., Pennington, B., & Hager- opment, processes, and intervention. Hillsdale, man, R. J. (1993). The neurocognitive phe- NJ: Erlbaum. notype of female carriers of fragile X: Further Seung, H.-K., & Chapman, R. (2000). Digit span evidence for specificity. Journal of Development in individuals with Down syndrome and in and Behavioral Pediatrics, 14, 328–335. typically developing children: Temporal as- McDuffie, A., Chapman, R. S., & Abbeduto, L. pects. Journal of Speech, Language, and Hearing (in press). Language development in adoles- Research, 43, 609–620. cents and young adults with Down syndrome Shatz, M. (1983). Communication. In J. Flavell & or fragile X syndrome. In J. E. Roberts, S. F. E. M. Markman (Eds.), Handbook of child psy- Warren, & R. S. Chapman (Eds.), Speech and chology (pp. 841–889). New York: Wiley. language development and interventions in Down Simon, J. A., Keenan, J. M., Pennington, B. F., syndrome and fragile X syndrome. Baltimore: Taylor, A. K., & Hagerman, R. J. (2001). Dis- Brookes. course processing in women with fragile X Mervis, C. B., & Robinson, B. F. (1999). Meth- syndrome: Evidence for a deficit establishing odological issues in cross-syndrome compari- coherence. Cognitive Neuropsychology, 18, 1– sons: Matching procedures, sensitivity (Se), 18. and specificity (Sc). Commentary on Sigman, Tager-Flusberg, H. (2001). A reexamination of the M., & Ruskin, E. Continuity and change in theory of mind hypothesis of autism. In J. A. the social competence of children with au- Burack, T. Charman, N. Yirmiya, & P. R. Ze- tism, Down syndrome, and developmental lazo (Eds.), The development of autism: Perspec- delays. Monographs of the Society for Research in tives from theory and research. (pp. 173–193). Child Development, 64(No. 256), 115–130. Mahwah, NJ: Erlbaum. Murphy, M. M., & Abbeduto, L. (2003). Language Thorndike, R. L., Hagen, E. P., & Sattler, J. M. and communication in fragile X syndrome. In (1986). Stanford-Binet Intelligence Scale: Fourth L. Abbeduto (Ed.), International review of re- edition. Chicago: Riverside. search in mental retardation (Vol. 27, pp. 83– van der Meij, H. (1988). Constraints on question 119). New York: Academic Press. asking. Journal of Educational Psychology, 80, Murphy, M. M., & Abbeduto, L. (2005). Indirect 401–405. genetic effects and the early language devel- Whitehurst, G. J., & Sonnenschein, S. (1985). The opment of children with genetic mental retar- development of communication: A function-

᭧ American Association on Intellectual and Developmental Disabilities 229 VOLUME 113, NUMBER 3: 214–230 ͦ MAY 2008 AMERICAN JOURNAL ON MENTAL RETARDATION Noncomprehension of language L. Abbeduto et al.

al analysis. In G. J. Whitehurst (Ed.), Annals sity of Wisconsin-Madison as well as by funds of child development (pp. 1–48). Greenwich, from the Waisman Center and Friends of the CT: JAI. Waisman Center. The authors are deeply grateful Yirmiya, N., Erel, O., Shaked, M., & Solomonica- to the participating families for their time, pa- Levi, D. (1998). Meta-analyses comparing the- tience, and enthusiastic support. Thanks also to ory of mind abilities of individuals with au- Randi Hagerman and Marcia Braden for their ad- tism, individuals with mental retardation, and vice on recruitment and testing accommodations, normally developing individuals. Psychological Robert Nellis and Jill Roxberg for conducting the Bulletin, 124, 283–307. audiological evaluations, Pamela Lewis for assis- Zelazo, P. D., Burack, J. A., Benedetto, E., & Frye, tance with the autism diagnostic evaluations, and D. (1996). Theory of mind and rule use in Susan Kirkpatrick for assistance with genetic test- individuals with Down’s syndrome: A test of ing and interpretation. Brief summaries of prelim- the uniqueness and speciﬁcity claims. Journal inary analyses of some of the data reported here of Child Psychology and Psychiatry, 37, 479–484. have appeared in several chapters in edited vol- umes, including Abbeduto and Murphy (2004) Received 9/22/06, accepted 10/3/07. and Murphy and Abbeduto (2003), and were pre- Editor-in-charge: Elisabeth Dykens sented at the 2006 meeting of the Society for the Study of Behavioral Phenotypes, Dublin, Ireland. Preparation of this manuscript was supported by Requests for reprints should be sent to Leonard Grants R01 HD024356 and P30 HD003352 from Abbeduto, Waisman Center, University of Wis- the National Institutes of Health and by the Grad- consin-Madison, 1500 Highland Ave., Madison, uate School Research Committee of the Univer- WI 53705. E-mail: [email protected]

230 ᭧ American Association on Intellectual and Developmental Disabilities