Research in Developmental Disabilities 41–42 (2015) 13–21
Contents lists available at ScienceDirect
Research in Developmental Disabilities
Predictors of reading comprehension ability in primary school-aged children who have pragmatic language impairment Jenny Freed *, [1_TD$IF]Catherine Adams, Elaine Lockton School of Psychological Sciences, University of Manchester, Ellen Wilkinson Building, Oxford Road, Manchester, M13 9PL, UK
A R T I C L E I N F O
A B S T R A C T
Article history: Received 3 November 2014 Received in revised form 30 January 2015 Accepted 31 March 2015 Available online
Purpose: Children who have pragmatic language impairment (CwPLI) have difficulties with the use of language in social contexts and show impairments in above-sentence level language tasks. Previous studies have found that[2_TD$IF] typically[2_TD$IF] developing children’s reading comprehension (RC) is predicted by reading accuracy and spoken sentence level comprehension (SLC). This study explores the predictive ability of these factors and above-sentence level comprehension (ASLC) on RC skills in a group of CwPLI. Method: Sixty nine primary school-aged CwPLI completed a measure of RC along with measures of reading accuracy, spoken SLC and both visual (pictorially presented) and spoken ASLC tasks. Results: Regression analyses showed that reading accuracy was the strongest predictor of RC. Visual ASLC did not explain unique variance in RC on top of spoken SLC. In contrast, a measure of spoken ASLC explained unique variance in RC, independent from that explained by spoken SLC. A regression model with nonverbal intelligence, reading accuracy, spoken SLC and spoken ASLC as predictors explained 74.2% of the variance in RC. Conclusions: Findings suggest that spoken ASLC may measure additional factors that are important for RC success in CwPLI and should be included in routine assessments for language and literacy learning in this group. ß 2015 Elsevier Ltd. All rights reserved.
Keywords: Pragmatic language impairment Social communication disorder Literacy Reading comprehension
1. Introduction 1.1. Pragmatic language impairment Pragmatic language impairment (PLI) is a developmental language disorder in which children show a disproportionate difficulty with pragmatics and social communication compared to relatively good ability in the structural aspects of language such as grammar, vocabulary and phonology (Bishop, 2000). Children who have pragmatic language impairment (CwPLI) are typically verbose, have difficulty staying on topic and possess poor turn-taking skills (Adams, 2001; Bishop & Adams, 1989). They have difficulties with above-sentence level language tasks including narrative organisation, inferential
* Corresponding author. Tel.: +44 161 306 1672; fax: +44 161 275 3373. E-mail addresses: [email protected] (J. Freed), [email protected] (C. Adams), [email protected] (E. Lockton). http://dx.doi.org/10.1016/j.ridd.2015.03.003 0891-4222/ß 2015 Elsevier Ltd. All rights reserved.
14
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
comprehension and comprehension of discourse (Botting & Adams, 2005). The objective of the present study was to identify predictors of reading comprehension in CwPLI and explore the contribution of above-sentence level tasks (an area of difficulty for this group) to reading comprehension ability. PLI is a heterogeneous disorder in which there are wide variations in social communication and language skills. Children present with some of the features of language impairment observed in specific language impairment (SLI) including receptive language difficulties, word finding difficulties and semantic errors as well as some of the pragmatic communication difficulties of high functioning autism (Bishop, 2000; Bishop & Norbury, 2002). However, CwPLI can typically be distinguished from children with high functioning autism because they do not have impairment in all three aspects of the autism triad due to the absence of restricted and repetitive behaviours in the former group (Bishop, 1998; Bishop & Norbury, 2002; Gibson, Adams, Lockton, & Green, 2013; Leyfer, Tager-Flusberg, Dowd, Tomblin, & Folstein, 2008). The term ‘social (pragmatic) communication disorder’ (SCD) has recently been used in DSM-5 (American Psychiatric Association, 2013) to describe this group of children. SCD is categorised in DSM-5 as a subtype of language impairment with a similar set of features as those described by Bishop (2000) in her definition of PLI. In the current study, the term ‘children who have PLI (CwPLI)’ will be used to describe children with a pattern of pragmatic and social communication needs similar to those described by Bishop (2000) and the SCD profile in DSM-5. The diagnostic overlaps