Introduction
Stuttering occurs when a person shows disruptions or blockages during speech. It usually presents as repeating a part of the word, stretching the voice, and locking [1]. The first signs of developmental stuttering are seen between 3 and 5 years, followed by a relatively fluent speech period [2]. On average, the prevalence of stuttering in preschool children is 3% to 5% and is 3 to 4 times higher in boys than girls [3]. In addition to the mentioned symptoms, some secondary behaviors are added to stuttering, which causes negative attitudes and feelings towards speaking [4]. Over time and as a result of these consequences, a person experiences various emotions such as failure, frustration, fatigue, and fear. These negative emotions are added to the early symptoms and reduce a person’s social participation in various social situations [5]. Currently, most clinical experts and specialists pay attention only to the external and visible aspects of stuttering, and the important internal aspect of a person with stuttering, especially in children, is neglected.
In treating hidden aspects of stuttering, like other preventative measures, the skills must first be turned into reliable information. Given that the negative impact of stuttering on social relations, emotions, and behaviors of a person with stuttering has been proven in various studies [3, 9], the first test to assess the attitude of children with stuttering was made by Silverman in 1970 in which the individual’s attitude and strategies for communication were evaluated [10].
One of the common tools for measuring attitude in preschool children with stuttering is Kiddy CAT, which directly assesses their communication attitude. Kiddy CAT examines the hidden characteristics of stuttering and is designed to measure a person’s emotional, behavioral, and cognitive reactions to stuttering [13]. 
Among the studies done in Iran in this field, we can mention the study of Yadegari et al. (2005), but with a different age group from ours. In this study, the communication attitude test was translated into Persian, and its psychometric properties were assessed among 48 students with stuttering and 312 students without stuttering, aged from 8 to 15 years, and natives of the Persian language. According to the results, the Cronbach α coefficient was equal to 0.83, the Guttmann split-half reliability coefficient of the test was equal to 0.76, and the Spearman-Brown correlation coefficient was equal to 0.78. The results proved that children with stuttering had more negative communication attitudes than children without stuttering [18]. 
Therefore, in the first place, we need a suitable and valid tool to identify and measure these adverse effects and then take the necessary actions to eliminate or reduce them. Considering the inquiries done by the researchers of the present study, unfortunately, there is no suitable tool for evaluating and examining these aspects for children with stuttering in Iran. So, we must provide such a tool for this group of children in the country [9].
Materials and Methods
This study is a methodological and validation study and was conducted for 6 months in the spring and summer of 2018 in Isfahan City, Iran. Sixty subjects included the parents of preschool children with stuttering who were literate, and the age of their children ranged from 3 to 5 years and 11 months old, 36 boys and 24 girls who went to speech therapy clinics at this level in Isfahan.
It should be noted that these subjects had been referred to the speech therapy clinic for the first time, and the treatment interventions had not yet started for them; they had no previous history of speech therapy. They participated in our study after doing the necessary assessments and ensuring stuttering and their families’ consent. The study procedure included test design, validity review, reliability review, and data analysis. The inclusion criteria were being 3 to 6 years old, receiving no previous treatment for stuttering, the presence of stuttering symptoms in a child with average intelligence, normal hearing, lacking perceptual language and speech disorders except stuttering, and the natural speech processing based on the history taking and language and speech pathologist’s evaluation. Also, the parents should be literate, and written consent from the parents was required before the study. The exclusion criteria were the incidence and confirmation of natural speech impairment of the child during sampling and the unwillingness of the participants to continue the study. This study consisted of several stages as follows.
