Child Friendly Spaces: a Systematic Review of the Current Evidence Base on Outcomes and Impact

Ager et al.

Child friendly spaces: a systematic review of the current evidence base on outcomes and impact

Alastair Ager, Janna Metzler, MarisaVojta & Kevin Savage

Child friendly spaces are widely used in Introduction emergencies as a mechanism for protecting children International standards, currently being from risk, as a means of promoting children’s developed, de¢ne a child friendly space psychosocial wellbeing, and as a foundation for (CFS) programme as one that ‘supports strengthening capacities within communities for the resilience and well-being of children child protection. A systematic review of published and young people who have experienced and ‘grey’ literature identi¢ed 10 studies that disasters through community organised, met speci¢ed inclusion criteria. Each study was structured activities conducted in a safe, reviewed with respect to the potential protective, child friendly, and stimulating environment’ (Child ProtectionWorking Group (CPWG), promotive, and mobilising impacts of the inter- 2012). Programmes are typically hosted in vention. All 10 studies documented reports of a tent, or other temporary structure, and positive outcomes of child friendly spaces, particu- operate as part of a short to medium term larly with respect to psychosocial wellbeing. response (UNICEF, 2009). CFSs often However, major weaknesses in design constrain provide the opportunity for communities to the ability to robustly con¢rm change over time mobilise towards enhanced child protection (only three studies reported pre intervention base- and support capacities long past the onset lines) or attribute any such change to this inter- of disasters. vention (only two studies utilised a comparison Since its use in the 1999 Kosovo crisis, CFS with communities without child friendly spaces). programing to support the protection and Analysis suggests that: greater commitment to psychosocial wellbeing of children a¡ected documentation and measurement of outcomes and by situations of humanitarian crisis is wide- impactsisrequired;morestandardisedandrigorous spread (UNICEF, 2009). There is growing measurement of processes, outputs, outcomes and interest and adoption of CFSs as a prime impacts is necessary; evaluation designs need to intervention strategy, as evidenced by its more robustly address assessment of outcomes reference in a number of agency and inter- without intervention; there is a need to sustain agency documents guiding humanitarian engagement of children within the context of response (Kostelny & Wessells, 2008; evaluations; and long term follow-up is critical to Mad¢s, Martyris, & Triplehorn, 2010; Save establishingevidence driven interventions. the Children, 2008, 2009; Save the Children Sweden, 2010; UNICEF, 2009; World Vision Keywords: child friendly spaces, child International, 2006). protection, emergencies, evaluation, huma- In 2012 alone, Relief Web listed well over nitarian, outcome, psychosocial 100 programmes across the world utilising

133 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147

CFSs in emergency contexts (ReliefWeb, CFSs are seen as a key vehicle for mobilising 2013). These included: programmes for communities around the protection and Syrian refugees in Lebanon, Jordan and wellbeing of children, and strengthening Iraq; responses to the refugee and Internally community protection mechanisms (Global Displaced Persons (IDP) situations in South Protection Cluster et al., 2011). Sudan; and interventions for the Democratic The evidence base for the outcomes and Republic of the Congo refugees in Rwanda impact of CFSs is generally considered to and Uganda. In addition to these con£ict be limited. As e¡orts are made to develop related crises, CFSs were also utilised in the standards and international guidelines to wake of a number of natural disasters, support CFS work in emergencies, it is including: £oods in Assam, India; tropical important to develop and consolidate evi- storm Kai-Tak, Hong Kong and Philippines; dence regarding the protective, promotive and typhoons Washi and Bopha in the and mobilising e¡ects CFSs have onchildren Philippines. Agencies involved included: and youth. World Vision International, ACTED; INTERSOS; Lutheran World a global agency with a major commitment Federation; Mercy Corps; Plan; Save the to child protection in emergencies, and Children; SOS Children’sVillages; UNFPA; Columbia University, an institution with a UNICEF; and War Child. World Vision strong tradition of applied ¢eld research alone established CFSs in emergency res- in humanitarian contexts, have initiated ponses in the Philippines, India, Lebanon, a series of structured evaluations of CFS Uganda, DRC, Niger and South Sudan over interventions as part of a wider CPWG the course of 2012. agenda regarding CFS and related com- There are a number of factors that have munity based child protection support. contributed to the frequent adoption of a To ensure that these studies are fully CFS model in humanitarian emergencies. informed by existing knowledge of CFS out- These include: potential for rapid deploy- comes and impacts, a systematic review of ment; relatively low costs; and scalability the current literature was completed. and adaptability of activities to diverse contexts (UNICEF, 2009). The inherent Methodology £exibility of a CFS model, although From April to July of 2012, the authors originally intended for children aged 7 to undertook a systematic review of literature 13, potentially accommodates children of describing CFSs, or equivalent inter- all ages (Global Protection Cluster et al., ventions, within humanitarian contexts. 2011; UNICEF,2009). Inclusion criteria for the review were:1) the Guidance on CFSs generally suggests publication referenced CFSs or equivalent such interventions being of value with interventions within an emergency context; respect to three major objectives. First, CFSs 2) the publication provided data relevant are seen to serve as a protective mechanism, to outcomes and impacts of CFSs (either protecting children from abuse, exploitation baseline information and/or some assess- or violence. Second, CFSs are considered ment of outcomes); and 3) the publication as a means to provide psychosocial support was published within the last 15 years in to children, strengthening their emotional the English language. To supplement this wellbeing, social wellbeing, and/or skills review of published sources, we solicited and knowledge (Ager et al., 2011a). Third, ‘grey literature’ (unpublished agency reports