between PLI and SLI and autism remain unresolved. Therefore, we are not seeking to position PLI as a diagnostic category or ‘clinical condition’ but argue that there are children [6_TD$IF]presenting to speech and language therapy services whose primary difficulty is in pragmatics (with or without some relatively mild structural language problems). These children require an intervention approach that is distinct from children with purely structural language difficulties. The persistent social and communication difficulties experienced by CwPLI extend beyond spoken language into academic attainment; for example, Freed, Adams and Lockton (2011) reported that a significant proportion (41%) of CwPLI have reading comprehension (RC) difficulties and these children generally require long-term support for learning in the classroom (Botting & Conti-Ramsden, 1999). Establishing an experimental group of CwPLI is therefore important to ascertain information about predictors of RC in this group given the potential educational implications of persistent poor RC. This may, in turn, indicate appropriate directions for intervention. The present study therefore sought to identify what factors function as predictors of RC in this group. 1.2. Reading comprehension RC ability is thought to be underpinned by a number of language and cognitive skills including reading accuracy (the ability to decode or recognise individual words), spoken language comprehension, vocabulary and processing skills (Cain & Oakhill, 2006; Oakhill, Cain, & Bryant, 2003). According to the Simple View of Reading, RC is the product of two skills, reading accuracy and linguistic comprehension (Gough & Tunmer, 1986). Therefore previous research has examined the influence of skills such as reading accuracy and spoken sentence level comprehension (SLC) on RC. Reading accuracy is an important predictor of RC and the two skills are highly correlated in children with typical language development (CwTLD) (Gough, Hoover, & Peterson, 1996; Nation & Snowling, 1997) and children with SLI (CwSLI) (Botting, Simkin, & Conti-Ramsden, 2006). If reading accuracy is slow or inaccurate then a child may have to devote most of his/her attention to extracting the meaning from individual words. This may use up most of the reader’s processing capacity, leaving few resources for higher level comprehension skills such as integration and inference (Bishop, 1997; Snyder, Caccamise, & Wise, 2005). However, children with poor reading accuracy (as seen in children with dyslexia) can have good RC by employing their semantic and syntactic skills to support ineffective decoding strategies (Bishop & Snowling, 2004; Snowling, 2000). Conversely, studies have also shown that children with a specific deficit in RC (often referred to as ‘poor comprehenders’) can have decoding skills within the normal range (Nation & Norbury, 2005). Researchers have found strong links between RC and tests of spoken SLC such as the Test for Reception of Grammar, 2nd Edition (TROG-2; Bishop, 2003b) in CwTLD (Oakhill et al., 2003; Stothard & Hulme, 1992) and CwSLI (Botting et al., 2006). In children with autism spectrum disorders (CwASD), who are generally reported to have good reading accuracy skills, RC is associated with poor oral language skills including SLC (Nation, Clarke, Wright, & Williams, 2006; Nation & Norbury, 2005). Poor comprehenders were found to be impaired relative to a control group, matched for age and reading accuracy, on the TROG-2 as well as measures of semantic skills, sentence recall and past tense elicitation (Nation, Clarke, Marshall, & Durand, 2004). The Simple View of Reading (Gough & Tunmer, 1986) proposes that poor RC is due to poor word reading, poor oral language or both. The predictive ability of reading accuracy and spoken SLC on RC has not been examined in CwPLI, although Freed et al. (2011) reported a significant correlation (r = .81) between reading accuracy and RC in CwPLI. Therefore, the first aim of the present study is to investigate the proportion of variance in RC that can be explained by reading accuracy and spoken SLC in CwPLI and to examine whether this is in line with research with CwTLD. It has been suggested that tests that assess language at the word or sentence level may underestimate children’s language difficulties in more complex language tasks such as telling stories (Bishop, 2004). Whilst previous research has explored the influence of reading accuracy and spoken SLC on RC, there has been a lack of research into the influence of above-sentence level comprehension (ASLC) skills for language impaired groups and particularly in CwPLI. Therefore the second aim of the present study is to examine the predictive ability of ASLC to RC in CwPLI.
