1
 CHAPTER ONE 
INTRODUCTION 
 
1.1 Introduction 
 
Inclusive education in South Africa focuses on ?acknowledging and respecting that all 
people have different learning needs which are equally valued? (Landsberg, 2005, 
p.18). With the adoption of a more equitable educational policy of inclusion in South 
Africa, emphasis is placed on a holistic understanding of factors that may serve as 
barriers to optimal learning and development, and may thus need to be accommodated 
and included in a single education system (Donald, Lazarus, & Lolwana, 2002).  
 
As such, the current challenge within South African schools is to address a wide range 
of learning needs that may be present. However, transformation is slow, and 
adjustments within schooling systems indicate gradual change at present. 
Consequently, many schools in South Africa face a transitional phase, which straddles 
both existing and incoming educational structures. Resultantly, while schools attempt 
to address a range of possible barriers to learners, segregation of learners with special 
educational needs within schools is still commonplace. This suggests that the 
recognising and addressing of barriers to learning differs across those students with 
special educational needs, be they internal or external, and those learners placed in 
?mainstream? classes. 
 
The successful implementation of inclusive education in South Africa thereby calls 
for recognition and accommodation of a variety of personal and environmental 
stressors that put learners at risk for behavioural, emotional, and learning difficulties 
(Landsberg, 2005). In light of the current educational context in South Africa, the 
study is embedded in a body of literature that highlights the cognitive, behavioural 
and emotional manifestations and effects of learned helplessness, and related 
constructs, which may act as barriers to optimal learning. The study explored 
differences in academic achievement, learned helplessness and depression among 
samples of learners with varying educational needs. Moreover, the study explored the 
correlation between learned helplessness, depression and academic achievement, in 
line with existing literature in the field.   
 2
 1.2 Research Aims 
 
The primary aim of this study was a comparative analysis of the prevalence of learned 
helplessness, depression and academic achievement across two differing samples- a 
sample of learners in a mainstream class, and a sample of learners presenting with 
barriers to learning placed in a small class inclusion programme (SCIP).  In light of 
this, several secondary aims were purported. The study aimed to identify and 
describe, using separate test scales, the prevalence of learned helplessness and 
depression, within the chosen samples. In turn, data yielded from the tests were used 
to explore the differences in the prevalence of learned helplessness and depression 
between the sample groups.  Furthermore, the study aimed to examine the correlations 
between learned helplessness, depression, and academic achievement, both within and 
between the sample groups. 
 
In accordance with existing research in the field (Abramson, Seligman, & Teasdale, 
1978; Peterson, Maier, & Seligman, 1993; Seligman, 1995), the study aimed to 
provide analysis of current trends and hypotheses within the field, based on 
comparative data yielded from differing sample groups. In turn, results from the study 
were used to direct researchers towards priorities for further research in light of 
learned helplessness, depression, and academic achievement, within the current South 
African context. 
 
1.3 Rationale 
 
The issue of special educational needs in light of the current educational context in 
South Africa suggests the relevance of this study to the South African context at 
present. Following the transformation to an inclusive educational system, the term 
?special educational needs? emphasises the view that learning and behavioural 
difficulties may arise as a reciprocal product of both intrinsic and/or extrinsic factors. 
The reformulation of definitions of learning and behavioural difficulties not only 
broadens existing understanding of such difficulties, but also serves to clarify the high 
incidence of special educational needs and learning difficulties in the South African 
school-going population at present, and further accounts for the extent and range of 
special needs in the South African context (Landsberg, 2005). The reviewing of 
 3
 educational policies and practices, as outlined in Education White Paper 6 
(Department of Education, 2001), resulted in the realisation that a range of needs 
exists among learners in South Africa. These needs were conceptualised as ?barriers 
to learning and development? which are defined as ?any factor, either internal or 
external to the learner, which causes a hindrance or barrier to that person?s ability to 
benefit from schooling? (Donald et al., 2002, p. 4). 
 
 As suggested in Landsberg (2005), a large number of barriers to learning arise from 
an interaction of factors within external and internal systems, especially within South 
Africa at present. Many of these barriers can be attributed to the transformations 
within the educational system at present, which, despite many positive changes, do 
not preclude existing inadequacies and deficiencies of the system to meet increasing 
learner demands.  This suggests that a number of factors impact on learning and 
development, and not only those factors associated with impairment and disability. It 
is thus evident that there are a number of factors that may affect optimal learning, 
academic achievement, and emotional development.  
 
The learned helplessness model provides a useful account of the possible 
psychological consequences that children may experience following difficulties in 
gaining control over their learning. The resultant cognitive, behavioural and emotional 
manifestations and effects of learned helplessness may serve as a possible hindrance 
to optimal learning and development. Several longitudinal studies have examined the 
link between explanatory style, depression, and poor school achievement, as 
suggested by the learned helplessness model (Nolen-Hoeksema, Girgus, &Seligman, 
1986; Sweeney, Anderson & Bailey, 1986). For the most part, such studies have 
supported the learned helplessness model as a useful account for depression, as all 
essential features of the model are all clearly present in depression cases (Seligman, 
1995). Furthermore, several studies have indicated the application of learned 
helplessness to school achievement (Dweck, 1975; Fincham, Hokoda, & Sanders, 
1989). As highlighted in the literature, ?helpless? behaviour negatively affects 
motivation and performance, which in turn, influences academic achievement over 
time (Cole, Martin, Peeke, Seroczynski, & Fier, 1999). 
 
 4
 It has been suggested that despite subsequent remedial effort, learning disabled 
children, or those who experience existing barriers to learning, may present as low in 
persistence and efforts, with increased frustration and anxiety, even at appropriate 
ability levels (Thomas, 1979). Studies suggest that specific learning disabilities may 
be addressed by the learned helplessness phenomenon (Thomas, 1979). Similarly, as 
highlighted in Fincham et al. (1985), in the absence of appropriate intervention, 
learned helplessness may be a response pattern that retards learning, especially when 
learning involves material that presents as challenging and difficult. Thus, learned 
helplessness influences and reinforces academic achievement patterns over time. It is 
for this reason that it has been suggested that learned helplessness is more prevalent in 
populations of children with learning disabilities or barriers to learning, subsequent to 
experiences of repeated failures and difficulties, and decreased motivation (Thomas, 
1979). Further to existing literature in the field, the study hypothesised that learners 
placed in the SCIP class would be more at risk for learned helplessness, thereby 
suggesting that the prevalence of learned helplessness, and related cognitive, 
emotional and behavioural deficits, would be greater in the sample of learners with 
barriers to learning.  
 
The rationale for this study thus lies in the possible recognition of learned 
helplessness as a barrier to learning. In light of the current educational context within 
South Africa, it is necessary to examine the possible differences in explanatory style, 
and resultant cognitive-behavioural manifestations and effects, as a possible hindrance 
to optimal learning and development. The recognition of differing emotional reactions 
and subsequent styles of coping has important implications for examining differences 
in causal attribution, emotional coping, and academic performance across learners 
with varying learning needs (Wiggins, 2003). 
  
Against a backdrop of transformation within the educational context at present, the 
study endeavoured to explore a relevant, but under-researched field in South Africa at 
present. 
 
 
 
 
 5
 CHAPTER TWO 
LITERATURE REVIEW 
 
2.1 Overview 
 
This chapter presents a review of the theories on which the research is based. Previous 
research conducted in the field relevant to the study has been included, in order to 
provide a contextual understanding for the present study. The link between 
explanatory style, depression, and poor school achievement, as suggested by the 
learned helplessness model, has been explored in several longitudinal studies (Nolen-
 Hoeksema et al., 1986; Sweeney et al., 1986). The concept of learned helplessness 
originally referred to a learned non-contingency between behaviour and important 
outcomes, resulting in reduced persistence and depressive-type affect (Seligman, 
Maier, & Solomon, 1971). Later research revealed that older children exhibit learned 
helplessness with respect to achievement outcomes (Dweck, 1975). This 
achievement-based helplessness was associated with a tendency to attribute failure to 
ability rather than effort. The belief that successful performance is no longer 
contingent on actions, but rather ability, led to decreased persistence and performance 
(Burhans & Dweck, 1995).  Achievement-based helplessness thereby includes several 
cognitive, affective and behavioural components. Such theory suggests that concepts 
such as poor self-esteem, poor motivation, and depression can be viewed as having a 
direct effect on experiences of success or failure in a variety of learning situations 
(Beck, 1971; Phares, 1973, as cited in Thomas, 1979). 
 
2.2 The Learned Helplessness Model 
 
The concept of learned helplessness was developed by Overmier and Seligman 
(1967), following a series of animal studies. Observation of dogs repeatedly exposed 
to uncontrollable negative events (inescapable shocks), showed that the dogs 
eventually discontinued their efforts to escape, even after the situation changed so that 
escape was possible.  The dogs learned an independence between their behaviour and 
the aversive stimulation they received and were therefore helpless to effect change in 
their later circumstances (Martinko & Gardner, 1982). Maier and Seligman (1976) 
proposed ?learned helplessness? as a cognitive model to explain the behavioural, 
 6
 cognitive and emotional deficits evident due to the experience of uncontrollable 
events. 
 
Experiments designed to explore the learned helplessness hypothesis in humans 
followed the original experiments seen in animal studies. As in the animal studies, 
humans were found to display cognitive, behavioural, and emotional deficits 
following uncontrollability. These included subsequent failure to learn that responses 
can affect outcome, decreased response initiation, and depressed affect. It was further 
noted that the expected independence between response and outcome could be 
generalised to different situations (Abramson et al., 1978).     
 
The phenomenon of learned helplessness was used as a possible explanation for 
observed behaviours in human research. Later, due to observed similarities in 
symptoms of learned helplessness, and symptoms of depression, learned helplessness 
was proposed as a model of depression in humans (Peterson et al., 1993).   The model 
proposed that learned helplessness and depression had similar behavioural and 
physiological manifestations, as all essential features of the model are all clearly 
present in depression cases (Seligman, 1995).  According to the model, depression 
results from a non-contingency between expected outcomes and personal responses 
(Peterson et al., 1993).  As explained in Thompson (2002), this suggests that when a 
person is exposed to an experience in which outcomes are non-contingent upon 
responses, the perception of non-contingency results in apathy, depressed affect and a 
cognitive deficiency in the learning of new responses.  In turn, subsequent learning, 
motivation, and affect are affected by generalised perceptions of non-contingency, 
should this occur. Resultantly, ?helpless? response patterns were purported as similar 
to those seen in depression cases. As such, the learned helplessness model was 
proposed as a model of depression (Maier & Seligman, 1976). 
 
2.3 Reformulation of The Learned Helplessness Model 
 
Abramson et al. (1978) reformulated the model of learned helplessness and reviewed 
the implications of the model as an account for depression. In light of the 
shortcomings of the model, the original helplessness model was revised to include 
attribution theory, which focuses on the causal beliefs, or the reasons that a person 
 7
 attributes to behaviour. Attribution theory is a cognitive theory, which explores both 
internal and external attributions or causes (Abramson et al., 1978). The inclusion of 
attribution theory thereby focused on the habitual explanatory styles that people 
ascribe to events they experience (Peterson et al., 1993).     
 
