Differences in behavior problems and academic achievement between students with and without health impairments in Ethiopia

Abstract

Students with health impairments (SWHIs) are at greater risk of academic and behavioral difficulties in comparison to their peers without health impairments. However, school-based comparative studies with regard to differences in behavior problems and academic achievement between students with and without health impairments are quite rare. The purpose of this study was to investigate differences between behavior problems and academic achievement among students with and without health impairments; and examine whether behavior problems were related to academic achievement among SWHIs. A random sample of 483 students with and without health impairments have participated in the study. The total difficulties scale of the strengths and difficulties questionnaire (SDQ) was used to assess behavior problems and average aggregate marks were used to measure academic achievement of students. Descriptive statistics, one-way MANOVA and Pearson’s correlation coefficient were utilized to analyze the data. The result demonstrated significant differences in behavior problems and academic achievement between students with and without health impairments. Behavior problems had a significant negative relationship with academic achievement of SWHIs. SWHIs in Addis Ababa are at increased risk for higher rates of behavior problems and lower scores of academic achievement. Implications of the findings and suggestions for future research are discussed.

IMPACT STATEMENT

SWHIs in Addis Ababa have higher rates of behavior problems and lower scores of academic achievement compared to SWOHIs. The findings of this study suggest the need to develop and implement specific interventions targeted at reducing behavior problems and increasing academic achievement in SWHIs.

Keywords:

• Academic achievement
• behavior problems
• students with health impairments
• students without health impairments

REVIEWING EDITOR:

• Daryl O’Connor, University of Ulster: Ulster University, United Kingdom

SUBJECTS:

• Health Psychology
• Psychological Methods & Statistics
• Comparative Psychology
• Educational Psychology
• Mental Health

Introduction

Students with health impairments (SWHIs) are more prone to behavioral and academic difficulties than are students without health impairments (SWOHIs). Factors that seem to increase difficulties in school functioning experienced by SWHIs include restrictions in school activities (e.g. engaging in classroom activities, participating in sports and attending co-curricular activities) and socialization with peers, interruption in regular school attendance and disruption of typical developmental processes (Chan et al., 2005; Gorodzinsky et al., 2011). However, the academic and behavioral aspects of SWHIs have received relatively little attention in the empirical literature. Therefore, it is essential to understand the differences in academic achievement and behavioral problems in students with and without health impairments to identify targets for school-based behavior and academic prevention and intervention programs.

Several empirical studies have shown that SWHIs display more behavioral problems compared to peers without health impairments (e.g. Boekaerts & Röder, 1999; Hysing et al., 2009; Zashikhina & Hagglof, 2007). Furthermore, Pinquart and Shen (2011) reported that SWHIs showed higher rates of internalizing than externalizing behavioral problems. SWHIs have also been found to increase the risk of peer victimization and develop a sense of disaffection from schoolmates (La Greca, 1990). Another study also showed that SWHIs are more likely to experience higher levels of depression, anxiety, somatic complaints and social withdrawal (LeBlanc et al., 2003). Irwin and Elam (2011) revealed that many teachers do not have adequate knowledge of managing SWHIs in the classroom. This may further exacerbate school difficulties associated with SWHIs.

SWHIs are also subject to more academic difficulties than SWOHIs. For instance, Fowler et al. (1985) revealed that SWHIs have lower examination scores than SWOHIs. Pinquart and Teubert’s (2011) meta-analysis study showed that SWHIs perform poorly in academic activities such as writing and mathematics compared to peers without health impairments. SWHIs have also a greater likelihood of falling behind in academic work, repeating grades and permanently leaving school (Shaw et al., 2015). SWHIs often miss more school compared to their classmates due to their illness symptoms and medical care (Chan et al., 2005; Gorodzinsky et al., 2011; Thies, 1999). Moreover, Shaw and McCabe (2008) indicated that school attendance problems could reduce the time spent on learning and socializing with peers and may contribute to poor academic achievement.