Table 1. Search terminology used in review by core theme Key search term Child Friendly Spaces (CFSs) Evaluation Humanitarian

Synonyms Safe spaces Outcome Emergencies Child centred spaces Impact Disasters Emergency spaces for children Con£ict Safe play areas War Child protection centres Refugee Psychosocial spaces Displaced Psychosocial intervention(s) Both British and American spelling variations were used.

and other documents) and reviewed them related to ‘Child Friendly Spaces’, ‘Evaluation’ within the same inclusion criteria. and ‘Humanitarian’ (see Table 1 for synonyms Table 1 summarises the search terminology of search terms used). These searches used to identify CFS studies, and Figure 1 identi¢ed a total of 7,225 items, with 5,220 details the selection process of papers duplicates, that represented a literature of through di¡erent stages of review, using 2,005 articles. these criteria. We identi¢ed relevant litera- Abstracts of all 2,005 articles were reviewed ture by searching structured bibliographic for relevance by the ¢rst author, which sources, including Medline, PubMed, Psy- identi¢ed 53 papers as potentially ful¢lling chINFO and Scopus, using the search terms inclusion criteria. Full versions of these

7225 studies 22 documents identified through provided by NGO keyword search contribution

22 documents 5220 duplicates 2005 studies selected selected for abstract excluded for abstract review review

1952 studies 53 studies selected 18 documents 4 documents excluded on basis of for full text review on selected for full text excluded on basis of abstract review the basis of abstract review abstract

3 studies and 7 NGO 11 NGO documents 50 studies excluded documents included excluded upon full upon full text review upon full text review text review

Figure 1: Overview of selection of papers during review process. NGO (nongovernmental organisation).

135 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147

papers were obtained, detailed review of Africa (Kostelny & Wessells, 2008; Desse- which led to three of these studies being mie, 2010), and one each in the Middle con¢rmed as meeting inclusion criteria. East (Save the Children, 2011) and Oceana To identify relevant ‘grey’ or unpublished and the Caribbean (Mad¢s et al., 2010). literature, over 60 NGOs active in the Of the remaining papers, one addressed use of CFS in emergency contexts were a CFS intervention in a Serbian refugee set- contacted by email through relevant huma- ting (Ispanovic-Radojkovic, 2003); another nitarian networks (including the Global targeted con£ict a¡ected communities CPWG, the Mental Health and Psychosocial in the occupied Palestinian territories (oPt) Support (MHPSS) Network, and the Child (Loughry et al., 2006); while the last Protection in Emergencies (CPiE) learning addressed South Sudanese returnees moving network. These networks specialise in the through the Kosti Way Station of North coordination of humanitarian practice Sudan (Gladwell, 2011). in the sectors of child protection, mental The majority of papers described CFS health and psychosocial support. Based on interventions for both children and adole- current and previous work in emergency scents, covering ages from four up to the settings worldwide, agencies were invited late teens (Gladwell, 2011; Loughry et al., to provide documents that reported on (a) 2006; Mad¢s et al., 2010; Save the Children, CFS or equivalent interventions (b) within 2011). For most interventions, children and humanitarian contexts that (c) included youth were separated into di¡erent activity data relevant to the working of CFS (either shifts according to age and developmental baseline information, or some assessment abilities. Three studies addressed interven- of outcomes). Twenty-two documents were tions with a narrower age span: Ispanovic- provided by this means. Documents were Radojkovic (2003) evaluated youth clubs reviewed with respect to the same inclusion for adolescents between the ages of 15 and criteria as used for published papers. 18, while Demessie (2010) and Kostelny & This resulted in the selection of a further Wessells (2008) evaluated CFSs for pre seven documents. The body of literature school aged children under six. Three that provides the basis for this review thus papers did not indicate the age range of comprises a total of ten documents, three targeted participants (Arus, 2008; Sabina, identi¢ed through formal bibliographic 2012; TANGO International, 2009). search and seven identi¢ed through agency consultation. Evaluation design adopted Only three papers reported both baseline Findings and follow-up data related to CFS Characteristics of interventions studied (Ispanovic-Radojkovic, 2003; Loughry, Of the ten papers reviewed (see Table 2), 2006; Mad¢s et al., 2010). Of these three, six addressed CFSs established in con£ict only one paper provided information a¡ected areas, while four examined CFS related to a comparison group of children interventions taking place in areas a¡ected at both baseline and follow-up times by natural disasters. Seven of the ten (Loughry et al., 2006). One paper reported papers addressed work with IDP commu- no baseline data, but assessed impact nities; three in Asia (Arus, 2008; Sabina, by comparing across groups who had 2012; TANGO International, 2009), two in received, or not received, the intervention

