The Implementation of a Rational-Emotive Educational Intervention for Anxiety in a 3rd Grade Classroom: An Analysis of Relevant Procedural and Developmental Constraints
Cristea, Ioana-Alina, Benga, Oana, Opre, Adrian, Journal of Cognitive and Behavioral Psychotherapies
We tested the efficiency of a rational-emotive behavioral intervention to reduce the level of anxiety (emotional and behavioral) and remedy the irrational thinking in children (ages 9-10). The participants were 63 schoolchildren (3 classes), boys and girls. They were initially evaluated with the Spence Anxiety Scale (for the general and specific anxiety level) and the CASI questionnaire (for the level of irrational beliefs). Their parents completed the Spence Anxiety Scale (parent version). There were 3 groups (classes): rational-emotive behavioral education (REBE), sham intervention (a Placebo type group), no intervention. The intervention lasted for 20 sessions and we assessed the level of irrational beliefs (CASI) and the general and specific anxiety (Spence Anxiety Scale) before and after the intervention. Results did not show a significant improvement of the REBE group compared to the others, neither in measures of anxiety, nor in those of irrationality. The level of REBE specific knowledge (tested with a knowledge questionnaire) after the intervention was significantly higher in the REBE group than in the other two groups. Parents' evaluations differed from children's own evaluations: they tended to overlook the existence or frequency of anxiety symptoms in their children. Possible implications and explanations are discussed. Implications envisage the efficiency of REBE in reducing the anxiety and irrational thinking of school-children and possible problems regarding its applications in the classroom.
Keywords: anxiety, children, rational-emotive behavioral education, irrational beliefs
Growing up is turning out to be increasingly hard. In a world expanding its complexity in an amazing rhythm, lots of children find themselves having to deal with "adult" problems, such as socio-economic problems, abuse, neglect, relational problems. Their developmental equipment (their level of cognitive, emotional and social development) is not sufficiently advanced to keep up with the racing complexity and difficulty of their everyday world (Vernon, 2004). It has become clear that we need to provide children with instruments that would supplement and sometimes compensate the abilities they have due to their developmental paths (typical or atypical), in order to ensure their efficient adaptation to this complexity. These instruments can be regarded as skills (emotional, cognitive) that children might acquire from an educational intervention.
Rational-emotive and behavioral education: general principles and empirical data for anxiety problems in children
It is generally accepted that the prevention of problems is an easier task to tackle than the intervention in cases where the problems have already set in. In other words, we don't have to and should not wait for the onset of clinical or subclinical psychological problems (anxiety) to intervene. Rather we should teach children the abilities they need to attenuate the risk of developing these problems.
The present paper is focalized on such an intervention, namely rationalemotive and behavioral education (REBE), designed to provide children with an equipment of abilities for emotional and cognitive regulation. This refers to a rational, less distorted way of thinking instead of the irrational one that leads to dysfunctional negative emotions, which in their turn can lead to psychological conditions, such as anxiety or depression. REBE endorses the principles of rational-emotive behavior therapy (REBT), initiated by Albert Ellis at the middle of the last century, and from which it was derived. These principles can be synthesized as the ABC cognitive model (for a more detailed description of these principles and the ABC model, see David, 2006). Briefly, the ABC model states that it's not the event itself that causes our emotional states, but the interpretations we form regarding event. In the model, "A" refers to the activating event (internal and external stimuli). The "B" in the model represents the beliefs the person holds about the event. Among these beliefs, special emphasis is placed on evaluations, as a great amount of research distinguishes between two key evaluation "styles": irrational and rational. Irrational evaluations are not logical, don't have factual support in reality (are not concordant with reality) and hinder the person from achieving his/hers goals. Rational evaluations are their complete opposite, being logical, concordant with reality and helping the person achieve his/hers goals. According to the ABC model, our evaluations of an external or internal event directly cause the "Cs" (emotional and behavioral consequences). The Cs can consist of dysfunctional emotions and behaviors, which are brought about by irrational evaluations or functional ones, brought about by rational evaluations. While the valence of functional and dysfunctional emotions can be the same (e.g. both are negative when the event is the person finding out he/she has a serious illness), the differences lie in intensity and their impact upon behavior. The basic approach of rational-emotive behavioral education is to try to make irrational evaluations flexible, changing them into rational ones, thus correcting and/or preventing dysfunctional emotions and behaviors. When working with children and adolescents, this is not carried out directly, but through a series of activities and follow-up discussions. Vernon (2004) considers that, in contrast to other emotional education programs, REBE offers its beneficiaries the "power" to assume control over their own lives. This is achieved first through their understanding of the link between their thoughts, feelings and actions. Second, it is accomplished by the grasp that, even though they have no means of changing the other people or events in their lives, they can exert control over themselves, their thoughts, emotions and behaviors.
Moreover, studies show that approximately 70% of the young people who benefit from mental health services do so only at school, which turns the educational system into a privileged system for offering these types of services for young people (Farmer, Burns, Philips, Angold, & Costello, 2003; Gonzales et al., 2004). But as the number of young people that could benefit from such services increases, so does the need of implementing empirically sustained interventions in schools (evidence-based interventions) (Gonzales et al., 2004; Stoiber & Kratochwill, 2000). REBE is one such intervention, as we will detail later on.
Anxiety disorders are among the most prevalent forms of psychopathology in children. Studies with community samples showed that approximately 8-12% of children meet the criteria for a form of anxiety disorder that is serious enough to interfere with their daily functioning (Anderson, Williams, McGee, & Silva, 1987; Costello, 1989; Spence, 1998). Anxiety problems in children are associated with a range of negative consequences in terms of school, social and personal adaptation (Messer & Beidel, 1994; Spence, 1998). There is evidence showing that anxiety problems are not merely temporary problems for children. If left untreated, they can persist in adolescence and adulthood (Keller et al., 1992; Pfeffer, Lipkins, & Plutchik, 1988; Spence, 1998) and can also be predictors for other clinical disorders (e.g. depression).
