A cross-sectional study on the impact of parental educational level and gender differences on emotional and behavioural difficulties in adolescents in Qatar

ABSTRACT Heterogeneous findings on the prevalence of emotional and behavioural difficulties (EBD) have been reported in Middle Eastern countries. Moreover, the prevalence of EBD among adolescents in Qatar is unknown, making it difficult to implement the necessary educational or psychosocial programs. This study used a cross-sectional methodology to assess emotional and behavioural problems among 533 high school students from the public government schools. Only 10.6% of the students were found to have EBD, and the prevalence varied significantly in different EBD subscales, as identified using the Strengths and Difficulties Questionnaire (SDQ). The reliability of the scales was good. The population of this study had lower mean EBD scores than those reported in three other Middle Eastern countries. There were significant differences in EBD between males and females, with a higher prevalence in females. No effect of parental education was observed, suggesting that in a society with high economic stability and an excellent social structure, parental education may not be an independent factor influencing the EBD in adolescents. More studies are needed to confirm this hypothesis; however, the findings of the current study will be of importance in establishing the use of the SDQ as a tool for gauging EBD and prioritising necessary interventions.


Introduction
Emotional and Behavioral Difficulties (EBD) are a set of mental and behavioural disorders that usually occur during childhood and adolescence (World Health Organization 2016). Adolescence is associated with the emergence of a group of EBD, such as aggression, disruption, impulsivity, and oppositional conditions, requiring specialised evidence-based psychological and educational services (Kauffman and Landrum 2017;Zaheer et al. 2019). In educational settings, EBD can adversely affect the ability to learn or build positive relationships; therefore, EBD are linked to academic challenges, and the paths between behavioural and academic performance can be bidirectional (Van der Ende, Verhulst, and Tiemeier 2016;Mitchell, Kern, and Conroy 2019).
Early identification of EBD can help families find specialised educational and medical assistance to stop or prevent the development of psychological disorders (Green et al. 2013). The prevalence of EBD can vary depending on parenting techniques and gender (Schweder and Raufelder 2019;Serpell et al. 2009). Studies investigating the impact of cultural factors on EBD among adolescents are abundant in Western countries (Chowdhary et al. 2014;Nicolas et al. 2009;Peterson, Villarreal, and

Parental education and EBD in adolescents
Studies have indicated that parental factors play a crucial role in the development of EBD in children (Butterfield et al. 2021). The vicious cycle of parental impunity and children's emotional and behavioural psychopathology originates from improper child-rearing techniques. When children are exposed to several risk factors and persistent psychosocial stresses, such as poverty, their behavioural problems tend to deteriorate (Teekavanich et al. 2017). Therefore, it stands to reason that children from less-than-ideal backgrounds -those marked by poor parental education, financial status, or a history of trauma -are more likely to develop EBD (Harland et al. 2002). One of the essential parental factors influencing children's emotional disorders is parental style (Carlson et al. 2020). The pattern of parental care, whether participatory or single, directly impacts children's behaviour (Amato, Kane, and James 2011). Parents' education directly affects the pattern of care they provide to their children (Hadjicharalambous and Demetriou 2020). Parental education is an indicator of parents' role in promoting their children's positive development (Morris et al. 2020). Jiang (2022) concluded that parental education is a beneficial intervention to prevent EBD in children because it helps parents create a supportive home environment for their children. 2019). Adolescent females are more likely to withdraw, have bodily issues, experience depression, and behave aggressively (Kim et al. 2014). Several studies have stated that females are more likely to seek help from mental health services than males (Bener and Ghuloum 2011;Bildt and Michélsen 2002;Walton et al. 2021). Despite the tendency of females to experience mental health problems, studies conducted in the Arab world have revealed that when dealing with mental health problems, females are more likely than males to adhere to superstitious beliefs and seek the advice of traditionally accepted healers from their cultures and religions (Bener and Ghuloum 2011).