Attribution theory postulated that human behaviour is not controlled by ?the schedule 
of reinforcement in the environment?, but rather by an internal mental state- the 
explanations made by people for why the environment has scheduled such 
reinforcements. Learned helplessness is thus inextricably linked to explanatory style. 
Learned helplessness functions from the premise that ?whatever you do doesn?t 
matter?. It can thus be deduced that a largely pessimistic explanatory style belies the 
development of learned helplessness (Seligman, 1990, p. 41).  
 
Abramson et al. (1978) proposed that people prone to learned helplessness make 
causal explanations for the uncontrollable events they encounter. These causal 
explanations affect self-esteem, and the generality of deficits. This suggests that a 
person need not actually experience repeated events in order for them to occur- but 
rather, the person needs to expect that events will be uncontrollable. Expected non-
 contingency between responses and outcomes results in causal explanations for the 
expected non-contingency. This explanation influences expectation of future non-
 contingency, which in turn determines the nature of helpless deficits (Peterson et al., 
1993).  
 
In the reformulation, Abramson et al. (1978) propose three dimensions of causal 
explanations ? internal-external causes; stable-unstable causes; and global-specific 
causes.  Internal attributions are associated with personal helplessness, as 
uncontrollability is attributed to individual factors within the person. External 
explanations are associated with universal helplessness, as uncontrollability is 
attributed to situational or circumstantial factors that would likely have an effect on 
anyone who experienced such factors. Thus loss of self-esteem with internal factors is 
more likely than with external attributions. 
 
Within the second dimension, stable causes are those that are persistent and enduring, 
while unstable factors are those that are temporary and sporadic. Stable causes are 
 8
 more likely to result in passivity and helpless deficits, as the causes are seen as 
constant. Thirdly, global causes are those that can be generalised to several situations 
or outcomes, while specific causes are those that remain particular to a given situation 
or outcome.   
 
As explained in Abramson et al. (1978), these three dimensions interact to determine 
a person?s particular attributional style, which accounts for how people ?habitually 
explain events?and allows for individual variation in response to uncontrollability 
(Peterson et al. 1993, p. 151)?. Attributional style is seen as relatively stable, which in 
turn suggests that the probability and duration of helplessness can be predicted in light 
of personal attributional style. Individuals at risk for cognitive, behavioural, and 
emotional effects of helplessness were hypothesised to be those who tend to make 
stable, global and internal attributions for failure (Thompson, 2002).    
 
Learned helplessness thus occurs when an individual learns that outcomes are 
uncontrollable by responses, and in turn, is seriously debilitated by this knowledge 
(Maier & Seligman, 1976). When an action seemingly has no change or effect on a 
desired outcome, events are seen as beyond control, despite efforts. This occurs to the 
extent that efforts are ceased, task persistence drops, and emotional accompaniments, 
such as passivity and anxiety may occur (Thornton & Jacobs, 1971; Maier & 
Seligman, 1976). 
 
Research studies (Hiroto & Seligman, 1975; Gatchel & Proctor, 1976; Miller & 
Seligman, 1975, as cited in Martinko & Gardner, 1982) have supported the notion that 
learned helplessness is a fundamental type of learning, as both the original and the 
reformulated theory of learned helplessness focus on cognitive processes (Overmier & 
Seligman, 1967). The attributions made by a person for non-contingency between acts 
and outcomes determine subsequent expectation of future non-contingency, which in 
turn determines the type of helplessness symptoms (Valas, 2001a, 2001b).  Indeed, 
learned helplessness has been demonstrated to be a reasonable explanation for 
depressive behaviour, and motivational deficits in the classroom (Seligman, 1975; 
Dweck, 1975).  
 
 
 9
 2.4 Learned Helplessness as a Model for Depression 
 
The research is embedded in a body of literature that explores the possible 
psychological consequences that children may experience following difficulties in 
gaining control over their learning environment. As suggested in the literature, the 
resultant cognitive, behavioural, and emotional manifestations of learned helplessness 
might serve as reasonable explanations for depressive symptoms, and learning 
deficits.  
 
Some of the research emanating from the learned helplessness model has ?confirmed 
the existence of a depressive attributional style that attributes uncontrollable bad 
events to internal, stable and global factors? (Mitchell, 2001, p. 448). The results of 
such research have confirmed the applicability of a characteristic attributional style 
for depression, and have further suggested a more general applicability. Studies 
(Sweeney et al., 1986) have confirmed the generalization suggested by attribution 
theory, which holds that Internality, Stability, and Global Attributions are 
characteristically higher for positive events than negative events in individuals who 
do not present with depressive symptoms. It was hypothesized that there would be an 
association of negative attributional style and depressive symptoms only in the 
presence, but not in the absence, of negative life events (Abramson et al., 1989). Thus 
a pre-existing cognitive style, in which internal, stable, global attributions are made 
following experience of negative events, purports a predisposition to depression. 
 
2.5 Cognitive Models of Depression 
 
As highlighted by Hirsch and Conner (2006), cognitive theories of depression have 
highlighted the critical importance of cognitive styles and patterns. Much of this 
research has focused on cognitive styles that promote the development of adverse 
behavioural and affective responses. Cognitive models of depression suggest the 
tendency to commit negative cognitive errors about personally relevant events, which 
contribute to the onset and maintenance of affective problems, as seen in both child 
and adolescent populations (Kaslow, Rehm, Pollack & Siegel, 1984; Weis, Weiss, 
Wasserman & Rintoul, 1987). In the context of continuing interest in the determinants 
of depression, there has been a focused interest in an explanation of depression based 
 10
 on the reformulated learned helplessness model (Abramson et al., 1978). Learned 
helplessness, as a cognitive-behavioural approach, emphasizes the role of beliefs and 
behavioural contingencies in causing depression. Beck (1976) proposed that children 
with depression hold maladaptive thoughts and beliefs that distort the way in which 
they process information. Incoming information is thus interpreted according to these 
negative perceptions, such that everyday events are viewed as hopeless, sad and 
aversive (Kronenberger & Meyer, 2001). Similarly, Seligman?s model of learned 
helplessness supports the notion that depressive attributional style has been found to 
be related to depressive symptomatology in children. Maladaptive attributions focus 
on internal, stable, and global attributions for negative life events (Kronenberger & 
Meyer, 2001).   
 
According to the reformulated learned helplessness model, depression results from 
uncontrollable aversive events. The nature and consequences of depression are 
determined by causal attributions or explanations given for such events. Internal 
attributions resultantly lead to depression in which a loss of self-esteem is 
experienced, while stable attributions predict long-lasting symptoms of depression. 
Furthermore, if the uncontrollable events are attributed to causes that are felt to be 
present in a variety of situations (global attributions) as opposed to specific 
attributions, the resulting depression is predicted to be pervasive (Abramson et al., 
1978). 
 
2.6 Learned Helplessness and Depression 
 
In a sample of children, Nolen-Hoeksema et al. (1986) assessed the contribution of 
attributional style and negative life events to depression. In two of four instances, the 
interaction between pessimistic attributional style and negative life events 
significantly predicted future levels of depression. Studies have shown that 
individuals with a pessimistic explanatory style show increased depressive symptoms 
following exposure to negative events (Priester & Clum, 1992; Gillham et al., 2001, 
as cited in Hirsch & Conner, 2006).  
 
According to Abramson et al. (1978), the generality of the depressive deficits will 
depend on the globality of the attributions of helplessness. The chronicity of the 
 11
 depression deficits will depend on the stability of the negative attributions. Changes in 
self-esteem will depend on the internality of these negative attributions. 
 
Review of the literature suggests that more cross-sectional than longitudinal studies 
have been conducted to examine the relationship between explanatory style and 
depression. Results of these studies (Peterson & Seligman, 1984; Sweeney et al., 
1986) have shown that the cognitive tendency to attribute negative events to internal, 
stable, and global causes is associated with severity of concurrent and future 
depression. Furthermore, in a meta-analytic review of more than 100 explanatory 
style studies, Sweeney et al. (1986) reported evidence for the predicted relationship 
between explanatory style and depressive symptoms. These studies further found that 
the severity of depressive symptoms (as measured by the Beck?s Depression 
Inventory) is often correlated with attribution of internal, stable, and global causes to 
bad events. 
 
According to Abramson et al. (1989), in conducting research on the correlation 
between explanatory style and depression, it would be useful to select subjects on the 
basis on the presence or absence of an attributional style predisposed to depression, 
and to compare those groups. Accordingly, in a study conducted by Alloy, Lipman 
and Abramson (1992), it was found that currently non-depressed students with a 
depressive attributional style had higher rates of specific depressive symptoms in the 
past two years than currently non-depressed students with a non-depressed 
attributional style. Further research suggests that children with depression hold more 
negative beliefs about interpersonal relationships, and present with weaker social 
problem-solving skills (Kronenberger & Meyer, 2001).  Moreover, research on the 
cognitive correlates of depression reveals that depressed children selectively recall 
events that are negative, evaluate such events more harshly, and over- generalize 
about such events (Cole et al., 1999).  
 
The above-mentioned studies have clearly demonstrated a link between attributional 
style, as suggested by the learned helplessness model, and depressive symptoms/ 
depression. However, it is evident that while some of these studies (Nolen-Hoeksema 
et al., 1986) have been conducted with child and adolescent populations, there is a 
need for further research in South Africa with child and adolescent populations, which 
 12
 are applicable to the South African educational context at present. Furthermore, in 
light of the move towards inclusive education in South Africa, there is an increased 
need for research that focuses on cognitive styles that may foster and promote 
improved behavioural and affective functioning.  
 
2.7 Learned Helplessness and Academic Achievement 
 
As explained by Seligman (1990), traditional views of achievement attribute success 
to a combination of talent and desire. Similarly, it can be deduced that failure results 
from absence of talent and desire. However, Seligman (1990) posits that the presence 
of a third factor- optimism or pessimism- acts as a significant contributor to measures 
of achievement.     
 
As postulated by Dweck (1975), the theory of attributional determinants can be used 
to explain academic performance and achievement. The cognitive (attributions), 
behavioural (motivation) and emotional (self-esteem and depression) effects on a 
child, who has a cognitive processing style predisposed to learned helplessness, have 
been correlated with poor academic achievement. Dweck asserts that children can be 
categorised as ?helpless? or ?mastery-oriented? in their attributions to academic 
success or failure.  ?Helpless? children are those who attribute failure to lack of 
ability despite possible prior success. In this regard, the attribution of failure to 
internal and stable causes, leads to decreased persistence and increased negative 
feelings.   
 
Children?s idea of ability becomes more differentiated with age. It is only at early 
adolescence that some form of formal operational thought is developed, which 
suggests that ability is fully differentiated from effort, and is conceptualised as an 
entity unaffected by effort.  Furthermore, as children enter secondary school they 
experience systematic changes in activities, organisation and evaluation practice, 
which also contribute to shifts in their ability judgements (Stipek & MacIver, 1989 as 
cited in Cole et al., 1999).  School becomes more formal, more evaluative and more 
competitive, while the focus shifts from the learning process to evaluation of the 
learning outcomes. Such changes are coupled with increased academic demands and 
pressure, and changes in motivation, expectations, and effort. Students? self-
 13
 confidence in mastering tasks declines with age and experience (Cole et al., 1999).   
Thus it is expected that older students will be more likely to attribute failure to 
internal, stable and uncontrollable factors more often than younger students, which in 
turn may affect helpless expectations, helplessness, self-esteem, and depression 
(Valas, 2001a, 2001b).  
 