136 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Ager et al. in critical and analytical thinking, increased knowledge and skills, sense of happiness, loss of traumas, pride, creativity, self-con¢dence, and improved peer relations. CCPC: No clear data related to e¡ectiveness of trainings and workshops for community leaders, communities, local government members, etc. supervision for younger children (age2^6).PSS:Inappropriately age-targeted activities observed. Parent and animator responsibility ill de¢ned. CCPC: Community- based systems of protection, PTAs, and awareness raising in the community less emphasised in CFS. improvements in psychosocial wellbeing. CCPC: Limitations clearly stated, calling for improvement in parent and community participation in CFS PSS: Self-reported improvements PC: General lack of adequate PSS: Anecdotal evidence suggests and key informant interviews with programme sta¡, community facilitators, etc. question FGD with animators, supervisors, community- based child protection committees, and PTA leaders. lasting10 days. Analysis of training records and CFS monthly reports. KII and FGD with 31children (aged 8^20 years),7 mothers and 10 educators. FGD with bene¢ciaries Unstructured, open-ended Mixed methods evaluation per week lasting for hours1.5 based on international guidelines on international guidelines international guidelines CFS intervention 3-4 days CFS intervention based CFS intervention based on Bantul District post tsunami Indonesia and Alsalam IDP camps in North Darfur, Sudan North Sudan Table2.Summaryofstudiesidenti¢ed StudyArus et al. (2008) 4 sub-districts of Location Intervention Evaluation method Major ¢ndings Demessie (2010) Alfasher Tawilla, Gladwell (2011) KostiWay Station,

137 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147 self-respect and improvements in peer relations. Signi¢cant decrease in psychological problems observed, particularly withdrawal and anxiety-depression in male refugees and withdrawal and social problems in female refugees. and sexual exploitation and heightened sense of safety in children relative to comparison group. PSS: Signi¢cant improvements in psychosocial wellbeing relative to comparison group. PSS: Anecdotal reports of increased PC: Decreased incidence of rape administered students to 1,106 between the ages of 15 and 18 pre and post intervention (follow-up period: academic school year) caregivers, community members, single mothers, widows, and camp leaders, CCS sta¡, Child Activity Leaders and ChildWell Being Committee members. Semi- structured questionnaire administered to 294 households. Analysis of comparison groups performed. Semi-structured questionnaire Eight FGDs held with 92 elderly involving 90-minute group meetings 1^2 times per week after school for 6 months 4.5 hours per day for 5daysaweek ot ClubYouth programme CFS intervention lasting in Belgrade, Serbia IDP camps in Gulu, Uganda Boarding high schools Unyama and Paicho Radojkovic (2003) Wessells (2008) StudyIspanovic- Location Intervention Evaluation method Major ¢ndings Kostelny & Ta bl e 2. ( C o n t i n u e d )

