Level and Profile of Internalizing Behaviours Problem among Preschool Children : Evidence from Malaysia

Internalizing problems such as depression and anxiety disorders were prevalent in early childhood and these problems could lead to increasingly poor outcomes during school age then later in adult life. The main objectives of this paper were to examine the distribution and identify the level of internalizing behaviours problem among preschool children in Malaysia. Also, to come out with a profile of the children who were in the borderline and clinical status of internalizing problems. A total of 551 (46%) out of 1198 children age four and below were conveniently selected for this study that consist of those who were in the borderline and clinical range of having internalizing problems as rated by parents using the Child Behavioural Checklist for Age 1 1⁄2-5 (CBCL/1 1⁄2 –5). Descriptive statistics such as mean, standard deviation, t-test, ANOVA and correlation tests were applied in data analysis. Results revealed that level of internalizing problems among preschool children in Malaysia was high whereby 338 in clinical and 213 in borderline which represent about 28% and 18% of the total study samples respectively. Overall, female children with the age 3 years old, first born with 2 siblings and live with both parents were the most having internalizing problems. Most of these children had parents in the 30s age range and majority had secondary higher education background. Findings from this study provide evidences that early detection and intervention is crucially needed to prevent continuity of internalizing problems into school age and later to adulthood.


Introduction
The rapid developments in Malaysia have increased the interest in the early childhood development and create tremendous opportunities in education of preschool children.Population of children aged four years and below as stated in Malaysian Census 2010 was 3,291,816 with about 52% were boys and 48% girls (DOS Malaysia, 2011).Preschool education in Malaysia is provided by Ministry of Education (MoE), Jabatan Kemajuan Masyarakat (KEMAS) (Community Development Department), Jabatan Perpaduan Negara dan Integrasi Nasional (JPNIN) (Department of National Unity and Integration) and private operators.The percentage of children aged 4+ and 5+ who enrolled in preschools education had increased from 72.4% in 2010 to 90.7% in 2014 (EPU Malaysia, 2015).
Recognizing how important this development in charting a child from early childhood to school age then adulthood, many researchers focus on studying factors that contribute to child outcomes in the aspects of children behavioural and emotional problems (Caspi, 2000;Caspi et al., 2003;Sourander, 2001;Poulou, 2015).All children misbehave, it is sometime a norm to some and to others a disorder, however behaviours that are a disorder are much more serious.In addition to that, these problems could lead to increasingly poor outcomes in adulthood such as depressive disorder, fearfulness, academic underachievement and other disrupting behaviour problems (Poulou, 2015;Siu, 2008).
It has also been known that emotional and behavioural problems start at a young age (Sourander, 2001;Egger & Angold, 2006;Poulou, 2015).These problems can be categorized into two namely, internalizing and externalizing problems.The internalizing problem consists of problems that are mainly within the self or inner-directed and symptoms include emotionally reactive, anxiety/depressed, somatic complaints and withdrawal (Achenbach & Rescorla, 2000;Poulou, 2015).As such it can affect psychological and emotional condition.Meanwhile the externalizing problems are outer-directed, mostly encompass conflicts with other people and expectations of the child, such as attention problems and aggressive behaviour (Achenbach & Rescorla, 2000;Kristensen, Henriksen, & Bilenberg, 2010;Poulou, 2015).Symptoms of externalizing would include noticeable noncompliance, aggression toward peers, high activity level, and poor regulation of impulses (Campbell, Shaw, & Gilliom, 2000).
Both of these behaviour problems usually can be recognized during toddlerhood and will continue to persist into school age and adulthood unless proper intervention or treatment is given (Heberle, Krill, Briggs-gowan, & Carter, 2015).The children with co-occurring internalizing and externalizing problems are those having high chance to show persisting problems (Basten et al., 2015;Poulou, 2015) and preschool children experience emotional and behavioural difficulties with similar occurrence as older children (Egger & Angold, 2006).Once these behaviours are established, there is high possibility to be followed by an increased pattern of withdrawal, depressive disorders and related emotional behaviours in adolescence and young adulthood.
Internalizing problems have generally received less research attention as compare to externalizing problems (Mesman, Bongers, & Koot, 2001) due to their covert and non-intrusive nature (Siu, 2008).In many societies, these problems will go unnoticed or unidentified by parents and teachers.For instance, quiet, fearfulness, mild expressions of anxiety and dependency are behaviours considered normal for girls and are not seen as problematic as aggression, defiance and overactive (Keenan & Shaw, 1997;Siu, 2008;Basten et al., 2015).However, the World Health Organisation (WHO) predicts that, by 2030, internalizing problems (depression) will be second only to HIV/AIDS in burden of diseases (Mathers & Loncar, 2006).
Findings of sex differences in behaviour problems especially internalizing behaviour are mixed, as some studies report no child sex differences, while others have found significant differences in sex.During the early preschool years boys and girls appear to show similarities of behaviour problems (Keenan & Shaw, 1997).According to Poulou (2015), sex differences prior to age-4 are negligible and a study by Prior, Virasinghe and Smart (2005) highlighted that parents reported significantly higher on total behavioural problems for boys than girls because girls tend to become less aggressive or competitive in their preschool years.Also girls had more somatic complaints and their feelings were easily hurt (Sourander, 2001).Socioeconomic status (SES) was commonly measured using parent's occupational and highest education level.Parents and teacher reported that low SES groups showed significantly higher levels of problem behaviour than middle and high groups (Prior, Virasinghe, & Smart, 2005;Sourander, 2001).The low income households with few resources to buffer the effects of financial hardships, could experience more difficulties in the management of current and subsequent stressful life events and as such could increase the level of psychological distress and parenting (Bradshaw & Ellison, 2010;Kahn & Pearlin, 2006).According to Burlaka, Bermann and Graham-Bermann (2014) lower maternal education was related with children's internalizing behaviour problems.However, a parental high education level was related to a lower level of internalizing and externalizing symptoms (Sourander, 2001).Sourander explained that these discrepancies in many studies may be due to the differences in the measurement, cultural issue and biological issue or may reflect the difference in parental expectation.
The main objectives of our study were to examine the distribution and identify the level of internalizing behaviour problem.Also to come out with the profile of the children who were in the borderline and clinical status of internalizing problems by child and parents characteristics, while at the same time to investigate the magnitude of association between internalizing behaviour problems and four other subscales among preschool children in Peninsular Malaysia in order to understand and address the issues pertaining to such problems.In the present study, Child Behavioural Checklist for Age 1 ½-5 was used.