Previous studies, as suggested in Valas (2001a, 2001b), show a link between 
constructs of learned helplessness and academic outcomes. Research findings (Cole et 
al., 1999; Valas, 2001a, 2001b) suggest that past academic achievement influences the 
patterns of attributions. Thus, children who have a history of poor performance are 
more likely to attribute failure to low ability. As past academic performance may also 
affect attributional patterns indirectly through academic self-concept, children who 
experience continuous academic under-achievement, make attributions of academic 
incompetence, in which failure is attributed to lack of ability (Valas, 2001a, 2001b). 
 
 In a study conducted by Valas (2001a, 2001b), mathematical and verbal performance 
were found to be significantly related to attributional style, expected outcomes, and 
symptoms of helplessness. It was found that high achieving students attributed 
success more often to effort and more seldom to ability than low achieving students. It 
was also found that high achieving students had more positive expectations and 
showed less helplessness than low achieving classmates. Significant relations were 
found between achievement and attributional style, highlighting that achievement is 
negatively related to attributional style, which attributes outcomes to abilities rather 
than effort. In another study conducted by Peterson and Barrett (1987), as cited in 
Martinez and Sewell (2000), university freshmen with a pessimistic explanatory style 
tended to achieve lower grade point averages in their first year of college.  
 
It is evident that academic adjustment, generally demonstrated through achievement, 
motivation and perceived competence, is widely perceived as important for continued 
success (Shahar et al, 2006). Overall, students? academic adjustment seems to decline 
in secondary school. As suggested above, these declines are often thought to be 
associated with transitions within the schooling environment, exacerbated by 
developmental changes (Shahar et al., 2006). However, not all children experience 
 14
 these declines, and researchers have begun to focus on the factors associated with 
individual differences in this area. 
 
2.8 Learned Helplessness, Academic Achievement and Learning Difficulties 
 
Learned helplessness has been vastly studied in light of children with learning 
difficulties (Fincham et al., 1989; Martinko & Gardner, 1982; Thomas, 1979) Many 
children who experience learning difficulties may experience school as a negative 
experience, over which they feel little or no control, despite effort.  Often, the self-
 images of these children are particularly at risk, following repeated academic 
difficulty and/or failure. Furthermore, learned helplessness affects motivation to learn 
(Valas, 2001a, 2001b). 
  
Results from studies (Fincham et al., 1989; Shahar et al., 2006; Valas, 2001a, 2001b) 
suggest that children with a learning difficulty experience failure in core subject, 
possibly despite considerable work. In turn, these students will attribute failure to 
uncontrollable causes- lack of ability. Furthermore, research findings, as highlighted 
in Valas (2001a, 2001b), suggest that past academic performance influences the 
pattern of attributions. Often, difficulties and failures are attributed to internal, stable, 
and global factors, such as low ability. This leads to an expected non-contingency 
between work and results, which gives rise to symptoms of personal helplessness. In 
turn, these may lead to lowered self-esteem and increased depressive symptoms 
(Valas, 2001a, 2001b). 
 
In comparative studies between children with and without learning difficulties, it has 
been found that learned helplessness might be more prevalent in children with 
learning disabilities (Shahar et al., 2006; Valas, 2001a, 2001b). However, in a 
longitudinal study conducted by Fincham et al. (1989), learned helplessness was 
found to a reliable predictor of achievement in a sample of children without learning 
disabilities.  
 
 
 
 
 15
 2.9 Academic Achievement and Depression 
 
As highlighted in Marcotte, Levesque and Fortin (2006), several recent studies have 
reported high rates of depressive symptoms in school populations. Resultantly, 
impaired school performance has been noted as one of the short-term consequences of 
a depressive episode. Despite this observed consequence of depression, impaired 
academic functioning has rarely been the focus of studies.  As such, the association 
between depression and academic performance rarely has been addressed in the 
literature. Traditionally it is the impact of learning and conduct disorders that have 
been associated with a lower academic performance. However, Cheung (1995) notes 
that few studies have reported that depressive symptoms may be associated with 
academic difficulties and lower performance. Furthermore, a higher level of 
depressive symptoms was reported in students with learning difficulties. Chen, Rubin, 
and Bo-Shu (1995) further reported that a decline in academic performance was 
associated with depressive affect; however, academic difficulties were predictive of 
later depression only for children from families in which a conflictual marital 
relationship was presenting.  Further studies concluded that depressive symptoms in 
children predicted lowered academic performance in both genders, while low 
academic performance predicted depressive symptoms for girls only (Kellam, Rebok, 
Mayer, Ialongo, & Kalodner, as cited in Marcotte et al., 2006). 
 
It is evident that despite some research in the field of academic performance and 
depression, further research is needed to explore the association between school 
difficulties and depression. As a result of this, it remains uncertain whether depressive 
symptoms manifest as a result of academic difficulties, or if such symptoms preceded 
the decline in academic performance (Marcotte et al., 2006).  
 
2.10 The impact of inclusive education in promoting optimal functioning and 
adjustment- The move towards a positive psychology 
 
The individualistic, medical approach has thus far ignored the range of systemic and 
broader socio-economic factors that are influential in the development and 
perpetuation of behavioural and emotional problems, and learning difficulties. 
Inclusive education thereby prescribes a shift from the traditional child-deficit, 
 16
 medical approach towards an ecological and multi-levels systems approach, 
suggesting a wider scope of analysis and action. The adoption of inclusive education 
in South Africa calls for recognition of a variety of personal and environmental 
stressors that put learners at risk for behavioural, emotional, and learning difficulties 
(Landsberg, 2005).  With an increased focus on contextual factors, an inclusive 
approach to education focuses on developing learner strengths, to overcome 
difficulties, thus empowering and enabling learners to participate actively and 
critically in the learning process (Engelbrecht, 2004).  
 
In line with this, the positive psychology movement has focused on the promotion of 
psychological strengths and environmental conditions that support the development of 
strengths. This movement seeks to expand the focus of psychology from repairing 
deficits to building positive qualities in people (Lopez, Snyder, & Rasmussen, 2003; 
Seligman & Csikszentmihalyi, 2000). Karver and Bickman (2002, as cited in Lewis, 
Huebner, Reschly, & Valois, 2009) highlight the importance of examining both 
maladaptive and positive functioning, in order to enhance understanding of well-being 
and development.  Thus, positive psychology upholds the principles of inclusive 
education by placing emphasis on both intrinsic and extrinsic factors that foster and 
sustain improved functioning and development.   
 
Previous research has shown that social relationships with parents, peers, and 
teachers, as well as school bonds and neighborhood contexts, are related to the social, 
behavioral, and emotional adjustment of students.  Research findings have suggested 
associations between children?s relational?contextual experiences and social?
 emotional adjustment (Murray & Greenberg, 2006; Talbott & Fleming, 2003).  
Murray and Greenberg (2006) highlight the role of caregiver, teacher, and peer 
relationships in influencing adjustment and well-being.  Positive aspects of contexts 
are suggested to be crucial in maintaining optimal functioning and development.  The 
findings suggest that an accumulation of positive or negative relationships and 
experiences may prove more influential in the overall health and well being of 
children and youth than any single variable on its own. This alludes to the importance 
of examining the dynamic interplay of various systems that contribute to overall 
functioning in children.  
 
 17
 A move towards a positive psychology incorporates analysis of positive experiences 
and individual personality traits that promote well-being and emotional functioning 
(Duckworth, Steen, & Seligman, 2005). It is suggested that such analysis serves to 
enhance factors that foster adjustment and learning, and may thereby counteract 
barriers to learning. Thus the move towards a positive psychology in central to the 
implementation of inclusive education in South Africa, as it serves to highlight and 
encourage learner strengths, and to build on those to promote development and 
learning. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 18
 CHAPTER THREE 
RESEARCH DESIGN AND METHODOLOGY 
 
3.1 Research Design 
 
A research design serves as a plan for minimising the risk of extraneous variables that 
may hinder, and thus impact the validity of the results (Mouton & Marais, 1996). 
Quantitative methods, measuring the frequency and differences in processes, were 
employed in this study. The use of descriptive, correlational statistics was identified 
as the most suitable method in eliciting required data for comparative analysis 
(Breakwell, Hammond, & Fife-Shaw, 1998). The study yielded quantitative data, 
which was used to examine the differences in learned helplessness, depression, and 
academic achievement between the sample groups. The data gathered and analysed 
were thus descriptive, as the study was concerned with describing the existing 
distribution of variables and comparing the frequencies between the different groups 
(Grimes & Schultz, 2002).  Furthermore, the study was exploratory and correlational 
in nature as it aimed to explore the relationship between learned helplessness, 
depression, and academic achievement across samples. The data were analysed and 
interpreted to establish if a correlation exists between learned helplessness, depression 
and academic achievement. Correlational studies examine associations between 
exposures and outcomes in populations (Grimes & Schultz, 2002). This study further 
examined specific correlations between the outcomes of two population groups ? of 
the mainstream learners, and learners with barriers to learning. 
 
3.2 Research Questions  
 
The following research questions were used to guide the study: 
 
1.  Is there a significant difference in the prevalence of learned helplessness 
between a sample of mainstream learners and a sample of learners with 
barriers to learning? 
 
 19
 2.  Is there a significant difference in the prevalence of depression between a 
sample of mainstream learners and a sample of learners with barriers to 
learning? 
 
3.  Is there a significant difference in the academic achievement of a sample of 
mainstream learners and a sample of learners with barriers to learning? 
 
4.  What is the correlation between learned helplessness, depression, and 
academic achievement, both within and between sample groups? 
 
3.3 Research Sample 
 
A theoretical sampling method was employed in selecting participants for the study 
(Neuman, 1997). The subjects of the study constituted a purposive sample, 
comprising the learners from the intersen phase (grades 5, 6 & 7) of the chosen 
school. As outlined in Burhans and Dweck (1995), the validity of the learned 
helplessness model has been documented in middle to late grade school and junior 
high school children, as certain aspects of the model appear not to be applicable to 
younger children. The choice of sample was thus based on existing literature which 
suggests that attributions associated with learned helplessness may not result in 
learned helpless behaviour until the middle elementary school years and the 
simultaneous emergence of a more stable self-concept (Fincham et al., 1989). As 
such, most of the available data on learned helplessness in school-aged children 
pertain to learners in fourth and fifth grade and upwards (Fincham et al, 1989). 
 
 The subjects of the study included learners from a private co-ed primary school in 
Johannesburg. The school offers both a mainstream class per grade, as well as a 
bridging class for learners who experience barriers to learning. Both classes follow 
the same curriculum and prescribe the same learner outcomes, with the intention of 
possible mainstreaming of learners with barriers to learning. As such, only those 
learners who may benefit from the small class inclusion programme (SCIP) are placed 
in classes for learners with barriers to learning. In light of this, these learners are not 
globally delayed in functioning, and present with low average-to-average IQ scores. 
The range of barriers to learning accommodated at the school includes mild (non-
 20
 clinical) emotional and scholastic difficulties that may affect optimal learning within 
the classroom. These learners receive both in-class and extra-curricular support. 
 
  The learners attending the school range from middle to middle-upper socio-
 economic class families and multicultural backgrounds. The medium of language 
taught at the school is English, and the use of chosen instruments is thus appropriate 
for the sample.  
 