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
15
Tasks of ASLC, including RC, require constructive processing (Bishop, 1997), that is, the participant must progressively integrate information across sentences and draw inferences from world knowledge. In the present study we distinguish between two different types of ASLC task as predictors of RC: spoken and visual. Spoken ASLC tasks refer to those where the participant listens to the stimulus material, such as a spoken story and then answers comprehension questions related to the stimulus material. Spoken ASLC tasks rely on sentence-level linguistic skills in addition to constructive processing skills in order to process information at the above-sentence level. Visually presented ASLC tasks refer to those where the stimulus material is presented pictorially, for example, the participant is shown a storyboard of pictorial cues depicting a story, and then answers questions about the story. In contrast to spoken ASLC tasks, measures of visual ASLC do not rely on sentencelevel linguistic skills but rely solely on constructive processing of meaning. Ability on RC and spoken ASLC tasks is closely related (Nation & Snowling, 1997). For example, poor comprehenders were impaired on a measure of spoken ASLC compared to age-matched peers, but performed at a similar level to a younger group matched for RC ability (Stothard & Hulme, 1992). Poor comprehenders are impaired in relation to good comprehenders on tests of inference skills including interpreting ambiguous sentences and figurative expressions such as metaphors and idioms (Nation et al., 2004) and so share some characteristics with CwPLI in terms of difficulties with higher level skills. Previous research has found inference ability to be an important predictor of RC (Oakhill et al., 2003). As this is an area of difficulty for CwPLI it will be interesting to see the contribution that these skills make to RC. There has been a lack of research into the relationship between visual ASLC and RC ability. Therefore[7_TD$IF] the use of both spoken and visual ASLC tasks within this study allows the exploration of tasks requiring constructive processing (including inferencing and integration) with and without the influence of linguistic processing. The ASLC measures may be more sensitive to the higher level language processing difficulties of CwPLI than measures of spoken SLC. Whilst poor spoken ASLC can be explained to some extent by poor SLC, this is not the case for visual ASLC. Therefore it is important to consider other factors that are important for constructive processing ability, and specifically those that may be relevant to CwPLI. The fact that CwPLI have above-sentence level language processing difficulties makes them an interesting group in which to explore the relationship between ASLC and RC. There may be two potential routes to ASLC difficulties in CwPLI. They might be influenced by language impairments, or ASLC might be affected by some degree of autistic cognition, such as difficulties with central coherence. A third possibility is that CwPLI have a ‘double hit’ of impairments, meaning it is not possible to separate out the contribution of these two underlying factors (Loucas et al., 2008). To summarise, previous studies have examined explanations for poor RC ability such as reading accuracy and spoken SLC. However, the authors are not aware of any studies that have examined these skills as predictors of RC in CwPLI. In addition, there is a lack of knowledge about the contribution of ASLC to RC in this group. As stated earlier, Freed et al. (2011), which included a proportion of the children involved in this study, reported that 41% of CwPLI had RC impairment. The present study aims to explore the possible underlying causes for this difficulty, and the relative contribution each skill makes to predicting RC ability in CwPLI. 1.3. Research questions The specific research questions are as follows: (1) What is the predictive ability of reading accuracy and spoken SLC to RC ability in CwPLI? Are these findings in line with CwTLD? Based on previous research with CwTLD and CwSLI (who share the higher level language difficulties of CwPLI) it is expected that reading accuracy and spoken SLC will be significant predictors of RC ability. (2) Do measures of spoken and visual ASLC make unique contributions to RC in CwPLI? It is pertinent to explore the influence of ASLC on RC in CwPLI as this is a well identified area of difficulty for these children. Both modalities of ASLC task rely on constructive processing, whereas the spoken SLC measure does not. It would be expected that there will be high correlations between the ASLC tasks and RC because these tasks rely on a number of common skills. The two ASLC measures will be used in order to allow comparisons between a task that relies on both constructive processing and linguistic skills (spoken ASLC) and one that relies on constructive processing of meaning only (visual ASLC). 2. Method Ethical approval for this study was granted from[8_TD$IF] the Northern and Yorkshire Research Ethics Committee. Parents gave informed consent on behalf of their children. 2.1. Participants CwPLI (n = 69, 60 males) were recruited from a cohort of 85 CwPLI taking part in the Social Communication Intervention Project (SCIP), a randomised controlled trial of intervention for CwPLI (Adams et al., 2012). All parents of children involved in SCIP were approached to seek consent to allow their child’s data to be used in the present study and 69 parents consented.1
1
The 69 participants in this study include the 59 participants who took part in Freed et al. (2011).