One of the first studies regarding rational-emotive education (REE - a form of education which was later on enriched with behavioral elements, thus resulting in rational-emotive behavioral education or REBE) was carried out by Knaus and Bokor (1975). They designed a pilot study to measure the efficiency of a REE intervention to influence children in developing a more positive selfconcept and reducing test anxiety. The results sustained the greater efficiency of REE compared to a program of enhancing self-esteem, but both approaches were superior to no intervention.
DiGiuseppe and Kassinove (1976) examined the effect of a REE intervention, with a duration of 15 weeks, on the emotional regulation mechanisms of 204 students, grades 4 to 8. The experimental group was compared both to a group that received alternative treatment and to a control group. The results sustained the idea that rational-emotive principles could be acquired by students and that this acquisition was considered to be conducive to a reduction in the anxiety and neuroticism scores.
Miller (1978) compared REE with a condition combining REE with behavioral reinforcers, another one combining REE and homework, and a control condition. The subjects were 96 children, with low and high IQ levels. The dependent variables were the children's knowledge, neuroticism and trait-anxiety. The results of the three experimental groups were significantly superior in comparison to the control group. Intelligence did not prove to have an effect on the results (see also Silverman, McCarthy, & McGovern, 1992).
Greenwald (1985) addressed his intervention to 4th grade students, aged between 10 and 12 years. They were randomized in 4 groups - one control and 3 experimental groups. These were as follows: REE, REE plus rational-emotive and behavioral bibliotherapy, REE plus rational-emotive and behavioral imagery. The results showed that students in the third experimental group (REE + imagery) displayed the most significant improvements in self-concept and rational thinking, compared to the other groups. The reduction in anxiety was greater for the first group (REE) than for the second one (REE + bibliotherapy). It was concluded that rational-emotive imagery could bring additional benefits to REE (see also Silverman, McCarthy, & McGovern, 1992).
Grassi (1985) investigated the efficiency of REE and self-instructional training on children with medium and high anxiety, grades 4 to 6 (36 4th grade, and 36 6th grade subjects). Emotional evolution was monitored through 2 questionnaires filled in by the children, a behavior evaluation scale, filled in by the parents, and another one completed by teachers. Compared to the control group, both experimental groups attained the content of the two types of training, but only the effects of REE were maintained in the follow-up phase. They both contributed to the reduction of anxiety, but REE was more efficient in reducing neuroticism.
Cardenal Hernaez and Diaz Morales (2000) studied the effect of three months of REE versus relaxation techniques on self-esteem and anxiety level, in 12-14 years old children from Spain. 93 students were randomized in the 2 experimental groups and the control group. The measures applied consisted of the Piers-Harris self-concept scale, a body attraction scale and STAI. Measures were carried out for pretest, posttest and at three months follow-up. Results showed that both experimental conditions equally contributed to the global increase in selfesteem and the reduction of anxiety.
Meta-analytical research concerning the efficiency of REBE for anxiety problems is not consistent. Gossette and O'Brien (1993), in a metaanalysis that took into account 33 unpublished dissertations, found an efficiency of 25%. DiGiuseppe and Bernard (1990), in a metaanalysis conducted on 23 studies, found an efficiency of 50%, whereas Hajzler and Bernard (1991), analyzing 21 studies, found an efficiency of 80% (see also Popa, 2004).
One of the most recent meta-analysis (Gonzales & al., 2004) regarding the efficiency of rational-emotive behavioral therapy (REBT) in a wider range of emotional and behavioral problems (including anxiety) highlights some important discoveries. It was carried out on 19 peer-reviewed studies and it analyzed 5 domains of results (disruptive behavior, impropriation of rationality, GPA -grade point average, self-concept and anxiety). Their first and more general conclusion is that, subsequent to a REBT intervention, the modal child or adolescent had better performances, regardless of the type of result considered, than approximately 69% of the control, no treatment groups. But another conclusion proves very interesting and can also act as a justification for the current study: REBT intervention seems to be efficient both for children and adolescents with an identified clinical problem, as well as for those without one. This motivates preventive interventions, targeting sub-clinical problems or problems that have not yet manifested, but for which we know the child to be vulnerable. Another important conclusion of this meta-analysis, which contributes even more to justifying the current study, states that the efficiency of a REBT intervention is much higher, IF it addresses younger children (primary school) than older ones (secondary school or high-school). Another conclusion warns about the danger of the intervention not being effective because of its too short duration: REBT efficiency is higher in conditions with a medium (675-770 minutes) or high (1200-2115 minutes) duration of the intervention. Regarding anxiety, effect sizes for these measures are presented in 6 of the 19 studies (12 effect sizes), leading to a .48 effect size.
Developmental considerations regarding children's emotional and behavioral problems and the implementation of REBE for these problems
As we mentioned before, many young people are not ?ready", developmentally speaking, to deal with the ?adult" challenges they often face. One of the factors contributing to this is the fact that their level of cognitive development predisposes them to irrational thinking, in the form of: 1) overgeneralizations, 2) demandingness, 3) low frustration tolerance, 4) awfullizing and 5) global evaluation (Vernon, 2004).
Bernard, Ellis and Terjesen (2006) express a similar opinion, drawing attention to the close relations that exist between children's emotional and behavioral problems and certain developmental problems in the domain of cognitive processing of emotional or social aspects. It becomes obvious that any therapeutic approach of children, be it preventive or corrective, has to take into account the fact that they are developmentally vulnerable to some cognitive processing errors (Bernard, et al., 2006). Many of these errors are due to the ontogenetic features of the cognitive development of 9-10 years old children.
