One often sees children playing ‘make believe’ and taking on the role of a schoolteacher, mother or doctor. Children generally have experience of taking on the role of another person in a different situation from a young age. Literature suggests that role-play is a natural method adopted by children to learn, as all children engage in some form of socio-dramatic play (Goldstein & Cisar, 1992; McSharry & Jones, 2000). In the field of speech-language pathology, therapists seek out evidence-based methods of learning to facilitate language development in children who require such intervention. Role-play holds potential as an effective method of learning for children (Clarke & Wales, 2005; Greenwood, Horton & Utley, 2002; Killen, 2006; Mason, 2006; Purvis, 2008). Its effectiveness as an approach to targeting pragmatic skills in children with language learning disability (LLD) was investigated in this research study.

The rationale for this study stems from personal clinical experience and observation, where it was noted that children in a special needs classroom were more involved in the therapy session and more easily retained new vocabulary when role-play was used. This exemplified literature regarding learners requiring more explicit intervention that supports generalisation and provides immediate feedback (Greenwood et al., 2002). Investigation into role-play as a learning strategy dates back at least 30 years (Ladousse, 1987; Van Ments, 1983), where the use of role-play was found to be effective in the education context. A recent change in approach to teaching and learning strategies has seen a rise in focus given to constructivism and active learning. These concepts are based on the tenet that effective learning occurs when the learner is actively involved in the construction of knowledge, as opposed to receiving knowledge from a third party (Brady, 2004). Various studies in the field of education advocate for the use of role-play as an active learning strategy (Brady & Skully, 2005; Clarke & Wales, 2005; Killen, 2006; Yehuda, 2006). A problem often encountered by speech-language therapists is that of a lack of generalisation of therapy aims to contexts outside the therapy environment. Role-play allows the therapy context to closely approximate natural interactions, and therefore results in more functional outcomes and increased generalisation (Killen, 2006). Even though role-play is used in certain areas of speech-language pathology, there is sparse literature documenting its method of implementation and effectiveness.

The experimental group received intervention during Phase 2, with the pre- and post-intervention assessments occurring in Phases 1 and 3 (see Figure 1). Results from Phases 1 and 3 were analysed and presented in Tables 1 and 2 below.

With regard to the control group, results from Phases 1 and 3 were analysed first (Table 2), and thereafter results from Phases 3 and 5 were analysed (Table 3). The control group received no intervention during Phase 2 and received intervention excluding role-play (the session plan included only the introduction, narrative, discussion and reflection) in Phase 4.

Session records for each participant were reviewed and summarised in the table below (Table 4). Data from session records were used to supplement and support quantitative data findings.

Results from Phases 1 and 3 revealed that the average increase in the mean of the experimental group for both assessment measures was greater than that of the control group. The CELF-4 PP and DCT scores of the experimental group increased by an average of 11 and 3.5, respectively (Table 1), while the control group CELF-4 PP and DCT scores increased by an average of 0.5 and 0, respectively (Table 2). Improvements in the experimental group post-intervention were further supported by participant-specific data from session records (Table 4).

It was found that improvements were noted post-intervention for both requesting for clarification and stylistic variation. Requesting for clarification was targeted for the first six group sessions and stylistic variation was targeted during the second six group sessions. Improvements in both these skills were more apparent in performance on the DCT, which assessed them directly. It was found that participants were able to grasp and apply the concept of requesting for clarification more easily than stylistic variation. Even though participants were already familiar with the facilitator and the components of the session, they took longer to independently identify pragmatically appropriate and inappropriate behaviour with regard to stylistic variation. This may be because stylistic variation is context specific, and is therefore more cognitively and linguistically demanding. It was also found that generalisation occurred to untrained skills as well, which was more apparent in participants who were actively involved in group sessions (Participants 2 and 3).

Phases 3–5 of the study involved the control group receiving intervention, while the experimental group received no intervention. However, the control group received the intervention without the role-play. The purpose of this was to allow the researcher to compare the effects of the intervention with and without the role-play component, thus establishing if it is in fact the use of role-play that is effective. Results indicated that the experimental group presented with a higher average increase in scores on the CELF-4 PP post-intervention (Mean increase: 11), as compared with the control group (Mean increase: 0.25). However, the control group presented with a higher average increase in score on the DCT assessment compared with the experimental group. The low increase in DCT scores of the experimental group appears to be as a result of two of the participants already achieving a high score on the DCT assessment pre-intervention. Analysis of session records revealed that the control group did make progress in therapy without the role-play component (Table 4). This progress appears to have reflected in their post-intervention DCT assessment. The progress, however, did not reflect in the post-intervention pragmatic profile assessment. The reason for this could be that progress was made in the therapy context, but did not generalise, and was therefore not observed when completing the profile during classroom and break time observation.