16
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
Children were aged between 6.00 and 10.08 years and were on National Health Service (NHS, the UK publicly funded medical and health services) speech and language therapy caseloads in the North West of England or South East Scotland. All children[9_TD$IF] were attending a mainstream school[10_TD$IF], were identified as having special educational needs (England) or [1_TD$IF]additional [12_TD$IF]support [13_TD$IF]needs (Scotland) and had English as their primary language of communication and learning. None of the children had severe emotional or behaviour needs, severely unintelligible speech, severe physical difficulties or hearing loss. All children were considered, by their speech and language therapist (SLT) to have PLI and to meet at least two of the following aspects of social communication impairment: (1) Difficulty understanding and interpreting social context and friendship, e.g. social roles, emotions. (2) Difficulty understanding and/or using nonverbal aspects of communication e.g. facial expression, intonation. (3) Difficulty with aspects of conversation e.g. beginning and ending, taking turns, giving relevant and sufficient information. (4) Making bizarre, tangential or inappropriate comments. (5) Difficulty using and understanding non-literal language in social interactions.
All children had a score in the communication impaired range (58, equivalent to a score
Contents lists available at ScienceDirect
Research in Developmental Disabilities
Predictors of reading comprehension ability in primary school-aged children who have pragmatic language impairment Jenny Freed *, [1_TD$IF]Catherine Adams, Elaine Lockton School of Psychological Sciences, University of Manchester, Ellen Wilkinson Building, Oxford Road, Manchester, M13 9PL, UK
A R T I C L E I N F O
A B S T R A C T
Article history: Received 3 November 2014 Received in revised form 30 January 2015 Accepted 31 March 2015 Available online
Purpose: Children who have pragmatic language impairment (CwPLI) have difficulties with the use of language in social contexts and show impairments in above-sentence level language tasks. Previous studies have found that[2_TD$IF] typically[2_TD$IF] developing children’s reading comprehension (RC) is predicted by reading accuracy and spoken sentence level comprehension (SLC). This study explores the predictive ability of these factors and above-sentence level comprehension (ASLC) on RC skills in a group of CwPLI. Method: Sixty nine primary school-aged CwPLI completed a measure of RC along with measures of reading accuracy, spoken SLC and both visual (pictorially presented) and spoken ASLC tasks. Results: Regression analyses showed that reading accuracy was the strongest predictor of RC. Visual ASLC did not explain unique variance in RC on top of spoken SLC. In contrast, a measure of spoken ASLC explained unique variance in RC, independent from that explained by spoken SLC. A regression model with nonverbal intelligence, reading accuracy, spoken SLC and spoken ASLC as predictors explained 74.2% of the variance in RC. Conclusions: Findings suggest that spoken ASLC may measure additional factors that are important for RC success in CwPLI and should be included in routine assessments for language and literacy learning in this group. ß 2015 Elsevier Ltd. All rights reserved.