138 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Ager et al. scores, externalising problem scores and internalising problem scores following intervention when compared with children in comparison group. existing threats noted. PSS: Positive changes in psychosocial functioning observed. for improvements in mood and general changes following intervention. Anecdotal support for improved relationships between children and animators. PSS: Lower CBCL total problem PC: Increased knowledge of PSS: Parental anecdotal support administered to 400 parents and children between the ages of 6 and 17receiving the intervention and 100 parents and children not receiving the intervention pre and post (1-year) intervention observation rubric completed for 10 children in Haiti and 10 children in the Solomon Islands pre and post (6 week) intervention Interviews with supervisors, leaders, business community, and teachers. Desk review conducted. Structured questionnaire Parent questionnaire and FGDs with parents and animators. involving daily after school activities and week long holiday camps at local recreational centres intervention based on international CFS guidelines international guidelines Child-focused intervention Emergency safe spaces CFS intervention based on We s t B a n kGaza, a oPt n d in post hurricane Noel Haiti and post tsunami Solomon Islands of Debhata, Tala, and SatkhiraSadar in Southern Bangladesh Communities from IDP children living (2006) (2010) StudyLoughry et al. Location Intervention Evaluation method Major ¢ndings Mad¢s et al. Sabina (2012) Flood a¡ected areas

139 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147 ¼ Parent and ¼ Focus Group Discussion. IDP ¼ secure and safe area. PSS: Informal reports of bene¢t, but some activities viewed as culturally inappropriate for girls. CCPC: Committee members continuing to engage on CP after closure of CFS psychosocial outcomes in children. CCPC: Ayeyarwady Division were much more likely to contribute to CFS than their counterparts inYangon Division PC: Children reported CFS as a PSS: Anecdotal support for improved Psychosocial Support (Objective 2). PTA ¼ Child Centred Space. FGD ¼ aged 6 to 18, and 2 participatory feedback sessions over the course of one week enumerators (unknown structure). FGDs with parents and children. 16 KII,12 FGDs with children Quantitative data collected by Protection of Children (Objective 1). PSS ¼ on international guidelines on international guidelines CFS intervention based CFS intervention based Community Child Protection Capacities (Objective 3). CCS ¼ Key Informant Interviews. PC ¼ Strengths and Di⁄culties Questionnaire. ¼ populations in Amran and Haradh, Ye m e n Ayeyarwady IDP populations in Myanmar post Cyclone Nargis IDP and host Ya n g o n a n d Child Behaviour Checklist. CCPC ¼ Children (2011) International (2009) Internally Displaced Persons. Association. KII Teacher SDQ StudySave the Location InterventionCBCL Evaluation method Major ¢ndings Ta n g o Ta b l e 2. ( C o n t i n u e d )

(Kostelny & Wessells, 2008).The remaining psychosocial wellbeing; and strengthening six papers documented only post interven- of community child protection capacities. tion outcome data, or data collected towards The building and strengthening of a the close of services for those receiving the protective environment for children vulner- interventions described (Arus, 2008; Desse- able to abuse, exploitation and/or violence mie, 2010; Gladwell, 2011; Sabina, 2012; Save is paramount to e¡ective CFS programing. the Children, 2011; TANGO International, Improvements in protection outcomes, 2009). such as increased sense of safety, and Structured or semi-structured question- decrease in sexual exploitation and rape, naires were developed for use among were documented in ¢ve studies (Gladwell, children, youth, and parents in four 2011; Kostelny & Wessells, 2008; Mad¢s studies, three of which were adapted from et al., 2010; Sabina, 2012; Save the Children, established tools used in other settings 2011). Other studies noted the decrease (Kostelny & Wessells, 2008; Loughry et al., in physical injuries from the start of a 2006; Mad¢s et al., 2010). Unstructured, CFS intervention (Dessemie,2010; Gladwell, focus group discussions were conducted in 2011; Kostelny & Wessells, 2008). seven studies, with various stakeholders All ten studies reported positive psycho- including children and youth, parents, social outcomes for children and/or the animators, CFS programme sta¡, com- wider community. Eight studies indicated munity based child protection committees, increases in social and emotional wellbeing PTA leaders, community members and/or of children, although in only four studies leaders, and educators (Arus,2008; Kostelny was this documented by di¡erences between & Wessells, 2008; Sabina, 2012; Save the baseline and follow-up (rather than through Children, 2011; TANGO International, retrospective judgments). In only one of 2009). Key informant interviews with these did the design allow such change to programme sta¡, child protection actors in be reliably attributable to CFS. There was other agencies, parents, teachers, Child generally little documentation regarding Protection Committee (CPC) members, sex di¡erences in social and emotional and children were conducted in four of the wellbeing of children. However, one study nine papers reviewed (Arus, 2008; Gladwell, reported girls having more di⁄culty 2011; Sabina, 2012; Save the Children, 2011). accessing CFS services, due to culturally Additional evaluation methods used include inappropriate activities, and the layout of the collection andanalysis of training records the CFS compound acting as a deterrent to (Gladwell, 2011), monthly ¢eld reports engagement (Sabina, 2012). (Gladwell, 2011;Sabina, 2012),and participa- The in£uence of CFS on facilitating tory feedback sessions (Save the Children, community capacity for the protection and 2011). support of children was seldom reported on in detail. Increased knowledge and awareness of child protection concerns and avail- Findings regarding outcome and impact able services was noted in three studies Outcomes and impact of CFS programm- (Gladwell, 2011; Mad¢s, 2010; Sabina, 2012). ing are discussed in relation to the three There is evidence to support community objectives described in the introduc- engagement and/or involvement in CFS tion: protection from risk; promotion of activities, such as cleaning, cooking and

