The pathways to the documented increased social and emotional difficulties in individuals with Developmental Language Disorder (DLD) are unclear. We explored whether differences in social evaluation could account for social and emotional difficulties in adolescents with DLD using a computerized social evaluation task. Twenty-four adolescents with DLD were matched with twenty-six adolescents with typical language development (TLD) (Mage = 13.5 years, SE = 2.38; n = 18 female). They completed the Social Evaluation Learning Task (SELT; Button et al., 2015) which measures how quickly people learn the computer likes or dislikes either them or someone else. Adolescents and parents reported social and emotional functioning. Adolescents with DLD had poorer social understanding, in that they took longer to learn that the computer disliked them. They learned similarly to their TLD when the computer liked them and someone else. Adolescents with DLD also had higher self-reported anxiety and more parent reported emotional and peer problems; however, there was no mediational effect of social evaluation on socioemotional difficulties. This study demonstrates that adolescent with DLD have specific difficulties in interpreting cues that they are disliked by others but are just as good at understanding when they are liked. The differences seen in their social evaluation skills did not account for their increased socioemotional difficulties. This social evaluation bias might explain previous findings of good self-rated social competence while other ratings indicate social difficulties. Future research is necessary to investigate the implications of this finding further.
The supportive impact of social relationships on mental health is well established in typically developing (TD) populations (van Harmelen et al., 2017) and good communication skills are a key component for developing and strengthening social connections (Gallagher, 1993). However, approximately 7% of the population may struggle in these areas because they have Developmental Language Disorder (DLD), a difficulty with receptive and/or expressive language not related to cognitive or auditory impairments or an autism spectrum diagnosis (Bishop et al., 2016). Children and adolescents with DLD are consistently rated as having significantly higher levels of social and emotional difficulties than their TD peers (Yew & O’Kearney, 2013). These difficulties reflect the inherently social nature of communication, illustrating that DLD does not simply manifest as a difficulty with language but also impacts on social and emotional functioning as well. Contrasting findings have emerged from studies investigating the predictors of these socioemotional difficulties, ranging from pragmatic skills (St Clair et al., 2011) to peer problems (Forrest et al., 2021; Wadman et al., 2011a, b). An alternative view is to consider the driving force behind these predictors: Children and young people with DLD may have a fundamental difference in how they evaluate social situations, such as the way they perceive and interpret social cues, which impacts their social and emotional experiences. Social evaluation encompasses how we perceive and interpret social cues from others, with a particular focus on interpreting what others think about ourselves or someone else. The current study proposes that differences in social evaluation may explain this relationship and presents preliminary findings from an interactive task that examines social evaluation as a potential mediator of the relationship between DLD status and social and emotional difficulties in adolescents. To the authors knowledge, this is the first study looking at social evaluation differences in DLD or neurodiversity more generally, although differences in social evaluation have been investigated across a range of mental health conditions (Button et al., 2015; Hoffmann et al., 2024).
DLD and Social and Emotional Difficulties
Compared to their TD peers, individuals with DLD are at risk for increased social difficulties, such as poor social skills (Lloyd-Esenkaya et al., 2020), difficulties making friends (Durkin & Conti-Ramsden, 2007), and victimisation (Redmond, 2011). Individuals with DLD also experience higher levels of emotional difficulties than their TD peers, such as feelings of anxiety (Beitchman et al., 2001; Burnley et al., 2023) and depression (Conti-Ramsden & Botting, 2008). These negative outcomes persist throughout the lifespan (Beitchman et al., 2014), with social difficulties increasing in adolescence (St Clair et al., 2011), highlighting the need for more research in this age group. Despite the established connection between DLD and social and emotional problems, these difficulties are not always directly predicted by poor language abilities (Hart et al., 2004). Examining alternative pathways, such as associated social evaluation difficulties, may provide a better understanding of how these associated socioemotional difficulties develop and could help with designing more targeted interventions. Given research in the DLD population which has shown a mediating effect of peer problems on emotional problems (Forrest et al., 2021; Wadman et al., 2011a, b) and the moderating effect of victimization on internalising problems (Kilpatrick et al., 2019) it seems pertinent to explore the effect of social evaluation differences. There has been very limited research on social evaluation differences in individuals with DLD, but previous research on social cognition (which encompasses social evaluation amongst other skills) has focused on children with DLD. The current study extends the literature by focusing on adolescents with DLD, looking specifically at social evaluation skills.
DLD and Social Cognition
Social cognition is an umbrella term that refers to an individual’s understanding of social interactions. Social cognition draws on skills such as Theory of Mind (ToM) or ‘mentalizing’, which are the ability to understand others’ thoughts, feelings and motives (Frith & Frith, 2003). These skills are necessary for successful social interactions and are predicted by language ability (Astington & Jenkins, 1999; Jenkins & Astington, 1996). According to the usage-based theory of language acquisition (Tomasello, 2009), children learn language while at the same time strengthening their social cognition abilities through conversational interactions with their caregiver (i.e. joint attention and understanding speaker intentions). Indeed, ‘mentalizing’ language used by mothers predicted better ToM skills in typically developing children (Ruffman et al., 2002) and deaf children of hearing parents demonstrated poorer ToM skills than deaf children of deaf parents, who performed similarly to hearing children (Schick et al., 2007). These studies highlight the integral relationship between language and social cognition; children need exposure to their native language to be able to parse others’ body language and non-verbal cues as internal thoughts and also need to be able to communicate easily with conversation partners, without any barriers. Children and young people with DLD may therefore be at a disadvantage due to their language difficulties and may not be exposed to the same level of mentalising vocabulary and social interactions. Indeed, a recent study indicated that deficits in receptive vocabulary play an important role in predicting ToM deficits and when controlled for vocabularly group differences between children with DLD and TD was non-significant (Bulgarelli et al., 2022). Other studies show children with DLD have poorer social cognition abilities than their peers (Farmer, 2000; Ford & Milosky, 2003); however, the number of studies is limited, and the majority of the research is conducted on children (Nilsson & de Lopez, 2016; Vissers & Koolen, 2016).
Social evaluation is an element of social cognition that focuses on how we evaluate other people’s intentions and judgements. In some mental health conditions, such as depression and social anxiety, individuals often have a negative interpretation bias that skews their interpretation of social situations and focuses their attention on negative elements (Chen et al., 2020; Suslow et al., 2020). As these conditions are more common in individuals with DLD (St Clair et al., 2023; Voci et al., 2006), it is important to investigate whether there are any social evaluation differences associated with DLD. A recent study investigating adolescents (using the same sample as this paper) found that performance on the Social Attribution Task was poorer in adolescents with DLD than in their matched TD peers (Forrest et al., 2023). This study indicated that adolescents with DLD had fundamental differences in how they interpreted the interactions of the animated shapes within the Social Attribution Task, in particular attributing fewer psychological attributes when asked to think of the shapes as people. They also used fewer emotional state words when describing the actions of the shapes. To our knowledge, there is no other research looking at DLD and social evaluation differences.