Typical cognitive errors (according to Bernard, et al., 2006) include: 1. drawing arbitrary inferences (conclusions that are not based on evidence or that contradict the evidence); 2. selective abstraction (focusing on a detail, taken out of context, ignoring essential characteristics of the situation); 3. maximization/minimization (errors in evaluating the significance of the event); 4. personalization (the tendency to relate external events to themselves when there is no basis for making this connection); 5. overgeneralization (drawing a conclusion based on limited and isolated elements); 6. dichotomous thinking (tendency to place events in opposite categories, e.g. good-bad).
These errors become even more poignant and can develop into dysfunctional processing styles, when the information to be processed has an increased emotional valence. A relevant example to sustain the idea that cognitive development cannot be ignored in the study of emotional and behavioral problems comes from the studies of social cognition. Research in this domain shows that family factors (e.g. exposure to problematic parental factors) that predict adaptation problems, also predict social cognition deficits (Barahal, Waterman, & Martin, 1981; Downey & Walker, 1989; Pettit, Dodge, & Brown, 1988; Smetana, Kelly, & Twentyman, 1984,). This suggests that social cognition abilities could mediate the relationship between family risk factors and child's adaptation. On the other hand, if children from high-risk families are exposed to competent models, they can develop social cognition abilities that can compensate the increased risk for adaptation problems, related to parental psychopathology and abuse (Downey & Walker, 1989). Thus the developmental level of social cognition is an important element in the relation between the risk factors the child is exposed to and the emergence of emotional disorders. We can notice a fairly transparent symmetry with irrational beliefs (the Bs in the ABC model, for details about the model, see David, 2006) and the emergence of dysfunctional negative emotions (which by repetition can turn into a dysfunctional emotional pattern, the premise for the development of emotional and behavioral problems, such as those from the anxiety spectrum). An interesting research topic would envisage the conceptual and empirical relations between these constructs (irrational beliefs and social cognition).
One of the fundamental reasons for considering REBE as a privileged modality to approach children's emotional and behavioral problems is the fact that it uses a perspective consistent with that of developmental psychology. More precisely, REBE satisfies the existent criteria for determining a theory's developmental potential (Bernard, et al., 2006; Holmbeck & Updegrove, 1995). Among these criteria, according to Bernard, et al., (2006), we can mention: keeping up-to-date to the most recent discoveries in the developmental literature; taking into account the critical developmental periods and tasks, relevant for the child's problem; flexible prioritizing of the symptoms, as a function of the degree in which each symptom is atypical from a developmental standpoint. In addition, REBE follows a developmental perspective also because of its evaluation and intervention modalities, which are projected so they would take into account the developmental level (physical, cognitive, emotional, and social) of the child or adolescent (Bernard & Joyce, 1984; Bernard, et al., 2006). Many of its techniques and activities are specific ones, developed together with educators and teachers, who work in direct interaction with young people.
The main objective of the current study is the evaluation of a REBE educational intervention in children of 9-10 years of age for emotional (e.g. intense worry) and behavioral problems (e.g. avoidance of anxiety-inducing stimuli) from the anxiety spectrum. Our goal is to see whether through REBE interventions we can improve the mental (reduction of the irrationality level) and emotional (reduction of the anxiety level) functioning in these children. Also we attempt to clarify the efficiency and problems that can arise in implementing a specific REBE program in the ecological context of the classroom.
The comparative evaluation refers to determining the efficiency bonus that a specific REBE intervention can bring over: 1) a sham intervention (where the improvement would be caused by Placebo mechanisms, the results being due to the mere presence of an intervention and not to the technique) and 2) no intervention. The improvements regard the reduction of the irrationality level and the anxiety level, measured by specific instruments.
We must mention that the study has an exploratory character. Therefore we cannot formulate precise hypotheses, so the following are more likely suppositions that should be taken with some degree of caution.
1. Children who benefit from the REBE intervention will present a more significant reduction of irrationality than those who benefit from the sham or no intervention.
2. Children who benefit from the REBE intervention will present a more significant reduction of anxiety (global score as well as specific problem categories) than those who benefit from the sham or no intervention.
Subjects: Subjects were 63 children, aged between 9 and 10 years from three 3rd grade classes (n1= 18, n2= 23, n3= 22), selected from 2 schools in Cluj- Napoca. 36% of all children were girls and 64 % boys. Participation in the program was voluntary and school, teachers' and parents' agreement for the program were previously secured. We could not in this case ensure a random selection and distribution in groups. Even if the schools were randomly selected, in order to carry out the intervention we depended on the availability of the school-principle, teachers and parents. Moreover, we were interested in seeing how the intervention works in an ecological environment, because that will be the setting for structured mental health programs dealing with children's emotional difficulties. To prevent some of the problems that arise from the lack of randomized selection, we controlled for the existence of significant differences between the 2 groups on measures of anxiety and irrational beliefs.
Experimental design: The research method was quasi-experimental, in the form of a pre- and posttest group design, because we have no means of controlling the various environmental influences children are subjected to, outside the limited weekly duration of the intervention. The independent variable consisted in the type of intervention and had 3 modalities (rational-emotive intervention, sham intervention, no intervention), which will be detailed in the procedure section. The dependent variables circumscribed the level of irrational beliefs, the anxiety level (both general level and specific types) and the degree of rational-emotive knowledge following the intervention. These were measured by specific tests and a knowledge questionnaire.