141 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147

paying levies (TANGO International, 2009). More standardised and rigorous measurement of The enhancement of mechanisms to receive processes, outputs, outcomes and impacts is required and respond to reports of abuse, neglect, Studies reviewed suggested that signi¢cant exploitation or violence against children, development is required in both the such as child protection committees, referral standardisation and rigor of measurement. systems, and PTAs, were considered in three Regarding the speci¢cation of CFS itself, studies (Arus, 2008; Gladwell, 2011; Save it is clear that the composition of, and the Children, 2011), with a lack of robust emphasis on, speci¢c activities may di¡er documentation of impact generally reported. dramatically organisation to organisation, leading to confusion over programme Discussion goals and objectives among programme Greater commitment to documenting outcomes and staf and community members (Gladwell, impacts is required 2011; UNICEF, 2009). Without an agreed Given the widespread use of CFS as an set of activities and ‘shared vision’ of what intervention strategy to address children’s constitutes a CFS, it is di⁄cult to de¢ne needs in humanitarian emergencies, the and measure a set of ‘standardised’ outputs. review indicates a remarkably small evi- Output indicators, such as the number of dence base. That only ten studies could children attending the programme, are a be identi¢ed suggests a failure, either to mainstay of psychosocial interventions commit to conducting evaluations indicat- (Arus et al., 2008; Gladwell, 2011; Sabina, ing impacts in the lives of children and 2012; Save the Children, 2011). They are their families, or failure to disseminate such relatively easy to measure, often including evaluations to the broader humanitarian some measure of quality of care, and yet community, or both. Over 60 nongovern- appear far from standardised among mental organisations were contacted by practitioners (Ager et al., 2011a; Dessemie, way of three inter-agency working groups 2010; Mad¢ s et al., 2010). in this search for unpublished or agency While an improvement in documentation speci¢c documents related to CFS outcomes and measurement of processes and outputs and impacts. Only 22 documents were is important, arguably the most critical returned, half of which were submitted by a requirement is an appropriate focus on single agency. relevant outcomes and impacts. The collec- This lack of evaluation may result from tion of output data alone cannot reasonably limitations relatedto sta¡ capacity and other validate programme impact or e¡ectiveness. realities of the ¢eld (Mad¢s et al.,2010).Lack Focus group discussions and self-reports of expertise in monitoring and evaluation can contribute to the documentation of (M&E) methods, including indicator devel- CFS outcomes, providing useful insight opment in the midst of a humanitarian crisis, into local perceptions and encouraging may often result in the low prioritisation of participation throughout the evaluation baseline data collection, a key foundation (Kostelny & Wessells, 2008). However, self- for most robust evaluation designs. Training reports are limited in their ability to relay and additional M&E support needs to be accurate information on wellbeing, as they made available to programme sta¡ in order rely on the individual’s ability to remit to encourage robust M&E designs in the sensitive information, usually related to feel- future (Ager et al., 2011a). ings or attitudes (Duncan & Arntson, 2004).

