The effectiveness and contribution of art therapy work with children in 2018 -what progress has been made so far? A systematic review

ABSTRACT This systematic literature review is a companion to our review of the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018) and aims to contribute towards the ongoing discussion in the field of art therapy on the effectiveness of art therapy in a wide range of child-aged clients. For this purpose, four major electronic databases were searched for quantitative articles relating to outcome measures in the field of art therapy with children from 2000 to 2017. A total of 13 articles responding to the inclusion criteria were identified and divided into three levels of evidence (Case-Smith, 2013). The results are organised into five clinical categories: trauma, special education and disabilities, non-specific difficulties, medical conditions and juvenile offenders. The potential benefits of art therapy in these five clinical populations is discussed and suggest that art therapy can be effective with children of the described categories. The limitations of this review and the current state of affairs are presented, alongside recommendations for future research to promote art therapy effectiveness research. Plain-language summary This article brings together all the research studies conducted so far on how helpful art therapy is for children. We use electronic databases to look at all the research published from 2000 to 2017. We found 13 articles which were relevant and classified under the following 5 headings: Art therapy with children dealing with traumatic events in their past. Four studies examining effectiveness have been conducted since 2002. These studies strengthen the claim that art therapy may help alleviate post-traumatic symptoms in children. Art therapy with children with special educational needs and disabilities. Four studies on effectiveness have been conducted since 2001. Most of the studies in this category suggest that art therapy has a positive effect on children with special educational needs and disabilities. Art therapy with children with no specific diagnosed difficulty. Three studies on effectiveness have been conducted since 2000. All three articles support the claim that art therapy may help children who are not diagnosed with specific difficulties but are faced with a variety of challenges in life. Art therapy with children dealing with medical conditions. Only one study fell into this category and addressed children coping with persistent asthma. This article lends some weight to the claim that art therapy may help children dealing with medical conditions, and specifically persistent asthma. Art therapy with juvenile offenders. Only one study fell into this category. This article supports the claim that art therapy may help juvenile offenders.In comparison to our recent review of research on how well art therapy works with adult clients (Regev & Cohen-Yatziv, 2018), we found far few studies in relation to children. Thus, we recommend further research on how well art therapy works with children.