All learners within Grades five, six and seven were invited to participate in the study. 
The sample used for the study, however, only included those learners who had 
returned signed parental consent forms, and were thus legally entitled to participate in 
the study. In total, the sample included 57 learners from the intersen phase.  
 
3.4 Measurements and Instruments 
 
3.4.1 Children?s Attributional Style Questionnaire (CASQ) 
 
Learned helplessness is measured using the CASQ questionnaire, which was designed 
to measure attributional style in children. The CASQ is a forced-choice instrument 
created by Seligman, Peterson, Kaslow, Tanenbaum, Alloy, &Abramson (1984), 
following noted difficulty experienced by young children when completing the adult 
ASQ. As such, the CASQ is a 48 item standardized questionnaire designed for 
children aged eight to 13. Hypothetical good or bad events involving the child are 
followed by two possible explanations. For each event, one of the explanatory 
dimensions is varied, while the other two are held constant. 16 questions pertain to 
each of the three dimensions; half refer to good events and half refer to bad events 
(Peterson et al., 1993).  
 
Assigning a 1 to each internal or stable or global response, and a 0 to each external or 
specific response scores the CASQ. Scales are formed by summing the three scores 
across the relevant questions pertaining to each of the three dimensions, separating 
good and bad events. While the CASQ subscales are only moderately reliable, 
satisfactory reliabilities can be obtained from combining subscales, separately for 
good and for bad events. In a study conducted in South Africa by Mayer (1999, as 
 21
 cited in Thompson, 2002), the internal reliability of the CASQ, as measured by 
Cronbach?s alpha, was found to be satisfactory (? = 0.75) with a sample of black 
children. The CASQ scales and composites are consistent in showing explanatory 
style to be a somewhat coherent individual difference among children (Peterson et al., 
1993). 
 
In order to gain an overall composite explanatory style score, the composite score for 
bad events is subtracted from the composite score for good events. Resultantly, the 
overall style score (G-B) is used as a measure of learned helplessness.  An overall 
score of less than five is considered indicative of a pessimistic explanatory style, and 
thus learned helplessness (Nolen-Hoeksema, et al., 1986).  
 
The CASQ was employed in a previous study in South Africa (Thompson, 2002), in 
which certain items were modified in order to make the CASQ more suitable to the 
South African context.  While the constructs measured by each item were held 
constant, vocabulary and content were altered on certain items, in order to ensure 
applicability of the CASQ to the South African context. This modified version was 
employed for the purposes of this study. An example of an original item (item 37) is 
given below. 
 
 You go to an amusement park and have a good time.              
 ____ A. I usually enjoy myself at amusement parks. 
 ____ B.  I usually enjoy myself. 
 
As the term ?amusement park? is not commonly referred to in the South African 
context, this item was adapted to refer to a specific amusement park in South Africa. 
Thus the modified version reads as: 
  
You go to Gold Reef City and have a good time.              
 ____ A. I usually enjoy myself at Gold Reef City. 
 ____ B.  I usually enjoy myself. 
  
 
 
 22
 3.4.2 Children?s Depression Inventory (CDI) 
 
As outlined by Seligman (1996), there is no conclusive way to assess depression in 
children short of a clinical diagnostic interview. However the CDI serves as a reliable 
approximation of depressive symptoms in children aged between six and 17. The CDI 
(Kovacs, 1992, as cited in Kronenberger & Meyer, 2001), a downward extension of 
the Beck Depression Inventory, is a questionnaire used to measure the intensity of 
each of the four clusters of depressive symptoms, and is thus a useful tool in 
highlighting possible depressive behaviour. It is a 27-item self-report measure of 
cognitive, affective, behavioural, and social symptoms of depression. Each item 
consists of three statements, one of which represents nondepressed symptomatology 
(scored as 0), one of which represents moderately depressed symptomatology (scored 
as 1), and one of which represents severely depressed symptomatology (scored as 2). 
The child chooses the statement that best describes experiences over the past two 
weeks.  
 
Items are then added to give a total score and five subscale scores, each measuring 
Negative Mood, Interpersonal Problems, Ineffectiveness, Anhedonia, and Negative 
Self-Esteem. A raw score of 19 is commonly used as the cut-off, indicating clinical 
levels of depression. Lower cut-off scores prescribe greater sensitivity, but lower 
specificity (Kronenberger & Meyer, 2001). For the purposes of this study, item 9, 
which enquires about suicidal ideation and intent, was omitted from the questionnaire, 
as it was felt that the sensitive nature of this question was not conducive to the 
purposes of the study, and was not appropriate for the nature of the chosen sample. 
Scoring thereby assumed a total possible score based on 26 items.  
 
While the CDI has not been standardised in South Africa, the questionnaire has good 
psychometric properties with high internal consistency, and moderate test-retest 
reliability (Kronenberger & Meyer, 2001).  It must be noted that the CDI was not 
designed to reflect symptomatology as reflected in DSM-IV-TR, and thus may not be 
adequate in identifying children with DSM-IV-TR defined depression. Therefore, the 
CDI should only be considered as a self-report measure of depression-anxiety, with 
most items being face-valid for depression.  
 
 23
 3.4.3 Academic Marks 
 
Academic marks of the participants were obtained from the class teachers in the form 
of English and Maths marks as indicated on the previously completed end-of-term 
school report. These marks were used as subjective measures for academic 
achievement, reflecting standards held by the school, and prescribed outcomes as 
required by the Gauteng Department of Education. As suggested by Sun (2006), the 
reliability of the scores was dependent on teachers? standards and expectations while 
validity was attained relative to each child?s unique performance (within-subjects 
comparison) and to the class overall (between-subjects comparison). 
 
3.5 Data Analysis 
 
Statistical methods were employed to analyse the data. Descriptive statistics were 
used to highlight the prevalence of learned helplessness, depression, and academic 
achievement within the sample groups. Descriptive statistics were further used to 
ascribe a numerical value to raw scores gleaned from the sample data (Breakwell et 
al., 1998). As the study was concerned with describing the existing distribution of 
variables and comparing the frequencies between the different groups, the data 
gathered and analysed were descriptive (Grimes & Schultz, 2002). Descriptive 
statistics allowed for the graphic (tabulated) representation of scores, calculation of 
means and other measures, and analysis of distribution of scores (Howell, 1997). 
 
 As the study was comparative in nature, statistical measures were used to assess the 
differences between groups. Moreover, as the study investigated three dependent 
variables (learned helplessness, depression, and academic achievement), which may 
be related to the other variables assessed in the study, a multivariate analysis of 
variance (MANOVA) was required in order to analyse the data. Thus, rather than 
using separate analyses of variance to understand each variable, it was considered 
prudent to assess all of the related dependent variables in one analysis (Devlin, 2006). 
An assumption required for the correct use of a MANOVA is the assumption that the 
scores in each population are normally distributed around the population mean, or that 
the sampling distributions between means are normal (Howell, 2008). Normal 
 24
 distribution of each of the variables was thus tested, through the use of a Shapiro-
 Wilk?s test of normality.   
 
A MANOVA involves one analysis, thus avoiding the need to use separate t-tests, 
thereby reducing the possibility of Type 1 error, or the possibility that significant 
results are due to performing multiple statistical tests, rather than to a true difference. 
A MANOVA was used to assess whether there is a significant linear combination of 
the dependent variables in terms of the independent variables (the two sample 
groups). The univariate effects, or between-subject effects, were then examined 
separately for each of the dependent variables (Devlin, 2006). These were used to 
assess the difference of each of the dependent variables between the sample groups. 
 
As the study also aimed to explore the correlation between each of the dependent 
variables, a Pearson-r correlation was used to examine if there is a systemic 
relationship between the variables (Devlin, 2006). The aim of a correlational design is 
to show that levels of one variable are associated with another, and as such, no causal 
relationship can be inferred from correlational analyses (Breakwell et al., 1998).   
 
3.6 Procedure 
  
This research study abided by ethical codes of conduct according to the ethical 
research guidelines established by the committee for Research on Human Participants 
(Humanities) of the University of Witwatersrand (Wits). Ethical clearance was 
granted (protocol number MEDP/09/006 IH). Furthermore ethical guidelines, as 
outlined by Babbie and Mouton, (2001), were employed for this study. Permission to 
conduct the study was obtained form the school principal. Full information regarding 
the research study was provided (See Appendix A). 
 
The parents/legal guardians of the invited participants were required to provide 
written informed consent to allow their children to participate in the study, stipulating 
their understanding and acceptance of the research aims and process (refer to 
Appendix B). The learners, who accepted the invitation to become participants in this 
research project, further completed assent forms, stipulating in writing their voluntary 
involvement in this project (refer to Appendix C).  
 25
 Questionnaires were administered on a group basis by the researcher in order to 
ensure consistency of process and content. Questionnaires were read aloud by the 
researcher, in order to account for any reading difficulties that may have been 
encountered by the participants. Group exercises followed administration of the 
questionnaires to serve as debriefing. Following administration and completion of the 
questionnaires by the learners, the class teachers were given the telephone number of 
the Emthonjeni Centre, should further referrals for counselling or assessment have 
been deemed necessary. The class teachers were further encouraged to contact the 
Emthonjeni Centre or school-based psychologist should it have been noted that any of 
the learners require further intervention.  
 
In order to ensure that confidentiality was maintained throughout the research process, 
all respondent data was coded and kept in a locked cupboard by the researcher, 
ensuring privacy in the best way possible.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 26
 CHAPTER FOUR 
RESULTS 
 
4.1 Overview of Chapter 
 
This chapter presents an overview of the statistical analyses conducted in the study, 
and the result thereof. Results are reported in terms of the aims and research questions 
guiding the study, as specified. Descriptive statistics comprise an introduction to this 
chapter, as they serve to highlight the existing distribution of variables between the 
sample groups. Similarly, descriptive statistics allow for the calculation of means and 
analysis of distribution of scores. Descriptive statistics are presented in tables 1, 2, 
and 3 below.  
 
Further to this, this chapter presents the statistical results pertaining to analysis of the 
differences in each of the measured constructs (learned helplessness, depression, and 
academic achievement) between sample groups. Results of a multivariate analysis of 
variance (MANOVA) are reported in order to determine significant statistical 
differences between the mainstream group, and learners placed in small-class 
inclusion programmes (SCIP). As the correct use of a MANOVA requires 
assumptions of normality, results of the Shapiro-Wilk?s test of normality are 
presented as an introduction to the presentation of the MANOVA results.  
 
The final part of this chapter presents an analysis of the correlation between the 
measured constructs- learned helplessness, depression, and academic achievement- 
both within and between the sample groups. Results of a Pearson-r correlation are 
reported to determine the relationship between the measured constructs. These are 
initially reported as separate analyses of the correlations within each of the sample 
groups. However, in light of the research questions guiding the study, results 
pertaining to the whole sample are presented, in order to gain understanding of any 
significant relationships between the measured constructs.  
 
 
 
 
 27
 4.2 Descriptive Statistics of Sample Groups 
 
Tables 1, 2, and 3 below show the means, standard deviations, and minimum and 
maximum scores of each of the measured constructs. These are tabulated separately 
for each of the sample groups (Tables 1 and 2) as well for the full sample group 
(Table 3). 
 
Academic Achievement is recorded as an overall percentage gleaned from English 
and Maths marks. According to the standards maintained by the school and the 
Department of Education, an overall average equal to or greater than 50% is 
considered acceptable.   
 