We decided to include a sham intervention group so that, should the intervention be efficient, we could extract some information about the mechanisms that lead to its efficiency. This group benefited from an intervention with the same duration and the same person as the REBE group, children and their parents being told that the person is certified in such an intervention. The activities and discussions had the same structure and set of rules (non-evaluative). What differed was the content of the sessions (astronomy - information about universe, stars, planets), constructed in such a way that the mechanisms assumed to operate in the rational-emotive intervention are, in as much as possible, inactive. Thus, the mentioning of people, their emotional problems, cognitions, or relationships was suppressed from the content of the sham intervention. If this particular REBE intervention is efficient and operates on the basis of the general mechanism presented by the REBT theory, then it should lead to a reduction of anxiety significantly more substantial than what could have been achieved by the mere maturation of the subjects (check control group) or on the basis of a different intervention, in which the REBT change mechanism is kept inactive (check sham group).
Procedure: The testing phase was individual for all subjects. The same instruments were used in the pre- and posttest phase (after the intervention). The intervention phase lasted for 3-4 months, with 2 regular sessions of about 45 minutes per week (20 sessions of actual intervention for the REBE group and the sham group).
The REBE group benefited from an intervention structured in 4 modules:
1. Emotions (development of vocabulary and knowledge about emotions)
2. Beliefs and behaviors (understanding what beliefs are and how they determine our emotions and behavior - the ABC model)
3. Self-acceptance (learning to accept themselves and others as imperfect human beings, with positive and negative features, avoiding global evaluations).
4. Problem solving (developing problem solving strategies and approaching specific problems for anxious behavior)
The material used consisted of the book ?Programul de dezvoltare a inteligenNei emoNionale prin educaNie raNional-emotiva si comportamentala, clasele I-IV"/ Thinking, Feeling, Behaving. An emotional education Curriculum for Children (author Ann Vernon, translated and adapted in Romanian by Opre, David, Baltag, & Vaida, 2004). Each session comprised an activity part (stories, games or other activities), followed by discussions.
The sham intervention group benefited from an intervention of the same frequency and duration and with the same person. The content however envisaged the enrichment of the knowledge about Earth and Universe and avoided as much as possible discussions involving animate beings (humans, their relations, beliefs, emotions). The structure was identical to that of the REBE group, images, texts and games were used, interactions were encouraged, the discussions stage was present, but these were all focused on the specific information presented. In the no intervention group, there was only a pre- and posttest phase.
For measuring anxiety we used to the following instruments:
* The Spence Children' Anxiety Scale - SCAS (Spence, 1994). The scale is composed of 38 items, 6 filler items and an open question. The child is asked to read each statement and appreciate how often that particular thing happens to him on a 4-point scale. The questionnaire offers a global anxiety score, as well as scores for specific clusters of anxiety related problems. These clusters are represented by the subscales of the SCAS: panic attack and agoraphobia, separation anxiety, physical injuries fear, obsessive-compulsive behavior, generalized anxiety. It is constructed following the DSM-IV criteria, which enhances its precision in accurately identifying anxiety problems and it is meant to be an indicator of the number and severity of anxiety symptoms. The authors also establish cutoff points, circumscribing three problematic categories in which the subject could be placed: at risk (16% of the population), borderline clinical (7%), clinical (2-3%). The SCAS has been adapted for the Romanian population (Benga, 2006, in progress), and the preliminary date indicate good reliability, both for the global scale and for its subscales. Data from other populations (German, Dutch) indicated very good reliability for the scale and its subscales and good discriminate validity, using a clinical anxiety diagnosis as criterion (Spence, 1998).
* The Spence Children' Anxiety Scale (SCAS) - Parent version (Spence, 1994). The content, scoring and interpretation are almost identical to the SCAS. It consists of 38 items and an open question. There are fewer studies regarding it, but the data indicate satisfactory to very good reliability for the scale and its subscales and good discriminant validity regarding the clinical anxiety diagnosis (except for the generalized anxiety subscale), Nauta et al., 2004. It is also being adapted for the Romanian population (Benga, 2006, in progress).
For measuring irrational beliefs we used:
* The Child and Adolescent Scale of Irrationality - CASI (Bernard & Laws, 1999): It is addressed to children and adolescents between 9-18 years of age and is comprised of 28 items, formulated as statements about which the subjects to express their agreement on a 5-point Likert scale (1- strongly disagree, 5- strongly agree). CASI overcomes the problems presented by the other existent irrational beliefs scales, as the theoretic model it was based on takes into account the recent theoretical and empirical discoveries in REBT and REBE research. Also the items are exclusively cognitive ones. The scale was adapted on Romanian population (Popa, 2006). The validation study used factorial analysis and 4 factors were identified (consistent with the ones in the initial validation study). These are: low tolerance to frustration brought on by rules, global evaluation of the self, demands for fairness, low tolerance to frustration brought on by work (Popa, 2006). The scale has good global fidelity (a Cronbach= 0.84) and its subscales have satisfactory to good fidelity (Popa, 2006).
We also resorted to an evaluation of the knowledge attained subsequently to the rational-emotive behavioral intervention by means of a knowledge questionnaire. This comprised of 20 questions, formulated from the content of the REBE lessons and it was administrated to all 3 groups. We wanted to see whether there were any significant differences after the intervention in the level of declarative knowledge between the 3 groups, in other words to see whether or not at least declaratively (even if not expressed in a reduction of anxiety or irrational beliefs), the REBE group had learned things not acquired by the other groups.
The table below displays the means and standard deviations for the anxiety measures (pre and post-test).