This article brings together all the research studies conducted so far on how helpful art therapy is for children. We use electronic databases to look at all the research published from 2000 to 2017. We found 13 articles which were relevant and classified under the following 5 headings: (1) Art therapy with children dealing with traumatic events in their past. Four studies examining effectiveness have been conducted since 2002. These studies strengthen the claim that art therapy may help alleviate post-traumatic symptoms in children.
(2) Art therapy with children with special educational needs and disabilities. Four studies on effectiveness have been conducted since 2001. Most of the studies in this category suggest that art therapy has a positive effect on children with special educational needs and disabilities. (3) Art therapy with children with no specific diagnosed difficulty. Three studies on effectiveness have been conducted since 2000. All three articles support the claim that art therapy may help children who are not diagnosed with specific difficulties but are faced with a variety of challenges in life. (4) Art therapy with children dealing with medical conditions. Only one study fell into this category and addressed children coping with persistent asthma. This article lends some weight to the claim that art therapy may help children dealing with medical conditions, and specifically persistent asthma. (5) Art therapy with juvenile offenders. Only one study fell into this category. This article supports the claim that art therapy may help juvenile offenders.
In comparison to our recent review of research on how well art therapy works with adult clients (Regev & Cohen-Yatziv, 2018), we found far few studies in relation to children. Thus, we recommend further research on how well art therapy works with children.
adults does not accurately represent the fact that art therapists treat a much larger proportion of children (Eaton, Doherty, & Widrick, 2007). The current article thus focuses on the effectiveness of art therapy with children. Most reviews that have addressed art therapy in children have considered children who experienced traumatic events, such as one by Eaton et al. (2007) that summarises the literature in the field and covers ten case studies and two quantitative studies. They discuss the difficulties involved in generalising these studies. First, half of the studies had fewer than ten participants. Second, many articles did not specify the length of the therapeutic intervention although therapy did not appear to last for more than a few weeks. Third, the diagnoses were not always clearly stated in the articles reviewed. More recently, van Westrhenen and Fritz (2014) reviewed all research related to arts therapies on children with traumatic life experiences and reported that 73% utilised visual art therapy. Although their search identified 38 publications between 2002 and 2012, they noted that it was still difficult to generalise the contribution of art therapy to this population because the vast majority of these articles involved case studies rather than quantitative research.
A very recent article published by McDonald and Drey (2018) examined primary-school-based art therapy. This article reviewed two controlled (non-randomised) studies and two randomised controlled studies (RCT). Three of these were composed of sample groups of approximately 30 participants (including the experimental group and the control group). Although the authors concluded that these interventions appear to positively assist clients, they raised serious questions about the lack of research in this field, particularly since the demand for child emotional therapy is high; for example in Israel (Snir et al., 2018), the UK (McDonald & Drey, 2018) and other countries.
This review, and its companionour review on the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018), deals with research measuring the effectiveness of art therapy. It addresses two major challenges, the first of which is the definition of the term 'effectiveness'. Hill, Frappier-Davignon, and Morrison (1979) defined effectiveness as 'the attribute of an intervention or maneuver that results in more good than harm to those to whom it is offered' (p. 1203). This review adheres to the positivist perspective (Holton, 1993) to examine the assessment of effectiveness reported in quantitative studies conducted in the field. Since the field of art therapy is still relatively new, there are only a small number of studies of differing quality. This makes it hard to generate a comparative review that presents the state of the art in the field and draw exhaustive conclusions. For these reasons, this review implements the Case-Smith (2013) categorisation, that divides studies into three levels of evidence. Level 1 refers to randomised controlled trials (RCTs), level 2 refers to nonrandomized twogroup studies, and level 3 refers to nonrandomized one-group studies.
The second challenge has to do with the definition of art therapy. We applied the standard definition provided by the British Association of Art Therapists: Art therapy is a form of psychotherapy that uses art media as its primary mode of expression and communication. Within this context, art is not used as diagnostic tool but as a medium to address emotional issues which may be confusing and distressing … . Although influenced by psychoanalysis, art therapists have been inspired by theories such as attachmentbased psychotherapy and have developed a broad range of client-centered approaches such as psychoeducational, mindfulness and mentalization-based treatments, compassion-focused and cognitive analytic therapies, and socially engaged practice (The British Association of Art Therapy [BAAT], 2018).
This definition makes it clear that art therapy is a process that takes place in the presence of a certified art therapist, and indicates different areas where an effect or outcome in therapy can be expected as a result of this form of treatment.
Thus, the research question was formulated according to 'PICOS' components (Participants, Interventions, Comparisons, Outcomes, and Study design) (Moher, Liberati, Tetzlaff, & Altman, 2009): is art therapy effective for children as measured in results published from 2000 to 2017, in various quantitative studies corresponding to Levels 1, 2, 3 (Case-Smith, 2013)? These studies assessed the effectiveness of art therapy on variety of indices including symptoms, quality of life, health or mental health assessments. These indices were typically evaluated through questionnaires and occasionally by projective drawings. By posing this question, this systematic review aims to contribute to the ongoing discussion in the field on the degree of effectiveness of art therapy with children. This forms part of the academization process in the field of art therapy, which involves attempting to relate intervention techniques in the field to their significance for theoretical research.

Method
Searches were conducted in January 2017 by accessing four main electronic databases: Medline, PsycInfo, Scopus and Web of Science. The term 'art therapy' combined with the terms 'Effectiveness', 'Efficacy', 'Outcome', 'Measurement', 'Treatment' and 'Intervention' was searched. The search was restricted to articles published in English from the year 2000 onward, for reasons of relevancy and recency. We also consulted other reviews (such as the ones addressed above) to locate additional articles applicable to this study.
In the initial screening stage, both authors (who are certified art therapists) reviewed the abstracts to eliminate those that did not meet the research objectives. This resulted in 151 articles (see Figure 1). Then, these articles were read completely. Articles were selected for this study if they met the following criteria ( Figure 1): . They utilised a quantitative measure of the effectiveness of art therapy on a sample of clients. All case studies, method descriptions, qualitative analyses, and literature reviews that did not meet this criterion were rejected. A total of 80 articles were eliminated at this stage. . They assessed the specific impact of art therapy. All articles that described a combination of therapeutic intervention techniques, or a variety of art modalities simultaneously were eliminated. A total of 14 articles were eliminated at this stage. . Art therapy took place in an ongoing manner and in the presence of a qualified art therapist. Articles describing artistic intervention techniques that were not implemented in a therapeutic context or were implemented in a single art therapy session, or during sessions in which a qualified art therapist was not present were omitted. A total of 17 articles were eliminated at this stage.
All the articles that met the inclusion criteria were defined as articles that examined 'effectiveness' and quantified the impact of art therapy in a measurable way. In total, 37 studies were identified in 40 articles (three studies were published in two different articles). Of these 40, 13 articles related to art therapy with children. This article categorises these 13 articles in terms of the levels of evidence proposed by Case-Smith (2013).