Step 1: Questionnaire design
This stage was qualitative, and its purpose was to gather information about children’s behaviors and attitudes about stuttering. In the questionnaire design stage, the participants were selected from the parents of children with stuttering based on purposive sampling. Ten parents participated in this stage. Unstructured and in-depth interviews were used to collect information. The sampling process continued until no new data appeared during the data acquisition. In other words, the data became saturated. Then, based on the results of the interviews and the opinions of experts, as well as questions that are very similar to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) diagnostic categories, the following scales were obtained. The concepts were divided into the dimensions of social, emotional, and behavioral disorders, and the main constructs of the questionnaire included the structure of “uncompromising/hyperactive behavior,” “social skills,” “communication skills,” “aggression,” “fear,” and “separation anxiety.”
Step 2: Determining the validity of the questionnaire
In the second stage, three methods were employed to evaluate the validity of the questionnaire, including face validity (qualitatively), content validity (content validity ratio and content validity index), and construct validity using confirmatory factor analysis (Analysis of moment structures).
Step 3: Determining the reliability and stability of the questionnaire
In the third stage, 60 parents of preschool children were asked to fill in the questionnaire to determine the internal consistency using the Cronbach α. To evaluate the stability in the interval of two weeks to one month in the first test, it was repeated for 30 of the prototype volumes. The results were calculated by the test-retest method and comparing scores of the two stages through the cluster correlation index.
Results
In this study, test design, validity review, reliability review, and data analysis were performed as follows. In the test design stage, 61 initial items in 6 scales of the questionnaire were reduced to 40 after merging the expressions with overlapping concepts. The tool contains very low “1,” low “2,” medium “3,” high “4,” and very high “5” options. Because of the non-uniformity of the questions and prevention of the parents’ being directed about the child’s disorders, we attempted to express the questions in forwarding and inverse ways. Questions 2, 11, 12, 14, 15, 17, 18, 36, and 40 are scored in reverse, so the answers “I completely agree” and “I completely disagree” got scores of 1 and 5, respectively. For example, in question 2, “He takes the turn” is a reverse question, and therefore the answer “I completely disagree” from the parent indicates the existence of a disorder in the child. In contrast, there are direct questions, so “I strongly agree” gets a score of 5, and “I strongly disagree” gets a score of 1. For example, question 1, “It disturbs others,” will prove a sign of a disorder in the child if the parents answer “I completely agree.” It should be mentioned that unanswered questions will be given a score of zero. The minimum total score of this questionnaire is 0, and the maximum total score is 200.
These items are grouped in 6 constructs or subtests: “uncompromising/hyperactive behavior (containing 12 items)”, “social skills” (containing 6 items), “communication skills” (containing 8 items), “aggression” (containing 5 items), “fear” (containing 4 items),” and “separation anxiety” (containing 5 items). The tool has very low “1,” low “2,” medium “3,” high “4,” and very high “5” options. Based on the results, the values of content validity ratio (CVR) and content validity index (CVI) of the whole questionnaire were 0.76 and 0.90, respectively, but the validity of some questions was less than the minimum which they were removed or corrected according to the comments received from the experts. 
In the construct validity index, according to the output of AMOS software, the calculated value of χ2 equals 631.25. The presence of lower χ2 indicates the proper fit of the model (P=0.000) because the lower the value of χ2, the better the proposed model. The following results are obtained from the software output (Table 1).