If we compare these scores to the normative values (from validation studies carried out on other populations), we can see that for the REBE group, the pretest mean values for separation anxiety (cut-off points: 7-9), physical injuries fears (cut-off points: 5-6), as well as global anxiety (cut-off points: 40 - 51), can be placed in the at risk category for clinical anxiety problems (according to the cut-off values previously described). Also the pretest mean value for obsessive compulsive behavior for this group can be located in the borderline clinical domain (cut-off points: 11-12). For the sham group, none of the pretest mean values can be located in the at risk, borderline clinical or clinical spheres. For the control group, only the pretest mean values for physical injuries fears (cut-off points: 5-6) and obsessive compulsive behavior (cut-off point: 10) can be located in the at risk area. None of the posttest mean values, for any of the three groups, can be located in the at risk, borderline clinical or clinical domains. Where more than one value is displayed for cut-off points, it is because the values are different for boys and girls.
In table 2, we present the measures for irrational beliefs. As the normative values reported in the initial validation study (Popa, 2006) are general ones for young people with ages between 9 and 17 years, we didn't view them reliable enough for reporting the mean values obtained.
It is clear that, given the developmental differences among children of these ages, the normative values for high and very high irrationality cannot be indifferent to age.
We used paired samples t tests. Significant values are marked with an asterisk (p<.05). Effect sizes (ES) for significant differences are included.
We can see that the effect of the intervention in each group concerning irrational beliefs is practically insignificant. Regarding the anxiety level, we must first notice that the REBE group displays significant improvements on the panic attack and agoraphobia subscale, improvements that are also present in the sham group, but not the control group. An interesting result is that both the REBE group and the sham group show significant improvements on the measures of generalized anxiety, again not present in the control group. Also all groups show significant improvements on the obsessive compulsive disorder subscale and in the global anxiety score. Perhaps not surprinsingly the level of specific REBE knowledge has only significantly improved in the REBE group. Regarding effect sizes, their values are generally small (<.0.4), with a small to medium effect size for the changes in the level of REBE knowledge.
We must note that in the pretest phase the differences among the 3 groups are not significant at p<.05 for irrationality (F=2.66) and anxiety (F=3.20) both as global scores and subscale scores. However, we must acknowledge that the means of the REBE group (for anxiety) are consistently higher (although not significantly so) than those in the other groups, which was also an ethical consideration that oriented us to using that particular group as the target group. At posttest, there are still no significant differences at p<.05 among the 3 groups for anxiety (F=3.67) or irrationality (value for F=0.53). All the above, correlated with the intra-group comparison data, allow us to conclude that in the particular case of these classes of students, the REBE intervention did not have a significant, consistent, transparent effect on irrationality and anxiety. However, when we look at the data regarding the REBE knowledge, we notice that at posttest they are significantly different in the REBE group from the sham group (F=3.70, p<.05) and the no intervention group (F=5.26, p<.05).
Data from parents
A number of 42 parents from all three groups completed the parent version of the Spence scale (at pretest). We computed Spearman correlation coefficients for the evaluations of parents and those of their children. The parents' evaluation of the anxiety problems of their children and the children's own evaluations differ in the sense that the parents tend to overlook the existence or severity of anxiety problems. We have found positive medium correlations at the subscales of separation anxiety (r=0.463, p<.05) and physical injuries fears (r=0.488, p<.05). The correlations for global anxiety and the other anxiety subscales were not significant at p<.05.
We should mention that, even if some intra-group comparisons are statistically significant, effect sizes are small, which reveals small clinical and ecological significance.
Regarding anxiety, we should note that all three groups show statistically significant improvements for their global anxiety. The effect size is higher in the REBE group that in the two other groups, but still in the small size range, which makes this result not relevant from a clinical standpoint.
The REBE group displays significant improvements on the panic attack and agoraphobia subscale, improvements that are also present in the sham group, but not in the control group. Effect size is a little higher in the REBE group than in the sham group on this measure, but differences are not that relevant as they remains in the small effect size range. Therefore, we are not justified to assert a specific effect of the REBE intervention on this problem category. We must also mention that for this subscale, the initial mean values were very low (even much more so for the sham and control groups), so the significant difference can be an epiphenomenon of these small means, reported to which any small change can count. Both the REBE and the sham group show significant improvements on the measures of generalized anxiety. A possible explanation could be that, for children this age, a big part of the generalized anxiety problems are generated by school pressure, evaluation, friendship or competitive relations with peers. Both the REBT and the sham intervention indirectly approach these problems by using activities in which children are not evaluated, trying to get everyone involved, making them work in groups. Another possible explanation is procedural and refers to the fact that the items measuring generalized anxiety are more general and vague (e.g. I feel scared), so they could be interpreted differently by children in different moments, which could partly explain the fluctuation of answers. Also all groups show significant improvements on the obsessive compulsive disorder subscale, but with small effect sizes. The differences could be due to the natural evolution of the subjects (maturation). Another explanation could be a procedural one: the items for obssesive compulsive behavior on the Spence scale are more difficult to understand for the child (e.g. ?I have to think of special thoughts to stop bad things from happening (like numbers or words)"). It might be that children don't really understand what these items refer to or that at posttest they are more familiar with them and with the way they should be interpreted and so their answers would reflect reality more accurately.
Regarding intra-group comparisons for irrational beliefs, the significant result at for global evaluation of self and global irrationality in the sham group (not present in the other groups) could be accounted for by the evolution of subjects or Placebo-type mechanisms. However it is more likely that they are accounted for by procedural aspects regarding the problematic application of the CASI questionnaire with children this age (will be detailed in the limits section) or social desirability (children remembered having filled in the scale before and feel they are expected to offer different answers, even though in this case they were specifically instructed that the second application has nothing to do with the first). The fact that there is a significant increase in the level of REBE knowledge only in the REBE group can allow us to safely assume that, at least at a declarative level, the children in the REBE group have impropriated some of the specific knowledge, even more so as this process seems to be absent in the other groups.
Efficiency of the program (intra-group and inter-groups comparisons)
Starting off from this data, we came up with 2 major interpretative ways: a procedural perspective and a developmental perspective.