Despite the heightened concerns of parents and teachers on the reduced school functioning of SWHIs (Gorodzinsky et al., 2011; Groenewald et al., 2020), studies comparing behavioral problems and academic achievement of students with and without health impairments are limited. The dearth of population-based data on normative rates of behavioral problems and other indicators of school functioning makes it difficult to compare and interpret findings of behavioral and academic difficulties in SWHIs. Therefore, a comparison of academic achievement and behavioral problems between students with and without health impairments is worthy of further investigation.

A large body of research has documented the link between behavioral problems and academic achievement in the general student population in schools. For example, many studies have consistently found that behavioral problems have an inverse relationship with academic achievement (e.g. Williams & McGee, 1994; Rapport et al., 2001). Longitudinal studies have also indicated that behavioral problems have a negative relationship with academic performance and these associations endured over time (e.g. Kremer et al., 2016; Malinauskiene et al., 2011; Roeser et al., 1998). A study has also revealed that behavior problem precedes and causes academic underachievement and not vice versa (McIntosh et al., 2008). Moreover, several studies have established behavioral problems as a significant negative predictor of academic achievement (e.g. Malecki & Elliot, 2002).

Although several studies have highlighted the association between behavior problems and academic achievement among general school student samples, few have investigated this relationship in SWHIs. Therefore, there is a pressing need to examine this relationship in this specific population of students and to provide empirical evidence for developing and implementing effective intervention programs to reduce the rates of behavior problems and academic difficulties targeting these students. Therefore, this study was conducted to address two objectives. The first objective of this study was to investigate possible differences in behavioral problems and academic achievement between students with and without health impairments. The second objective was to examine the relationship between behavioral problems and the academic achievement of SWHIs. Specifically, the study attempted to address the following two research questions:

1. Are there significant differences in behavior problems and academic achievement between students with and without health impairments?

2. Is there significant relationship between behavior problems and academic achievement among SWHIs?

Methods

Participants

The study participants included SWHIs (i.e. students with diabetes mellitus (DM) and students with heart disease [HD]) and a comparison group of SWOHIs enrolled in regular primary, secondary and preparatory schools in Addis Ababa. DM and HD are two of the most common chronic health conditions in Ethiopia (Awoke et al., 2014). SWHIs were recruited from Tikur Anbessa Specialized Hospital and considered eligible if they were: (a) aged 12 − 19; (b) receiving treatment and follow-up services at the diabetes or cardiac outpatient clinics; and c) enrolled in upper primary schools (Grade 5th–8th) secondary schools (Grade 9th–10th) and preparatory schools (Grade 11th–12th) in Addis Ababa. SWHIs were ineligible if they were newly diagnosed with diabetes or cardiac illness for less than six months, had other diagnosed disabilities, and if they were attending schools outside of Addis Ababa.

In total, 629 participants met the eligibility criteria. Of these, 299 were students with DM and 330 with HD. Using Cochran sample size determination formula (1977) with a 10% rate of attrition rate, a sample size of 262 SWHIs was determined and selected using stratified sampling technique. The strata were created on the basis of participants’ specific health impairment types. A simple random sampling technique was used to select the participants from each stratum.

Of 262 selected eligible participants of SWHIs, 233 (88.9%) participants made up the final sample (i.e. 115 students with DM and 118 students with HDs). The remaining 18 (6.9%) were not reachable, 5 (1.9%) declined and 6 (2.3%) were excluded from analyses due to substantial missing and outlier data. Univariate outliers were examined by analyzing frequency distributions of z-score (z ≤ ±3.29) of the data while multivariate outliers (both for academic achievement and behavior problems) were identified through Mahalanobis distance and Box plot method. The ages of the participants ranged from 12 to 19 years with a mean of 15.5 years (SD = 2.32). The sample consisted of 110 (47.2%) males and 123 (52.8%) females. In terms of grade level of the participants, 114 (48.9%) were in the upper cycle of primary school (grade 5–8), 58 (24.9%) were in secondary school (Grade 9–10) and 61 (26.2%) were in college preparatory schools (grade 11–12). Students with DM and HD accounted for 49.4% and 50.6% of the sample respectively. With respect to perceived severity level of their health impairments, 32.6% of the participants had mild condition, 48.5% had moderate severity and 18.9% had severe impairment.