Findings
The findings described in this article emerge from the 13 studies that met the inclusion criteria. The decision to present these studies as a review rather than as a meta-analysis is due to the emergent nature of the field of art therapy. There is little research in the field, and the differences between studies and the indices are so great that it would have been impossible to produce a meta-analysis that would yield meaningful results (a conclusion also reached by .
The authors discussed the clinical categorisation of each article until full agreement was reached. This categorisation was chosen to enable the reader to access knowledge in the field in a way that would allow and encourage researchers to conduct further research. The next section presents the findings classified into five clinical categories.
Different research methods were used in the 13 articles reviewed: five used a randomised comparison group (Level 1), two included a comparison group without randomised allocation (Level 2), and six used an intervention group without a comparison group (Level 3). Ten of the 13 articles did not define the research population by gender, and the remaining three studies only included female participants. No studies were found that exclusively addressed male participants, among the studies that met this article's inclusion criteria.

Category 1: children dealing with traumatic events
The first category consists of art therapy with children dealing with traumatic events in their past (see Table  1). Four studies examining effectiveness have been conducted since 2002 of which two (Pifalo, 2002;Pifalo, 2006) were conducted without a control group (Level 3) and two (Lyshak-Stelzer, Singer, Patricia, & Chemtob, 2007;Pretorius & Pfeifer, 2010) included control groups and random allocation (Level 1). Two of these dealt with children and adolescents, one exclusively with children and one exclusively with adolescents. One article focused on female participants alone. The total sample sizes ranged from 13 to 41 clients, all of whom received short-term therapy lasting eight to 16 sessions in group therapy settings.
The studies made use of questionnaires that assessed the children's symptoms of trauma. In one study, researchers also used a drawing of a human figure. All these studies strengthen the claim that art therapy may help alleviate post-traumatic symptoms in children.

Category 2: children with special educational needs and disabilities
The second category consists of children who had special educational needs or disabilities (see Table 2). Four studies on effectiveness have been conducted since 2001, two of which (Epp, 2008;Pachalska et al., 2001) without a control group (Level 3) and the other two (Freilich & Shechtman, 2010;Regev & Guttmann, 2005) which included control groups and random allocation (Level 1). Two articles focused solely on children and two other articles on both children and adolescents. The total sample sizes ranged from 14 to 109 clients in different experimental groups. Three studies described group therapy and the fourth one individual therapy. The therapeutic process lasted four to eight months.
Most of the studies in this category suggest that art therapy has a positive effect on children with special educational needs and disabilities. There was a wide variety of research instruments used in these studies including the Auditory Dysarthria Scale, and questionnaires evaluating self-esteem, and behavioural and social measures. However, one study in this category (Regev & Guttmann, 2005) found no effect of art therapy on the sample that was tested; namely, children with learning disorders. This study administered validated questionnaires in which the children reported their emotional state. It is possible that these research indices were not sufficiently sensitive to the changes that occurred during the 25-session period. Alternatively, it is also possible that it is difficult for children to look closely enough at the process they underwent during this period. The authors concluded: 'Obviously, there are countless possible relevant dependent variables, and some may prove more affected by art activity than others. It is for further research to examine these.' (p. 310).

Category 3: children with no specific diagnosed difficulty
The third category consists of children who received art therapy but were not diagnosed with any specific difficulty (see Table 3). Three studies on effectiveness have been conducted since 2000, one of which (Saunders & Saunders, 2000) without a control group (Level 3), one (Wallace et al., 2014) with a control group (Level 2), and one (Bazargan & Pakdaman, 2016) with a control group and random allocation into groups (Level 1). Two studies had a sample of both children and adolescents, and one only examined adolescents. One article dealt solely with female participants. The total sample sizes ranged from 30 to 94 clients in various experimental groups; some clients received individual therapy and others group therapy. The therapeutic process was very short, ranging from three to an average of 16 sessions.
The measurement indices that were selected for these studies included a variety of questionnaires measuring symptoms, psychological and social functioning, and aspects of the therapeutic relationship. All three articles reinforce the claim that art therapy may help children who are not diagnosed with specific difficulties but are faced with a variety of challenges in life, as reflected in the indices examined.