Social Cognition and Social and Emotional Difficulties
Impairments in social cognition associated with DLD could have significant consequences for individuals’ social and emotional functioning. Deficits in ToM abilities make it harder to pick up on a speaker’s intentions and may lead to faux pas due to misinterpreting social situations. For example, studies show children with DLD found it harder to navigate conflict resolution tasks than their peers, displaying aggressive or withdrawal behaviours instead of using speech (Bakopoulou & Dockrell, 2016; Marton et al., 2005). These negative behaviours may result in peer rejection, as evidenced by poor performance on ToM tasks, such as the eyes task and strange stories task, predicting lower levels of friendships in adolescents with DLD (Botting & Conti-Ramsden, 2008) and children with DLD receiving more “least liked” ratings from classmates (Andres-Roqueta et al., 2016). Total social cognition score (a composite of unexpected contents, change of location and strange stories tasks) explained 11% of the variance in least liked ratings, emphasising that poor language skills alone do not account for the lower social functioning seen in children with DLD (Andres-Roqueta et al., 2016). Emotional adjustment may also be impacted by poor social cognition skills as evidenced by lower social self-esteem in children with DLD (Marton et al., 2005). Additionally, social cognition and prosocial behaviour accounted for 44% of the variance in teacher-ratings of socioemotional difficulties in children with DLD (Bakopoulou & Dockrell, 2016). However, a recent study using the same sample in this paper found that performance on the Social Attribution Task did not explain additional variance in social and emotional outcomes after accounting for differences related to group (DLD/TLD; Forrest et al., 2023). Further investigation into the link between social cognition, and in particular social evaluation, and socio-emotional outcomes is needed given these conflicting findings.
Current Study
The current study had three main aims to address the gaps in the literature when examining the pathways between DLD and socioemotional difficulties. Firstly, we aimed to expand the literature by using an interactive task, novel to this population, to measure social evaluation, an element of social cognition. The current study used the dynamic Social Evaluation Learning Task (SELT; Button et al., 2015) to measure adolescents’ responses to judgements to either themselves or someone else in computerised social situations. This allowed the measurement of ‘online’ interpretation of social cues as participants were updating their judgements while interacting with the computer characters. The choice of task was also important for our second aim of investigating effects within an adolescent sample. Previous research has focused on social cognition abilities of children with DLD (excepting Conti-Ramsden and Botting (2008) who sampled adolescents with DLD, Clegg et al. (2005) who studied adults with DLD and Forrest et al. (2023) who evaluate a different social cognitive task in this adolescent sample). Given that social cognition abilities are still developing in adolescence (Blakemore & Choudhury, 2006) and adolescence is generally a risk period for the development of psychiatric disorders (Kim-Cohen et al., 2003), it is crucial to study social understanding during this period of development. Thirdly, we aimed to expand the literature to examine the effect of social evaluation, including self evaluation, on emotional as well as social outcomes.
Adolescents with DLD were compared to age- and sex-matched peers with typical language development (TLD) on measures of socioemotional functioning. Performance on the SELT was evaluated as a mediator between DLD and socioemotional difficulties. Based on previous research, we expected that the participants with DLD would experience higher levels of social and emotional difficulties. Due to the research discussed above regarding social cognition and social evaluation differences as a potential comorbid difficulty in children and young people with DLD, we predicted the DLD group would perform poorly on the SELT, reflected by a difficulty learning whether they were “liked” or “disliked” and incongruent social global ratings given after each condition. Finally, we expected that social evaluation abilities would mediate the relationship between DLD and socioemotional difficulties, with a poor performance on the SELT predicting poor socioemotional functioning.
Method
Ethics
Ethical approval was granted by the University of Bath Psychology Ethics Committee (Ref: 15-245).
Participants
Figure 1 displays participant flow. Participants were recruited directly or through screening. Individuals with a history of DLD were directly recruited into the DLD group through referrals from speech and language therapists and Special Educational Needs Coordinators (SENCos) within schools, or via flyers posted in online support groups for DLD. A screening procedure advertised via flyers in local schools, at the University of Bath and on social media was used to recruit the TLD comparison group matched on age (within six months) and sex. Screening packs consisting of background questionnaires for parents/carers and self-report measures of language and communication skills for participants were sent to 258 adolescents, of which 109 were completed by both parent/carer and young person. Those with identified language difficulties were included in the DLD group.
Participants were from mainstream schools aged 11–18 years old, although three participants attended a specialised language unit within a mainstream school. Adolescents were not eligible for inclusion if parents/carers reported hearing impairment, intellectual disabilities or diagnosis of autism spectrum disorder (ASD). Further exclusionary criteria consisted of exceeding the cut-off score on the Autism Quotient (AQ; Baron-Cohen et al., 2006; Baron-Cohen et al., 2001b) and performance on the nonverbal IQ measure indicating global intellectual delay in the testing phase. Our exclusion criteria based on the AQ was due to being cautious about possible of misdiagnoses of DLD rather than language disorder associated with autism. As this study was part of a large study investigating social cognitive differences associated with DLD, we were cautious and excluded adolescents with high autistic traits as well as an autism diagnosis. This was appropriate, given the well established social cognitive and ToM deficits associated with autism.
Forty individuals with DLD profile were initially identified for the study. Four individuals did not respond to invitation emails. Five adolescents with parent-report or self-report of language difficulties were included in the DLD group. Thirteen participants were not eligible (n = 1 hearing impairment, n = 12 autistic or exceeded cut-off on the AQ) and one participant withdrew from the study.
Twenty-seven participants were invited to the assessment as part of the DLD group, with twenty-seven matched controls forming the TLD group. One participant was excluded from the DLD group after scoring more than 2 SDs below the mean on the nonverbal measure. Technical problems resulted in missing SELT data (n = 1 DLD, n = 1 TLD). Another participant in the DLD group reported significant emotional problems therefore the testing phase was terminated in order to follow the safety protocol. This resulted in a total sample of 24 adolescents with DLD and 26 TLD participants matched on age and sex.
Table 1 shows approximately 36% of the sample were female, with an average age of 13;6 years (SE = 2.38 months, range = 12;1 years to 17;9 years). Participants were native English-speakers although three participants spoke a second language (n = 1 DLD, n = 2 TLD). As expected, the DLD group were significantly more delayed in reaching early speech and language developmental milestones compared to the TLD group and were more likely to have received speech and language therapy. There was a significantly lower level of parental education in the DLD group, and a significantly lower socioeconomic status, as measured by the Income Deprivation Affecting Children Index (IDACI) Rank (see Measures for details). The IDACI Rank ranged from 303 to 32,662 in the DLD group and from 13,021 to 32,489 in the TLD group, demonstrating higher deprivation in the DLD group.
Table 1
Demographics of sample
DLD (n = 24)
TLD (n = 26)
Total (n = 50)
DLD vs. TLD
Mean age (SE)
13;6 (3.39)
13;6 (3.41)
13;6 (2.38)
–
Female %
33.3
38.5
36.0
–
Mean IDACI Rank (SE)
19713.95 (1711.32)
24583.85 (1119.98)
22351.81 (1042.32)
– 4869.89(– 8983.10,– 756.69)*
Language spoken
English only %
95.8
92.3
94.0
English plus other %
4.2
7.7
6
Motor development
n.s
Delayed %
29.2
3.8
16
Typical %
54.2
80.8
68
Fast %
16.7
15.4
16
Speech & language development
0.03(0.005, 0.14)^***
Delayed %
75
7.7
40
Typical %
25
69.2
48
Fast %
--
23.1
12
Self-help development
0.13(0.03, 0.55)^**
Delayed %
45.8
3.8
24
Typical %
45.8
84.6
66
Fast %
8.3
11.5
10
Speech & Language Therapy
X232.37***
Yes %
83.3
3.8
42
No %
16.7
96.2
58
Biological parents
n.s.
Yes %
91.3
100
95.9
No - adopted %
8.7
--
4.1
Parental marital status
n.s.
Married/Cohabiting/Civil Partnership %
65.2
88.5
77.6
Separated %
17.4
7.7
12.2
Divorced %
17.4
3.8
10.2
Parental psychological distress
n.s.