A control group of SWOHIs was recruited from three schools in Addis Ababa. The sample size for the comparison group was determined by matching (1:1) with age, sex and grade level of the SWHIs group samples. To ensure that this comparison group did not include SWHIs and other disabilities, the demographic questionnaire administered to this group included an additional question asking whether the student had a chronic illness and/or other known disabilities. SWOHIs were considered ineligible if they had chronic health conditions and/or other disabilities. The SWOHIs were selected using a stratified sampling technique from three government schools in Addis Ababa. The schools were selected because of their ease of accessibility to the researcher and the representation of a broad range of demographic compositions of students in Addis Ababa.

In total, 262 SWOHIs were selected for comparison. Among the selected 262 SWOHIs, 254 (97%) participated in the study. Eight (3%) of the selected SWOHIs did not participate in the study mainly because of lack of interest and failure to obtain a signed parental consent form. After excluding of four (1.5%) cases due to missing values and outliers, the data of 250 samples of SWOHIs were used for final analysis. The comparative sample consisted of 120 boys and 130 girls with ages ranging from 12 to 19 years (M = 15.3, SD = 2.27). SWOHIs were from grades 5 to 12 in grade enrollment with the following distribution: grades 5–8 upper cycle primary school (n = 117), grades 9 and 10 secondary school (n = 65) and grades 11 and 12 preparatory schools (n = 68).

Measures

Data was collected through self-report questionnaire and compiling school records.

Demographic questionnaire

A demographic questionnaire was developed by the researcher and information regarding age, sex, grade level, religion, and family income level was obtained from students with and without health impairments. SWHIs also reported the type and perceived severity of their health impairment.

Strengths and difficulties questionnaire

Behavioral problems were assessed using the total difficulties scale of the Strengths and Difficulties Questionnaire (SDQ), self-reported by adolescents (Goodman et al., 1998). The SDQ consists of 25 behavioral attributes. Each item is scored on a 3-point scale with ‘not true’, ‘somewhat true’ and ‘certainly true’. These 25 items were divided into five subscales with items each: emotional symptoms, conduct problems, hyperactivity, peer problems and pro-social behavior. Responses are rated from 0 to 2 for negatively worded items and are rated inversely from 2 to 0 for positively worded items. Thus, higher scores indicate more problematic attributes. All items were added, except for the items about pro-social behavior, to generate a total difficulties score with a range of 0–40. The internal consistency of the total difficulties scale of the self-report version of SDQ was 0.82 (Goodman et al., 1998). In this study, alpha of 0.83 was acquired for the total difficulties scale.

Academic achievement

Academic achievement was measured as the average of the students’ aggregate results. School records were used to obtain information on the students’ academic achievement. Higher scores indicate higher academic achievement.

Procedures

Approval from the institutional ethical committee of Tikur Anbessa Specialized Hospital and permission from the heads of the schools were obtained. Parents and students were informed about the purpose of the study, that participation was voluntary, and that they could withdraw from the study at any point in time if they desired without any consequences. The participants were assured of complete confidentiality and anonymity of the data as no identification was used in the research study. Written informed consent from parents and assent from the participants were obtained before data collection. Data were collected with three well-trained research assistants.

Data analyses

Data were analyzed using SPSS statistical package version 23.0 (SPSS Inc., Chicago, IL). Descriptive statistics, including means, standard deviations and percentages were used to describe the demographic characteristics of the samples and the study variables. One-way MANOVA was used to analyze group differences in behavioral problems and academic achievement. Pearson’s moment correlation coefficients were used to examine the univariate associations between behavioral problems and academic achievement. A probability value (p value) less than .05 was considered significant.