From a procedural perspective, explanations should be searched in problems specific to working in a classroom, which can be confounding variables and could account for these results. The context is an ecological one and can offer an accurate image of the way such an intervention could really work, in an educational program implemented in schools (increased external validity). But problems arise with regard to the internal validity, due to the fact that we can't exclude the distorting influence of other variables than those of interest. Many studies have discussed the role of parents' and teachers' irrational beliefs in inducing and reinforcing an irrational thinking style in the child (Bernard, Ellis, & Terjesen, 2006). Parents and their behaviors may play a double role, both as models and reinforcers. Their action could go in the opposite direction to the REBE intervention. We tried to diminish this source of error by explaining to the parents, in a preliminary meeting, that they should not contradict and instead try to encourage the new ideas and behaviors the child acquires during the intervention. Still, we can't exclude the fact that some of them have acted, consciously or not, in the opposite direction of our intervention. A frequently documented example is that of perfectionism: the child is taught that he/she can't possibly do things perfectly and that's it's not sound to ask that from himself. A better version is for him to try as much as he can to do everything very well, but without thinking that a potential failure would make him a worse human-being. However, many parents demand perfection from their children and punish them if they don't succeed in doing everything perfectly (for example always get the highest grades, at all subjects) (Bernard et al., 2006). It is also evident that the parents' influence on their children is much greater than any influence we would hope to achieve in weekly sessions, during a few months.
The teachers could represent another source of error, similar in its action mechanism. They too have got more time with the children and more control and reinforcement possibilities to make their influence more consistent than that of the intervention. We tried to control this problem since the teacher of the REBE group had attended REBE training. Still, she was present during all the sessions, although we specifically requested her not to intervene. It is however possible that her presence could have acted as an inhibitor for the adequate involvement of children in the proposed activities (especially those that required personal examples).
However, procedural aspects are not by themselves enough to explain our results. The activities used were taken from a manual and were specifically conceived for group work. We need to consider explanations at a deeper level; therefore it would be useful for us to look at the developmental characteristics of the children involved in this program. So the second perspective for data interpretation is a developmental perspective. We will start off from the observation that in the REBE group, although there is no effect in the direction of ameliorating anxiety and irrational beliefs, there is a significant and consistent effect on REBE knowledge. Children seem to have acquired the knowledge being discussed, but this doesn't seem to impact their way of thinking and their emotional problems.
The main objective of our intervention referred to the fact that, subsequently to the activities and discussions, children would extract ideas about certain concepts (emotions, beliefs, behaviors) and then generalize and apply these concepts in their daily lives. However, as we have pointed out in the theoretical part, their cognitive development is impinged on by some typical processing errors (Bernard et al., 2006). It may be that, in the context of this intervention, all these translated into a limited capacity of transferring the acquired information in real life situations, especially when dealing with emotionally loaded content. For example a cognitive error such as that of selective abstraction (focusing on details and ignoring essential features of the situation) (Bernard et al., 2006) could lead children to see the activities used as simple games, without extracting general principles (which was the real purpose of the activity). Even in the cases when they did extract some regularity, another cognitive error specific to their point of cognitive development is the situated, localized nature of their inferences and concept application (a concept's area of application is circumscribed to the context it was learned in). In other words, it could be that what is learned in the classroom is only applied in the classroom and not transferred to other life situations (e.g. family problems, problems with peers). These issues could be even more significant as the similarities between contexts (class situation - other life situations) are not really transparent to children. The problem situations that are outside the actual intervention sessions may not automatically activate the idea of applying the learned concepts. Even though they have the declarative knowledge, it is possible that children cannot explore the benefits of this knowledge because of their developmental particularities. A mental health educational program should take these issues into account.
An additional observation should also be made. It regards the specific action mechanisms of rational-emotive education: the modifications of irrational cognitions lead to the correction of dysfunctional emotions. The present research does not offer enough data to extract inferences about the validity and applicability of this mechanism in the case of anxiety. Anxiety problems in children this age have, as we have previously said, a resilient behavioral component (Keller et al., 1992; Spence, 1998). It could be that a general action mechanism, such as the one postulated by the REBT theory, may not be sufficiently efficient in the case of anxiety. Intervention might have to be specifically targeted on the particular aspects of anxiety behaviors.
Data from parents
The results are consistent with other data obtained using the Spence scales with other populations. Nauta et al. (2004) indicate inter-correlations in the range of 0.41-0.66 in the group of children with diagnosed anxiety disorders and in the range of 0.23-0.60 in the group without diagnosed anxiety disorders (our results fit in that range). The highest degree of agreement is met for subscales that envisage behaviors easily observable (Nauta et al., 2004). In our case, the significant, positive correlations were medium sized and obtained in the cases of separation anxiety and physical injuries fears, which include observable behaviors.
A series of studies from other domains, such as temperament research, raise the issues of the credibility and accuracy of parental evaluations. Studies reveal the different biases and errors that can mark the parental evaluations: in their answers, parents project the image they have constructed about the child and his/her actual behaviors (Benga, 2002); they don't understand the items or the instructions; they don't know the child's behaviors and their significance; their recollections can lack accuracy (Benga, 2002; Rothbart, Chew, & Gartstein, 2001); they want to offer socially desirable answers. All of these can be possible explanations for the reduced correlations between the parents' evaluations and the child's self-evaluations in the current research.
