Introduction
Despite advancements in assistive hearing technologies, cochlear implants (CI), intervention services, and increased awareness of the effect of hearing impairment (HI) on speech and language development, language deficits remain a significant challenge for children with HI. Studies have demonstrated that children with HI acquire language at a rate of 50% to 60% of that of their normal-hearing peers. When these children enter school, they often have a language delay of 1 year or more compared to normal children [1]. Hearing difficulty influences various linguistic aspects, including vocabulary, phonology, and pragmatics. The morphosyntactic domain is an area that is considerably affected by HI. Studies have shown that children with HI have difficulty comprehending and expressing complex levels of grammar [1]. Furthermore, the mean length of utterance (MLU) is shorter in children with HI than in normal-hearing children [2]. The vocabulary of children with HI is also quantitatively and qualitatively inferior to that of their normal-hearing peers. Research indicates that the acquisition of new words in normal-hearing children is correlated with the size of their vocabulary reservoir rather than their age. Additionally, word learning strategies in children with HI develop at a later age, and in most of these children, no concept has a specific label, at least for a while [3]. Consequently, children with HI exhibit a delay in vocabulary knowledge compared to their normal-hearing peers. Among vocabulary concepts, words that do not refer to concrete concepts or activities, but rather are used to describe desires, feelings, emotions, and beliefs, are among the words that are acquired later in a child’s speech development [4]. Those words that represent an individual’s states, beliefs, and thoughts are called mental state words (MSW) in studies. These words are limited and less diverse in the speech of children with HI [5]. Acquiring these words can enhance self-awareness and cognitive control in such children [6، 7]. Among MSW, mental verbs begin to develop in children with normal development at the age of 3- and 5-year-olds have a full comprehension of these verbs [8]. However, the development of these verbs in children with HI is observed at the age of 7-8 years. Therefore, children with HI are at least 3-4 years behind in the milestones of this developmental-cognitive domain [9]. Since MSW occurs in only 10% of everyday conversations, and consequently, there is less verbal stimulation in this area for the child, learning these words is challenging. Additionally, MSW often has multiple meanings, and determining the correct meaning for a given phrase requires an underlying cognitive skill. This is why the development of MSW occurs after concrete words, and the learning path for MSW is longer. Maclean et al. investigated the development of MSW in children and believed that the development of these words should be enhanced with increasing age, being directly related to the frequency of using these words by others [10]. In children with HI, the development of MSW does not follow this pattern and generally shows differences. Various studies have cited various reasons for these differences, such as the difference in the development of cognition and theory of mind (TOM) in children with HI [11]. Moreover, various factors, such as the age of onset of hearing, chronological age, receptive vocabulary reservoir, and the frequency of exposure to these words can also be mentioned as factors affecting the emergence of MSW in children with HI [6]. What is regarded as the underlying cognitive skill for the acquisition of MSW is the ability of TOM [6، 12]. The cognitive capacity to understand the motivations behind human behaviors, beliefs, thoughts, and desires is defined as TOM [13-17]. Deaf children whose parents are deaf and use sign language and are significantly exposed to sign language function better in this ability compared to deaf children whose parents are hearing and who are delayed regarding oral language skills [9، 17].