The Survey
The total samples for the survey were 1198 of preschool children age up to four years old.The participants were selected by using multistage random sampling technique at regional, state and preschool level.Total population of children up to four years old were selected at preschool level.Since this is a quantitative, descriptive, as well as correlational study, a set of questionnaires were used.Data were collected in 2013 and these questionnaires were given to the children parents or guardian to answer.The questionnaire contains two sections.The first section had questions related to background profile of the children, parents and teacher and the second section had questions asking child's behavioural and emotional problems.
For the purpose of this study, only 551 respondents were selected and analysed.Selection sample was based on the children who were in the borderline and clinical range of having internalizing behavioural problems.This contributed almost 46% of total study samples.
To measure internalizing problems, four subscales were included which consisted of emotionally reactive, anxious/depressed, somatic complaints and withdrawn.These problems reflect conflicts within self.Higher scores indicate greater problem behaviour whereas lower scores indicate otherwise.Normalized T-scores converted from the raw scores were used and T scores ≥ 60 were regarded as having internalizing problems in the child and can be further categorized as borderline (60 ≤ T ≤ 63) or clinical at T ≥ 64.

Demographic Background
Demographic information about parents and their children were provided by the parents themselves.These information were child gender, age, race, birth sequence and total sibling, parents' marital status, parents' income, age and level of education.

Statistical Analyses
Descriptive statistics was used to describe the major variables in this study.Percentages were used for categorical variables and for continuous variables means and standard deviations were used to describe the distribution.Independent sample t-test and ANOVA were used to examine the differences between major study variables with internalizing behaviour problem score.All these were carried out using the SPSS version 21.

Internalizing Scale and Status
Table 1 provides the information on the distribution of internalizing and its' subscales.Overall, the internalizing score ranged from 60 to 90 with a mean (M=65.68,SD=4.64) and 5% trimmed mean was 65.34, indicated in a clinical status (T ≥ 64).The median was 65.00 suggested that 50% of the respondents were in the clinical status.
Internalizing scale was measured using 4 subscales and the distribution of each subscales were shown below.The highest mean was emotional reactive, followed by somatic complaints, withdrawn and anxious/depressed.As for the subscales, when compared borderline with clinical status, three subscales had higher proportion of borderline which were emotional reactive, anxious/depressed and somatic complaints with 42.8%, 22.5% and 29.4% respectively.Only withdrawn had higher proportion of clinical (21.14%) as compared to borderline (14.7%) status.All four subscales, findings on one-way ANOVA test indicated that there were statistically significant differences in internalizing score among all three internalizing status groups.