The current study
Middle Eastern cultures prefer to manage EBD concerns through community-based solutions led by family or religious organisations rather than seeking professional therapy (Mohammadzadeh, Awang, and Mirzaei 2020). Therefore, it is suspected that many adolescents in Qatar suffer from EBD but do not seek professional care. Nonetheless, the accurate identification and treatment of psychological and behavioural disorders in adolescents are essential for effective management. Furthermore, information on EBD must be shared in the community and schools so that parents are aware of these difficulties early in their children's development and can intervene to avoid more severe psychological problems (Green et al. 2013). This study aimed to determine the prevalence of EBD among adolescents in Qatar, focusing on parental education and gender disparities. The secondary objective of this study was to examine the reliability of the SDQ and its subscales in adolescents in Qatar. The findings of this study can assist educators, stakeholders, and policymakers in identifying students who require assistance and treatment with suitable psychological and social services. This approach is in line with the concept of early intervention prevention, which advocates a proactive strategy to address behavioural anomalies as soon as they are identified.

Participants
The participants in this study were 533 students who were randomly selected from eight public schools in Qatar. The average age (M) was 16.3 years (range = 15-19 years, SD = 1.1). In terms of gender, 367 participants were male (63.7%), and 166 were female (36.3%). Owing to the diversity of Qatari schools, the nationalities of the study participants varied, including 38% Qataris and 62% Arabs of other nationalities (Palestine, Syria, Iraq, Oman, Saudi Arabia, and Bahrain). The respondents self-rated their parents' education. The percentage of fathers who did not complete a university education was 43.2% (N = 249), whereas 43.7% (N = 233) had a university education. Among mothers, 47.9% (N = 276) had a below-university education, and 45.1% (N = 260) had a university education.

Data collection procedures
This research project was approved by the ethical review board at Qatar University as well as the Ministry of Education and Higher Education in Qatar. Consent forms were sent to all participants to obtain their permission. All participants had parental consents and agreed to participate themselves. Electronic consent forms were sent to all participants to obtain their permission. Secondary schools in Qatar are classified based on gender into 'all-male' and 'all-female' schools. All the schools were selected from Doha, Qatar's capital city. To obtain a representative sample of the general population, schools were chosen from different regions within the Al-Doha municipality, which contains more than 80% of the general population. Two schools-one all-female and one all-male -were selected from each city in Doha.
The students in the eight schools selected for this study were all in grades 10, 11, and 12. Students were given questionnaires through a link in their learning management system (LMS). Due to the COVID-19 outbreak at the time of the study, schools were put on lockdown, and all learning was conducted online. The Ministry of Education and Higher Education sent text messages to students as reminders to complete the questionnaires and submit them through the link on the LMS.

Instruments
An Arabic version of the Strengths and Difficulties Questionnaire (SDQ) (Goodman 1997) was used for adolescents in Qatar. The SDQ is a well-known self report screening tool available in various languages and has been used in several countries (Alyahri and Goodman 2006;Du, Kou, and Coghill 2008;El-Keshky and Emam 2015;Woerner et al. 2004). The demographic survey obtained information on gender, age, study strand (science, arts, and technology), self-rated parental education, and socioeconomic status. The Arabic-Qatari version of the SQD was developed in 2012 (Al-Hendawi and Keller 2012). The results suggest a high level of validity and reliability indicators that justify the scale's use in this study. This study used a five-scale structure because evidence from other Gulf-Arabic regions supports the use of this structure (Alyahri and Goodman 2006;El-Keshky and Emam 2015).
The SDQ consists of 25 items divided into five subscales of five items each. The SDQ describes children's and adolescents' positive and negative attributes that can be assigned to emotional problems, conduct problems, hyperactivity, peer problems, and prosocial behaviour subscales. Each item in each subscale is self-rated on a 3-point scale, with 0 indicating 'not true,' 1 indicating 'somewhat true,' and 2 indicating 'certainly true.' Subscale scores were calculated by adding the selfrated scores for each item of the scale (after recoding reversed items; range, 0-10). Higher scores on the four difficulty subscales for emotional problems, conduct, hyperactivity, and peer problems indicated difficulties. In contrast, a higher score on the prosocial subscale indicates strength. The SDQ and sociodemographic questionnaires were provided in Arabic.