Results

Descriptive statistics

As shown in Table 1, descriptive analysis showed that the mean score for behavior problems of SWHIs was 0.63 (SD = 0.31) and the mean score for SWOHIs was 0.53 (SD = 0.28). The mean score for academic achievement was 60.43 (SD = 9.12) for the SWHIs and 63.32 (SD = 10.04) for the SWOHIs. Visual inspection of the means showed that SWHIs recorded higher mean scores on behavioral problems (M = 0.63) than SWOHIs (M = 0.53), while the mean scores of academic achievement (60.43) for SWHIs were lower than those for SWOHIs (63.32). To test the significance of differences in the mean values for composite scores of behavioral problems and academic achievement, a one-way MANOVA was performed.

Table 1. Descriptive statistics of the variables.

Variables	SWHIs		SWOHIs
	Mean (M)	Standard deviation (SD)	Mean (M)	Standard deviation (SD)
Academic achievement	60.43	9.12	63.32	10.04
Behavior problems	0.63	0.31	0.53	0.28

Differences in behavioral problems and academic achievement

To examine differences in the outcome variables of academic achievement and behavioral problems of students as a function of their health impairment status, a one-way MANOVA was performed. Assumptions of MANOVA, including normality, linearity, collinearity, and homoscedasticity (Levene’s test of error variance and Box’s test of homogeneity of variance-covariance matrices were non-significant, p > 0.05) were satisfied. Therefore, for the MANOVA test statistic, the Wliks’ lambda (λ) test was used, as it is considered the preferred measure in research when assumptions of MANOVA are not violated compared to other commonly used test statistics, such as Pillai’s trace, Hotellings’s trace, and Roy’s largest root (Hair et al., 2010; Tabachnick & Fidell, 2013).

As shown in Table 2, the multivariate test results using the Wilks’ lambda (λ) test statistic demonstrated that there were significant differences between SWHIs and SWOHIs on combined measures of outcome variables as a function of the health impairment status of the students (λ = 0.964, F (2, 480) = 8.897, p < 0.001). The partial eta-squared value (η² = 0.036) explained that only 3.6% of the variances in the total scores of academic achievement and behavior problems were explained by the group.

Table 2. MANOVA results for the effect of health impairment status on behavior problems and academic achievement.

Effect	Wilks’ lambda value (λ)	F	Hypothesis df	Error df	Sig	Partial eta-squared (η^2)
Intercept	0.017	13956.299	2.000	480.000	0.000	0.983
HI status	0.964	8.897	2.000	480.000	0.000	0.036

Following a significant multivariate effect, univariate test results (ANOVAs) were observed to identify the effects of the health impairment status of the students on each dependent variable of academic achievement and behavioral problems separately. The results revealed significant group differences in academic achievement, F (1, 481) = 10.920, p < 0.01, η² = 0.022), and behavioral problems, F (1, 481) = 12.168, p < 0.01, η² = 0.025), indicating that academic achievement was significantly lower while behavioral problems were higher among SWHIs than SWOHIs. As shown in Table 3, the magnitude of the effect of students’ health impairment status on academic achievement (η² = 0.022) and the behavior problems (η² = 0.025) was small, with health impairment status accounting 2.2% and 2.5% of the variations in academic achievement and behavioral problems, respectively.

Table 3. Univariate test results for the effects of health impairment status on academic achievement and behavior problems.