The results of non-standard estimation software demonstrated that the measurement model is suitable because its Chi-square and RMSEA values are less than 0.07, and GFI and AGFI values are more than 0.8. The values showed a significant level so that only questions 39 and 18 were excluded from the model. This means these two questions do not have a significant coefficient in the desired dimension and should be removed from the model. Because of the limitation in the volume of the article, Table 2 presents only the results of estimating the path coefficients, standard deviation, and significance of questions 18 and 39, which lacked the desired significance level.


Internal consistency of the questionnaire was equal to 0.82 before considering the validity of the structure using the Cronbach α coefficient. In the second stage, after doing factor analysis, the Cronbach α of the questionnaire in general and in all dimensions (by omitting two questions, 39 and 18 in factor analysis) was obtained: the results are presented in Table 3.


Stability assessment was repeated by the test-retest method using the Intracluster Correlation Index (ICC) criterion with an interval of two weeks to one month for 30 prototypes; the results are presented in Table 4.


Discussion and Conclusion 
In the content validity section, the CVR and the CVI were calculated, which led to the omission of question 2 from the dimension of uncompromising/hyperactive behavior. Questions 4, 5, 12, 20, and 21 were also changed in spelling. Hyrkas et al. (2003) recommended a score of 0.79 and above for accepting items based on a CVI score [26]. In the next step, based on the mean scores of the content validity index of all expressions of the questionnaire, the Average Content Validity Index (S-CVI/Ave) of the questionnaire was calculated. Polit and Beck (2006) recommended a score of 0.90 or higher for S-CVI/Ave acceptance [16]. The S-CVI/Ave of the questionnaire was also calculated, which was 0.9; considering that it is more than 0.79 and based on the content validity sources, the present questionnaire has content validity. Based on the findings, the S-CVI/Ave of the questionnaire had a desirable level (0.90), except for question 32, which was transferred to the separation anxiety dimension due to lack of relevance and having a score less than 0.90. Also, the CVR for the whole questionnaire was 0.76. According to the Lawshe method, when 15 experts are used to determine the content validity, the CVR should be 0.51 and higher. So, the validity index of the social, emotional, or behavioral disorders questionnaire of preschool children with stuttering is confirmed. The strength of this study compared to other studies is that a quantitative method has been used to determine the validity of the content [24]. The construct validity of the questionnaire was assessed by confirmatory factor analysis on the remaining 39 questions. The results showed that the values of the Chi-square and RMSEA were less than 0.07, and the values of GFI and AGFI were more than 0.8, so the model had a good fit (P<0.001). On the other hand, estimation of path coefficients showed that questions 18 and 39 with path coefficients of 0.35 and -0.07, respectively, could not remain in the model and were therefore excluded from the study (P>0.05). Therefore, the confirmation of the validity of the questionnaire by confirmatory factor analysis and according to the results of the model and coefficients of the paths and by removing the above two questions proved that these study results are consistent with Dadsetan’s study [6]. The psychometric properties of the social and emotional skills questionnaire were examined in normal preschool children, while the present questionnaire has reported a specific and detailed study of disorders among children with stuttering; so far, no attention has been paid to such disorders as stuttering in children. 
A value of more than 0.7 of the Cronbach α coefficient indicates good and desirable coherence of questions; a value of zero of this coefficient indicates unreliability, and +1 indicates complete reliability [25]. Therefore, the Cronbach α coefficient of the presently designed questionnaire equals 0.89, indicating the internal consistency above the instrumental expressions. It confirms the reliability of the questionnaire for measuring social, emotional, and behavioral disorders in preschool children with stuttering. This study is consistent with the study of Shahim (2008) [20], who used the Cronbach α coefficient to measure the reliability of the preschool children’s behavioral problems questionnaire, which were 0.89, 0.80, and 0.70 for the three factors of aggression, inattention and childish behaviors, and isolation and anxiety, respectively. It is also similar to the study of Yadegari (2005) [18], which translated the communication attitude test into Persian, and its psychometric properties were assessed among 48 students with stuttering and 312 students without stuttering, aged from 8 to 15 years and natives of Persian language. The Cronbach α was equal to 0.83. One of the differences between this study and Shahim (2008) [20]and Yadegari (2005) [18] is the study population. In Shahim’s study, the population was normal preschool children, while in ours, the study population was preschool children with stuttering. On the other hand, the population of the present study consisted of 3 to 6 years old preschool children with stuttering, but the statistical population of Yadegari’s study was 8 to 15 years old preschool children with stuttering [18]. Also, the stability of the present questionnaire was evaluated using the test-retest method, which was the criterion for measuring the stability of the questionnaire of the ICC. If this index is higher than 0.70, the stability rate is desirable. The closer the number to 1, the more desirable it is [25]. According to the results, the correlation between the questionnaire clusters was statistically significant (P<0.001). The results of this study are consistent with the study of Yadegari et al. (2005), in which the reliability of the communication attitude test among 8 to 15 years old primary school children with stuttering, obtained through the Guttman split-half coefficient and Spearman-Brown correlation coefficient, were equal to 0.76 and 0.78, respectively [18]. Considering the results, this questionnaire has good stability over time (not limited to a specific time), so the results have sufficient stability, too. It means that if 3 to 6 years old preschool children with stuttering are sampled twice in the intervals of two weeks to one month, there is no significant difference between the first and the second samplings, and the results are highly correlated.
The study results show that the questionnaire has good internal consistency, test stability, and, in general, good reliability. The analysis also proved that all the test questions have good reliability and validity and can be used in clinical and research applications. According to the obtained results, the Persian version of the questionnaire for assessing social, emotional, and behavioral disorders in preschool children (especially parents) has adequate validity and reliability to determine the effect of negative feelings and emotions on preschool children and their parents. It can be used as a suitable and useful tool. This way, it helps evaluate the child before treatment and obtain more information about stuttering. It can also be employed as a suitable tool to measure the treatment results and determine the therapeutic effects.
The limitations of this study are the small sample size, difficulty in finding samples with the necessary conditions, the lack of similar tools in Persian to check the concurrent validity of the questionnaire, and the study setting, which was only in one city (Isfahan).