Keywords: Pragmatic language impairment Social communication disorder Literacy Reading comprehension
1. Introduction 1.1. Pragmatic language impairment Pragmatic language impairment (PLI) is a developmental language disorder in which children show a disproportionate difficulty with pragmatics and social communication compared to relatively good ability in the structural aspects of language such as grammar, vocabulary and phonology (Bishop, 2000). Children who have pragmatic language impairment (CwPLI) are typically verbose, have difficulty staying on topic and possess poor turn-taking skills (Adams, 2001; Bishop & Adams, 1989). They have difficulties with above-sentence level language tasks including narrative organisation, inferential
* Corresponding author. Tel.: +44 161 306 1672; fax: +44 161 275 3373. E-mail addresses: [email protected] (J. Freed), [email protected] (C. Adams), [email protected] (E. Lockton). http://dx.doi.org/10.1016/j.ridd.2015.03.003 0891-4222/ß 2015 Elsevier Ltd. All rights reserved.
14
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
comprehension and comprehension of discourse (Botting & Adams, 2005). The objective of the present study was to identify predictors of reading comprehension in CwPLI and explore the contribution of above-sentence level tasks (an area of difficulty for this group) to reading comprehension ability. PLI is a heterogeneous disorder in which there are wide variations in social communication and language skills. Children present with some of the features of language impairment observed in specific language impairment (SLI) including receptive language difficulties, word finding difficulties and semantic errors as well as some of the pragmatic communication difficulties of high functioning autism (Bishop, 2000; Bishop & Norbury, 2002). However, CwPLI can typically be distinguished from children with high functioning autism because they do not have impairment in all three aspects of the autism triad due to the absence of restricted and repetitive behaviours in the former group (Bishop, 1998; Bishop & Norbury, 2002; Gibson, Adams, Lockton, & Green, 2013; Leyfer, Tager-Flusberg, Dowd, Tomblin, & Folstein, 2008). The term ‘social (pragmatic) communication disorder’ (SCD) has recently been used in DSM-5 (American Psychiatric Association, 2013) to describe this group of children. SCD is categorised in DSM-5 as a subtype of language impairment with a similar set of features as those described by Bishop (2000) in her definition of PLI. In the current study, the term ‘children who have PLI (CwPLI)’ will be used to describe children with a pattern of pragmatic and social communication needs similar to those described by Bishop (2000) and the SCD profile in DSM-5. The diagnostic overlaps between PLI and SLI and autism remain unresolved. Therefore, we are not seeking to position PLI as a diagnostic category or ‘clinical condition’ but argue that there are children [6_TD$IF]presenting to speech and language therapy services whose primary difficulty is in pragmatics (with or without some relatively mild structural language problems). These children require an intervention approach that is distinct from children with purely structural language difficulties. The persistent social and communication difficulties experienced by CwPLI extend beyond spoken language into academic attainment; for example, Freed, Adams and Lockton (2011) reported that a significant proportion (41%) of CwPLI have reading comprehension (RC) difficulties and these children generally require long-term support for learning in the classroom (Botting & Conti-Ramsden, 1999). Establishing an experimental group of CwPLI is therefore important to ascertain information about predictors of RC in this group given the potential educational implications of persistent poor RC. This may, in turn, indicate appropriate directions for intervention. The present study therefore sought to identify what factors function as predictors of RC in this group. 1.2. Reading comprehension RC ability is thought to be underpinned by a number of language and cognitive skills including reading accuracy (the ability to decode or recognise individual words), spoken language comprehension, vocabulary and processing skills (Cain & Oakhill, 2006; Oakhill, Cain, & Bryant, 2003). According to the Simple View of Reading, RC is the product of two skills, reading accuracy and linguistic comprehension (Gough & Tunmer, 1986). Therefore previous research has examined the influence of skills such as reading accuracy and spoken sentence level comprehension (SLC) on RC. Reading accuracy is an important predictor of RC and the two skills are highly correlated in children with typical language development (CwTLD) (Gough, Hoover, & Peterson, 1996; Nation & Snowling, 1997) and children with SLI (CwSLI) (Botting, Simkin, & Conti-Ramsden, 2006). If reading