CDI scores are represented to indicate the overall levels of depression amongst the 
sample groups. A score equal to or greater than 19 is considered indicative of clinical 
levels of depression. Cut-off scores equal to or greater than 12 may indicate greater 
sensitivity to depression, with lower specificity (Kronenberger & Meyer, 2001). 
 
The overall score used as a measure of learned helplessness (gleaned from the G 
minus B scale), as represented by the CASQ score, indicates the attributional style of 
the child. Overall, scores lower than 4 indicate that the child attributes bad events to 
internal, stable and global causes, and thus has a largely pessimistic attributional style. 
Similarly, scores equal to or greater than 5 indicate that good events are attributed to 
external, unstable, and specific causes, highlighting an overall optimistic attributional 
style (Nolen-Hoeksema, et al., 1986).   
 
Mainstream Sample 
Variable N Mean Std Dev Min Max 
Academic Achievement 28 78.04 9.79 55.00 92.00 
CDI score 28 6.53 8.64 0.00 22.00 
CASQ score 28 5.79 8.73 0.00 16.00 
 
Table 1: Descriptive Statistics of Mainstream Sample (N=28)  
 
 28
 The mean Academic Achievement score of the mainstream sample is 78.04. This 
highlights that the overall academic achievement of this group is well above the 
acceptable average.  The mean CDI score of the mainstream sample is 6.53. This 
mean indicates that this group does not display moderate or significant levels of 
depression. Moreover, a mean score of 5.79 on the CASQ indicates that this group has 
a more optimistic than pessimistic attributional style.  
 
SCIP Sample 
Variable N Mean Std Dev Min Max 
Academic Achievement 27 60.43 9.83 41.50 77.50 
CDI score 27 8.46 8.72 1.00 20.00 
CASQ score 27 6.24 8.72 0.00 16.00 
 
Table 2: Descriptive Statistics of SCIP Sample (N=27) 
 
The mean Academic Achievement score of the SCIP sample is 60.43. As expected, 
this mean is lower than that of the mainstream group. However, a mean of 60.43 
indicates that the overall academic achievement of this group is still well above the 
acceptable average.  The mean CDI score of the mainstream sample is 8.46. This 
mean indicates that this group does not display moderate or significant levels of 
depression, despite the mean score?s being higher than that of the mainstream group. 
The mean CASQ score of the SCIP group is 6.24. This indicates that this group has a 
largely optimistic attributional style.  
 
Full Sample 
Variable N Mean Std Dev Min Max 
Academic Achievement 55 69.39 13.17 41.5 92.00 
CDI score 55 7.45 8.68 0.00 22.00 
CASQ score 55 6.01 8.73 0.00 16.00 
 
Table 3: Descriptive Statistics of Full Sample (N=55) 
 
 29
 The mean Academic Achievement score of the full sample is 69.39. This highlights 
that the overall academic achievement of this sample is well above the acceptable 
average.  Overall, the mean CDI score of the full sample is 7.53, indicating that this 
group does not display moderate or significant levels of depression. Moreover, a mean 
score of 6.01 on the CASQ indicates that overall, this group has a more optimistic 
than pessimistic attributional style.  
 
4.3 Results of the Shapiro-Wilk?s test of normality 
 
Tests of normality, such as the Shapiro-Wilk?s, are used to assess whether the scores 
in each sample group are normally distributed around the sampling mean, or that the 
sampling distribution between means is normal (Howell, 2008). Normality thus needs 
to be checked for each of the independent variables. Tests of normality were 
conducted for each of the sample groups. A p-value equal to or greater than 0.05 is 
considered to be an acceptable measure of normality.  
 
 Results of the initial tests of normality are shown in Table 4 below. 
 
Group Variables Shapiro-Wilk 
Mainstream Academic Achievement w=0.86 
 
p=0.001 
 CDI score w=0.87 p=0.002 
 CASQ score w=0.89 p=0.006 
SCIP Academic Achievement w=0.96 p=0.379 
 CDI score w=0.98 p=0.793 
 CASQ score w=0.95 p=0.207 
 
Table 4: Shapiro-Wilk?s test of normality: p- values for independent variables  
  
The p-values, as shown in Table 4, indicate that the test statistics for the mainstream 
group are significant at p=0.05 for all the dependent variables. This rejects the null 
hypothesis that the data from this group represents a normal distribution. However, 
 30
 the p-values gained from the SCIP group show that this data represents a normal 
distribution.  
 
Observation of the raw scores for the mainstream group highlighted an outlying value 
for the overall academic achievement score (31.5%). Similarly, this outlier presented 
among the highest CDI score (19) and lowest CASQ score (2). This outlier was 
therefore removed from the calculations, in order to examine possible effects thereof 
on the normality of the dependent variables within the mainstream group. Table 5 
shows the result of re-calculation of p-values within the mainstream group, following 
removal of the outlying observation.  
 
Group Variables Shapiro-Wilk 
 
Mainstream Academic w=0.93 
 
p=0.075 
  CDI score w=0.88 p=0.004 
  CASQ score w=0.89 p=0.008 
 
Table 5: re-calculated p-values of dependent variables subsequent to removal of 
outlying observation within the mainstream group 
 
Table 5 indicates that the removal of the outlying observation substantially increased 
the p-values of the academic achievement within the mainstream group. Hence, scores 
of academic achievement are approximately normally distributed, with p=0.075. 
However, re-calculated p-values for scores on the CDI and CASQ are still 
significantly low, indicating that the data for these scores are not normally distributed. 
 
Thus, in order to conduct further analyses of these scores, it was necessary to apply 
transformations in order to normalise the variables as measured by the CDI and 
CASQ. The log transformation was used for this purpose. Despite the initial 
normative distributions evidenced in the SCIP sample, log transformations were 
applied to the same constructs as measured in this sample, in order to maintain 
homogeneity among the groups in terms of statistical values. The results of the log 
transformations are presented in Table 6.  
 31
 Group Variables Shapiro-Wilk 
Mainstream Academic Achievement w=0.93 
 
p=0.075 
  LogCDI score w=0.95 p=0.3 
  LogCASQ score w=0.93 p=0.073 
 
SCIP Academic Achievement 
 
w=0.96 
 
p=0.379 
  LogCDI score w=0.99 p=0.975 
  LogCASQ score w=0.93 p=0.065 
 
Table 6: results of log transformations applied to CDI and CASQ variables 
 
The p-values in this table indicate that each of the dependent variables is now 
assumed to be normally distributed. Further analysis was done based on these 
variables- Academic Achievement, LogCDI score (depression) and LogCASQ score 
(learned helplessness). 
 
4.4 Results of MANOVA 
 
A MANOVA (multivariate analysis of variance) was used to determine significant 
differences between the Mainstream and SCIP groups compared simultaneously on 
Academic Achievement, CASQ score and CDI score. Multiple measure of 
association, known as Wilk?s Lambda, are used for this purpose. Results of this are 
shown in Table 7 below. 
 
 
 
 
Table 7: Wilk?s Lambda Values 
 
The computed value of Wilks' Lambda is found to be 0.53 and p<0.0001 indicating a 
significant difference between the sample groups pertaining to at least one of the 
dependent variables- Academic Achievement, CDI score or CASQ score. Thus, 
 Wilk?s Lambda p- value 
0.5282109         p<.0001 
 32
 univariate tests were conducted to determine which of these differences is significant. 
Results of the univariate tests are represented in Table 8 below. 
 
Variables Source DF f-value p-value 
Academic Achievement Full Sample 1 44.32 <.0001 
CDI score Full Sample 1 2.19 0.1452 
CASQ score Full Sample 1 0.18 0.6719 
 
Table 8: results of univariate tests 
  
Results indicate that no significant difference exists between sample groups with 
regards to depression, as measured by the CDI score with f=2.19 and p=0.145. 
Similarly, results of learned helplessness, as indicated by the CASQ score, are not 
significant with f=0.18 and p=0.672.  However, there was a significant difference in 
the academic achievement between the two groups, with f=44.32 and p<0.0001.  The 
mean value of Academic Achievement for the mainstream group is 78.04 and that of 
the SCIP group is 60.43.  
 
4.5 Results of the Pearson-r correlation 
 
Correlations indicate the type of relationship that exists between each of the variables. 
A Pearson correlation coefficient was calculated in order to investigate the 
relationship between academic achievement, depression (as measured by the CDI 
score), and learned helplessness (as measured by the CASQ score). This was 
conducted for the sample group as a whole (N=55), as well as within each of the 
sample subgroups (mainstream and SCIP groups). Further hypothesis testing was 
conducted in order to determine whether the correlations are significant. 
 
4.5.1 Results within the mainstream sample 
 
Table 9 indicates results of the Pearson-r correlation within the mainstream sample.  
 
 
 33
   Academic LogCDI score LogCASQ score 
Academic 1     
LogCDI score - 0.55** 1   
LogCASQ score 0.23NS - 0.2NS 1 
* indicates significance at the 0.05 level (p<.05)  
** indicates significance at the 0.01 level (p<.01) 
NS indicates no significance 
 
Table 9: correlations between the independent variables within the mainstream 
sample 
 
Table 9 indicates a moderate negative correlation between academic achievement and 
CDI score (r = -0.55).  The results suggest that there is a moderate relationship 
between academic achievement and depression within this sample group.  Other 
correlations between the independent variables are not significant. 
 
4.5.2.  Results within the SCIP sample 
 
Table 10 below indicates results of the Pearson-r correlation within the SCIP sample. 
 
  Academic LogCDI score LogCASQ score 
Academic 1     
LogCDI score 0.1NS 1   
LogCASQ score 0.15NS - 0.45* 1 
* indicates significance at the 0.05 level (p<.05) 
** indicates significance at the 0.01 level (p<.01) 
NS indicates no significance 
 
Table 10: correlations between the independent variables within the SCIP sample 
 
Table 10 indicates a moderate negative correlation between CASQ score and CDI 
score (r = - 0.45). The results suggest that there is a moderate relationship between 
 34
 learned helplessness and depression within this sample group.  Other correlations 
between the independent variables are not considered significant. 
 
4.5.3.  Results within the full sample 
 
Table 11 below indicates results of the Pearson-r correlation within the full sample. 
 
  Academic LogCDI score LogCASQ score 
Academic 1     
LogCDI score - 0.33* 1   
LogCASQ score 0.1 NS - 0.29* 1 
* indicates significance at the 0.05 level (p<.05) 
** indicates significance at the 0.01 level (p<.01) 
NS indicates no significance 
 
Table 11: correlations between the independent variables within the full sample 
 
Table 11 indicates a low negative correlation between academic achievement and CDI 
score (r = -0.33). Similarly, a low correlation between CASQ score and CDI score (r 
= -0.29) is evident. These results suggest a weak relationship between depression, and 
each of the other variables.  Other correlations between independent variables are not 
significant.  
 
 
 
 
 
 
 
 
 
 
 35
 CHAPTER FIVE 
DISCUSSION 
 
5.1 Overview of Chapter 
 
This study comprised a comparison between learned helplessness, depression and 
academic achievement between two sample groups- learners placed in a mainstream 
class, and learners placed in a small class inclusion programme to accommodate 
barriers to learning. The aim of this study was thus to explore the difference in 
prevalence of each of the measured constructs between the sample groups. Moreover, 
the study aimed to examine the correlation between learned helplessness, depression 
and academic achievement, both within and between the sample groups. 
   