Category 4: children dealing with medical conditions
The fourth category consists of children suffering from various medical conditions (see Table 4). Only one study fell into this category (Beebe, Gelfand, & Bender, 2010) and addressed children coping with persistent asthma. The study was conducted with randomised participants in a trial group and a control group (Level 1), and covered seven weeks of group art therapy. The study implemented questionnaires that dealt with subjects such as quality of life and the emotional state of the participants, as well as the 'Person Picking an Apple from a Tree' (PPAT; Gantt & Tabone, 1998) art-based assessment tool. This article lends weight to the claim that art therapy may help children dealing with medical conditions, and specifically persistent asthma, as reflected in the indices examined.

Category 5: juvenile offenders
The fifth and final category consists of juvenile offenders who received art therapy (see Table 5). Only one study fell into this category (Hartz & Thick, 2005). It implemented two approaches to art therapyart psychotherapy and art therapy, but had no control group (Level 3). The research tools included questionnaires that assessed self-perception and self-esteem. This article supports the claim that art therapy may help juvenile offenders, as reflected in the indices examined.

Discussion
The purpose of this review was to assess whether art therapy is effective in a wide range of child-aged clients, as documented in quantitative studies published from 2000 to 2017. Since the review conducted by Reynolds et al. (2000), research on adults has made significant progress and there has been an increase in the number of high quality studies (Level 1) and the number of participants in such studies (Regev & Cohen-Yatziv, 2018). However, there is still a dearth of studies on children: we only located 13 studies conducted since 2000 that have examined the effectiveness of art therapy with children and met this article's inclusion criteria. Most had small samples and only five included randomised allocation to an experimental group and a control group (Level 1). The advantage of such RCT studies lies in the likelihood of fewer errors than in studies that do not include a control group, or studies that have a control group but lack random allocation to groups. The use of randomisation is the best way to ensure that differences in outcome measures between the experimental group and the control group result from the intervention itself and not from other factors, thereby decreasing biases.
Analogous to our conclusion in our review of the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018), the shortage of studies that relate to each population and the considerable discrepancies in the duration and nature of the therapeutic process, the proposed interventions and the indices make it difficult to draw conclusions or create a database from which a meaningful meta-analysis could be carried out. Thus, more research needs to be conducted on art therapy for children to examine and evaluate its effectiveness.
This raises many questions in that art therapy clearly emerges as pertinent and highly suitable for working with children; in many countries, children constitute the majority of the clients (Eaton et al., 2007). However, the challenge may be associated with the difficulty of evaluating children as research participants since they cannot always complete questionnaires, and it may also be connected to the onerous ethical considerations involved in research with children.
An important issue the arose is the necessity to define the profession, in order to evaluate therapeutic interventions. Despite major differences between cultures in relation to the status of the art therapy profession and the legislation in the field, in order to maintain standards we need to define, clarify and specify what art therapy is and what it is not; and specifically to clarify that this type of therapy must be composed of ongoing sessions and be conducted by a certified art therapist who meets the criteria Only intervention group (N = 13).
An approach that integrates art therapy and group process to target areas of need when working with child and adolescent victims of sexual abuse.
Trauma-focused group treatment using the combined modalities of AT and CBT.
One hour weekly for eight weeks.
Baseline, immediately after. Adolescents with Posttraumatic Stress Disorder Symptoms. Inclusion criteria: Ages 13-18, were able to sustain a school programme for 2 weeks running, and were expected to stay at the hospital for at least 16 weeks.
Groups consisted of 2-5 participants. One hour group sessions over 16 weeks.
The UCLA PTSD Reaction Index for DSM-IV, Child Version.
There was a significant effect of treatment over time for both groups across treatment conditions. There was a significant treatment by condition interaction showing that TF-ART was significantly more effective in reducing trauma symptoms from pre-treatment to post-treatment. Level 1 Group art therapy with sexually abused girls (Pretorius & Pfeifer, 2010) The Solomon fourgroup design was used to investigate the efficacy of the intervention.
Sexually abused girls aged 8-11 years. A group art therapy intervention designed by the authors aimed at reducing depression, anxiety, sexual trauma and low self-esteem. The programme was based on existentialhumanistic, Gestalt, client-centered and abuse focused principles.
The programme consisted of four themes with eight sessions.
The Trauma Symptom Checklist for Children (TSCC), The Human Figure Drawing (HFD).
The results indicated that the experimental groups improved significantly compared to the control groups with regard to anxiety and depression. Level 1 Art therapy with elements of logopedic therapy.
The Auditory Dysarthria Scale.
On the Auditory Dysarthria Scale, the largest improvements were noted in overall intelligibility; less improvement, though still statistically significant took place with respect to volume, tempo, and control of pauses. The least improvement was noted in the pronunciation of vowels and consonants (the most purely motor of the measured parameters). All the children also exhibited significant improvement in fluency of speech. Primary-school children with learning disorders.
The intervention programme included 25 weekly encounters, each of 45 min.