Statistical analyses
Statistical analyses were conducted using Statistical Package for Social Sciences (SPSS) software version 26.0. To assess the quality of SDQ responses, descriptive statistics of the SDQ subscale scores, reliability, internal consistency, and correlation coefficients were used. Descriptive statistics of demographic variables, inferential statistics, and mean differences were calculated using ANOVA and MANOVA to substantiate the impact of parental education and gender on the SDQ.

Results
SDQ distribution and scores were computed for the SDQ subscales (emotional problems, conduct problems, hyperactivity, peer problems, and total difficulties). Cronbach's alpha reliability scores (presented along with the mean scores and standard deviations in Table 1) were calculated using the scores of each item for each subscale. The difficulty scores were within normal values when all 20 items that made up the difficulty subscales (emotional problems, conduct problems, hyperactivity, and peer problems) of the SDQ were used. With the exception of the Peer Problems Subscale, which had an α = 0.4, Cronbach's alpha reliability scores were above 0.5.
For the standardised SDQ, all interpreted means were within the normal ranges. Peer problems had the highest mean score among the difficulty subscales (emotional problems, conduct problems, hyperactivity, and peer problems). Conduct problems had the lowest mean scores. Based on the mean score, adolescents were classified as normal (difficulties score: 0-15), borderline (difficulties score: 16-19), or abnormal (difficulties score: 20-40) ( Table 2). Overall, 62 adolescents (10.6%) were classified as abnormal; the most frequently observed abnormal interpretations were related to emotional problems, accounting for 10.1% (n = 58) of the adolescents. This was followed by hyperactivity (n = 50, 8.6%), conduct problems (n = 21, 3.7%), and peer problems (n = 18, 3.2%).
Furthermore, the data were stratified according to gender. Females had a higher percentage of emotional problems (n = 33, 15.7%). Hyperactivity was observed in 29 males (7.9%) and 21 females (10.2%). In contrast, 4 (1.1%) males and 8 (3.9%) females had abnormal conduct problems. Parental education was recoded into one variable by combining fathers' and mothers' education and then classified into three categories based on the distribution: high, middle, and low parental education.
A MANOVA was initially conducted on the four dependent SDQ variables (emotional problems, conduct problems, hyperactivity, and peer problems). Three major MANOVA assumptions were evaluated; however, only one of the three major multivariate normality assumptions was satisfied. The first MANOVA assumption was that there was a sufficient sample size within the crossed levels (2 × 3 MANOVA design); this was satisfied at 20% and higher of the sample at each level. However, a test of normality using the Kolmogorov-Smirnov test showed statistical significance for all dependent variables (p < 0.001), which violated normality assumptions. Furthermore, multicollinearity between the dependent variables was tested. Mahalanobis' Distance (MD) measures the extent of multivariate outliers. The maximum MD value was 25.5 (df = 3). The critical value of MD with df = 3 is 16.27 (Penny, 1996); thus, the data violated one of the main assumptions of the MANOVA, implying that several dependent variables were outliers.
The second option was to perform two-way ANOVA for each dependent variable. First, the sample size was adequate within crossed levels (2 × 3). The second assumption of ANOVA, homogeneity of variance, was met using the Levine test of homogeneity, which tests the null hypothesis that the error variance of the difficulty subscales is equal between parental education levels and the gender of participants. The Levine F-values were all insignificant, implying that the hypothesis was accepted and that there was equal variance in the dependent variable across combinations of independent factors. An interaction combination was used to test the third assumption of data normality. Several outliers were identified when the levels of the independent variables of gender and parental education were combined. Although one assumption was violated and the other two were acceptable, the study proceeded with the analysis because of the robustness of ANOVA statistics.
The means and standard deviations of each level of analysis (Table 3) were examined before performing the ANOVA. The first section of Table 3 presents the mean values for parental education. The hyperactivity subscale had the highest mean score (M = 3.4, SD = 2.28) for parents with a middle education. The conduct problem subscale had the lowest mean (M = 1.7, SD = 1.68) for parents with a middle education. The differences between male and female adolescents are presented in Table 3. Females had the highest mean for hyperactivity (M = 3.38, SD = 2.22), whereas males had the lowest mean for conduct problems (M = 1.7, SD = 1.52). Females had a higher mean for difficulties (M = 11.62, SD = 6.12) than males (M = 10.22, SD = 5.89).  Table 4 presents the results of the ANOVA. The two-way ANOVA factorial design used parental education (3-levels) and gender (2-levels) to examine their impact on each SDQ subscale (emotional health, conduct problems, hyperactivity, peer relationship problems, and difficulties). The impact of parental education was insignificant for all four subscales. Conversely, gender was statistically  significant (p < 0.05) for emotional and conduct problems. Females had higher mean scores than males and scored higher on the overall measure of difficulty. The ANOVA did not reveal any differences in the interaction effects.