Yes %
17.4
15.4
16.3
No %
82.6
84.6
83.7
Parental education
0.22(0.07, 0.68)**
Secondary school %
39.1
19.2
28.6
Diploma %
13
--
6.1
Undergraduate %
39.1
42.3
40.8
Postgraduate %
8.7
38.5
24.5
Statistics are b coefficients or odds ratio where stated ^ (95% confidence interval).
a As measured by parent-report of mental health difficulties in the background questionnaire.
* p <.05, ** p <.01, *** p <.001
Measures
Parent/Carer Questionnaires
The background questionnaire consisted of questions about the child’s early development, family background and history of speech and language problems (suspected or diagnosed). See Table 1 for details. Postcodes were gathered to generate an Income Deprivation Affecting Children Index (IDACI) Rank, a measure of socioeconomic status. The IDACI Rank is based on the percentage of children living in families that are income deprived in Lower-layer Super Output Areas across England, where 1 = most deprived neighbourhood and 32,844 = least deprived neighbourhood. School postcodes were used when home postcodes were missing (n = 5). Two participants in the DLD group did not have either postcode available.
The Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997) consists of 25 items which form five scales (Peer Problems; Emotional Problems; Hyperactivity Problems; Conduct Problems and Prosocial scale), the first four of which are totalled to produce the Total Difficulties score. The SDQ is a well-established measure and has a test-retest reliability of 0.85 (Goodman, 1999). The scales of interest were the Emotional Problems and Peer Problems subscales, each consisting of five items rated on a scale of Not True (0), Somewhat True (1) and Certainly True (2). Total scores for each subscale range from 0 to 10.
The adolescent version of the Autism Spectrum Quotient (AQ) was administered to parents of children aged 12–15 years old (Baron-Cohen et al., 2006). Cut-off scores of 30 or more were applied as exclusionary criteria.
Adolescent-Reported Measures
The Communication Checklist Self-Report (CC-SR; Bishop et al., 2009) consists of 70 questions about communication abilities. Items are rated on a scale of 0– Less than once a week (or never) to 3– Several times a day (or all the time) and form three composite scales. The Structural Language composite describes aspects of language such as grammar and meaning (e.g. “I mix up ‘he’, ‘she’, ‘it’ and ‘they’”). The Pragmatic Skills composite measures language use in social contexts (e.g. “I give detailed information when a more general comment would be fine”). Finally, the Social Engagement composite describes nonverbal communication and social functioning (e.g. “I find it hard to know when people are upset or annoyed”). Positive items are reverse scored and the mean scaled score is 10 (SD = 3). A scaled score lower than 5 on the Structural Language composite and greater than 7 on the Pragmatic Skills composite is indicative of DLD (Bishop et al., 2009). Internal consistency for each of the composites is greater than 0.85 (Bishop et al., 2009). The CC-SR was used in the screening sample to identify or rule out DLD.
The adult version of the AQ was completed by participants aged 16 years or over (Baron-Cohen et al., 2001). Cut-off scores of 32 or above were applied as an exclusion criterion.
The Warwick Edinburgh Mental Well-being Scale (WEMWBS; Tennant et al., 2007) consists of 14 items focusing on positive mental wellbeing. Participants rate statements on a scale of 1 None of the time to 5 All of the time, according to how they felt in the past two weeks. Scores range from 14 to 70 with a higher score reflecting a better state of mental wellbeing.
The Perceived Social Support– Friendship questionnaire (PSS-Fr; Procidano & Heller, 1983) consists of 20 statements about friendship quality and measures how well participants believe their support needs are met by their relationships with friends. Participants answer Yes, No or Don’t Know and the number of Yes answers is summed to create a total score (0–20, M = 15.15). A higher score indicates better social support.
The Revised Children’s Manifest Anxiety Scale (RCMAS; Reynolds & Richmond, 1978) consists of 28 statements regarding feelings of anxiety. Participants rate each item according to how often they have felt or acted that way in the past two weeks (Never, Sometimes, Mostly, or Always). Ratings are assigned scores of 0–3 respectively and summed. Total scores range from 0 to 84 with a higher score denoting more symptoms of anxiety.
The Moods and Feelings Questionnaire (MFQ; Angold & Costello, 1987) consists of 33 statements that measure depressive thoughts and feelings. Participants rate each item according to how often they have felt or acted that way in the past two weeks (Never, Sometimes, Mostly, or Always). Ratings are assigned scores of 0–3 respectively and summed. The total score ranges from 0 to 99 with a higher score denoting more depressive symptoms.
The Social Anxiety Scales for Adolescents (SAS-A; La Greca & Lopez, 1998) consists of 18 statements measuring how the participant feels in social situations and four filler statements (e.g. “I like to read”). Items are rated on a scale of 1 Not at all to 5 All the time. Three subscales (Fear of Negative Evaluation (FNE), Social Avoidance and Stress Specific to New Situations (SAD-New) and Social Avoidance and Stress in General (SAD-General) are generated along with a total score. Scores range from 8 to 40 for FNE, 6–30 for SAD-New, 4–20 for SAD-General and 18–90 for the total score. The total score was used in the current study. A total score of 50 or above reflects High Social Anxiety, while Low-Socially Anxious participants report a score of 36 or below.
Cognitive Measures
Participants completed the Recalling Sentences and Word Classes subtests from the Clinical Evaluation Language Functioning– 4 (CELF-4; Semel et al., 2006). The Recalling Sentences subtest measures expressive language and requires participants to listen to sentences of increasing length and complexity and repeat verbatim. The Word Classes subtest measures receptive language and requires participants to pick two words out of a list of four, providing a definition for how the two words are similar. Both subtests have an excellent rating of reliability (internal consistency coefficient of 0.92 and 0.91 respectively; Semel et al., 2006) and Recalling Sentences is a strong marker of DLD (Conti-Ramsden et al., 2001).
The Block Design subtest from the Wechsler Intelligence Scale for Children– 4th edition (WISC-IV) was administered to provide a measure of nonverbal ability (Wechsler, 2004). Participants are required to use 3D blocks to recreate 2D patterns of increasing complexity. Block Design is a measure of spatial awareness and contributes to fluid reasoning.
Social Evaluation
The Social Evaluation Learning Task (SELT; Button et al., 2015) is a computerised probabilistic learning task, designed to measure how well participants can learn whether a computer character likes them or not (see Fig. 2 for an example). Participants are presented with positive and negative word pairs and are asked to pick which of the words best describes what the computer character (“Charley”, “Sam”, “Chris” or “Bobby”) thinks of the participant (self-referential condition), or another character called “George” (other-referential condition). For example, “I think [you are] / [George is] witty– dull”. The original version used adult participants and the word pairs were matched using similar ratings of “likeableness” from Anderson (1968). The list of word pairs was modified to account for the age and language ability of participants in the current study. Given the difficulty in recruiting the target sample of adolescents with DLD, two TLD adolescents piloted the original word list and any items that were questioned were replaced with new words that retained similar “likeableness” ratings (Anderson, 1968). For example, “inquisitive” was replaced with “curious” and “candid” was replaced with “direct”. Participants receive feedback after each trial (“Correct” or “Incorrect”) in order to learn the “rule” of each condition. The two rules (“Like” or “Dislike”) and two conditions (“Self” or “Other”) produce four blocks: “Self-Liked”; “Self-Disliked”; “Other-Liked”; “Other-Disliked”. Each block has 32 trials with an 80 − 20 split on feedback. For example, in the Like rule, when participants should be learning that the computer has a favourable view, participants receive Correct feedback for selecting positive words 80% of the time, and Incorrect feedback for positive words 20% of the time. In the Dislike rule, participants receive Correct feedback for selecting positive words 20% of the time and Incorrect feedback for positive words 80% of the time. In general, individuals trend towards a more optimistic outlook, perceiving themselves in a good light. This ‘positive self-bias’ is evident in the difference between errors when learning the Dislike rule compared to learning the Like rule. Participants’ understanding of the rule (Like or Dislike) is measured by the Errors to Criterion outcome, which is the number of errors made before reaching the criterion of eight consecutive responses consistent with the prevailing Like or Dislike rule, regardless of occasionally misleading feedback. This score therefore indexes the number of trials taken to acquire the correct rule, demonstrating the learning process. At the end of each of the four conditions the participant must rate how much they think the computer character liked them or George on a scale of 0 Not at all to 10 Very Much. This forms the Global Rating outcome (mean “likeability”), demonstrating how the participant has interpreted the feedback following each trial.