Mixed method approaches are thus required demonstrate ‘added value’. To demonstrate for a more robust measurement of CFS out- positive change is insu⁄cient, if there is comes and impacts, and are well represented evidence of such positive change being amongst the stronger papers reviewed secured without focused programmatic (Kostelny & Wessells, 2008; Loughry et al., interventions. 2006; Mad¢s et al., 2010). As part of a There is a range of methods available to Child Centred Spaces (CCS) initiative in address this (Ager et al., 2011a). Comparison Northern Uganda, Kostelny & Wessells groups provide an opportunity to view the (2008), for example, utilised locally derived improvement of children’s wellbeing pre indicators of child wellbeing identi¢ed and post intervention, relatively indepen- through focus group discussions with care- dent of outside factors. Examining the coun- givers and programme sta¡. This consensus ter factual [] provides baseline information driven approach allowed for a culturally related to positive outcomes not attributable relevant interpretation and adaptation of to the intervention, while still collecting an established ‘western’ tool; the Strengths valuable information relative to programme and Di⁄culties Questionnaire (Kostelny & impact. Outcome mapping1 provides an Wessells, 2008). This participatory feedback alternative approach to assessing pro- loop also helped to inform the analysis of gramme performance based on changes in CFS impact on the social and emotional factors, such as ‘behaviours, relationships, actions wellbeing of children. Loughry et al. (2006) or activities of the people, groups, and organisations used qualitative interviews to comple- with whom a development programme works ment and elaborate their analysis of child directly’ (Overseas Development Institute wellbeing based on the Child Behaviour (ODI), 2012). Measuring these ‘factors’ Checklist. provides a solid basis to measure programme change, and ultimate e¡ects, on bene- Evaluation designs need to more robustly address ¢ciaries ^ particularly when other agencies assessment of outcomes without intervention are working in similar programme areas, It is not only that measures need to be among the same population. more rigorous, but also that the evaluation designs within which they are deployed need to be strengthened. In particular, There is a need to sustain engagement of children attributing positive outcomes requires within the context of evaluations evaluation approaches that allow some esti- Such robust evaluation designs should not mation of likely outcomes without a CFS be seen as a basis to exclude the active intervention. This is particularly important participation of children in the development given the acknowledgement of child and of measures, and the implementation of community resilience in contexts of humani- evaluation studies. Rather prioritising tarian emergencies (Reed et al., 2008).With participation strengthens the robustness studies documenting the recovery of chil- of an evaluation. Acknowledging children, dren following humanitarian emergencies, youth and community members as active through individual and community partners throughout the design, monitoring e¡orts and without programmatic support, and evaluation process is essential, and (Ager et al., 2011b; Ager et al., 2010) it is should be made explicit early on in the important for CFS (or any intervention) to process.

143 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147

Increasingly, participatory methods (prim- Long term follow-up is critical to establishing arily focus group discussions) are being evidence driven interventions used as a core evaluation technique in There is emerging recognition that CFSs emergency settings (Dessemie, 2010; Sabina, have the potential to lay the groundwork 2012; Save the Children, 2011; TANGO for post disaster, formal educational systems, International,2009).Focus group discussions as well as link in and support indigenous potentially provide an excellent way to systems of protection. Unfortunately, there raise awareness among the community, is little documentation regarding the long encourageparticipantstovoicetheiropinions lasting e¡ects of CFS programming follow- and provide feedback related to the pro- ing the close of services. Of the ten studies gramme. The participation of caregivers, reviewed, only three provide pre and post community members and children them- intervention data, none of which docu- selvesinfocusgroupsprovidesanopportunity ment e¡ects beyond one year after the to engage, and build, lasting relationships programme’s close. Establishing evidence critical to long term sustainability of driven interventions requires long term systemsofeducationandprotection.However, follow-up directed towards impacts on the focus groups may neither constitute a repre- wellbeing of children and youth, as well as sentative sample of programme bene¢ciaries, at understanding these community based nor an e¡ective means of capturing emotions systems of protection and support. or internal processes. Other participatory Longitudinal studies would explore the methods, such as interactive games and role- nature and casual pathway of CFSs as the play, have been applied in South Sudan to interventionbridges to more sustainable out- explore children’s attitudes, attendance rates lets.These studies may be costly and rigorous and major problems in the KostiWay Station in nature, but create the potential to assess (Gladwell, 2011). Demassie (2010) used lasting change. Complications arise from dancing, picture drawing and storytelling interpreting causal links between objectives, (among several other techniques) to encou- and their subsequent impacts, in the midst rage a participatory evaluation process with of other humanitarian programming that children. However promising these inno- may a¡ect the wellbeing of children and vative approaches may be, methods used youth (Mad¢s et al., 2010; Loughry et al., for analysis of data gained through such acti- 2006). With proper planning and robust vities are typically not reported in su⁄cient design, active learning can commence, detail to allow replication. even in the midst of anhumanitarian crisis. Mad¢s et al. (2010) acknowledges children as more than ‘passive recipients of services’ Limitations suggesting that programmes should make a This review is an appraisal of evidence better e¡ort to engage children, emphasise related to outcomes and impacts of CFSs in their role andtheir‘capacitytoprotectthemselves’ humanitarian settings, through a structured (p. 857). Genuine participation goes beyond document review process. One major a token engagement and actively works constraint concerns the comprehensiveness with children and youth, as well as the of the search process and inclusion criteria community, in designing e¡ective monitor- used. While the systematic structure of ing and evaluation strategies that are both this review is likely to identify most, if not relevant and respectful. all, of the published corpus of literature,