CSCS -The Piers -Harris
Children's Self-Concept Scale, IARQ -The Intellectual Achievement Responsibility Questionnaire, CS -The Children's Sense of Coherence Scale, LSDQ -The Loneliness and Social Dissatisfaction Questionnaire.
Children in the art-therapy group did not score better than those in any other group on any of the dependent variables. Level 1 Outcome-Based evaluation of a Social Skills Programme using Art Therapy and Group Therapy for children on the autism spectrum (Epp, 2008) Only intervention group -(N = 66) Children ages 6-12 with ASD.
Group of approximately six children of similar age and social communication ability.
One-hour group therapy sessions held once a week -October till Mai.
The Social Skills Rating System (SSRS)was constructed to screen and classify children suspected of having social behaviour problems.
Scores revealed a significant improvement in assertion scores, coupled with decreased internalising behaviours, hyperactivity scores, and problem behaviour scores in the students. Level 3 (Continued )   Group art therapy included specific art therapy tasks designed to encourage expression, discussion, and problem-solving in response to the emotional burden of chronic illness.
60-minute art therapy sessions once a week for 7 weeks.
Baseline, immediately after, and 6 months after the final art therapy.
The Pediatric Quality of Life (PedsQL) -Asthma Module (child and parent report), The Beck Youth Inventories -Second Edition, The Draw a Person Picking an Apple from a Tree evaluation -(FEATS).
Immediately after art therapy there was a reduction in parent-reported and childreported worry scores from the PedsQL questionnaires, a reduction in the anxiety score and an increase in the selfconcept score from the child-reported Beck Inventories, and improvements in the colour, logic, and details scores from the FEATS in the intervention group compared with the control group. Six months after completion of the therapy there was improved parentreported and child-reported worry and total scores from the PedsQL questionnaires, a lower anxiety score from the Beck Inventories, and higher colour and detail scores from the FEATS persisted remained in the intervention group. Level 1 Six core groups were randomly divided so that half of them received one type of art therapy and half received the other.
Ten 1-1/2-hour art therapy sessions were conducted with each group during a 12-week period.
The Self-Perception Profile for Adolescents (SPPA) + The Hartz Art Therapy Self-Esteem Questionnaire (Hartz AT-SEQ).
The art psychotherapy group showed a significant increase in the domains of close friendship and behavioural conduct whereas the art as therapy group did so in the domain of social acceptance. Level 3 defined for the profession (BAAT, 2018). This process could reduce the large number of articles conducted in the field by uncertified art therapists or in a single therapeutic session, as found in this review and previously documented on its companion on the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018). The first category presented here covered art therapy for children dealing with traumatic events. There is a growing body of research indicating that art therapy can benefit these children, and that art therapists play a significant role in helping them cope with a variety of traumatic experiences and can contribute to improving their quality of life and their ability to cope with a variety of emotional issues. Our review complements previous ones (Eaton et al., 2007;van Westrhenen & Fritz, 2014) which have shown that art therapy with children dealing with traumatic events is one of the most frequently studied topics in this field.
The second category covered art therapy with children with special educational needs and disabilities. In some countries, children with special needs are legally entitled to art therapy within the education system (Snir et al., 2018). The findings in most of these studies indicate an improvement in a variety of measures resulting from art therapy interventions. However, the differences between participants and assessment tools make it difficult to summarise and generalise the findings.
The third category covered art therapy with children with no specific diagnosis. The decision to examine clients in the general population reflects the unique nature of art therapy and its suitability for individuals who are dealing with a wide range of difficulties, or as mentioned in the definition of the profession by the British Association of Art Therapists: 'Art therapists work with children, young people, adults and the elderly. Clients may have a wide range of difficulties, disabilities or diagnoses. These include emotional, behavioural or mental health problems, learning or physical disabilities, life-limiting conditions, neurological conditions and physical illnesses.' (BAAT, 2018). In the reality of everyday life, there are many frameworks in which therapy is offered to a wide variety of children who are coping with an array of different difficulties and who remain undiagnosed. However, the vast differences between such clients make it difficult to find relevant indicators for evaluation, and as such there are considerable variations in the indices between studies.
The fourth category covered art therapy clients dealing with various medical conditions. It was striking that there was only one study on effectiveness, dealing with children coping with persistent asthma, suggesting that conclusions relating to this category should be made with caution. The findings indicated that art therapy interventions had a beneficial effect on this population, as shown by the indices measured. It is surprising that there are so few quantitative studies in the field. By comparison, in our review on the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018), six works (published in seven different articles) were found on clients who have cancer (Bar-Sela, Atid, Danos, Gabay, & Epelbaum, 2007;Monti et al., 2006;Monti et al., 2012;Öster et al., 2006;Öster, Magnusson, Thyme, Lindh, & Åström, 2007;Svensk et al., 2009;Thyme et al., 2009) and three others on clients coping with a variety of other medical conditions (Feldman, Betts, & Blausey, 2014;Sela et al., 2011;Sudres et al., 2013). Since the field of art therapy with clients dealing with various medical conditions is of vital significance and research conducted with adults clearly points to its contribution (Regev & Cohen-Yatziv, 2018), it is important to continue developing research in this field.
The fifth and final category covered art therapy with juvenile offenders. Only one study was found and indicated that art therapy interventions had a beneficial effect for this population, as expressed in the indices measured. This result extends the important work of David Gussak who has investigated the effectiveness of art therapy with adult prison inmates (Gussak, 2004(Gussak, , 2006(Gussak, , 2009a(Gussak, , 2009b and reported its beneficial effects. Nevertheless, conclusions concerning this category should be treated carefully since it is based on a single level 3 study.