Accordingly, based on previous studies, TOM in children with HI is more related to language skills rather than chronological age [18]. There are three hypotheses regarding the reasons for the relationship between language disorders and the function of children with HI in TOM tasks. First, children with HI may not have adequate verbal experience to refer to mental states. Second, the use of these words is also dependent on the ability to use certain levels of syntactic skills, which are disrupted in children with HI [7]. Third, these children may not have sufficient language skills to comprehend tests and instructions [19]. On the other hand, which language domain may be more closely related to TOM skills is unclear. Some studies have suggested a relationship between a child’s ability to comprehend and use grammar and their skills in TOM skills [18]. A part of language skills that may be related to the TOM domain is vocabulary, particularly MSW. TOM is indeed a foundation for learning and comprehending words that refer to other individuals’ emotions and intentions [8، 20]. If a child, due to language deficiencies, is unable to communicate with others or if their communication with others is limited, they will have a weaker TOM and consequently, less developed MSW [6، 13، 21، 22].
Various studies have resulted in the relationship between linguistic features and TOM. Different aspects of the impact of TOM skills and their relation with various linguistic factors can be examined. Papafragou, et al. (2007) investigated the underlying processes of learning mental verbs by children. According to the results of their study, for mental verbs to enter a child’s vocabulary reservoir, the underlying mental concepts must first be acquired by the child and TOM is one of the most significant underlying abilities impacting the acquisition of mental verbs in children [23]. In a meta-analysis study to explore the correlation between false-belief understanding and language skills in children, Milligan et al. concluded that understanding the relationship between TOM and language skills was complex, and vocabulary skills had a weaker correlation with TOM tasks compared to overall language ability [24]. Comprehending MSW plays an important role in various language skills. The semantic recognition of these verbs and using them in sentences with more complex syntax strongly contribute to false-belief tasks [25]. Grazzani and Ornaghi investigated the relationship between language and TOM by assessing 110 elementary school students. They examined language skills, MSW, comprehension of metacognitive language, false-belief understanding (a subset of TOM) and comprehension of emotions. The results of their study demonstrated a moderate relationship between the development of TOM and the expression of MSW and a strong relationship between comprehension of MSW and the development of TOM. In addition, metacognitive language was the best predictor of function in false-belief tasks [7]. In another study, Yu et al. (2020) explored the relationship between language skills and TOM abilities in children with HI in 84 children aged 4-6 years with HI. According to the results of their research, children with more advanced language skills had also a higher level of TOM skills due to the long time they used assistive hearing devices. This finding demonstrates the importance of the relationship between TOM and language [26]. In addition, Delkhah et al. investigated the predictive role of language in TOM skills in normal children and those with CI. The results revealed that children with CI scored lower in all parts of the TOM test. Also, in 5-6-year-old children with CI, comprehension of mental verbs was a predictor of 53% of basic TOM skills. Moreover, in 8-9-year-old children, comprehension of dependent clauses in sentences and comprehension of mental verbs were predictors of 61% of basic TOM changes and 73% of advanced TOM changes, respectively [25]. 
It is still unclear which specific linguistic domain is more strongly correlated with TOM skills in children with HI, and this topic requires further investigation. Hence, the current study investigates the relationship between grammatical skills and vocabulary, particularly MSW, with TOM skills in children with HI.