Ethical Considerations
Compliance with ethical guهdelines
The ethics code of this research is IR.MUI.RESEARCH.REC.1393.055.

Funding
This study was carried out with the support of Isfahan University of Medical Sciences research department and project code is 393828.

Authors' contributions
All authors equally contributed to preparing this article.

Conflict of interest
The authors declared no conflict of interest.


References

1. Clark CE, Conture EG, Frankel CB, Walden TA. Communicative and psychological dimensions of the KiddyCAT. Journal of Communication Disorders. 2012; 45(3):223-34. [DOI:10.1016/j.jcomdis.2012.01.002] [PMID] [PMCID]
2. Onslow M. Treatment of stuttering in preschool children. Behaviour Change. 2004; 21(04):201-14. [DOI:10.1375/bech.21.4.201.66104]
3. Proctor A, Yairi E, Duff MC, Zhang J. Prevalence of stuttering in African American preschoolers. Journal of Speech, Language and Hearing Research. 2008; 51(6):1465-79. [PMID]
4. Bluemel CS. Primary and secondary stammering. Quarterly Journal of Speech. 1932; 18(2):187-200. [DOI:10.1080/00335633209379870]
5. Boey RA, Van de Heyning PH, Wuyts FL, Heylen L, Stoop R, De Bodt MS. Awareness and reactions of young stuttering children aged 2-7 years old towards their speech disfluency. Journal of Communication Disorders. 2009; 42(5):334-46. [DOI:10.1016/j.jcomdis.2009.03.002] [PMID]
6. Dadsetan P, Asgari A, Rahimzade S, Bayat M. [Kindergarten Inventory of Social/Emotional Tendencies: A cross-validation study (Persian)]. Developmental Pschology. 2010; 7(25):7-27. [Link]
7. DeVon HA, Block ME, Moyle-Wright P, Ernst DM, Hayden SJ, Lazzara DJ, et al. A psychometric toolbox for testing validity and reliability. Journal of Nursing Scholarship. 2007; 39(2):155-64. [DOI:10.1111/j.1547-5069.2007.00161.x] [PMID]
8. Guitar B. Stuttering: An integrated approach to its nature and treatment. Philadelphia: Lippincott Williams & Wilkins; 2013. [Link]
9. Yaruss JS, Quesal RW. Overall Assessment of the Speaker’s Experience of Stuttering (OASES): Documenting multiple outcomes in stuttering treatment. Journal of Fluency Disorders. 2006; 31(2):90-115. [DOI:10.1016/j.jfludis.2006.02.002] [PMID]
10. Silverman EM. Communication attitudes of women who stutter. Journal of Speech and Hearing Disorders 1980; 45(4):533-9.  [PMID]
11. Guitar B, McCauley RJ. Treatment of stuttering: Established and emerging interventions. Philadelphia, PA: Lippincott Williams & Wilkins; 2010. [Link]
12. Vanryckeghem M, Brutten GJ. The BigCAT: A normative and comparative investigation of the communication attitude of nonstuttering and stuttering adults.Journal of Communication Disorders. 2011; 44(2):200-6. [DOI:10.1016/j.jcomdis.2010.09.005] [PMID]