Source	Dependent variable	Type III sum of squares	Df	Mean square	F	Sig	Partial eta square
Corrected model	AA	1007.733	1	1007.733	10.920	0.001	0.022
	BPs	1.086	1	1.086	12.168	0.001	0.025
Intercept	AA	1,846,981.69	1	1,846,981.69	20015.18	0.000	0.977
	BPs	160.970	1	160.970	1803.125	0.000	0.789
HI status	AA	1007.733	1	1007.733	10.920	0.001	0.022
	BPs	1.086	1	1.086	12.168	0.001	0.025
Error	AA	44,386.230	481	92.279	–	–	–
	BPs	42.940	481	0.089	–	–	–
Total	AA	1,897,708.50	483	–	–	–	–
	BPs	204.265	483	–	–	–	–
Corrected total	AA	45,393.964	482	–	–	–	–
	BPs	44.026	482	–	–	–	–

Relationship between behavioral problems and academic achievement

Pearson’s moment correlation coefficients were computed between students’ behavioral problems and academic achievement. Behavioral problems were found to be significantly and negatively correlated with academic achievement (r = −0.34, p < 0.01). This finding shows that the more behavioral problems are related to lower academic achievement. The correlation matrices of the variables are presented in Table 4.

Table 4. Intra and inter-correlations among variables and sub-scales (n = 233).

Variables	Subscales	1	2	3	4	5	6
1. Behavior problems	Total	1
	2. Emotional symptoms	0.72**	1
	3. Conduct problems	0.72**	0.36**	1
	4. Hyperactivity	0.72**	0.32**	0.37**	1
	5. Peer problems	0.72**	0.36**	0.35**	0.34**	1
6. Academic achievement	Total	−0.34**	−0.34**	−0.17**	−0.19**	−0.26**	1

**p < 0.01.

The analysis also indicated that all dimensions of behavioral problems (emotional symptoms, conduct problems, hyperactivity and peer problems) had significant negative correlations with academic achievement (rs = −0.17 to −0.34**; p < 0.01; Table 4), indicating that those with higher levels of behavioral problems had lower academic achievement scores.

Discussion

The results revealed a significant difference in behavioral problems between students with and without health impairments. SWHIs reported significantly more behavior problems than SWOHIs. This implies that SWHIs tend to exhibit more behavioral problems than SWOHIs. One possible reason for the higher levels of behavioral problems observed in SWHIs may be frustrations resulting from limitations of regular life routines, perceived inability to manage symptoms or the course of the disease, peer isolation and treatment side effects (Pinquart & Shen, 2011). In addition, the cumulative effects of many sources of stress across various settings (e.g. home, school and hospital) may induce a sense of despair and consequently associated with increased behavioral problems. This result is congruent with the large body of empirical evidence showing elevated rates of behavioral problems in SWHIs (Hysing et al., 2009; Pinquart & Shen, 2011; Zashikhina & Hagglof, 2007).

The results demonstrated a significant difference in academic achievement between SWHIs and SWOHIs. This shows that SWHIs have lower academic achievement scores than their school counterparts with no health impairments. Several factors may contribute to the lower academic achievement of SWHIs. First, the presence of physical symptoms of chronic illnesses, such as fatigue, lethargy and irritability may reduce self-confidence and motivation in academic activities and affect academic achievement (Chan et al., 2005; Gorodzinsky et al., 2011; Shaw et al., 2010). Second, chronic illnesses and medications used for treatment may result in neuro-cognitive deficits, such as memory difficulties, sight problems, fine motor skills limitations and decreased concentration (Thies, 1999). Third, their frequent school absences and missed instructional times may limit their participation in school activities and decrease readiness to learn, and they may fall behind their peers in school work (Shaw et al., 2010). This finding is consistent with previous research reporting lower scores on academic achievement in SWHIs than in their school counterparts without health impairments (Fowler et al., 1985; Pinquart & Teubert, 2011; Thies, 1999).

Correlation coefficient analysis showed a significant negative relationship between behavioral problems and academic achievement. This relationship suggests that an increase in behavioral problems is associated with a decrease in academic achievement. One possible mechanism for this finding may be that students with higher behavioral problems could distract them from concentrating and giving full attention to classroom activities, which in turn could lead to low academic achievement (Roeser et al., 1998). This finding is consistent with many previous studies showing that behavioral problems are negatively correlated with academic achievement (e.g. Kremer et al., 2016; Rapport et al., 2001; Williams & McGee, 1994).