accuracy is slow or inaccurate then a child may have to devote most of his/her attention to extracting the meaning from individual words. This may use up most of the reader’s processing capacity, leaving few resources for higher level comprehension skills such as integration and inference (Bishop, 1997; Snyder, Caccamise, & Wise, 2005). However, children with poor reading accuracy (as seen in children with dyslexia) can have good RC by employing their semantic and syntactic skills to support ineffective decoding strategies (Bishop & Snowling, 2004; Snowling, 2000). Conversely, studies have also shown that children with a specific deficit in RC (often referred to as ‘poor comprehenders’) can have decoding skills within the normal range (Nation & Norbury, 2005). Researchers have found strong links between RC and tests of spoken SLC such as the Test for Reception of Grammar, 2nd Edition (TROG-2; Bishop, 2003b) in CwTLD (Oakhill et al., 2003; Stothard & Hulme, 1992) and CwSLI (Botting et al., 2006). In children with autism spectrum disorders (CwASD), who are generally reported to have good reading accuracy skills, RC is associated with poor oral language skills including SLC (Nation, Clarke, Wright, & Williams, 2006; Nation & Norbury, 2005). Poor comprehenders were found to be impaired relative to a control group, matched for age and reading accuracy, on the TROG-2 as well as measures of semantic skills, sentence recall and past tense elicitation (Nation, Clarke, Marshall, & Durand, 2004). The Simple View of Reading (Gough & Tunmer, 1986) proposes that poor RC is due to poor word reading, poor oral language or both. The predictive ability of reading accuracy and spoken SLC on RC has not been examined in CwPLI, although Freed et al. (2011) reported a significant correlation (r = .81) between reading accuracy and RC in CwPLI. Therefore, the first aim of the present study is to investigate the proportion of variance in RC that can be explained by reading accuracy and spoken SLC in CwPLI and to examine whether this is in line with research with CwTLD. It has been suggested that tests that assess language at the word or sentence level may underestimate children’s language difficulties in more complex language tasks such as telling stories (Bishop, 2004). Whilst previous research has explored the influence of reading accuracy and spoken SLC on RC, there has been a lack of research into the influence of above-sentence level comprehension (ASLC) skills for language impaired groups and particularly in CwPLI. Therefore the second aim of the present study is to examine the predictive ability of ASLC to RC in CwPLI.
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
15
Tasks of ASLC, including RC, require constructive processing (Bishop, 1997), that is, the participant must progressively integrate information across sentences and draw inferences from world knowledge. In the present study we distinguish between two different types of ASLC task as predictors of RC: spoken and visual. Spoken ASLC tasks refer to those where the participant listens to the stimulus material, such as a spoken story and then answers comprehension questions related to the stimulus material. Spoken ASLC tasks rely on sentence-level linguistic skills in addition to constructive processing skills in order to process information at the above-sentence level. Visually presented ASLC tasks refer to those where the stimulus material is presented pictorially, for example, the participant is shown a storyboard of pictorial cues depicting a story, and then answers questions about the story. In contrast to spoken ASLC tasks, measures of visual ASLC do not rely on sentencelevel linguistic skills but rely solely on constructive processing of meaning. Ability on RC and spoken ASLC tasks is closely related (Nation & Snowling, 1997). For example, poor comprehenders were impaired on a measure of spoken ASLC compared to age-matched peers, but performed at a similar level to a younger group matched for RC ability (Stothard & Hulme, 1992). Poor comprehenders are impaired in relation to good comprehenders on tests of inference skills including interpreting ambiguous sentences and figurative expressions such as metaphors and idioms (Nation et al., 2004) and so share some characteristics with CwPLI in terms of difficulties with higher level skills. Previous research has found inference ability to be an important predictor of RC (Oakhill et al., 2003). As this is an area of difficulty for CwPLI it will be interesting to see the contribution that these skills make to RC. There has been a lack of research into the relationship between visual ASLC and RC ability. Therefore[7_TD$IF] the use of both spoken and visual ASLC tasks within this study allows the exploration of tasks requiring constructive processing (including inferencing and integration) with and without the influence of