Results of the study, as outlined in the previous chapter, will be interpreted and 
discussed in this chapter, in line with the literature reviewed in chapter two.   
 
5.2 Differences in the measured constructs between sample groups 
 
5.2.1 Difference in the prevalence of learned helplessness between the groups 
 
The Children?s Attributional Style Questionnaire (CASQ) was used as a measure of 
learned helplessness. The findings of the study, as indicated in Table 8, show that 
there is no statistically significant difference in the prevalence of learned helplessness 
between the sample groups (p=0.672). Results of the statistical analyses, as indicated 
in Tables 1 and 2, indicate that the prevalence of learned helplessness in each of the 
groups is relatively low (N=28, mean=5.79; N=27, mean=6.24).   
 
Achievement-based helplessness is associated with a tendency to attribute failure to 
ability rather than to effort. Thus, attributions of failure to ability foster the belief that 
successful performance is no longer contingent upon action and effort. In light of this, 
Benson and Kennelly (1976) assert that for learned helplessness to occur, experience 
with failure is necessary. While it is evident that the SCIP sample (N=27, 
mean=60.43), as indicated in Table 2, is performing lower on academic tasks than the 
mainstream group (N=28, mean=78.04), as indicated in Table 1, the overall academic 
 36
 achievement of the SCIP group is not indicative of academic failure. This may 
therefore serve to explain why the prevalence of learned helplessness within the 
sample groups is low (Tables 1, 2, and 3). 
 
Furthermore, the literature suggests that effort attributions become increasingly 
important with age, such that effort may be more highly valued than outcome itself in 
children aged between 10 and 12 (Cole et al., 1999). Moreover, Dweck (1975) asserts 
that children, who attribute outcome to effort, rather than ability, are more likely to 
work harder or longer. Dweck further posits that some learning-disabled children, 
despite academic difficulty, or even failure, retain a positive sense of competence 
through successful experience in other areas. Results of these studies suggest that 
negative attributions about ability become less stable following such successes, which 
further counteracts the effects of school difficulty and failure (Dweck, 1975).  
 
Similarly, it is evident that a number of contextual factors at play have contributed to 
the positive adjustment of learners faced with barriers to learners into the mainstream 
of the school. It is hypothesised that it is the contribution of these factors that have 
successfully buffered the development of learned helplessness in the SCIP sample, (as 
well as in the mainstream sample). In a study conducted by Peterson and Barrett 
(1987) examining attributional style as a predictor of performance in students with 
disabilities included in a college setting, it was found that the relation between 
attributional style and performance existed for the overall sample, regardless of the 
presence or absence of a disability. Similarly, Martinez and Sewell (1996) assert that 
with regards to explanatory style, college students with and without disabilities may 
in fact be more alike than different. This alludes to the presence of extraneous factors 
that may serve influential in the development of attributional style, be it optimistic or 
pessimistic. The findings of the study suggest that despite differences in the overall 
academic achievement of the sample groups, similarities exist in the attributional style 
of the learners within these groups. While no conclusive findings can be reported, it is 
hypothesised that it is the structuring of the learning environment within the school 
that has successfully buffered against the development of an adverse attributional 
style, thereby minimizing the prevalence of learned helplessness amongst the learners.  
 
 37
 In line with the principles underlying inclusive education, the extensive restructuring 
of the education system in South Africa has called for the development of learner 
strengths, through which learners are empowered and enabled to participate actively 
and critically in the learning process (Department of Education, 2001). This has 
prescribed the need for a positive teaching and learning culture, which has placed 
emphasis on positive attributes that foster optimal learning and development. The 
philosophy of inclusion promotes acceptance, which is crucial to social and emotional 
development (Engelbrecht, 2004). The structure of the school in which the study was 
conducted fosters a sense of belonging, in which each of the learners is valued as an 
integral and equal member of the school community. Inclusion in South Africa 
recognizes that every child can learn and belongs in the mainstream of both school 
and community life (Department of Education, 2001). The school nurtures this 
philosophy, by providing learners in the SCIP class with the necessary support to 
overcome barriers to learning. As the study was conducted in a private school, it is 
suggested that both the school itself and the parents of learners within the school, are 
thus able to afford a range of services that enhance learning and development. These 
services include access to private tutors, speech therapists, occupational therapist, 
psychologists, and even classroom assistants.  
 
Moreover, in a study conducted by Yssel et al. (2007) it was found that South African 
parents generally reported that their children encountered positive social experiences 
in inclusive settings. This was evidenced in the present study, as it was found that 
learners in the SCIP class maintain close friendships with peers in the mainstream 
class. Thus it is hypothesized that the range of contextual factors at play has 
successfully ameliorated the development of learned helplessness within the sample 
groups. It is further suggested that such factors have counteracted the development of 
depression within the learners.  
 
5.2.2. Difference in the prevalence of depression between the groups 
 
The Children?s Depression Inventory (CDI) was used as a measure of depression. The 
findings of the study, as indicated in Table 8, suggest that there is no statistically 
significant difference in the prevalence of depression between the sample groups 
(p=0.145). Results of the statistical analyses, as shown in Tables 1 and 2, indicate that 
 38
 the prevalence of depression in each of the groups is relatively low (N=28, 
mean=6.53; N=27, mean=8.46).   
 
Given the low prevalence of learned helplessness within the sample groups, and the 
average to above-average overall academic achievement of learners within these 
groups, this is an expectable result.  However, much of the literature suggests that 
children with learning difficulties are a vulnerable population, who are at a heightened 
risk for experiencing difficulties such as peer rejection, depression, anxiety, 
behavioral and conduct problems, poor academic adjustment, school dropout, and 
poorer long term outcomes than are children and youth without disabilities 
(Montague, Enders, Dietz, Dixon & Cavendish, 2008; Murray & Greenberg, 2006).  It 
is thus necessary to consider why results of the study seemingly contradict previous 
research findings, in which it is hypothesised that children faced with barriers to 
learning may present as more depressed than learners placed in a mainstream class.   
 
Developing greater understanding of the social and contextual lives of children with 
learning difficulties may therefore highlight the importance of social relationships and 
contexts as possible risk or protective factors in the development of adverse 
consequences (Greenberg, 2006; Talbott & Fleming, 2003; Werner, 2003). Previous 
research suggests that such contextual factors are associated with adjustment and 
development. Findings of such studies (Greenberg, 2006; Talbott & Fleming, 2003; 
Werner, 2003) indicate that positive social relationships with parents, teachers, and 
peers are related to social, behavioural and emotional adjustment of learners, 
especially those with learning difficulties. While this was not formally explored in the 
present study, it is suggested that the low prevalence of depression among the learners 
can be attributed to a range of contextual factors evidenced within the school 
environment that promote optimal emotional adjustment.  
 
Learners placed within the SCIP and mainstream classes feel a sense of belonging and 
connectedness to the school, fostered by engagement in a range of co-curricular and 
extramural activities. Moreover, the students appear to perceive the school as a safe 
and accommodating environment, geared towards providing all students with a 
positive learning experience. Furthermore, as suggested by Murray and Greenberg 
(2006), as children and youth spend a considerable portion of their day in schools, the 
 39
 role of teacher-learner relationships is important. Accordingly, it is suggested that 
emotionally warm relationships between teachers and students, characterized by open 
communication, support and involvement may provide learners with a sense of 
security within the school environment, which in turn, encourages social, emotional 
and academic competencies (Murray & Greenberg, 2006). Given the low teacher-
 learner ratios at the school in which the study was conducted, the learners benefit 
from individualized attention and support from the teachers. Further to this, constant 
communication between teachers and parents comprises an integral part of learner 
support within the school. Access to on-site resources, such as speech therapists, 
occupational therapists, counselors, psychologists, and teacher aids further provide 
ongoing support for learners at the school. It is hypothesized that perhaps the presence 
of a wide range of social, emotional and academic support fosters improved school 
adjustment, leading to a decrease in emotional difficulties, such as depression.  
 
Furthermore, peer relationships play an integral role in influencing the social, 
emotional, and academic adjustment of children and youth (Murray & Greenberg, 
2006; Talbott & Fleming, 2003). Learners placed in SCIP and mainstream classes 
have several opportunities to integrate and engage with each other, both during break 
time, and though combined structured learning activities. This has facilitated the 
development of many close friendships for the learners, both within and between the 
classes. Many of the learners have remained with their classmates throughout their 
schooling career, resulting in strong peer relationships. As suggested by Resnick et al. 
(1997), a strong sense of belonging in social contexts is associated with lower levels 
of emotional distress, and thus lowers levels of anxiety and depression. This may 
thereby account for the finding of the present study, in which levels of depression 
were identified as low across the sample. 
 
5.2.3. Difference in the overall academic achievement of the groups 
 
Measures of academic achievement comprised overall percentages gleaned from 
Maths and English marks, as reflected on end-of-term reports. The findings of the 
study, as indicated in Table 7, show a statistically significant difference in the overall 
academic achievement of the groups (P<0.0001). The mean value of Academic 
Achievement for the mainstream group is 78.04 (Table 1) and that of the SCIP group 
 40
 is 60.43 (Table 2). This is an expectable finding, consistent with the hypothesis of the 
study, considering that the groups are distinguished by academic ability and 
achievement.  
 
The learners in the SCIP class face barriers to learning that may affect optimal 
functioning in the classroom, thus the overall academic achievement is significantly 
lower than that of learners placed in the mainstream class. However, it is evident that 
the overall academic achievement within this group indicates that these learners are 
achieving learning outcomes, and are thereby effectively accommodated within the 
school setting. Results of the study support the notion that the small class inclusion 
programme fosters optimal development and adjustment of learners within the school, 
thereby upholding the underlying principles of inclusive education.  
 
5.3 Correlation between learned helplessness, depression and academic 
achievement 
 
The study aimed to examine the correlations between learned helplessness, 
depression, and academic achievement, both within and between the sample groups. 
Results of the Pearson-r correlation, as shown in Tables 9 and 10, highlight moderate 
relationships between two of the dependent variables within each of the sample 
groups, as will be delineated below. Further to this, results of the Pearson-r 
correlation conducted within the full sample indicate similar relationships between the 
variables as highlighted by within-sample results, although such correlations within 
the full sample were found to be significantly lower (Table 11). 
    
The results of the study, as shown in Table 10, suggest a moderate negative 
relationship between learned helplessness and depression within the SCIP group (r= - 
0.45).  Similarly, this correlation was found within the overall sample (r= -0.29), as 
seen in Table 11, although not as highly correlated as within the SCIP group. This 
finding thus suggests that higher scores on the CASQ, indicating low levels of learned 
helplessness, were reasonably associated with lower scores on the CDI, indicating low 
levels of depression. 
 
 41
 The results further suggest a moderate negative relationship between academic 
achievement and depression within the mainstream group (r = -0.55), as indicated in 
Table 9. Similarly, this relationship was further evidenced in the results of the full 
sample, although the correlation was markedly lower (r= -0.33), as shown in Table 
11. This finding thus indicates that higher academic achievement scores were 
moderately associated with lower levels of depression, as indicated by low scores on 
the CDI.  
 