Discussion
This study examined EBD among adolescents in Qatar and investigated the influence of gender and parental education on the different subscales of EBD, to provide baseline data and raise awareness among policymakers regarding psychiatric and mental health services for adolescents (Davidson et al. 2015). Excluding peer problems, Cronbach's alpha values were greater than 0.5 for all subscales.
The reliability of Qatari adolescents on peer problems was similar to that of other Arab adolescents in studies among Omanis (α = 0.3; Al-Mukhani, Bekker, and Cottrell 2018). The lower-than-anticipated reliability is most likely attributable to SDQ item 21, where almost equal numbers of adolescents responded that they were likely to get along with adults and were not likely to get along with adults.
Re-examining this question may increase the overall reliability of the instrument because it does not represent a problem with peers but rather culturally acceptable behaviour that may be acceptable in one culture but not another (Alyahri and Goodman 2006). Three studies from Middle Eastern countries were compared with findings from this study using a t-test for independent samples ( Table 5). The first was a study of Jordanian adolescents (Atoum, Alhussami, and Rayan 2018), the second was a study of Omani adolescents (Al-Mukhani, Bekker, and Cottrell 2018), and the third was a study of Saudi Arabian adolescents (El-Keshky and Emam 2015). Jordanian and Omani adolescents ranged in age from 11 to 6 years, while Saudi adolescents ranged in age from 6 to 8 years. In this study, adolescents from Qatar ranged in age from 16 to 9 years, with an average age of 16.1 years. Comparative analysis revealed that the Qatari sample had the lowest mean scores compared to the Jordanian, Omani, and Saudi Arabian samples. There were significant differences in the three subscales between Qatari and Jordanian adolescents: emotional problems, conduct problems, and hyperactivity.
The Jordanian sample exhibited lower mean scores on the peer problems subscale. Indeed, in Jordan, where the majority of the population is Palestinian, those who are 'urban' may feel stronger affiliations to their constituency and less to the boundaries of a profoundly traditional Bedouin community that makes up the rest of Jordanians. These dichotomies may have an impact on peer relationships. Jordan's recent economic meltdown has dampened hopes for future employment and well-being compared to affluent Qatar. Adolescents are forced to make difficult decisions because of perceived corruption and economic difficulties (Obermeyer 2015). Jordan's lower gross domestic product in relation to its vast population means that many young people in Jordan may experience a negative social and economic impact on their lives compared to Qatar. On the SDQ subscale, there were no significant differences in the mean scores between Qatari and Omani adolescents on any of the four subscales. Oman and Qatar are Arab Gulf countries with shared values and traditions, stable, prosperous economies, deep family ties, and communal solidarity. In Arab regions, where such cushions do not exist, adolescents' social and emotional problems are likely to be prominent.
Our study found abnormal difficulties scale for 10.6% of adolescents. In Egyptian adolescents, Mowafy et al. (2015) discovered that 13.7% of their participants experienced behavioural issues.Al-Mukhani, Bekker, and Cottrell (2018) study on Omani adolescents discovered that 9% of participants reported difficulties, whereas the prevalence of difficulty in Jordanian adolescents was 11.7%, as reported byAtoum, Alhussami, and Rayan (2018). Thus, with the highest percentage scores on emotional problems (highest level of abnormalities) and the lowest scores on peer problems (high levels of normality preceded by conduct problems), the findings of the Jordanian (Atoum, Alhussami, and Rayan 2018), Omani (Al-Mukhani, Bekker, and Cottrell 2018), and Egyptian (Mowafy et al. 2015) studies are consistent. The high prevalence of EBD in Egypt, as in Jordan, may be due to the social and economic constraints Egyptians face, which have seen a series of social upheavals and violence The symbol * indicates significance at the 1% level.
in recent years worsened by persistent economic problems. The most recent Arab 'Spring' socially split the nation and established a bleak future perspective among adolescents.
The central thesis of this study was that high levels of parental education could act as a buffer against EBD among adolescents. For example, parents with higher education can better cope with situational demands and manage their resources for their children than those with lower education. Stressful parents contribute to their children's EBD (Deater-Deckard 2008). Furthermore, parents' proclivity for high-stress levels due to combining work obligations and raising children may lead to emotional distress in children, which is reflected in their negative emotional conduct (Xuan et al. 2018). Parents with high levels of education may be able to organise and manage their personal lives in ways that support their children emotionally, intellectually, and socially. The findings of this study did not reveal any significant differences in the levels of parental education or EBD among adolescents. One reason for this anomaly could be that Qatar is a close-knit and collective society, with family life at the core; therefore, if children do not find emotional support at home, they may turn to other family members and close-knit communities. Furthermore, Qatar has high GDP per capita, and substantial family earnings may greatly ease the situational demands and personal resources of parents. When evaluating the parental effect on the emotional, behavioural, and subsequent consequences of adolescent achievement in life, future research must incorporate income and socioeconomic position as critical determinants.
Gender differences in overall difficulties were revealed by the ANOVA results, with females having higher mean scores than males. There were significant differences between females and males in emotional and conduct problems, with females scoring significantly higher than males did. In contrast, female participants had lower and insignificant scores on the hyperactivity and peer problems subscales than did male participants. Therefore, gender differences in the emotional problems subscale were highlighted. This is consistent with other studies in the Arab region, where Jordanian (Atoum, Alhussami, and Rayan 2018), Egyptian (Mowafy et al. 2015), Omani (Al-Mukhani, Bekker, andCottrell 2018), and Iranian (Mohammadi et al. 2013) female scored higher on the emotional problems subscale. Notably, these differences have also been observed in Asian studies. For example, Yao et al. (2009) discovered that female participants had higher scores on the emotional problems subscale than male participants in a sample of Chinese adolescents. In contrast, females scored lower on the conduct and the peer problems subscales than males. Emotional problems persisted among female adolescents, emphasising the importance of addressing overt expressions of feelings among parents, school counsellors, and school psychologists in a highly masculine and patriarchal society that creates a culture of vulnerability among females and produces negative emotions. Consequently, enabling young girls to express positive emotions is essential to their sensitivity and empowerment as productive and successful individuals.