restricting the search to English language earlier version of this paper appeared as a joint articles within the last15years may in£uence technical report of World Vision and Columbia the resulting number of studies identi¢ed. University. Furthermore, two-thirds of the documents gaining entry to the review were collected through inter-agency submissions. Inclusion References criteria for the submission of unpublished Ager, A., Ager, W., Stavrou, V. & Boothby, N. documents were quite speci¢c, and in (2011a). Inter-Agency Guide to the Evaluation of line with the criteria for published review. Psychosocial programing in Emergencies.NewYork: However, only a little over one-fourth of UNICEF. the total inter-agency submissions were deemed appropriate forinclusion. Ager, A., Akesson, B., Stark, L., Flouri, E., Okot, B., McCollister, F. & Boothby, N. (2011b). Conclusions and The impact of the school-based Psychosocial Recommendations Structured Activities (PSSA) program on con£ict-a¡ected children in Northern The evidence base for the outcomes and Uganda. Journal of Child Psychology and impact of CFSs is clearly limited. Out of Psychiatry, 52(11), 1124 -1133. the small number of studies identi¢ed, few presented well-designed and implemented Ager, A., Stark, L., Olsen, J., Wessells, M. & evaluations of CFSs in emergency settings. Boothby, N. (2010). Sealing the Past, Facing No doubt logistical constraints and sta¡ the Future: An evaluation of a program to capacity in£uence prioritisation and imple- Support the Reintergration of Girls and mentation of rigorous monitoring and Young Women Formerly Associated with evaluation techniques in the ¢eld. However, Armed Groups and Forces in Sierra Leone. as the international community continues Girlhood Studies, 3(1), 70 - 93. to support CFS work in emergencies, it is important to consolidate evidence as well Arus, Y., Pakpahan, D., Marpinjun, S. & as support evidence-based interventions Bramuntyo, D. (2008). Final Evaluation Report regarding the protective and restorative of Yogyakarta Earthquake Response and Recovery e¡ects CFSs have on children and youth. program of Plan Indonesia. Circle Indonesia: The development of inter-agency guidance Yogyakarta programme Unit. on minimum standards for CFS (CPWG, 2012) is to be welcomed, but such guidance Child ProtectionWorking Group (CPWG).(2012). needs to be informed by evidence of Minimum Standards for Child Protection in impact of interventions conforming to such Humanitarian Response. Geneva: CPWG. standards.

Dessemie, T. (2010). North Darfur program Acknowledgements Child Friendly Spaces: Assessment Report. This study was enabled by ﬁnancial support Stockholm: Save the Children Sweden. from Aktion Deutschland Hilft, World Vision Australia, World Vision Germany and World Vision UK. The views expressed in the paper Duncan, J., & Arntson, L. (2004). Children In are those of the authors alone, and do not Crisis: Good Practices In Evaluating represent the views of these organisations. An Psychosocial programing: The International

145 Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited. Child friendly spaces: a systematic review of the current evidence base on outcomes and impact Intervention 2013, Volume 11, Number 2, Page 133 - 147