Other possible reasons have to do with the specificity of some clusters of problems. The anxiety problems for which the evaluations are more concordant are separation anxiety and fears of physical injuries. These include behaviors that are more easily observable, more frequent and upon which parents are used to direct their attention (e.g. ?my child is scared of dogs", ?my child is scared if (s)he has to sleep on his/her own"). Moreover, these are behaviors that are usually verbalized or clearly expressed by the child (e.g. cries, yells, verbally protests to sleeping alone). However, parents are less equipped in noticing problem behaviors of a different nature (social phobia, obsessive-compulsive behavior). These are much less transparent and less frequently verbalized by the child because it is often difficult for him/her to identify the source of the problem or he/she simply does not perceive them as problems (e.g. a child who often repeats a number or phrase to himself/herself in order to prevent bad things might not see it as a problem).
Also we cannot omit the sociological explanation, which is the fact that parents spend less and less time with their children and thus don't have sufficient time to observe such problem behaviors. Actually, parents expressed this point in the preliminary discussions (e.g. ?I barely see my child, I don't now what he does most of the time").
IMPLICATIONS FOR FUTURE RESEARCH
Based on the data obtained and the analysis carried out, possible future research could approach:
* The modalities through which cognitive development particularities of children of different ages can be approached in educational programs for mental health and how this can enhance the efficiency of these programs (through the generalization and transfer of knowledge by children).
* The identification of specific mechanisms that operate in determining anxiety problems in children (especially discovering the mechanisms that come into action before these problems achieve clinical intensity) and how we can counteract these in preventive interventions.
* The development and adaptation of instruments of evaluating children's irrational beliefs, instruments with items that are more comprehensible and adequate for the particular age group.
Regarding the limitations, the first one refers to the lack of randomized selection and distribution of subjects. We detailed this in the methodology section, so we will not dwell on it again. Another limitation, resulting from this, is the fact that the anxiety means of the REBE group are consistently higher (although not significantly so) than those in the other groups. This could also have been a factor influencing the results of the intervention. We will also detail another important limitation which refers to the CASI and its adaptation on the Romanian population. We chose this scale because, among the ones that measure irrational beliefs, it is the most robust one (theoretically and empirically). However we can't ignore the problems presented by the Romanian version of the CASI, especially with children this age. These could have seriously impaired the results. In brief, some of the problems are: the use of a 5-point scale, as it is hard for children to operate with these distinctions and they usually go for the extreme values; negatively-worded items, that are difficult to interpret and the children have to resort to complicated logical deductions about denying a negative statement; the use of some terms that are hard to understand or vague, such as ?frustrated", ?desperate"; the lack of age differentiated norms.
The present research raises more questions than it gives answers. But if we were to quote Einstein ?formulating a problem is often more important than solving it". This research cannot offer clear and definitive answers on the ecological efficiency of REBE in reducing anxiety in 9 to 10 years old children. Yet we hope to have been able to provide some empirical data and interpretations that can contribute to a more exact formulation of the problem.
Anderson, J. C., Williams, S. M., McGee, R., & Silva, P. A. (1987). DSM-III disorders in preadolescent children: Prevalence in a large sample from the general population. Archives of General Psychiatry, 44, 69-76.
Barahal, R., Waterman. J., & Martin, H. (1981). The social cognitive development of abused children. Journal of Consulting and Clinical Psychology, 49, 508-516.
Benga, O. (2002). Temperamentul si bazele timpurii ale personalitaNii [Temperament and the early basis of personality]. In A.Opre (ed.), Noi tendinNe În psihologia personalitaNii. Modele teoretice (vol. I). Cluj-Napoca: Ed. ASCR.
Bernard, M. E., & Cronan, F. (1999). The Child and Adolescent Scale of Irrationality: Validation Data and Mental Health Correlates. Journal of Cognitive Psychotherapy, 2, 121-132.
Bernard, M., Ellis, A., & Terjesen, M. (2006). Rational Emotive Behavior Approaches to Childhood Disorders: History, Theory, Practice and Research. In A. Ellis, & M.E. Bernard (eds.), Rational Emotive Behavioral Approaches to Childhood Disorders. Theory, practice and research.
Bernard, M. E., & Joyce, M. R. (1984). Rational emotive therapy with children and adolescents: Theory, treatment strategies, preventive methods. New York: Wiley.
Cardenal Hernaez, V., & Diaz Morales, J. F. (2000). Self-esteem and anxiety modification by the application of different treatments (rational-emotive education and relaxation) in adolescents. Asiedad y Estres Spain: Sociedad Espanola para al Estudio de la Ansiedad y Estres, 6, 295-306.
Costello, R. J. (1989). Child psychiatric disorders and their correlates: a primary care pediatric sample. Journal of the American Academy of Child and Adolescent Psychiatry, 28, 851-855.
David, D. (2006). Psihologie clinica si psihoterapie: Fundamente [Clinical psychology and psychotherapy: Basics]. Cluj-Napoca: Editura Polirom.
DiGiuseppe, A. R., & Kassionove, H. (1976). Effects of rational-emotive school mental health program on children's emotional adjustment. Journal of Community Psychology, 4, 382-387.
Downey, G., & Walker, E. (1989). Social Cognition and Adjustment in Children at Risk for Psychopathology. Developmental Psychology, 5, 835-845.
Essau, C. A., Muris, P., & Ederer, E. M. (2002). Reliability and Validity of the Spence Children's Anxiety Scale for Child Anxiety Related Emotional Disorders in German Children. Journal of Behavior Therapy and Experimental Psychiatry, 1, 1-18.
Farmer, E. M., Burns, B., Philips, S. D., Angold, A., & Costello, J. E., (2003). Pathways into and through mental health services for children and adolescents. Psychiatric Services, 54, 60-66.
Gonzales, J. E., Nelson, R. J., Gutkin, T. B., Sauders, A., Galloway, A., & Shwery, C. S. (2004). Rational Emotive Therapy with Children and Adolescents: A Meta- Analysis. Journal of Emotional and Behavioral Disorders, 4, 222-235.