These results suggest a weak-to-moderate relationship between depression, and each 
of the other variables, namely learned helplessness, and academic achievement. Other 
correlations between independent variables are not significant. These findings were 
somewhat unexpected; as it was hypothesised that learned helplessness would be 
strongly correlated with academic achievement, as suggested by previous research 
findings. Thus the discussion explores possible reasons for findings of the study. 
 
5.3.1 Correlation between learned helplessness and academic achievement 
 
Learned helplessness was not found to be highly correlated with academic 
achievement, both within and between the sample groups, as indicated in Tables 9, 
10, and 11. Further to this, as learned helplessness was only moderately correlated 
with depression within the SCIP sample (Table 10), and weakly correlated with 
depression within the full sample (Table 11), reliability of the CASQ, as the measure 
of learned helplessness as employed in the study, is questioned. 
 
Previous studies in South Africa in which the CASQ was used as a measure of learned 
helpless (Mayer, 1999; Thompson, 2002), necessitate the modification of several 
items to the South African context. While a modified CASQ was employed in the 
study, thereby increasing applicability to the South African context, past studies of the 
questionnaire have indicated that several items on the questionnaire involve local item 
dependence (Higgins, Zumbo & Hay, 1999). For the CASQ, locally context-
 dependent item sets (Haladyna, 1992, as cited in Higgins et al., 1999), elicit person 
covariance and situation covariance. Hence, studies conducted by Higgins et al. 
(1999) suggest that individual variables such as attributional style cannot be separated 
from the contexts in which they occur. Thus, due to the ?projective? nature of 
 42
 attributional style measures, the reality of the situation may be irrelevant to individual 
differences in explanatory style. It is thus difficult to reduce the impact of situational 
variables on causal attributions.  
 
As the study focuses on learners in a school setting, it is important to consider that the 
CASQ does not measure school-based attributions, but rather focuses on general 
attributions. Thus it would be difficult to assess the correlation between school based 
academic achievement and learned helplessness, using only the CASQ. It is suggested 
that the employment of the Student Academic Attribution Scale (SAAS), as proposed 
by Bell and McCallum (1994) may have been more applicable in its examination of 
causal attributions for success or failure situations in school settings. The CAVE 
(Content Analysis of Verbatim Explanations) technique is employed as a qualitative 
measure of attributional style (Peterson, Luborsky & Seligman, 1983). This technique 
examines naturally occurring attributions, which are recorded (either in verbal or 
written form) and subsequently analysed to provide an understanding of attributional 
style. It is suggested that this technique, or a combination of both quantitative and 
qualitative procedures may have yielded more reliable results of learned helplessness, 
lending to improved analysis and discussion.  
 
5.3.2 Correlation between learned helplessness and depression 
 
Moreover, much of the literature focuses on learned helplessness as a predisposing 
factor to the onset of depressive symptomatology. Findings of the study suggest a 
weak relationship between learned helplessness and depression within the full sample 
(r= -0.29), as indicated in Table 11, and a moderate negative correlation between 
learned helplessness and depression within the SCIP sample (r= - 0.45), as shown in 
Table 10, as higher scores on the CASQ were associated with lower scores on the 
CDI. The results thus allude to the notion that the presence of a more optimistic style, 
as indicated by higher scores on the CASQ, is indicative of a decreased vulnerability 
to depression, as indicated by lower scores on the CDI.  These results indicate the 
influence of personal beliefs pertaining to academic and social efficacy on observed 
levels of depression within the sample groups (Locker & Cropley, 2004). Peterson et 
al. (1993) assert that depression is consistently predicted by prior depression, 
pessimistic explanatory style, and by adverse life events. Therefore, as the overall 
 43
 explanatory style of the sample was found to be more optimistic than pessimistic, it 
can be hypothesized that the group is at lower risk for depression. This was confirmed 
through low scores on the CDI, as evidenced in the research results.  
 
The literature (Nolen-Hoeksema et al., 1986; Peterson et al., 1993) supports the notion 
of a strong relationship between attributional style and depression, as negative affect 
is more common in those who attribute failure to internal, stable causes, such as lack 
of ability, to those who attribute failure to external stable causes, such as task 
difficulty (Bell et al., 2004).  This was found in the present study, in which more 
optimistic explanatory style of the group was evidenced. However, it must be noted 
that the CASQ fails to measure key dimensions of attributional style such as ability 
and effort. Further to this, Bell et al. (2004) assert that due to the transient nature of 
many of the items of the CASQ (such as having bad luck), hypothetical events as 
questioned by the CASQ may be perceived as less pervasive and permanent than 
attributions to ability and effort.  
 
In a study conducted on the relationship of school-based attributions to depression 
(Bell, McCallum, Doucette, 2004), it was found that school-related attributions are 
more highly related to depression in nonclinical intermediate-grade students than 
general attributions.  However, as mentioned above, the CASQ, as the most widely-
 used measure of attributional style, does not measure school-based attributions, and 
thus it is difficult to ascertain the relationship between school-based attributional style 
and depression, using only the CASQ as a measure of learned helplessness. Thus, it is 
suggested that the use of a school-based measure of attributional style may have 
yielded stronger correlations between learned helplessness and depression.   
 
5.3.3. Correlation between academic achievement and depression 
 
Depression was found to be negatively correlated with academic achievement, both 
within the mainstream sample (r = -0.55), as indicated in Table 9, and within the full 
sample (r= -0.33), as shown in Table 11. As highlighted in Marcotte et al. (2006), the 
association between depression and academic performance rarely has been addressed 
in the literature. Cheung (1995) asserts that few studies have reported that depressive 
symptoms may be associated with academic difficulties and lower performance. 
 44
 Conversely, it may be concluded that absence of depressive symptoms is associated 
with average academic performance, as found in the present study. Chen, Rubin, and 
Bo-Shu (1995) reported that a decline in academic performance was associated with 
depressive affect. As both groups do not indicate a decline in academic performance, 
it can hence be suggested that the average academic competence among the sample is 
correlated with low levels of depression. While higher level of depressive symptoms 
have been reported in students with learning difficulties (Cheung, 1995), it is 
suggested that the successful accommodation of learners in the SCIP class has 
counteracted a range of adverse consequences that may otherwise manifest. 
Moreover, further studies concluded that depressive symptoms in children predicted 
lowered academic performance (Kellam et al., as cited in Marcotte et al., 2006). 
Resultantly, impaired school performance has been noted as one of the short-term 
consequences of a depressive episode. As neither the mainstream nor the SCIP group 
present with marked depressive deficits, it is further postulated that the possible 
learning difficulties comprising barriers to learning within the SCIP class are not 
indicative of depressive symptomatology. Moreover, these barriers seem to be 
successfully managed within the school, resulting in decreased incidences of 
depression and learned helplessness among the learners.  
 
5.4 Summary 
 
Analysis of the results alludes to a range of possible extraneous factors that may have 
contributed to the findings of the study. As such, it is suggested that the results of the 
study are somewhat inconclusive. Due to the small sample size, it is difficult to 
ascertain whether the findings are consistent with current research and trends in the 
field. Further to this, given the specificity of the research sample, generalizability of 
the findings is limited. As such, discussion of the results is limited to the sample on 
which the study was conducted.   
 
The following chapter serves as a conclusion, highlighting strengths and limitations of 
the study. Directions for further research are delineated.  
 
 
 
 45
 CHAPTER SIX 
CONCLUSION 
 
6.1 Summary 
 
The study aimed to explore differences in learned helplessness, depression, and 
academic achievement between two sample groups, thereby exploring the differences 
in each of the constructs between learners placed in a mainstream class, and those 
placed in a bridging class for barriers to learning. Moreover, the study aimed to 
explore the correlations between learned helplessness, depression and academic 
achievement, both within and between the sample groups. Results of the study 
indicate that while a significant difference was noted in the overall academic 
achievement of the groups, the prevalence of learned helplessness and depression 
across both groups was relatively low. Moderate correlations between learned 
helplessness and depression, and depression and academic achievement were 
evidenced.  
 
Results of the study were somewhat unexpected as, as suggested in the literature, 
learning difficulties may manifest as a consequence of, or may result in, increased 
learned helplessness and depression. However, the results suggest that in contrast to 
previous findings, learned helplessness and depression was no more prevalent in a 
group of students with learning difficulties. Such findings provide valuable insight 
into the strengths and limitation of the study, and further serve to provide guidelines 
for future research. 
 
6.2 Strengths of the Study 
 
The findings of the study provide insight into the range of possible factors that may 
have contributed to the low prevalence of learned helplessness and depression within 
the sample groups. Contextual factors, such as social support, teacher support, and 
accommodation of a range of learner needs, may prove critical in overcoming barriers 
to learning and promoting optimal development and functioning. Such insight further 
fosters enhanced understanding of the effectiveness of inclusive education in South 
Africa in accommodating a variety of learning needs within the mainstream of 
 46
 education. In this regard, the study sheds light on the applicability of the learned 
helplessness model and related fields, to the examination of learning needs within 
South Africa at present. Given the previous research conducted in the field in South 
Africa, the study highlighted the importance of conducting further studies within the 
South African context.  Furthermore, the findings allude to the examination of 
positive factors that may serve to counteract adverse consequences of learning 
difficulties. This highlights the role of positive psychology, when exploring learning 
and development. In light of this, the study highlights possible gaps in the literature, 
in discussing factors that promote the management of barriers to learning.  The 
research findings are beneficial in motivating for further research in the field, 
especially against the backdrop of inclusive education in South Africa.  
 
6.3 Limitations of the Study 
 
Due to the small sample size, the generalizability and representivity of the study is 
limited. The sample comprises learners from a private, well-resourced school in 
Johannesburg. As such, it is difficult to ascertain whether the results of the study can 
be generalised to the wider population, given the range of socio-economic variables 
evidenced in South Africa. Further to this, while the sample comprises learners faced 
with barriers to learning, findings of the study cannot be generalised to groups of 
students faced with severe learning problems and disabilities. It is thus difficult to 
ascertain whether the findings are consistent across samples with varying learner 
needs. Moreover, as the study relied solely on quantitative data, the findings are 
inconclusive, as the study did not explore a range of contributing factors that may 
have affected the results.  
 
6.4 Recommendations for Future Research 
 
In light of the strengths and limitations outlined above, it is suggested that future 
research in the field is critical in providing further insight into the implications of 
inclusive education in South Africa. Thus, it is necessary that future samples comprise 
representative groups, including a range of demographic and socio-economic 
variables. Research within the field of severe learning difficulties and disabilities is 
also needed in order to provide further insight and understanding. This in turn, will 
 47
 foster improved comparison between groups. The importance of large sample sizes is 
necessitated.  
 
Given the range of factors that seem to have successfully counteracted the 
development of learned helplessness and depression within the groups, further 
research should serve to investigate and measure components of positive psychology 
that buffer against adverse reactions and consequences. The applicability of such 
studies to the South African context at present is evident.  
 
It is further suggested that both quantitative and qualitative methods be employed on 
future research, lending to increased reliability and validity of the findings, and richer 
discussion and analysis thereof.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 48
 REFERENCE LIST 
     
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in 
humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74. 
 
Alloy, L. B., Lipman, A. J.,  & Abramson, L. Y.  (1992). Attributional style as a vulnerability 
factor for depression: Validation by past history of mood disorders.  Cogn Ther Res, 
16, 391?407. 
 
Babbie, E. & Mouton, J. (2001). The practice of social research. South Africa: Oxford 
University Press. 
 