Materials and Methods 
This was a descriptive-analytical cross-sectional study targeting children with HI. The intended sample was selected from children with HI referring to rehabilitation centers in Kerman Province, Iran. A total of 38 children with HI were selected using a convenience sampling method over three months. The inclusion criteria were having an age range of 5-7 years, a normal nonverbal intelligence quotient according to the Wechsler intelligence scale [27] and an MLU of more than 3. Hearing difficulties in all of these children were in the moderate to severe range (50% to 70%), confirmed by behavioral-auditory tests [28]. Additionally, the verbal language was the primary mode of communication for the subjects, and their families were Persian-speaking. Ultimately, all children used assistive hearing devices (hearing aids or CI). Furthermore, based on the exclusion criteria, children who had previously received TOM training were excluded from the research. Children with HI who were confirmed by the relevant specialists to have sensory disorders, motor disorders, pervasive developmental disorders, and attention deficit hyperactivity disorder were not included in the study.
All participants were assessed for their comprehension of mental verbs using a researcher-developed questionnaire. To prepare the assessment form, a list of mental verbs was extracted from English sources and translated into Persian. The mentioned list was then sent to 10 speech-language pathologists experienced in working with children with HI to comment on and assess the content validity ratio (CVR) of the items so that if the selected words were appropriate for the research purpose, they could be used for the test preparation [16]. The Lawshe method was used to assess the CVR. Experts’ responses to each part were categorized into one of three options: Essential, useful but not essential, or not essential. According to the Lawshe criterion, items with a CVR >0.62 could be retained in the test. Since almost all verbs achieved a CVR of 0.62 and the number of items was high, according to the experts’ opinions, only verbs with a CVR >0.8 were ultimately used in the test. Then, using the words approved by the experts, the researcher developed a form consisting of 20 short texts. For each text, a question was designed that required the child to comprehend mental verbs correctly. The prepared form was sent to 10 speech-language pathologists for CVR assessment using the Lawshe method. Finally, 9 texts containing 9 mental verbs were selected by speech-language pathologists with a CVR >0.62, which is considered a satisfactory level of validity. One text was also used separately as a guide. Moreover, the test-re-test was used to assess the test-re-test reliability. Ten children with HI, with similar conditions to the participants in the present research, i.e. 5-7-year-old children with 50%-70% hearing difficulties, without any sensory deficits other thanHI, no attention deficit hyperactivity disorder, and no previous training regarding comprehension of mental verbs and TOM, were selected for this study and underwent the test. After two weeks, the same test was re-administered to the same children to determine the test-re-test reliability. The intraclass CC was employed to assess the test reliability based on the test-re-test results. The results demonstrated a strong positive correlation between the first and second measurement scores. In addition, the present assessment tool had acceptable reliability over time. The correlation coefficient (CC) between the mean scores of the test-re-test was examined. The CC was performed between the test-re-test mean scores. The CC was 0.80 for the first text, which is considered good reliability, and >0.90 for the remaining texts, which is considered excellent reliability. The test was administered in the following manner. Each text was read to the child and then the child was asked the intended question that required the child to comprehend the mental verbs to answer. Each text included one mental verb and the eighth text included two mental verbs. Verbs such as deciding, thinking, and wishing were among the mental verbs in these texts. A trained examiner read these texts to all children in the same manner and immediately asked the child a question about their comprehension of the mental verbs. In response to the question, the child’s semantic level and grammar were not considered. If the child correctly conveyed their meaning in any way, they received a score of one. The test duration for each child was approximately 30 min.
The TOM assessment was conducted using the Persian version of the TOM test adapted by Ghamarani. This test has satisfactory validity in the Persian language. The test consisted of 38 questions and the child had to answer several questions about stories and pictures. The child looked at the images, and then the examiner asked the questions. Each story also contained 3 or 4 sentences that were explained by the tester. If the child did not understand the story’s meaning after it was told, the examiner would repeat the story [29]. Each question that the child answered correctly received a score of one. The sum of all scores indicated the child’s overall performance on the TOM test [29]. The test-re-test measurement was employed [30] to assess the test reliability for children with HI [31]. The test reliability was 0.8 for the first question and 0.756 for the third question. The reliability of other questions was equal to 1. The test took approximately 30 min to administer to each child.
Speech samples to examine MLU for each child were collected in the following manner. After selecting the participants, language samples were collected individually in a quiet room with the help of the child’s mother. For this purpose, the language sample was collected and recorded during a 30-min period of free play between the child and the examiner, as well as during the description of images from a series. The samples were then transcribed. Finally, the MLU in words was calculated for each child by dividing the number of words by the number of utterances in samples containing 100 utterances [32]. Sampling for assessing MLU was postponed until the third session to ensure that the child’s unfamiliarity with the environment did not hinder their genuine verbal output. Research data were analyzed using the SPSS software, version 21. In addition, the Kolmogorov-Smirnov test was used to assess the distribution of data, and correlation tests were employed to assess the relationship among comprehension of mental verbs TOM and MLU.