Implications, limitations and future directions

The result of this study indicates that SWHIs are greatly at risk with significantly higher behavioral problems and lower academic achievement compared to peers without health impairments. The poor school-related outcomes, including increased behavior problems and decreased academic achievement of SWHIs, highlight both variables as targets for intervention. These findings suggest the need to comprehensively assess the academic and behavioral outcomes of these students and ensure academic and behavioral interventions targeting SWHIs. The findings of the study also clearly indicate that behavioral problems correlate negatively with students’ academic achievement. This finding further supports the importance of recognizing and managing behavioral problems, so as not to let them affect academic performance. Knowledge and understanding on this area could help many parties, such as teachers, special needs educators, counselors and school psychologists to design and develop proper intervention programs to reduce behavioral problems among SWHIs.

Although the results of the one-way MANOVA and ANOVAs analyses found significant differences in the mean scores of both behavior problems and academic achievement between the groups, the variance in scores that health impairment status accounts for is small. This may indicate that the situation for SWHIs is not as bad as it could be and gives hope that appropriate behavioral and academic interventions may help to bring SWHIs up to speed with SWOHIs.

Due to their functional restrictions and limitations, SWHIs encounter many challenges in the school environment. Collaboration among school personnel, health care providers and family members may benefit the child not to be excluded needlessly from the school curriculum. SWHIs should be provided extra support at school. For example, flexible timescales for returning homework and assignments may be employed when fatigue delays return. The learning and progress of SWHIs can be affected by frequent hospital care requirements. Thus, for lengthy absences, tutorials and make up classes may be arranged, allowing the student to bridge the gaps associated with absence and catch up with any missed work. Special programs may also be arranged to address missed tests and examinations.

This study had some limitations. First, the cross-sectional design employed in this study does not allow the establishment of a cause-effect relationship. Therefore, longitudinal studies are required to prove the causal relationship and test the association between behavioral problems and academic achievement. Second, the self-report questionnaire used to assess behavioral problems in the study might have elicited inflated or false responses. More extensive studies are required with different assessment strategies for students’ behavior problems. Third, the study findings cannot be generalized to the entire SWHIs population in Addis Ababa since the SWHIs sample involved only students with diabetes and HD who were drawn only from one specific site. Future research should include students from other chronic-illness groups and sites. Finally, the study did not assess the level of behavioral problems and academic achievement in students with specific groups of health impairment types (DM and HD) and levels of severity (mild, moderate and severe) and in comparison with SWOHIs. Future research should examine the influence of demographic variables such as type and severity level of health impairment might contribute to differences in academic achievement or behavioral problems.

Conclusions

SWHIs have considerably more behavioral problems and low academic achievement than SWOHIs. Behavioral problems of SWHIs are negatively associated with academic achievement. Thus, reducing a child’s behavioral problems may prove useful in improving the academic achievement of this particular group of students living with chronic health conditions.

Open Scholarship

This article has earned the Center for Open Science badges for Open Data, Open Materials and Preregistered. The data and materials are openly accessible at https://osf.io/xgjhd/?view_only=ca0db0a8d26b4aa79189b3743223a6b1, https://osf.io/xgjhd/?view_only=ca0db0a8d26b4aa79189b3743223a6b1 and No, any of the the project component was not registered before examination.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability

The datasets presented in this article are readily available.

Additional information

Funding

No funding available.

Notes on contributors

Birhanu Nebiyou Muluneh

Birhanu Nebiyou Muluneh is an assistant professor of Special Needs Education in the School of Psychology, Jigjiga University, Jigjiga, Ethiopia

Tekalign Deksissa Bejji

Tekalign Deksissa Bejji is an assistant professor of Special Needs Education in the Department of Psychology, Wolkite University, Wolkite, Ethiopia.