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International 
Universities Press.  
 
Bell, S. M., McCallum, R. S., & Doucette, J. A. (2004). Relationship of school-based 
attributions to depression. Journal of Psychoeducational Assessment, 22, 106-123.  
 
Benson, J. S., & Kennelly, K. J. (1976). Learned helplessness: The results of uncontrolled 
reinforcements or uncontrolled aversive stimuli. Journal of Personality and Social 
Psychology, 34, 138- 145.   
 
Breakwell, G. M., Hammond, S., & Fife-Shaw, C. R. (1998). Research methods in 
psychology.  London: sage Publications Ltd. 
 
Burhans, K. K. & Dweck, C. S. (1995) Helplessness in early childhood: The role of 
contingent worth. Child Development, 66, 1719-1738.  
 
Chen, X., Rubin, K. H., & Bo-shu, L. (1995).  Depressed mood in Chinese children: 
Relations with school performance and family environment. Journal of Consulting 
and Clinical Psychology, 63, 938-947. 
 
Cheung, S. (1995). Life events, classroom environment, achievement expectation, and 
depression among early adolescents. Social Behaviour and Personality, 23, 83-92. 
 49
 Cole, D. A., Martin, J. M., Peeke, L. A., Seroczynski, A. D., & Fier, J. (1999). Children?s 
over- and underestimation of academic competence: A longitudinal study of gender 
differences, depression, and anxiety. Child Development, 70, 459-473.       
 
Department of Education. (2001). Education White Paper 6: Special needs education: 
Building an inclusive education and training system.  Pretoria: Department of 
Education. 
 
Devlin, A. S. (2006). Research methods: planning, conducting and presenting research. 
USA: Thomson Wadsworth. 
 
Donald, D., Lazarus, S. & Lolwana, P. (2002). Educational psychology in social context (2nd 
ed.). South Africa: Oxford University Press. 
 
Duckworth, A. L., Steen, T. A., & Seligman, M. E. P. (2005). Positive psychology in clinical 
practice. Annual Review of Clinical Psychology, 1, 629-651.   
 
Dweck, C. (1975). The role of expectations and attributions in the alleviation of learned 
helplessness. Journal of Personality and Social Psychology, 31, 674-685. 
 
Engelbrecht, P. (2004). Changing roles for educational psychologists within inclusive 
education in South Africa. School Psychology International, 25, 20-29. 
 
Fincham, F. D., Hokoda, A., & Sanders, R. (1989). Learned helplessness, test anxiety and 
academic achievement: A longitudinal analysis. Child Development, 60, 138-145. 
  
Grimes, D. & Shultz, K. (2002). Descriptive studies: What they can and cannot do. The 
Lancet, 359, 145. 
 
Higgins, N. C., Zumbo, B, D., & Hay, J. L. (1999).  Construct validity of attributional style: 
Modeling context-dependent item sets in the Attributional Style Questionnaire. 
Educational and Psychological Measurement, 59, 804- 820. 
 
 50
 Hiroto, D. S. & Seligman, M. E. P. (1975). Generality of learned helplessness in man. 
Journal of Personality and Social Psychology, 31, 311-327.  
 
Hirsch, J. K. & Conner, K. R. (2006). Dispositional and explanatory style optimism as 
potential moderators of the relationship between hopelessness and suicidal ideation. 
Suicide & Life-Threatening Behaviour, 36, 661- 669. 
 
Howell, D. C. (1997). Statistical methods for psychology. USA: Wadsworth Publishing 
Company. 
 
Howell, D. C. (2008). Fundamental statistics for the behavioural sciences.  Canada: 
Thomson Wadsworth.  
 
Kaslow, N. J., Rehm, L. P., Pollack, S. L., & Siegel, A. W. (1988). Attributional style and 
self-control behavior in depressed and nondepressed children and their parents. 
Journal of Abnormal Child Psychology, 16, 163-175. 
 
Kronenberger, W. G. & Meyer, R. G. (2001). The child clinician?s handbook. U.S.A.: Allyn 
& Bacon.  
 
Landsberg, E. (Ed.). (2005). Addressing barriers to learning: A South African perspective. 
Pretoria: Van Schaik Publishers. 
 
Lewis, A. D., Huebner, E. S., Reschly, A. L., & Valois, R. F. (2009). The incremental 
validity of positive emotions in predicting school functioning. Journal of 
Psychoeducational Assessment, 27, 397-408.   
 
Locker, J. & Cropley, M. (2004). Anxiety, depression and self-esteem in secondary school 
children: An investigation into the impact of standard assessments tests (SATs) and 
other important school examinations. School Psychology International, 25, 333-345. 
 
Lopez, S. J., Snyder, C. R., & Rasmussen, H. N. (2003). Striking a vital balance: Developing 
a complementary focus on human weakness and strength through positive 
psychological assessment. In S. J. Lopez & C. R. Snyder (Eds.), Positive 
 51
 psychological assessment: A handbook of models and measures (pp. 3-20). 
Washington, DC: American Psychological Association. 
 
Maier, S. F. & Seligman, M. E. P. (1976). Learned helplessness: Theory and evidence. 
Journal of Experimental Psychology, 105, 3-46. 
 
Marcotte, D., Levesque, N., & Fortin, L. (2006). Variations of cognitive distortions and 
school performance in depressed and non-depressed high school adolescents: A two-
 year longitudinal study. Cogn Ther Res, 30, 211-225.  
 
Martinko, M. J. & Gardner, W. L. (1982). An alternative explanation for performance 
deficits. The Academy of Management Review, 7, 195-204. 
 
Martinez, R. & Sewell, K. W. (2000). Explanatory style as a predictor of college performance 
in students with physical disabilities. Journal of Rehabilitation, January, February, 
March, 30-36. 
 
Mitchell, J. V. (2001). Personality correlates of attributional style. The Journal of 
Psychology, 123, 447-463. 
 
Montague, M., Enders, C., Dietz, S., Dixon, J., & Cavendish, W. M. (2008). A longitudinal 
study of depressive symptomology and self-concept in adolescents. Journal of Special 
Education, 42, 67-78.  
 
Mouton, J., & Marais, H.C. (1996). Basic concepts in the methodology of the social 
 sciences. Pretoria: HSRC Publishers. 
 
Murray, C. & Greenberg, M. T. (2006). Examining the importance of social relationships and 
social contexts in the lives of children with high-incidence disabilities. Journal of 
Special Education, 39, 220-233.  
 
Neuman, L. W. (1997).  Social research methods: Qualitative and quantitative approaches. 
Boston: Allyn and Bacon. 
 
 52
 Nolen-Hoeksema, S. K., Girgus, J., & Seligman, M. E. P. (1986). Learned helplessness in 
children: A longitudinal study of depression, achievement and explanatory style. 
Journal of Personality and Social Psychology, 51, 435-442.  
 
Overmier, J. B. &  Seligman, M. E. P. (1967). Effects of inescapable shock upon subsequent 
escape and avoidance learning. Journal of Comparative and Physiological 
Psychology, 63, 28-33.  
 
Peterson, C. & Barrett, L. C. (1987). Explanatory style and academic performance among 
university freshmen. Journal of personality and Social Psychology, 53, 603-607. 
 
Peterson, C., Luborsky, L., & Seligman, M. E. P. (1983). Attributions and depressive mood 
shifts: A case-study using the symptom-context method. Journal of Abnormal 
Psychology, 92, 96-103.  
 
Peterson, C., Maier, S. F., & Seligman, M. E. P. (1993). Learned helplessness: A theory for 
the age of personal control. New York: Oxford University Press. 
 
Peterson, C., & Seligman, M. E. P. (1984). Causal explanations as a risk factor for 
depression: Theory and evidence. Psychological Review, 91, 347-374. 
 
Resnick, M. D., Bearman, P. S., Blum, R. M., Bauman, K. E., Harris, K. M.,  & Jones, J. 
(1997). Protecting adolescents from harm: Findings from the national longitudinal 
study on adolescent health. Journal of the American Medical Association, 278, 823?
 833. 
 
Seligman, M. E. P. (1990). Learned optimism. New York: Knopf.  
 
Seligman, M. E. P. (1995). The optimistic child: A proven program to safeguard children 
against depression and build lifelong resilience. U. S. A.: Houghton Mifflin 
Company. 
 
Seligman, M. E. P., & Csikszentmihalyi, M. (2002). Positive psychology: An introduction. 
American Psychologist, 55, 5-14. 
 53
  
Seligman, M. E. P., Maier, S. F., & Solomon, R. L. (1971). Unpredictable and uncontrollable 
aversive events. In F. R. Brush (Ed.), Aversive conditioning and learning. New York: 
Academic Press. 
 
Seligman, M. E. P., Peterson, C., Kaslow, N. J., Tanenbaum, R. J., Alloy, L. B., & 
Abramson, L. Y. (1984). Attributional style and depressive symptoms among 
children. Journal of Abnormal Psychology, 83, 235-238.   
 
Shahar, G., Henrich, C. C., Winokur, A., Blatt, S. J., Kuperminc, G. P., Leadbeater, B. J. 
(2006). Self-criticism and depressive symptomatology interact to predict middle 
school academic achievement. Journal of Clinical Psychology, 62, 147-155. 
 
Sun, N. (2006). Working memory and academic achievement in children: Visual and verbal 
working memory and academic achievement in second grade children: A 
correlational study. Unpublished honours research report, University of the 
Witwatersrand, South Africa. 
 
Sweeney, P. D., Anderson, K., & Bailey, S. (1986). Attributional style in depression: a meta-
 analytic review. Journal of Personality and Social Psychology, 50, 974-991. 
 
Talbott, E. & Fleming, J. (2003). The role of social contexts and special education in the 
mental health problems of urban adolescents. Journal of Special Education, 37, 111-
 123. 
 
Thomas, A. (1979). Learned helplessness and expectancy factors: Implications for research in 
learning disabilities. Review of Educational Research, 49, 208-221. 
 
Thompson, S. J. (2002). An attributional style training programme for pre-adolescents: a 
pilot study. Unpublished masters thesis, University of the Witwatersrand, South 
Africa. 
 
Thornton, J. W. & Jacobs, P. D. (1971). Learned helplessness in human subjects. Journal of 
Experimental Psychology, 87, 367-372. 
 54
  
Valas, H. (2001a). Learned helplessness and psychological adjustment: effects of age, gender 
and academic achievement. Scandinavian Journal of Educational Research, 45, 71-
 90. 
 
Valas, H. (2001b). Learned helplessness and psychological adjustment II: effects of learning 
disabilities and low achievement. Scandinavian Journal of Educational Research, 45, 
101-114. 
 
Weis, J. R., Weiss, B., Wasserman, A. A., & Rintoul, B. (1987). Control-related beliefs and 
depression among clinic referred children and adolescents.  Journal of Abnormal 
Psychology, 96, 58-63.  
 
Wiggins, J. D. (2003).  Academic and social coping styles of learning-disabled and non- 
learning-disabled adolescents. Unpublished doctor of philosophy dissertation, 
University of Houston, USA.  
 
Yssel, N., Engelbrecht, P., Oswald, M. M., Eloff, I., & Swart, E. (2007). Views of inclusion: 
a comparative study of parents' perceptions in South Africa and the United States. 
Remedial and Special Education, 28, 356-365.