Fig. 2
Example of a self-referential rule block from the Social Evaluation Learning Task.
Reprinted with slight adaptations from Button et al. (2015)
Parents/carers in the DLD group completed the consent form and questionnaires online or returned forms in a freepost envelope. Eligible participants completed the assessment in a quiet room either at the University, their school or their home. Participants recruited through the screening stream completed the CC-SR while parents/carers completed the consent form, brief background questionnaire, the AQ and SDQ either online or via paper copies. Adolescents with no reported language difficulties were matched on age (within six months) and sex to participants in the DLD group to form the control group. Parents/carers completed the remaining background questionnaire and participants completed the assessment wherever convenient. The assessment battery was administered as part of a larger study lasting approximately 90 min in total. The assent form, socioemotional questionnaires, on-screen instructions and word pairs for the SELT were read aloud to all participants. A glossary provided standardised definitions of words when requested. Participants received £15 on completion of the assessment and travel expenses were reimbursed. Entry into a prize draw to win a £50 voucher was offered as incentive to complete the screening stage. Brief reports of individuals’ results were sent to parents/carers and findings from the overall study were shared on completion.
Statistical Analysis
Data were analysed using Stata 14 (StataCorp., 2015). First, group differences in socioemotional functioning were analysed. Following tests for assumptions, regression was used to test for group differences in the wellbeing (WEMBS) and social support (PSS) scales. Due to skewed data, the anxiety (RCMAS), depression (MFQ) and social anxiety (SAS-A) scales were transformed using the square root function before regression analysis. Negative binomial regressions were used to test group differences in the SDQ Peer Problems and Emotional Problems subscales due to the most frequent response in each scale being zero. In the SELT, outcome variables were Errors to Criterion (mean number of errors made before 8 consecutive condition consistent responses) and Global Rating (mean likeability). Predictor variables were group (DLD vs. TLD); referential condition (Self vs. Other) and rule (Like vs. Dislike). A mixed ANOVA, controlling for IDACI Rank, was modelled with group as a between-subjects factor and referential condition and rule as within-subjects factors for each of the outcomes. Pairwise comparisons with Bonferroni corrections were used to examine interactions. The sgmediation2 command was used to test whether SELT performance (Errors to Criterion) mediated the relationship between DLD status and socioemotional outcomes. Case resampling bootstrapping was used to test the indirect effect, as recommended by (Zhao et al., 2010). Finally, a regression analysis tested the effect of language ability (Word Classes– Receptive and Recalling Sentences) on SELT performance (errors to criterion) for each group. IDACI Rank was entered as a covariate in all analyses.
Results
Cognitive and Language Measures
Table 2 illustrates the significant group differences on cognitive and language measures. As expected, the expressive and receptive language abilities of the DLD group were significantly worse than the TLD group and they performed below population norms of M = 10, SD = 3.
Table 2
Mean (SD) scaled scores from cognitive and Language tasks for the developmental Language disorder (DLD) group and typically developing (TLD) group.
DLD
TLD
DLD vs. TLD
Block design subtest
8.04(2.84)
11.81(2.48)
− 3.68(− 5.35, − 2.00)***
Recalling sentences subtest
4.75(2.98)
9.73(2.84)
− 4.87(− 6.72, − 3.03)***
Word classes– receptive subtest
5.79(2.89)
12.88(2.36)
− 6.58(− 8.19, − 4.97)***
Statistics are b coefficients (95% confidence interval), controlling for IDACI Rank
***p < 0.001
Social and Emotional Functioning
Table 3 shows the only significant group difference in self-reported social and emotional functioning was higher levels of anxiety in the DLD group (p =.04, Cohen’s d = − 0.66). The other self-reported measures followed a similar pattern but the comparisons were not significant, except both groups reported very similar levels of perceived social support on the PSS-Fr. Parent reports of emotional (p =.019, Cohen’s d = − 0.74) and peer problems (p =.003, Cohen’s d = −.06) were significantly higher in the DLD group than the TLD group.
Table 3
Mean (and standard error) ratings of social and emotional scales from self-report and parent-report
Self-report
DLD
TLD
DLD vs. TLD
p
Cohen’s d (95% CI)
Warwick Edinburgh Mental Wellbeing Scale (WEMWBS)
51.50 (1.94)
54.27 (1.30)
− 3.07 (− 8.26, 2.13)
0.241
0.34 (− 0.22,0.90)
Perceived Social Support– Friendship (PSS-Fr)
11.25 (0.79)
11.27 (0.53)
− 0.11 (− 2.06, 1.83)
0.908
0.01 (− 0.55,0.56)
Revised Children’s Manifest Anxiety Scale (RCMAS)
21.96 (3.18)
13.19 (2.09)
1.07 (0.04, 2.11)*
0.043
− 0.66 (−1.21, − 0.11)
Moods and Feelings Questionnaire (MFQ)
15.88 (2.87)
9.31 (1.86)
0.76 (−0.32, 1.83)
0.162
− 0.55 (− 1.11,0.01)
Social Anxiety Scale– Adolescents (SAS-A)a
44.55 (3.26)
39.31 (2.72)
− 0.14 (−0.07, 0.36)
0.190
− 0.36 (− 0.91,0.19)
Parent-report
SDQ Emotional Problems
3.63 (0.63)
1.69 (0.41)
0.85 (0.14, 1.56)*
0.019
− 0.74 (−1.33,-0.15)
SDQ Peer Problems
3.21 (0.49)
1.08 (0.31)
1.07 (0.37, 1.77)**
0.003
− 1.06 (− 1.71,-0.41)
Statistics are b coefficients (95% confidence interval), controlling for IDACI Rank
a. n = 22 in DLD group
*p < 0.05, **p < 0.01
Social Evaluation Learning Task
Table 4 shows the mean outcome scores of Errors to Criterion (learning of the rule) and Global Rating (likeability) from the SELT. A mixed ANOVA with Errors to Criterion as the outcome shows a significant main effect of group (F(1183) = 14.63, p <.001, ηp2 = 0.07), illustrating the DLD group took longer to learn the rule in general. Additionally, there was a main effect of rule (Like vs. Dislike) (F(1, 183) = 19.33, p <.001, ηp2 = 0.10), with participants taking longer to understand the social judgement of the computer character when the rule was Dislike. There was a significant interaction between group and rule (F(1183) = 10.63, p =.001, ηp2 = 0.06) as the DLD group took significantly longer to understand the Dislike rule than the TLD group (t = 5.02, 95% CI [3.08, 10.06], p <.001). Furthermore, there was a significant interaction between condition (Self vs. Other) and rule (Like vs. Dislike) (F(1183) = 5.06, p =.026, ηp2 = 0.03). More errors were made in the Self-Disliked condition than the Other-Disliked condition (t = 2.81, 95% CI [0.18, 6.94], p =.033) suggesting that participants in both groups found it harder to learn they were disliked than George. This highlights the ‘positive self-bias’ paradigm previously highlighted by Button et al. (2015), where individuals tend to view themselves more favourably than others. There was no effect of condition (Self vs. Other) (p =.087).