linguistic processing. The ASLC measures may be more sensitive to the higher level language processing difficulties of CwPLI than measures of spoken SLC. Whilst poor spoken ASLC can be explained to some extent by poor SLC, this is not the case for visual ASLC. Therefore it is important to consider other factors that are important for constructive processing ability, and specifically those that may be relevant to CwPLI. The fact that CwPLI have above-sentence level language processing difficulties makes them an interesting group in which to explore the relationship between ASLC and RC. There may be two potential routes to ASLC difficulties in CwPLI. They might be influenced by language impairments, or ASLC might be affected by some degree of autistic cognition, such as difficulties with central coherence. A third possibility is that CwPLI have a ‘double hit’ of impairments, meaning it is not possible to separate out the contribution of these two underlying factors (Loucas et al., 2008). To summarise, previous studies have examined explanations for poor RC ability such as reading accuracy and spoken SLC. However, the authors are not aware of any studies that have examined these skills as predictors of RC in CwPLI. In addition, there is a lack of knowledge about the contribution of ASLC to RC in this group. As stated earlier, Freed et al. (2011), which included a proportion of the children involved in this study, reported that 41% of CwPLI had RC impairment. The present study aims to explore the possible underlying causes for this difficulty, and the relative contribution each skill makes to predicting RC ability in CwPLI. 1.3. Research questions The specific research questions are as follows: (1) What is the predictive ability of reading accuracy and spoken SLC to RC ability in CwPLI? Are these findings in line with CwTLD? Based on previous research with CwTLD and CwSLI (who share the higher level language difficulties of CwPLI) it is expected that reading accuracy and spoken SLC will be significant predictors of RC ability. (2) Do measures of spoken and visual ASLC make unique contributions to RC in CwPLI? It is pertinent to explore the influence of ASLC on RC in CwPLI as this is a well identified area of difficulty for these children. Both modalities of ASLC task rely on constructive processing, whereas the spoken SLC measure does not. It would be expected that there will be high correlations between the ASLC tasks and RC because these tasks rely on a number of common skills. The two ASLC measures will be used in order to allow comparisons between a task that relies on both constructive processing and linguistic skills (spoken ASLC) and one that relies on constructive processing of meaning only (visual ASLC). 2. Method Ethical approval for this study was granted from[8_TD$IF] the Northern and Yorkshire Research Ethics Committee. Parents gave informed consent on behalf of their children. 2.1. Participants CwPLI (n = 69, 60 males) were recruited from a cohort of 85 CwPLI taking part in the Social Communication Intervention Project (SCIP), a randomised controlled trial of intervention for CwPLI (Adams et al., 2012). All parents of children involved in SCIP were approached to seek consent to allow their child’s data to be used in the present study and 69 parents consented.1
1
The 69 participants in this study include the 59 participants who took part in Freed et al. (2011).
16
J. Freed et al. / Research in Developmental Disabilities 41–42 (2015) 13–21
Children were aged between 6.00 and 10.08 years and were on National Health Service (NHS, the UK publicly funded medical and health services) speech and language therapy caseloads in the North West of England or South East Scotland. All children[9_TD$IF] were attending a mainstream school[10_TD$IF], were identified as having special educational needs (England) or [1_TD$IF]additional [12_TD$IF]support [13_TD$IF]needs (Scotland) and had English as their primary language of communication and learning. None of the children had severe emotional or behaviour needs, severely unintelligible speech, severe physical difficulties or hearing loss. All children were considered, by their speech and language therapist (SLT) to have PLI and to meet at least two of the following aspects of social communication impairment: (1) Difficulty understanding and interpreting social context and friendship, e.g. social roles, emotions. (2) Difficulty understanding and/or using nonverbal aspects of communication e.g. facial expression, intonation. (3) Difficulty with aspects of conversation e.g. beginning and ending, taking turns, giving relevant and sufficient information. (4) Making bizarre, tangential or inappropriate comments. (5) Difficulty using and understanding non-literal language in social interactions.
All children had a score in the communication impaired range (58, equivalent to a score