Results
The present research investigated the correlation among TOM, comprehension of mental verbs, and MLU in words in 5-7-year-old children with HI. The Kolmogorov-Smirnov test was used to examine the normality of the data distribution, and the results showed a normal distribution (P<0.05). Therefore, the Pearson CC was employed to assess the correlation between the variables (Table 1). 



Table 2 summarizes the descriptive statics for TOM, comprehension of mental verbs and MLU. To examine the correlation among comprehension of mental verbs, TOM and MLU in the study population, data from the assessment of these three skills were analyzed using the Pearson CC.



The relationship between TOM and comprehension of mental verbs
The relationship analysis using a 2-tailed test at a significance level of 0.05 revealed a significant relationship between TOM and comprehension of mental verbs in children with HI (Table 3).



The relationship between comprehension of mental verbs and MLU
The relationship analysis revealed a significant relationship between comprehension of mental verbs and MLU in children with HI.

The relationship between TOM and MLU
The relationship analysis revealed no statistically significant relationship between TOM and MLU in children with HI. The current study was conducted to investigate the relationship among comprehension of mental verbs, TOM, and MLU in children with HI. The results were examined using the Pearson CC.
According to the findings of the current study, there was a relationship between the scores of TOM and comprehension of mental verbs in children with HI. Various studies have also supported the link between language and TOM [19، 20]. Research has demonstrated that comprehension of mental states and the ability to respond to TOM tests require knowledge of MSW [20]. Discussing mental states enhances children’s TOM development. More precisely, language is an important tool for representing mental states and connecting them to reality [35]. Labeling can strengthen a child’s mental representation of intangible concepts [33]. A study by Barreto et al. (2016) showed that a child’s inclination to refer to mental states and TOM were two interconnected factors that influenced the child’s future communicative adequacy. This study reported a correlation between children’s TOM and speech about mental states [34].
Another finding of this study was the absence of a correlation between TOM and MLU as an indicator of syntactic development. A review of the existing literature reveals a reported connection between TOM and syntactic growth [33, 35]. A link between TOM and the use of relative clauses, which are among syntactic elements, has been observed in English [36, 37]. The use of compound sentences is associated with understanding explanations about mental beliefs [37]. Blake et al. [38]demonstrated that MLU is an indicator of grammatical complexity up to the age of 4-5 years, with an MLU exceeding four words at this age. In the present study, the children’s MLU was approximately four, indicating limited use of complex or compound sentences and, consequently, less frequent use of complex mental states in their utterances. Additionally, among linguistic and syntactic elements, some are more closely related to TOM development than others, with some studies preferring the general term language to describe the relationship with TOM without specifying a particular domain [33]. Recent studies have demonstrated that certain syntactic, semantic, and pragmatic linguistic elements are correlated with TOM until mid-childhood, but this correlation is not observed in adulthood [33]. Astington and Jenkins [12] argued that the ability to track and represent complex relationships between individual elements in a sentence might be the same ability that helps children understand the relationship between objects and the existence or non-existence of an object in a false-belief task. In support of this claim, they presented items from the test of early language development [4], uniquely predicting an individual’s skills on false-belief tasks [12].
Additionally, the ability to comprehend mental verbs was correlated with the participants’ MLU. During language acquisition, children learn to formulate utterances at various syntactic levels. This issue could indicate a semantic-syntactic relationship. When children begin learning and using MSW around the age of 3 years, they gradually acquire the necessary grammatical elements to employ these verbs in sentences as well [38]. Understanding and using MSW also culminate in a more complex development of children’s syntactic skills. When a child attempts to use MSW, they must also employ more complex syntactic structures. Thus, the growth of mental verbs can accompany syntactic growth in children’s language skills [39]

Conclusion
This study investigated the relationship among TOM, comprehension of mental verbs and MLU. According to the results, various linguistic aspects have differential effects on TOM skills and each other in children with HI. Given the relationship between comprehension of mental verbs and TOM, enhancing either skill may lead to improvements in the other. Moreover, considering the relationship between MLU, as an indicator of syntactic development, and the growth of mental verbs, different components of language likely exert a direct influence on each other. This issue could culminate in designing multifaceted goals for each interventional approach in these areas and saving time and cost in interventions for children with HI.

Study limitations
It was difficult to reach the intended number of children with HI at the assumed language level in the present study and it required time and coordination among several rehabilitation centers.

Future study recommendations
The relationship between MLU and TOM was not proven in this study, so more research is required with a larger sample size for more certainty. Meanwhile, examining the relationship between TOM and other syntactic indicators can be a good basis for further research. Also, the results of this research can lay the ground for clinical trials regarding the effect of TOM on strengthening comprehension of mental verbs.

Ethical Considerations

Compliance with ethical guidelines
This study was approved by the Ethics Committee of University of Rehabilitation Sciences and Social Health (Code: IR.USWR.REC.1402.094). The participants’ parents were provided with detailed explanations regarding the study objectives and were assured of the confidentiality of their data. The parents of participants who voluntarily agreed to participate in the study ultimately signed written informed consent forms.

Funding
This article is extracted from the research project (code: 3006), funded by University of Social Welfare and Rehabilitation Sciences.

Authors' contributions
Conceptualization and study design: Akbar Darouie and Parya Hajgholamrezaee; Methodology and data analysis: Fatemeh Hassanati and Zahra Hajgholamrezaee; Writing: Fatemeh Hassanati, Zahra Hajgholamrezaee and Parya Hajgholamrezaee; Editing, research and finalization: Fatemeh Hassanati and Zahra Hajgholamrezaee.

Conflict of interest
The authors declared no conflict of interest.