Table 4
Mean global rating scores and errors to criterion for referential condition and rule by group
Outcome
Condition
Rule
DLD
TLD
Errors to Criteriona
Self
Like (80%)
4.71
4.08
Dislike (20%)
13.21
6.54
Average
8.96
5.31
Positive Biasb
8.50
2.46
Other
Like (80%)
5.83
3.85
Dislike (20%)
9.42
3.42
Average
7.63
3.64
Positive Bias
3.59
− 0.43
Global Ratingc
Self
Like (80%)
7.13
7.15
Dislike (20%)
3.58
2.15
Average
5.36
4.65
Other
Like (80%)
6.46
7.04
Dislike (20%)
3.29
2.42
Average
4.88
4.73
a Mean number of errors made before 8 consecutive accurate responses (range = 0–24, lower score reflects better learning)
b Difference between errors made when learning Dislike rule compared to Like rule
c Mean rating of “How much did I like you/George?” at the end of each of the 4 conditions (ranging from 0 = Completely Dislike to 9 = Completely Like)
Rule: Like = 80% “correct” feedback for selecting positive word; Dislike = 20% “correct” feedback for selecting positive word
This positive self-bias is further illustrated in Fig. 3a and b, which show the cumulative mean accuracy for the different conditions plotted across each of the 32 trials. The DLD group was more accurate to begin with in the self-like condition (Fig. 3a), reflecting their tendency to choose positive words overall. Both groups were poorer at learning they were disliked to begin with, but the TLD group achieved higher accuracy over time, demonstrating that they were better at learning from the feedback provided. In the other-referential condition (Fig. 3b), the TLD group was better at understanding social cues in both the Like and Dislike rules. Again, they made more errors learning the Dislike rule to begin with but were very quick to learn that George was disliked and this slope was steeper than in the self-referential condition. The DLD group, however, struggled most with learning the Dislike rule, but not to the same extent as in the self-referential condition.
When asked explicitly what the computer thought of them the DLD group gave higher ratings than the TLD group, as Table 4 indicates. With global rating as the outcome, a mixed ANOVA showed a significant main effect of group (F(1183) = 4.12, p =.044, ηp2 = 0.02), with the DLD group giving higher ratings of ‘likeability’ overall. Higher ratings were given in the Like rule overall (F(1,183) = 244.40, p <.001, ηp2 = 0.57). Furthermore, there was a significant interaction between group and rule (F(1183) = 9.13, p =.003, ηp2 = 0.05); participants in the DLD group reported higher likeability ratings when the rule was Dislike (t = 3.55, 95% CI [0.33, 2.34], p =.003) but there was no difference for the Like rule (p = 1.00). There was no effect of condition (Self vs. Other) (p =.438) suggesting participants’ ratings were not influenced by whether the computer was judging participants themselves or George.
Fig. 3
a Cumulative mean accuracy over 32 trials for the “Self” condition. b Cumulative mean accuracy over 32 trials for the “Other” condition
Social Evaluation as a Predictor for Socioemotional Problems
In order to assess whether poor performance on the SELT predicted socioemotional problems we ran a mediation analysis with DLD status as the independent variable, overall errors to criterion as the mediator and socioemotional scales as the dependent variable. Parent-reported peer (b = 1.91, SE = 0.64, 95% CI [0.62, 3.20], p =.005) and emotional problems (b = 2.16, SE = 0.77, 95% CI [0.60, 3.72], p =.008) as well as self reported anxiety symptoms (b = 9.33, SE = 4.22, 95% CI [0.82, 17.84], p =.032) were modelled because there was a direct effect of DLD status on each of these outcomes. However, the mediation analysis revealed there was no mediation of SELT performance on peer problems (b = 0.39, SE = 0.33, z = 1.16, 95% CI [-0.27, 1.04], p =.245), emotional problems (b = 0.26, SE = 0.42, z = 0.62, 95% CI [-0.57, 1.10], p =.536), or anxiety symptoms (b = 3.61, SE = 2.41, z = 1.50, 95% CI [-1.11, 8.37], p =.134).
Language Ability and IQ as a Predictor of Social Evaluation
In order to test whether performance on the SELT was confounded by language ability, the Recalling Sentences subtest (expressive language) and the Word Classes– Receptive subtest (receptive language) were regressed against overall errors to criterion for each group. In the DLD group, there was no significant effect of expressive language (b = − 0.77, 95% CI [−1.66, 0.13], p =.089) or receptive language (b = − 0.28, 95% CI [− 1.24, 0.69], p =.556) on SELT performance. Similarly, SELT performance in the TLD group was not significantly predicted by expressive language ability (b = 0.31, 95% CI [− 0.20, 0.81], p =.225) or receptive language ability (b = − 0.61, 95% CI [− 1.23, 0.01], p =.053). Additionally, SELT performance in both the TLD and the DLD group was not predicted by IQ, as measured by the block design subtest (TLD: b = − 0.05, 95% CI [− 0.50, 0.41], p =.842; DLD: b = − 0.68, 95% CI [− 1.43, 0.06], p =.071).
Discussion
The current study aimed to provide a better understanding of the mechanisms involved in the relationship between DLD and increased socioemotional difficulties in adolescents. Social evaluation abilities were examined as a potential mediating factor using an interactive task, novel to the DLD population. The Social Evaluation Learning Task (SELT) measured online interpretation of social cues, based on the judgements of four computer characters, providing a closer insight into the social interactions and interpretations of adolescents with DLD. Participants updated their social perceptions using feedback from the computer before rating how much they thought the computer character liked them, or someone else. This provided a measure of how well adolescents with DLD can understand social inferences about themselves and others, a core skill for social interactions.
As expected, the DLD group reported higher feelings of anxiety and received higher parent-ratings of peer and emotional problems than their TLD peers. This is consistent with the literature (Conti-Ramsden & Botting, 2008). Conversely, the DLD group did not report more problems with depression, mental wellbeing or social support. This contrasts with previous findings, but it should be noted that many earlier papers investigated a much larger sample from the Manchester Language Study (MLS; e.g. Conti-Ramsden et al., 2013). The current findings of no significant difference in these domains may be due to the lower power associated with our lower sample size as well as the lower average age of our sample, which is just prior to increased incidence of depression that occurs in adolescence (Jones, 2013). Interestingly, Durkin and Conti-Ramsden (2007) reported that most participants from the MLS reported successful friendships, despite having significantly lower scores overall when compared to their TD peers, highlighting the heterogeneity of DLD. Similarly, Wadman et al. (2011a, b) found that adolescents with DLD perceived themselves to have adequate social functioning, although they still reported experiencing greater levels of social stress. In the current study, adolescents in both groups reported similar levels of social support from their friendships, suggesting that adolescents with DLD do not perceive themselves to experience the same level of peer difficulties that their parents report. This finding has been highlighted in previous studies by Forrest et al. (2021). Gough-Kenyon et al. (2021) found that the discrepancy between parent and adolescent report of wellbeing could be explained by social competence.