Child's Profile
Table 3 focusing on the distribution of child's profiles.Overall, there was slightly higher proportion of girls (51.5%) reported as having internalizing behaviour problem compared to boys (48.5%) with mean of 65.86 (SD=4.88)for girls and 65.48 (SD=4.38)for boys.Vice versa for borderline status group whereby boys proportion (51.6%) was higher than girls (48.4%).For clinical status group, girls (53.5%) had higher proportion than boys (46.5%).These findings were in line with Keenan and Shaw (1997) who stated that in the early preschool years boys and girls appear to show similarities of behaviour problems and again by Poulou (2015) who mentioned that sex differences prior to age-4 were negligible.
Overall the mean score recorded for respondent age was 3.02 years (SD=0.85)and 42.6% of the respondents were at the age of 3 years, followed by 4 years (32.2%) and 2 years (20.6%),whilst respondents whose age 1 year was the minority group (4.6%).Their means were 64.60 (SD=3.89)for 1 year old, 65.31 (SD=4.71)for 2 years old, 65.67 (SD=4.4.32) for 3 years old and 66.00 (SD=5.00)for 4 years old.Similar findings as overall could be seen for borderline and clinical status groups.From the above, the internalizing behaviour problems keep increasing from age 1 to age 4 and this is in accordance with Heberle, Krill, Briggs-gowan and Carter (2015) who mentioned the increasing of internalizing problem from toddlerhood to school age.At school age, when children have to go to school, the internalizing problems might become more obvious, while at younger ages these problems are seen as a normal developmental emotion (Keenan & Shaw, 1997;Siu, 2008;Basten et al., 2015) and tend to be identified as non-problematic (Keenan & Shaw, 1997).In term of birth sequence, overall almost half (45.6%) were first-born child (eldest), 29.8% were third and above and 24.6% were second child.We also found similar findings for both borderline and clinical status groups.As for total sibling, all status groups showed similar proportion whereby most of the respondents had 2 siblings, followed by the only child and 3 siblings with more than 20% while the rest were 5 and more and 4 siblings had only between 9% -12%.Lastly for all status group, majority (above 81%) of these children live with parents, followed by between 10%-12% live with single parent and less than 7% live with others.

Parent's Profile
Overall monthly household income (HHI) range from MYR200 to MYR10,728.52,with a mean of MYR2082.26 (USD475.40,1USD=4.38MYR),almost two fifths compared to the mean HHI for Malaysia in 2012 which was MYR5000 (USD1,141.55)(EPU Malaysia, 2015b).For borderline, the mean for monthly HHI was MYR2187.28 (USD499.38) and for clinical was MYR2015.31(USD460.12).We can conclude that most of these children come from lower income group.This is in accordance to the previous study by Prior, Virasinghe, & Smart (2005) whereby low socioeconomic status is highly associated with problem behaviour in children.The impact of economic hardship had resulted in a reduced capacity for support and involved parenting, which later resulted in a higher prevalence of emotional and behavioural problems among preschool children.As for the parent marital status, large majority (above 91.9%) of these children were those staying with parents who were still married for all three groups (overall, borderline & clinical).
The overall mean for mother age was 32.07 years (SD=6.13)and majority (87.3%) of the respondents' mother having the age of 39 years and below and only 12.7% having age 40 years and above.We also found similar findings for both borderline and clinical status groups.In term of mother's highest education received, the mean score was 11.35 (SD=1.27)with large majority (91.6%) of respondents' mother had secondary higher and above education meaning more than 9 years of formal schooling attended.Only 1.6% had primary education and 6.8% had secondary lower education.Similar findings as overall could be seen for borderline and clinical status groups with large majority of 91.9% and 91.4% respectively for secondary higher and above education.In comparison with father's profile, father's age had mean score of 35.51 years (SD=7.24),slightly older as compared to mother's mean score.For all status groups, results showed that more than half were between the age of 30-39 years with overall (56.9%), borderline (60.9%) and clinical (54.3%).Mean score for father's highest education received was 11.03 (SD=1.45)with majority (overall=85.8%,borderline=86.4%,clinical=85.3%) of fathers had secondary higher and above, lower percentage compared to mother's highest education received.For secondary lower, the proportion was between 10.3%-11.5% for all status groups, whilst primary school was between 2.1%-4.3%.
The findings on parental highest education received were in-line with Sourander (2001) but not with Burlaka, Bermann and Graham-Bermann (2014) who suggested that lower maternal education was related with children's internalizing behaviour problems.These discrepancies in the results across studies again highlight that there exist several issues related to definition and identification method of measurements, cultural and developmental differences, and different expectation by parents (Keenan & Shaw, 1997;Sourander, 2001;Basten et al., 2015;Poulou, 2015).