Limitations
This study had certain limitations that should be considered when interpreting the findings. First, I present the results of the self-rated SDQ. The nature of the self report questionnaire, particularly in terms of the validity of the measures, limits the complete interpretation of the analysis to a process in which teachers, parents, and participants provide a rating on the SDQ. Second, the participants in this study ranged in age from 15 to 19 years, excluding younger adolescents from the sample. Furthermore, the low response rate was a limitation caused by various reasons, including a lack of incentives (Fowler 2013) and a cultural reluctance to share personal information that might reflect negatively on the family, extended family, or the 'tribe' in the Arab Gulf context. This study did not compare the findings with those of Western countries because of the sociocultural disparities. Carefully constructed systematic reviews may be beneficial for comparing and contrasting findings in different countries and settings.
Furthermore, there is a lack of a research culture that emphasises the value of research and data, making adolescents the least likely to respond to questionnaires, particularly in their stage of human development. These limitations allow to delve deeper into providing a comprehensive profile of the existing emotional and behavioural abnormalities. In addition to parental education, other socioeconomic factors such as parental income and employment status may be significant predictors of emotional and behavioural problems. Reduced parental monitoring in children is associated with adverse emotional and behavioural outcomes. Thus, including parents' employment status, income, and the number of siblings in the analysis would provide a complete picture of children's diverse behaviours (Ary et al. 1999). Other limitations of this study include the lack of confirmatory factor analysis, which could have helped assure the validity of the instrument in this population. Another limitation of this study is the presence of a large expatriate population in Qatar, where adolescents attend international schools, many of whom were born in Qatar and have entirely different cultural and family upbringings than the local Arab and Qatari adolescents. Cultural predispositions to the SDQ will be detailed in a sample that includes and represents all adolescents in Qatar.