13. Vanryckeghem M, Brutten GJ. Communication attitude test for preschool and kindergarten children who stutter. San Diego, CA: Plural Publishing; 2007. [Link]
14. Anderson JD, Pellowski MW, Conture EG, Kelly EM. Temperamental characteristics of young children who stutter. Journal of Speech, Language, and Hearing Research. 2003; 46(5):1221-33. [PMID]
15. Guttormsen L, Kefalianos E, Næss KA. Communication attitudes in children who stutter: A meta-analytic review. Journal of Fluency Disorders. 2015; 46:1-14. [DOI:10.1016/j.jfludis.2015.08.001] [PMID]
16. De Nil LF, Brutten GJ. Speech-associated attitudes of stuttering and nonstuttering children. Journal of Speech and Hearing Research. 1991; 34(1):60-6. [PMID]
17. Bernardini S, Vanryckeghem M, Brutten GJ, Cocco L, Zmarich C. Communication attitude of Italian children who do and do not stutter. Journal of Communication Disorders. 2009; 42(2):155-61. [DOI:10.1016/j.jcomdis.2008.10.003] [PMID]
18. Yadegari F, Darooei A, Farazi M, Alipour Heidar M. Communication attitude in normal and stuttering students.  Journal of clinical psychology Andisheh va Raftar. 2005; 11(1):100-2. [Link]
19. Hossein Zadeh N, Shahbodaghi M, Jalaei S. [Reliability and validity of “ behavioral checklist “ and”communication attitude test” in stuttering children and comparison with non stutters at 6-11 years old (Persian)]. Journal of Modern Rehabilitation. 2010; 4(1-2):30-7. [Link]
20. Shahim S Yousefi F. [Preschool Behavioral Problems Questionnaire Special for parents (Persian)]. Social Sciences and Humanities of Shiraz University. 2008; 15(1):19-32. [Link]
21. Paul Klein. Easy Guide factor analysis. Translation solvent Sadrossadat, Asghari enamel. Tehran: The study of designing books Humanities University (left), 1380. 254 (1). [Link]
22. Knapp TR, Brown JK. Ten measurement commandments that often should be broken. Research in Nursing & Health. 1995; 18(5):465-9. [DOI:10.1002/nur.4770180511] [PMID]
23. Lawshe CH. A quantitative approach to content validity. Personnel Psychology. 1975; 28(4):563-75. [DOI:10.1111/j.1744-6570.1975.tb01393.x]
24. Polit DF, Beck CT, Owen SV. Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Research in Nursing & Health. 2007; 30(4):459-67. [DOI:10.1002/nur.20199] [PMID]
25. Saifpanahi S, Sobhani Rad D, Afzali M, Izanloo S, Mardani N, Gholamian M, et al. [An investigation of the correlation between phonological and visual working memory with severity of stuttering in 6-12 years-old children (Persian)]. Journal of Paramedical Sciences & Rehabilitation. 2015; 4(4):20-6. [DOI:10.22038/JPSR.2015.5344]
26. Hyrkas K, Schmidlechner KA, Oksa L. Validating on instrument for clinical supervision using an expert panel. International Journal of Nursing Studies. 2003; 40(6):619-25. [DOI:10.1016/S0020-7489(03)00036-1]