Despite no significant difference in self-report measures of social functioning there was a group difference in the understanding of social cues on the SELT. Compared to their TLD peers, adolescents with DLD were poorer at learning they were disliked. This finding is consistent with previous studies that demonstrate poor social cognition abilities in children and adolescents with DLD (Bakopoulou & Dockrell, 2016; Botting & Conti-Ramsden, 2008). Moreover, the DLD group tended to be more favourable in their ratings of how much they thought the computer character liked them (self-referential condition) or George (other-referential condition), as they chose more positive responses, particularly in the Dislike rule. This effect is somewhat surprising, as previous literature on social evaluation in adults indicates that higher anxiety relates to better learning from negative social cues (e.g., the dislike condition) (Button et al., 2012). However, even though the DLD group had higher anxiety they looked more like low anxiety adults from the previous literature, with good learning of the like condition and poorer learning in the dislike condition, particularly in the self referential tasks. This could reflect a general positivity bias in adolescents with DLD, even with high anxiety. This may also indicate that higher anxiety is related to other considerations in this population rather than a bias towards negative social cues.
Indeed, this positive outlook could be protective and may account for the lack of group difference in self-reports of wellbeing and social support if adolescents with DLD are poorer at recognising social problems. Indeed, previous qualitative research revealed that parents of young people with DLD believed their children’s withdrawal from social situations was an adaptive mechanism, allowing them to avoid potentially anxiety-inducing social interactions (Burnley et al., 2023). However, there was no association between SELT performance and socioemotional difficulties. This coincides with further research on this sample showing no associations between the Social Attribution Tasks and socioemotional outcomes (Forrest et al., 2023), but constrasts with other studies in this population which have demonstrated a relationship between social cognition deficits and poor social functioning (Andres-Roqueta et al., 2016; Bakopoulou & Dockrell, 2016; Botting & Conti-Ramsden, 2008). This may be due to the specific tasks that were used in the current study and our focus on social evaluation, both in this investigation and our previous study. While other aspects of social cognition might explain elevated levels of poor socio-emotional outcomes associated with DLD, it appears that social evaluation differences are not directly associated with these difficulties. Indeed, the findings in this paper indicate that the social evaluation differences may indeed be protective. Further research is necessary to elucidate the mechansism behind the social learning positivity bias we see in adolescents with DLD alongside higher anxiety.
Limitations
It is important to interpret these findings with caution, given the exploratory nature of the study. The sample size was small, which limits the power to detect significant differences (Button et al., 2013). Recruitment of the sample was difficult due to DLD being considered a childhood disorder, with more SLT services provided in the younger years. DLD is still relatively unknown compared to other developmental disorders (Bishop, 2010; Kim et al., 2023; Thordardottir & Topbaş, 2021) and the sample was further diminished by the exclusion of participants with high levels of autistic characteristics and technical problems resulting in missing SELT data. Despite this, significant group differences in parent-report of socioemotional problems, one self-report measure of anxiety and performance on the SELT were found.
The appropriateness of the SELT for this population, especially as it places a high demand on language, should be considered. To compensate for the high language demands, the list of word pairs was adapted for a younger sample and care was taken to ensure participants understood the task, such as reading aloud the instructions and providing definitions of the words. However, there was no effect of language ability on SELT performance. Additionally, the DLD group did not differ significantly from the TLD group when giving a “likeability” rating in the Like rule, indicating their difficulty was with interpreting the Dislike rule and not with understanding the concept of the SELT as a whole. This suggests that adolescents with DLD may have a difficulty accurately evaluating social cues separate to their language skills. However, examining this further with measures of syntactic or pragmatic language ability will add credence to this claim due to the importance of syntactic language for understanding mentalising vocabulary used in ToM tasks (Astington & Jenkins, 1999). The language measures in the current study were chosen because they are strong markers of DLD frequently cited in the literature (Conti-Ramsden et al., 2001) and we were cognisant of keeping the number of tasks relatively low so as not to fatigue participants. Future research should fully assess profiles of social pragmatic difficulties and how this particular profile, within the wider DLD profile, influences performance on the SELT.
Similarly, we did not include a full measure of IQ in this study for this same reason, instead focusing on a single subtest of the WISC-IV. While we did exclude individuals with very low scores on this subtest, we cannot exclude the possibility that a full scale IQ might have caught other DLD participants with very low global intelligence given our overall finding of a significant difference in NVIQ. However, the participant’s DLD diagnosis itself should have excluded participants with very low global intelligence. Additionally, similar to the language measures NVIQ did not predict SELT performance.
The SELT should be considered a form of linguistic social evaluation, where the participants are drawing on their semantic understanding of the words to determine what the computer thinks about them or “George”. While this is an innovative task to measure social evaluation, it does not encompass other aspects that are intricately involved in social evaluation, such as tone of voice, emotional expression and gestural or postural cues. All of these non-verbal cues are present in real-life interactions and may be processed differently by adolescents with and without DLD. The format of the SELT task could be used in future research integrating videos incorporating these cues, which could create a more naturalistic format in which to investigate social evaluation.
There is scope for future studies given the preliminary findings discussed in this paper and the lack of interventions to improve socioemotional functioning in adolescents with DLD (Arts et al., 2022). Inclusion of a social cognitive task that young people with DLD are known to have difficulty with, such as the Eyes Task or Strange Stories task, may help to evaluate the appropriateness of the SELT for this population. Additionally, a longitudinal design with performance on the SELT tested at the beginning and end of the school year could provide information about whether real-life social interactions influence social evaluation skills and socioemotional difficulties. The inclusion of peer ratings would also be interesting to examine whether the difficulty with understanding social cues that adolescents with DLD experience has an impact on others’ perceptions of them.
Conclusion
Better understanding of the mechanisms that are responsible for increased social and emotional problems in adolescents with DLD is crucial for parents and professionals to support young people effectively. This paper used a computerised social evaluation learning task (SELT) as a measure of social evaluation to explore the relationship between social understanding and socioemotional problems in adolescents with DLD. Although no link between SELT performance and socioemotional difficulties was found, adolescents with DLD struggled to understand they were disliked, demonstrating that adolescents with DLD have a significant difficulty learning negative social cues. However, they performed similarly to the TLD peers in comprehending positive social cues. The potential adaptive function of these findings warrants further investigation. Social evaluation is a necessary skill for navigating interactions with others. Drawing attention to these social differences highlights that DLD is not simply a difficulty with language, but it affects communication in all aspects of life.
Acknowledgments
This project was funded by a University Research Studentship Award from the University of Bath awarded to the first and last author.
Declarations
Conflict of interest
The authors disclose no conflicts of interest in relation to this manuscript.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Andres-Roqueta, C., Adrian, J. E., Clemente, R. A., & Villanueva, L. (2016). Social cognition makes an independent contribution to peer relations in children with specific Language impairment. Research in Developmental Disabilities, 49–50, 277–290.://WOS:000369117000025.PubMedCrossRef
Angold, A., & Costello, E. J. (1987). Mood and feelings questionnaire (MFQ). Durham: Developmental Epidemiology Program, Duke University.