Implications
Our study indicated that approximately 10% of adolescents had EBD. This information may serve as the primary input for national programs aimed at screening for EBD in schools, and will assist educators in developing individualised solutions for students with EBD. This strategy will aid in ensuring early counselling or psychiatric therapy for troubled students and contribute to an improvement in their academic performance. Analysis of this study underscores gender disparities in EBD among adolescents in Qatar. Policymakers, stakeholders, and the Ministry of Education Leadership should use the results and baseline data offered in this study to proactively assist female adolescents. The results of this study are also significant from an academic perspective. The fact that this study did not find parental education to be a factor that affects the incidence of EBD suggests that additional research into social support, parental financial stability, and general sociopolitical and economic stability is needed to examine the role of parental education. It should be noted that no effect of parental education was observed on any of the subscales. This factor is further emphasised by the fact that the incidence of EBD differs across Arabic regions, with economically and politically unstable Arab nations having a greater prevalence of EBD. These findings from Qatar, a nation with high per capita income and a strong social structure, might serve as a model for future research. Finally, pointing towards the need to examine the generalisability of the SDQ as a tool to gauge EBD and the underlying theoretical framework, the peer problem scale was found to have poor reliability. Such effects could be due to the ambiguous or inappropriate framing of some questions in this subscale. Other researchers will note this critical finding. In subsequent studies, the reliability of the peer problem scale in different contexts and settings will be examined, leading to the development of effective factor models and items that can be reliably used in different settings across all subscales.

Conclusion
This first study on EBD among adolescents in Qatar found an approximately 10% prevalence of EBD. The prevalence of various subscales in EBD varied greatly, with the emotional problems (close to 10%) having the highest incidence and the peer problems and the conducting problems having the lowest (both close to 3%). The reliability of the peer problem scale was lower than 0.5, which significantly questioned the adequacy of this subscale in this study. The prevalence of emotional problems, conduct problems, and difficulties were greater in females; however, parental education was not found to significantly influence any of the EBD subscales. A comparison with other Arab nations revealed that socioeconomic circumstances may influence the prevalence of EBD, with a higher prevalence in nations with economic and political instability. Further multicenter studies are needed to understand the influence of parental education on the prevalence of EBD. A more thorough examination of the factor structure of the SDQ and the reliability of different SDQ subscales in different settings is necessary before generalising the SDQ as a tool for measuring EBD in adolescents.