Intercorrelations between Major Study Variables
In this correlations analysis, Table 5 showed the correlation matrix between internalizing behaviour problems and four subscales (emotionally reactive, anxious/depressed, somatic complaints and withdrawn).Results showed that all correlations were statistically significant at 0.01 level of significance with positive relationships.The magnitude of relationships between internalizing behaviour problems and all subscales were between moderate to high.The highest magnitude was emotionally reactive (r=0.706)followed by anxiety/anxious (r=0.661),somatic complaints (r=0.601) and withdrawn (0.592).Correlation is significant at the 0.01 level (2-tailed).

Conclusions and Implications
Previous studies (Caspi, 2000;Caspi et al., 2003;Egger & Angold, 2006;Poulou, 2015;Sourander, 2001) has highlighted that social and emotional competencies in early childhood are indications of a range of development outcomes among children during middle childhood and even later in their adult life.The aims of this study were to examine the distribution and level of internalizing behaviour problems among preschool children in Malaysia.Also, to find out the profile of these children who were in the borderline and clinical status of internalizing problems by child and parents characteristics.Results of present study provided update information on incidence of internalizing behaviour problems among preschool children in Malaysia.
This study revealed that the level of internalizing problems among preschool children in Malaysia was high whereby 338 (28%) in clinical and 213 (18%) in borderline out of the total study samples.Overall, female children with the age 3 years old, first born with 2 siblings and live with both parents were the most having internalizing problems.Most of these children had parents in the 30s age range and majority of these parents had secondary higher education background.In terms of correlation between internalizing problems with its four subscales, all correlations had positive relationships and were statistically significant at 0.01 level of significance.The emotionally reactive had the highest magnitude of correlation and the lowest magnitude was withdrawn.
Throughout many studies and from this study, there was a gradual emergence of higher incidence of girls having internalizing problems compared to boys even though some studies among preschool children do not provide sufficient evidence for sex differences (Egger & Angold, 2006).As explained by Hay (2007), these differences emerge due to earlier maturation of girls whereby they are biologically more inclined than boys are to the social pressures that protect them against aggression and differences in social influences whereby girls develop distinct ways of manipulating the social world, which often include verbal taunts and indirect aggression, meanwhile boys play in ways that may promote aggression in their male peer groups.
The inconsistent results across studies in documenting the incidence and profiles of internalizing behaviour problems, and its association with demographic characteristics of child and parents among preschool children may be due to a measurement issue (definition and identification), cultural issue, developmental and biological differences, or different expectations/perceptions by parents (Keenan & Shaw, 1997;Sourander, 2001;Basten et al., 2015;Poulou, 2015).
In conclusion, findings in this study provide enough evidences that early intervention or preventive measure is crucially required in order to reduce the continuity of internalizing problems from toddlerhood into middle childhood and later to adulthood.Intervention targeting at influencing factors (relative contribution of child, parent, and environmental factors) during early years may prevent a child's internalizing behaviours from progressing further or severely and causing damage during times of increased social, behavioural and academic demands prior to school age and throughout formal schooling.The preschool period is known to be the best time to identify and intervening these early signs of problems before they become permanent.Thus, is important to properly identify the children who actually need early intervention especially during preschool period as they represent a key component in building a comprehensive prevention and early intervention programmes.

Table 1 .
Distribution of the internalizing scale and subscale (n=551) study were in the clinical level and represent about 28% of the total study samples (1198).Whereas, only 38.7% were in borderline level and represent about 18% of total study samples.An independent samples t-test conducted to compare the internalizing score for two groups (borderline and clinical) found that there was a statistically significant difference in internalizing scores for borderline (M=61.62,SD=1.17) and clinical (M=68.23,SD=4.17) with t(416.59)= -27.46,p=0.0001.

Table 2 .
Proportion and t-Test/ANOVA of the internalizing scale and subscales by status (n=551)

Table 3 .
The distribution of the major study variables (n=551) -child's profile

Table 4 .
The distribution of the major study variables (n=551) -parents' profile

Table 5 .
Intercorrelations between major study variables