Gossette, L. R., & O'Brien, M. R. (1993). Efficacy of rational emotive therapy with children: A critical reappraisal. Journal of Behavior Therapy and Experimental Psychiatry, 24, 15-25.
Grassi, R. (1985). Effects of self instructional training and rational emotive education on emotional adjustment in elementary school children [Abstract]. Dissertation Abstracts International, 46, 1730-1731.
Greenwald, K. E. (1985). Effects of rational emotive education, imagery and bibliotherapy on self concept, individual achievement responsibility, and anxiety in sixth grade children [Abstract]. Dissertation Abstracts International, 46, 979.
Hajzler, J. D., & Bernard, M. E. (1991). A review of rational-emotive education outcome studies. School Psychology Quarterly, 6, 27-49.
Holmbeck, G. N., & Updergrove, A. L. (1995). Clinical Developmental Interface: Implications of developmental research for adolescent psychotherapy. Psychotherapy, 32, 16-33.
Kazdin, A. E., & Weisz, J. (1998). Identifying and Developing Empirically Supported Child and Adolescent Treatments. Journal of Consulting and Clinical Psychology, 1, 19-36.
Keller, M. B., Lavoir, P., Wunder, J., Beardslee, V. R., Schwarts, C. E., & Roth, J., (1992). Chronic Course of Anxiety Disorders in Children and Adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 100-110.
Knaus, W. J., & Bokor, S. (1975). The effects of rational -emotive education lessons on anxiety and self-concept in sixth grade students. Rational Living, 11, 25-29.
Messer, S. C., & Beidel, D. C. (1994). Psychosocial correlates of childhood anxiety disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 33, 975-983.
Miller, N., & Kassinove, H. (1978). Effects of behavior rehearsal, written homework, and level of intelligence on the efficacy of rational emotive school mental health program. Journal of Community Psychology, 6, 366-373.
Nauta, M. H., Scholing, A., Rapee, R. M, Abbot, M., Spence, S. H., & Waters, A., (2004). A parent report measure of children's anxiety. Behaviour Research and Therapy, 7, 813-839.
Pettit, G., Dodge, K., & Brown, M. (1988). Early family experience, social problem solving patterns, and children's social competence. Child Development, 59, 107-120.
Pfeffer, C. R., Lipkins, R., & Plutchik, R. (1988). Normal children at risk for suicidal behavior: a two-year follow-up study. Journal of the American Academy of Child and Adolescent Psychiatry, 27, 34-41.
Popa, S. (2006, in press). Scala de iraNionalitate pentru copii si adolescenNi [CASI - The Child and Adolescent Scale of Irrationality].
Popa, S. (2004). EficienNa unui program de educaNie raNional emotiva În modificarea cogniNiilor iraNionale si inferenNiale la copii [The efficiency of a rational-emotive education program for changing irrational and inferential cognitions in children]. Romanian Journal of Cognitive and Behavioral Psychotherapies, 1, 53-67.
Rothbart, M. K., Chew, K., & Garstein, M. A. (2001). Assessment of temperament in early development. In L. Singer & P.S. Zeskind (Eds.), Biobehavioral assessment of the infant (pp.190-208), New York: Guilford.
Silverman, S. M., McCarthy, M., & McGovern, T. (1992). A review of outcome studies of rational emotive therapy from 1982-1989. Journal of Rational Emotive and Cognitive-Behavior Therapy, 10, 111-175.
Smetana, J. G., Kelly, M., & Twentyman, C. T. (1984). Abused, neglected, and nonmaltreated children's conceptions of moral and social-conventional transgressions. Child Development, 55, 277-287.
Spence, S. H. (1997). Structure of anxiety symptoms among children: A confirmatory factor-analytic study. Journal of Abnormal Psychology, 2, 280-297.
Spence, S. H. (1998). A Measure of Anxiety Symptoms Among Children. Behavior Research and Therapy, 5, 545-566.
Spence, S. H., Barrett, P. M., & Turner, C. (2003). Psychometric properties of the Spence Children's Anxiety Scale with young adolescents. Journal of Anxiety Disorders, 6, 605-625.
Stoiber, K. C., & Kratochwill, T. R. (2000). Empirically supported interventions and school psychology: Rationale and methodological issues. School Psychology Quaterly, 15, 75-105.
Vernon, A. (2004). Consilierea În scoala. Dezvoltarea inteligenNei emoNionale prin educaNie raNional- emotiva si comportamentala. Clasele I-IV [original: Thinking, Feeling, Behaving. An emotional education Curriculum for Children] (A. Opre, D. David, A. Baltag, & S. Vaida, Trans.) Cluj Napoca: ASCR.
Vernon, A. (2004). Rational Emotive Education. Romanian Journal of Cognitive and Behavioral Psychotherapies, 1, 23-37.
Weisz, J. R., Weiss, B., Han, S. S., Granger, D. A., & Morton, T. (1995). Effects of psychotherapy with children and adolescents revisited: A meta-analysis of treatment outcome studies. Psychological Bulletin, 117, 450-468.
Ioana-Alina CRISTEA*, Oana BENGA, & Adrian OPRE
Babes-Bolyai University , Cluj-Napoca, Romania
* Correspondence concerning this article should be addressed to: E-mail: firstname.lastname@example.org…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: The Implementation of a Rational-Emotive Educational Intervention for Anxiety in a 3rd Grade Classroom: An Analysis of Relevant Procedural and Developmental Constraints. Contributors: Cristea, Ioana-Alina - Author, Benga, Oana - Author, Opre, Adrian - Author. Journal title: Journal of Cognitive and Behavioral Psychotherapies. Volume: 8. Issue: 1 Publication date: March 2008. Page number: 31+. © A.S.C.R. PRESS Sep 2008. Provided by ProQuest LLC. All Rights Reserved.