Ethical clearance was obtained from the University of KwaZulu-Natal, Humanities and Social Sciences Research Ethics Committee (Reference number: HSS/0334/014M). Informed consent was obtained, as learners became participants in the study only if they provided verbal consent and their parent or caregiver provided written consent. Right to privacy or confidentiality was ensured by assigning a pseudonym to each participant for reference throughout the study and by not revealing any information about the participants and organisations involved. The raw data collected from the study are stored in password-protected electronic files and will be kept for at least 5 years. Only the researcher and supervisors have access to these data.

This study aimed to determine the effectiveness of role-play as a therapy approach to target stylistic variation and requesting for clarification in learners with LLD. It was found that there is limited research into effective methods of addressing pragmatic difficulties of learners with LLD. The need for such research is evident in the fact that learners with LLD typically present with difficulties in social communication (Funderburk et al., 2009; Hallahan & Kauffman, 2003), which negatively affects their social relationships, inclusion and quality of life (Diken, 2014). Results from both quantitative and qualitative data revealed that improvements in stylistic variation and requesting for clarification were observed post role-play intervention in the experimental group, with minimal changes in the control group.

Role-play as a therapy approach targeting pragmatic skills (stylistic variation and requesting for clarification) in learners with LLD was found to have a number of benefits that supported its effectiveness. These included that participants displayed increased interest and involvement when role-play was used; participants reported to enjoy the ‘acting’ and would enthusiastically decide who should play which role when it came to the role-play component of the session. During intervention with the control group (which excluded role-play), even learners who participated well began to lose interest and concentration before the end of the session. Research shows that if learners are not actively involved in the process of knowledge acquisition, they are less likely to make the necessary connections that make learning meaningful (Cuthrell & Yates, 2007). Role-play also allowed for peer learning to take place. It was found that stronger participants supported weaker participants, by offering prompts, modelling and giving examples and suggestions. This was noted during the role-play and reflection components of the group session. Participants responded well to support from their fellow learners and generally responded to the prompt or suggestion. Lastly, it was found that skills learnt appeared to generalise to outside the therapy context and were maintained after a period of 6 weeks of no intervention. Qualitative analysis of the CELF-4 PP (Semel et al., 2003) pre and post-intervention revealed that generalisation occurred to untrained skills as well as target skills. Role-play creates a ‘real-life’ type context for the learner (Killen, 2006; Van Ments, 1999), and practising a skill in realistic contexts increases the likelihood of generalisation of the target skill (Stewart, Carr & LeBlanc, 2007). Parent and teacher input during the pre- and post-assessment process, in the form of questionnaires or rating scales, is recommended to validate findings in future studies (see Gerber et al., 2012).

Limitations of the approach were also identified. During role-play intervention, it was found that the responsibility of creating an environment that supports active learning relies on the facilitator. The facilitator needs to maintain a role that is supporting and flexible (Killen, 2006; McDaniel, 2000), while ensuring support is graded according to the needs of each participant. The effectiveness of the intervention is therefore dependent on the facilitators’ skill in supporting child-centred learning. Role-play intervention can also be time consuming to plan and implement, and it relies on learner participation.

It can be concluded that all of the components in the session plan in role-play-based therapy were found to be necessary and beneficial for effective implementation of the intervention. However, further research is needed to fine-tune each component, in order to achieve the best outcomes. The recommendations for the implementation of role-play as a therapy approach derived from this study has been consolidated and presented in Figure 2 below. The figure presents the steps recommended in planning and implementation of a session, as well as the facilitator’s role in the process.

Role-play is an active learning strategy that closely mimics natural interactions, and therefore results in improved generalisation of skills (Killen, 2006). The combined use of positivist and interpretivist paradigms allowed the researcher to logically analyse the research data, while still considering the holistic view through observation and interpretation (Coolican, 2004; Weaver & Olsen, 2006). This was achieved through the use of an embedded mixed methods design. Qualitative data were used to support quantitative data, in order to view a complete picture and achieve data triangulation. The method of implementing role-play intervention was sourced from education literature and was found to be effective in its use as an intervention approach in speech-language pathology. It can be concluded that in this study role-play was found to be an effective approach to target stylistic variation and requesting for clarification in learners with LLD. Role-play as an approach to intervention may therefore be the way forward in ensuring generalisation of pragmatic skills. However, results of the study should be interpreted with the limitations in mind. The results of this study have also identified further areas of research regarding the use of role-play as a therapy approach and provided therapists with recommendations to inform their clinical practice. The results of this study have laid the foundation for future research and implementation of role-play as a therapy approach in speech-language pathology.