Arts, E., de Castro, B. O., Luteijn, E., Elsendoorn, B., & Vissers, C. T. W. M. (2022). Improving social emotional functioning in adolescents with Developmental Language Disorders: A mini review and recommendations. Front Psychiatry, 13. https://doi.org/10.3389/fpsyt.2022.966008CrossRefPubMedPubMedCentral
Astington, J. W., & Jenkins, J. M. (1999). A longitudinal study of the relation between Language and theory-of-mind development. Developmental Psychology, 35(5), 1311–1320. https://doi.org/10.1037/0012-1649.35.5.1311CrossRefPubMed
Bakopoulou, I., & Dockrell, J. E. (2016). The role of social cognition and prosocial behaviour in relation to the socio-emotional functioning of primary aged children with specific Language impairment. Research in Developmental Disabilities, 49–50, 354–370. https://doi.org/10.1016/j.ridd.2015.12.013CrossRefPubMed
Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The autism-spectrum quotient (AQ): Evidence from asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31(1), 5–17.PubMedCrossRef
Baron-Cohen, S., Hoekstra, R. A., Knickmeyer, R., & Wheelwright, S. (2006). The autism-spectrum quotient (AQ)—adolescent version. Journal of Autism and Developmental Disorders, 36(3), 343.PubMedCrossRef
Beitchman, J. H., Wilson, B., Johnson, C. J., Atkinson, L., Young, A., Adlaf, E., Escobar, M., & Douglas, L. (2001). Fourteen-year follow-up of speech/language-impaired and control children: Psychiatric outcome. Journal of the American Academy of Child and Adolescent Psychiatry, 40(1), 75–82. https://doi.org/10.1097/00004583-200101000-00019CrossRefPubMed
Beitchman, J. H., Brownlie, E. B., & Bao, L. (2014). Age 31 mental health outcomes of childhood Language and speech disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 53(10), 1102–1110. https://doi.org/10.1016/j.jaac.2014.07.006CrossRefPubMed
Bishop, D. V. (2010). Which neurodevelopmental disorders get researched and why? PLoS One, 5(11), e15112.PubMedPubMedCentralCrossRef
Bishop, D. V. M., Snowling, M. J.,Thompson, P. A., et al. (2016). CATALISE: A multinational and multidisciplinary delphi consensus study. Identifying language impairments in children. Plos One, 11(7). https://doi.org/10.1371/journal.pone.0158753
Bishop, D., Whitehouse, A., & Sharp, M. (2009). Communication checklist-self report (CC-SR). Pearson.
Blakemore, S. J., & Choudhury, S. (2006). Development of the adolescent brain: Implications for executive function and social cognition. Journal of Child Psychology and Psychiatry, 47(3–4), 296–312. https://doi.org/10.1111/j.1469-7610.2006.01611.xCrossRefPubMed
Botting, N., & Conti-Ramsden, G. (2008). The role of language, social cognition, and social skill in the functional social outcomes of young adolescents with and without a history of SLI. British Journal of Developmental Psychology, 26(2), 281–300. https://doi.org/10.1348/026151007x235891CrossRef
Bulgarelli, D., Testa, S., & Molina, P. (2022). Theory of Mind development in Italian children with specific Language impairment and autism spectrum disorder: Delay, deficit, or neither?? Journal of Autism and Developmental Disorders, 52(12), 5356–5366. https://doi.org/10.1007/s10803-021-05389-8CrossRefPubMed
Burnley, A., St Clair, M., Bedford, R., Wren, Y., & Dack, C. (2023). Understanding the prevalence and manifestation of anxiety and other socio-emotional and behavioural difficulties in children with developmental Language disorder. Journal of Neurodevelopmental Disorders, 15(1), 17. https://doi.org/10.1186/s11689-023-09486-wCrossRefPubMedPubMedCentral
Button, K. S., Browning, M., Munafò, M. R., & Lewis, G. (2012). Social inference and social anxiety: Evidence of a fear-congruent self-referential learning bias. Journal of Behavior Therapy and Experimental Psychiatry, 43(4), 1082–1087. https://doi.org/10.1016/j.jbtep.2012.05.004CrossRefPubMed
Button, K. S., Ioannidis, J. P., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S., & Munafò, M. R. (2013). Power failure: Why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365–376.PubMedCrossRef
Button, K. S., Kounali, D., Stapinski, L., Rapee, R. M., Lewis, G., & Munafò, M. R. (2015). Fear of negative evaluation biases social evaluation inference: Evidence from a probabilistic learning task. Plos One, 10(4). https://doi.org/10.1371/journal.pone.0119456CrossRefPubMedPubMedCentral
Chen, J., Short, M., & Kemps, E. (2020). Interpretation bias in social anxiety: A systematic review and meta-analysis. Journal of Affective Disorders, 276, 1119–1130. https://doi.org/10.1016/j.jad.2020.07.121CrossRefPubMed
Clegg, J., Hollis, C., Mawhood, L., & Rutter, M. (2005). Developmental Language disorders a follow-up in later adult life. Cognitive, Language and psychosocial outcomes. Journal of Child Psychology and Psychiatry, 46(2), 128–149. https://doi.org/10.1111/j.1469-7610.2004.00342.xCrossRefPubMed
Conti-Ramsden, G., & Botting, N. (2008). Emotional health in adolescents with and without a history of specific Language impairment (SLI). Journal of Child Psychology and Psychiatry, 49(5), 516–525. https://doi.org/10.1111/j.1469-7610.2007.01858.xCrossRefPubMed
Conti-Ramsden, G., Botting, N., & Faragher, B. (2001). Psycholinguistic markers for specific Language impairment (SLI). Journal of Child Psychology and Psychiatry, 42(6), 741–748. https://doi.org/10.1111/1469-7610.00770CrossRefPubMed
Conti-Ramsden, G., Mok, P. L. H., Pickles, A., & Durkin, K. (2013). Adolescents with a history of specific Language impairment (SLI): Strengths and difficulties in social, emotional and behavioral functioning. Research in Developmental Disabilities, 34(11), 4161–4169. https://doi.org/10.1016/j.ridd.2013.08.043CrossRefPubMedPubMedCentral
Durkin, K., & Conti-Ramsden, G. (2007). Language, social behavior, and the quality of friendships in adolescents with and without a history of specific Language impairment. Child Development, 78(5), 1441–1457. https://doi.org/10.1111/j.1467-8624.2007.01076.xCrossRefPubMed
Ford, J. A., & Milosky, L. M. (2003). Inferring emotional reactions in social situationsdifferences in children with Language impairment. Journal of Speech Language and Hearing Research, 46(1), 21–30. https://doi.org/10.1044/1092-4388(2003/002)CrossRef
Forrest, C. L., Gibson, J. L., & St Clair, M. C. (2021). Social functioning as a mediator between developmental language disorder (DLD) and emotional problems in adolescents. International Journal of Environmental Research and Public Health, 18(3), 1221.PubMedPubMedCentralCrossRef
Forrest, C. L., Lloyd-Esenkaya, V., Gibson, J. L., & St Clair, M. C. (2023). Social cognition in adolescents with developmental language disorder (DLD): Evidence from the social attribution task. Journal of Autism and Developmental Disorders, 53(11), 4243–4257. https://doi.org/10.1007/s10803-022-05698-6
Frith, U., & Frith, C. D. (2003). Development and neurophysiology of mentalizing. Philosophical Transactions of the Royal Society B: Biological Sciences, 358(1431), 459–473. https://doi.org/10.1098/rstb.2002.1218CrossRef
Gallagher, T. M. (1993). Language Skill and the Development of Social Competence in School-Age Children. Language, Speech, and Hearing Services in Schools, 24(4), 199–205. https://doi.org/10.1044/0161-1461.2404.199CrossRef
Goodman, R. (1999). The extended version of the strengths and difficulties questionnaire as a guide to child psychiatric caseness and consequent burden. Journal of Child Psychology and Psychiatry and Allied Disciplines, 40(5), 791–799. https://doi.org/10.1017/S0021963099004096CrossRefPubMed