Psychosocial Evaluation Committee and Save Sabina, N. (2012). Report of Emergency Response the Children Federation, Inc.Westport: SCF. Evalutation and Lessons Learned of Response in Satkhira Flood 2011. Dhaka, Bangladesh: Gladwell, C. (2011). Kosti Child Friendly Space: Sunderban Sub cluster on Child Protection in Evaluation Report: War Child Holland, Emergencies: Dhaka, Bangladesh. Sudan. Save the Children (2008). Child Friendly Spaces Global Protection Cluster, G. E. C., INEE, & in Emergencies: A Handbook for Save the IASC (2011). Guidelines for child friendly Children Sta¡. Retrieved from http://toolkit. spaces in emergencies. ineesite.org/toolkit/INEEcms/uploads/1064/ Child_Friendly_Spaces_in_Emergencies_EN. Goodman, R. (1997). The Strengths and PDF Di⁄culties Questionnaire: A Research Note. Journal of Child Psychology and Psychiatry, 38, Save the Children (2009). Child Friendly Spaces 581-586. Facilitator Training Manual. Retrieved from http://resourcecentre.savethechildren.se/con- Ispanovic-Radojkovic,V.(2003).Youth clubs: Psy- tent/library/documents/child-friendly-spaces- chosocial intervention with young refugees. facilitator-training-manual Intervention, 1(3), 38-44. SavetheChildren (2011).Community Based Child Protection Emergency Project: Internal Evalu- Kostelny,K.&Wessells,M.(2008).Theprotection ation Report. London: Save the Children. and psychosocial well-being of young children following armed con£ict: outcome research Save the Children Sweden (2010). Child Friendly on Child-Centered Spaces in Northern Spaces Handbook forAnimators (Volunteers). Uganda. Journal of Developmental Processes, 3(2), North Darfur program. Stockholm: Save the 13-25. Children Sweden.

Loughry, M., Ager, A., Flouri, E., Khamis,V., TANGO International (2009). Cyclone Nargis Afana, A. H. & Qouta, S. (2006).The impact Emergency Response (CNER) programme: of structured activities among Palestinian End of program (EOP) Evaluation Report: children in a time of con£ict. Journal of Child WorldVision. Psychology and Psychiatry, 47(12), 1211-1218. UNICEF (2009). A practical guide for developing Mad¢s, J., Martyris, D. & Triplehorn, C. (2010). child friendly spaces. NewYork: UNICEF. Emergency Safe Spaces in Haiti and the Solomon Islands. Disasters, 34(3),845-864. World Vision International (2006). Children in Emergencies Manual Chapter 10: Child Friendly Overseas Development Institute (ODI) (2012). Spaces in Emergency Situations. Geneva:WVI. Outcome Mapping Retrieved 11 June 2012, from http://www.odi.org.uk/rapid/tools/tool kits/Communication/Outcome_mapping.html 1 Outcome Mapping is a tool developed by ReliefWeb. (2013).Updates. Retrieved 27 February the International Development Research Centre, 2013from www.reliefweb.int Canada.

Alastair Ager is Professor of Clinical Population & Family Health with the Program on Forced Migration & Health at the Mailman School of Public Health, Columbia University. Janna Metzler is a graduate of the Program on Forced Migration & Health working as a research coordinator for the Evaluating Child Friendly Spaces in Emergencies program. MarisaVojta is Programme O⁄cer for Child Rights and Well-Being withWorldVision International in Geneva. Kevin Savage is Research Co-ordinator, Humanitarian and Emergency A¡airs, World Vision International, also based in Geneva. email: [email protected]

New manual available: Creating a safe school environment How can teachers develop a relationship with traumatised parents? How do you recognise psychological problems occurring in a school child? What are the most constructive methods for working with children a¡ected by collective violence? In this recently published manual, the War Trauma Foundation (WTF) shares its experiences of psychosocial programmes for schools in con£ict areas. Since 2004, theWTF has collaborated in providing psychosocial aid programmes to schools, with local organisations. WTF has been instrumental in training local aid providers, teachers and helpers in the Northern Caucasus, the Palestinian territories and Kosovo, aiming to enable them to recognise psychosocial problems and counsel children in a professional manner. The manualprovides information on how to set up apsychosocial school programme, what problems may be encountered, and which people may be involved in the process: local authorities, trainers, teachers, parents, children, as well as how to approach this variety of situations.The manual describes the working methods used, as well as the role of the local partner organisations, their approach to programmes and collaboration with the WTF. Furthermore, it provides a theoretical substantiation of the WTF approach. The manual also contains a number of methods for organisation and capacity development. Additionally,a large number of methodicaltips are included; for example, how children get throughto his or her parents when they are pre-occupied with the con£ict situation around them.The manual also contains inspiring examples of works, such as school children who wrote a letter to their own feelings during school time. It is a way to say goodbye to bad experiences and to let go. Discussing what certain sounds from nature symbolise ^ such as the sound of birds, rain, wind etc. ^ is another good way to induce children to speak about their feelings. The manual will also be translated in Russian. Youcan either order or download the manual at: www.wartrauma.nl