Conclusion
This review is a companion to our review of the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018) and documents research conducted in recent years. Similar to the conclusions drawn in earlier reviews (for example McDonald & Drey, 2018), we only found a small number of quantitative studies that relate to the effectiveness of art therapy with children, despite the growing need for this type of therapy. This review did not include qualitative studies, which are very common and acceptable in this field. Applying diverse approaches and a variety of research methods as a dependent measure of the research purpose and question, can help expand knowledge and understanding in this interdisciplinary field (Kapitan, 2017).
The current review has several limitations. First, similar to other reviews in the field (for example: Regev & Cohen-Yatziv, 2018), due to the small number of studies, various levels of quantitative studies were included. This variability makes it difficult to generalise across findings, but not mentioning these studies would have led to a total sample of an even smaller number of studies. Second, in many studies several indices of varying types (questionnaires, drawings) were used and on some occasions only some of these indices led to demonstrable indications of the effectiveness of art therapy. This complex picture of the findings, which is well reflected in the tables, only authorises partial conclusions at this stage. Further research in the field should make it possible to examine changes in similar research indices and draw more far-reaching conclusions. In so doing, future researchers should examine these specific studies closely before conducting further research on the same types of populations. In addition, due to the small number of studies in this field, we needed to combine various subjects in certain cases, make decisions and create artificial categories based on our professional knowledge and judgment. Future reviews, including a broader scope of research, could employ a different, possibly more specific, categorisation.
A number of recommendations for further research directions in this field can be made. First, art therapy is a very broad field that can be implemented with diverse populations, many of whom have never been examined in terms of treatment effectiveness. An exploration and investigation in these uncharted territories could point to meaningful implications and broaden our knowledge of the field and its scope of effectiveness. Second, based on the conclusions derived from this review and our previous review on the effectiveness of art therapy with adult clients (Regev & Cohen-Yatziv, 2018), future studies are needed on larger sample groups that include art therapy sessions led by qualified art therapists over an extended period of time. This accumulated research could enable a meta-analysis based on significant and comparable findings from the field within approximately a decade. Third, to raise the level of research in our field, it is important for researchers to devote time and thought to their research plans and objectives. Large sample groups on their own are not sufficient. Researchers should also consider randomised control trials (Level 1), and ways to ensure that the experimenters, the participants, the data compilers and the judges/raters are blind to the purpose of the study and the division of the study groups. Finally, most of the studies presented in this review used reliable and valid research indices. We encourage researchers to select valid and reliable research indices that have been used and documented extensively and also to review and consider previous results of studies in art therapy with similar populations that have used these indices in their work.
This long and complex search for articles underscores the importance of encouraging much needed research in the field of art therapy with children. We believe that more studies will emerge in the next few years and hope that this review will contribute to encouraging future studies.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributors
Liat Cohen-Yatziv, The School of Creative Arts Therapies, University of Haifa, Israel Dafna Regev, The School of Creative Arts Therapies and the Emili Sagol Creative Arts Therapies Research Center, University of Haifa, Israel