Gough Kenyon, S. M., Palikara, O., & Lucas, R. M. (2021). Consistency of parental and Self-Reported adolescent wellbeing: Evidence from developmental Language disorder [Original Research]. Frontiers in Psychology, 12(503). https://doi.org/10.3389/fpsyg.2021.629577CrossRef
Hart, K. I., Fujiki, M., Brinton, B., & Hart, C. H. (2004). The relationship between social behavior and severity of Language impairment. Journal of Speech Language and Hearing Research, 47(3), 647–662. https://doi.org/10.1044/1092-4388(2004/050)CrossRef
Hoffmann, J. A., Hobbs, C., Moutoussis, M., & Button, K. S. (2024). Lack of optimistic bias during social evaluation learning reflects reduced positive self-beliefs in depression and social anxiety, but via distinct mechanisms. Scientific Reports, 14(1), 22471. https://doi.org/10.1038/s41598-024-72749-6CrossRefPubMedPubMedCentral
Jenkins, J. M., & Astington, J. W. (1996). Cognitive factors and family structure associated with theory of Mind development in young children. Developmental Psychology, 32(1), 70–78. https://doi.org/10.1037/0012-1649.32.1.70CrossRef
Kilpatrick, T., Leitão, S., & Boyes, M. (2019). Mental health in adolescents with a history of developmental Language disorder: The moderating effect of bullying victimisation. Autism & Developmental Language Impairments, 4, 2396941519893313.CrossRef
Kim, J. H., Davies, B., & Xu Rattanasone, N. (2023). Have you heard of developmental Language disorder?? An online survey. Communication Disorders Quarterly, 44(4), 228–238. https://doi.org/10.1177/15257401221115822CrossRef
Kim-Cohen, J., Caspi, A., Moffitt, T. E., Harrington, H., Milne, B. J., & Poulton, R. (2003). Prior juvenile diagnoses in adults with mental disorder: Developmental follow-back of a prospective-longitudinal cohort. Archives of General Psychiatry, 60(7), 709–717. https://doi.org/10.1001/archpsyc.60.7.709CrossRefPubMed
La Greca, A. M., & Lopez, N. (1998). Social anxiety among adolescents: Linkages with peer relations and friendships [journal article]. Journal of Abnormal Child Psychology, 26(2), 83–94. https://doi.org/10.1023/a:1022684520514CrossRefPubMed
Lloyd-Esenkaya, V., Russell, A. J., & St Clair, M. C. (2020). What are the peer interaction strengths and difficulties in children with developmental Language disorder?? A systematic review. International Journal of Environmental Research and Public Health, 17(9), 3140.PubMedPubMedCentralCrossRef
Marton, M., Abramoff, B., & Rosenzweig, S. (2005). Social cognition and Language in children with specific Language impairment (SLI). Journal of Communication Disorders, 38(2), 143–162. https://doi.org/10.1016/j.jcomdis.2004.06.003CrossRefPubMed
Nilsson, K. K., & de Lopez, K. J. (2016). Theory of Mind in children with specific Language impairment: A systematic review and Meta-Analysis. Child Development, 87(1), 143–153. https://doi.org/10.1111/cdev.12462CrossRefPubMed
Procidano, M. E., & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal of Community Psychology, 11(1), 1–24. https://doi.org/10.1007/BF00898416CrossRefPubMed
Redmond, S. M. (2011). Peer victimization among students with specific Language impairment, Attention-Deficit/Hyperactivity disorder, and typical development. Language Speech and Hearing Services in Schools, 42(4), 520–535. https://doi.org/10.1044/0161-1461(2011/10-0078)CrossRefPubMed
Reynolds, C. R., & Richmond, B. O. (1978). What I think and Feel - Revised measure of childrens manifest anxiety. Journal of Abnormal Child Psychology, 6, 271–280. https://doi.org/10.1007/Bf00919131CrossRefPubMed
Ruffman, T., Slade, L., & Crowe, E. (2002). The relation between children’s and mothers’ mental state Language and Theory-of‐Mind Understanding. Child Development, 73(3), 734–751. https://doi.org/10.1111/1467-8624.00435CrossRefPubMed
Semel, E., Wiig, E., & Secord, W. (2006). Clinical evaluation of Language fundamentals-UK (CELF-4 UK). In: Pearson Education.
St Clair, M. C., Pickles, A., Durkin, K., & Conti-Ramsden, G. (2011). A longitudinal study of behavioral, emotional and social difficulties in individuals with a history of specific Language impairment (SLI). Journal of Communication Disorders, 44(2), 186–199. https://doi.org/10.1016/j.jcomdis.2010.09.004CrossRefPubMed
St Clair, M. C., Horsham, J., Lloyd-Esenkaya, V., Jackson, E., Gibson, J., Leitão, S., & Botting, N. (2023). The engage with developmental Language disorder (E-DLD) project: Cohort profile. International Journal of Language & Communication Disorders, 58(3), 929–943. https://doi.org/10.1111/1460-6984.12835CrossRef
StataCorp. (2015). Stata statistical software: Release 14. StataCorp LP.
Suslow, T., Hußlack, A., Kersting, A., & Bodenschatz, C. M. (2020). Attentional biases to emotional information in clinical depression: A systematic and meta-analytic review of eye tracking findings. Journal of Affective Disorders, 274, 632–642. https://doi.org/10.1016/j.jad.2020.05.140CrossRefPubMed
Tennant, R., Hiller, L., Fishwick, R., Platt, S., Joseph, S., Weich, S., Parkinson, J., Secker, J., & Stewart-Brown, S. (2007). The Warwick-Edinburgh mental Well-being scale (WEMWBS): Development and UK validation. Health and Quality of Life Outcomes, 5, 63–63. https://doi.org/10.1186/1477-7525-5-63CrossRefPubMedPubMedCentral
Thordardottir, E., & Topbaş, S. (2021). How aware is the public of the existence, characteristics and causes of Language impairment in childhood and where have they heard about it? A European survey. Journal of Communication Disorders, 89, 106057.PubMedCrossRef
van Harmelen, A. L., Kievit, R., Ioannidis, K., Neufeld, S., Jones, P. B., Bullmore, E., Dolan, R., Fonagy, P., Goodyer, I., & Consortium, N. (2017). Adolescent friendships predict later resilient functioning across psychosocial domains in a healthy community cohort. Psychological Medicine, 47(13), 2312–2322.PubMedPubMedCentralCrossRef
Vissers, C., & Koolen, S. (2016). Theory of Mind deficits and social emotional functioning in preschoolers with specific Language impairment. Frontiers in Psychology, 7. https://doi.org/ARTN
Voci, S. C., Beitchman, J. H., Brownlie, E. B., & Wilson, B. (2006). Social anxiety in late adolescence: The importance of early childhood Language impairment. Journal of Anxiety Disorders, 20(7), 915–930. https://doi.org/10.1016/j.janxdis.2006.01.007CrossRefPubMed
Wadman, R., Botting, N., Durkin, K., & Conti-Ramsden, G. (2011a). Changes in emotional health symptoms in adolescents with specific Language impairment. International Journal of Language & Communication Disorders, 46(6), 641–656. https://doi.org/10.1111/j.1460-6984.2011.00033.xCrossRef
Wadman, R., Durkin, K., & Conti-Ramsden, G. (2011b). Close relationships in adolescents with and without a history of specific Language impairment. Language, Speech, and Hearing Services in Schools, 42(1), 41–51. https://doi.org/10.1044/0161-1461(2010/10-0003)CrossRefPubMed
Wechsler, D. (2004). The Wechsler intelligence scale for Children - fourth edition UK. Pearson.
Yew, S. G. K., & O’Kearney, R. (2013). Emotional and behavioural outcomes later in childhood and adolescence for children with specific Language impairments: meta-analyses of controlled prospective studies. Journal of Child Psychology and Psychiatry, 54(5), 516–524. https://doi.org/10.1111/jcpp.12009CrossRefPubMed
Zhao, X., Lynch, J. J. G., & Chen, Q. (2010). Reconsidering Baron and kenny: Myths and truths about mediation analysis. Journal of Consumer Research, 37(2), 197–206. https://doi.org/10.1086/651257CrossRef
