Open AccessLanguage, Speech, and Hearing Services in SchoolsResearch Article2 Oct 2020

Narrative Intervention: Principles to Practice



    Narrative interventions are a class of language interventions that involve the use of telling or retelling stories. Narrative intervention can be an efficient and versatile means of promoting a large array of academically and socially important language targets that improve children's access to general education curriculum and enhance their peer relations. The purpose of this tutorial is to supply foundational information about the importance of narratives and to offer recommendations about how to maximize the potential of narrative interventions in school-based clinical practice.


    Drawing from decades of cognitive and linguistic research, a tutorial on narratives and narrative language is presented first. Ten principles that support the design and implementation of narrative interventions are described.


    Clinicians can use narrative intervention to teach story grammar, complex language, vocabulary, inferencing, and social pragmatics. Storytelling, as an active intervention ingredient, promotes the comprehension and production of complex language.


    When narrative intervention is implemented following a set of principles drawn from research and extensive clinical experience, speech-language pathologists can efficiently and effectively teach a broad set of academically and socially meaningful skills to diverse students.

    Narrative intervention is one of the most powerful approaches to language intervention that school-based speech-language pathologists (SLPs) can use. We define narrative intervention as any language intervention that involves children telling or retelling stories that have specific language-related features purposefully targeted by the interventionist (Petersen, 2011). In the last 10 years, there has been an increased focus on narrative intervention in research. Consistently, effect sizes for improving various aspects of language in children with and without language disorder have been large (e.g., S. L. Gillam et al., 2014; Spencer et al., 2018). In addition to yielding powerful effects on language, the research documents that narrative intervention is also highly feasible and flexible. While our purpose in this article is not to critically review the extant literature, we provide empirical evidence, share clinical insights, and outline how narrative intervention is versatile, potent, and academically and socially relevant. We have organized the available empirical and practical guidance in a manner that facilitates its transfer to clinical practice. As we strive to help clinicians maximize their effective implementation of narrative intervention, we first present a brief tutorial on narratives. Our understanding of narratives and their characterizations emerge from a depth of clinical experience and empirical studies conducted over several decades. The major points are summarized in this article to provide the reader with an introduction to narrative language (see Boudreau, 2008, and Johnston, 2008, for other summaries). Following the tutorial, we offer 10 principles that can guide the strategic use of narrative intervention in schools.

    Introduction to Narratives

    In the academic context, narratives are most often the monologic telling or retelling of a real or imaginary event (R. B. Gillam & Ukrainetz, 2006). Such narratives are goal directed, where an agent's efforts to resolve problems or complications and the resulting consequences are described. Canonical elements of stories include an initiating event (usually a problem), an attempt to solve the problem, and a consequence (Mandler, 1987; Stein & Glenn, 1979). Together, these form a basic, minimally complete episode, and the inclusion of additional story grammar elements such as an explicit plan to solve the problem and a resolution increases a story's episodic sophistication. When attempts to solve a problem are ineffective, additional episodes extend the story until an eventual resolution is achieved, as is common in novels.

    Stories are not stories without causal and temporal relations between events. It is the causal links between the initiating event (or problem), attempt, and consequence that reveal a character's goal-oriented behavior (Peterson & McCabe, 1983). If story events are presented in an unexpected order, comprehension of the causal connections can be diminished. Nonetheless, specific sequences and temporal and causal devices are culturally grounded. They are taught through extensive exposure to a linguistic community's storytelling patterns, which, over years of exposure, shapes the understanding and production of narratives in a certain way.

    There can be cultural differences in narrative structure (T. B. Champion, 1998; Gee, 1989; S. L. Gillam, Fargo, et al., 2012; Gorman et al., 2011; Mills et al., 2013; Tappe & Hara, 2013; C. E. Westby, 1994). For example, internal responses tend to be emphasized to a greater extent in narratives produced by Latino, Spanish-speaking children (Castilla-Earls et al., 2015). Native American stories often do not unfold in a linear manner (Sharifian, 2002), with story grammar that does not always follow Stein and Glenn (1979) conventions. Worth and Adair (1972) noted that Native Americans often focus considerably more on setting information than stories found in academic settings, with a major focus on detail and implicit information as opposed to explicitly stated temporal and causal connections (Basso, 1990; C. Westby et al., 2002). Japanese children also tend to have a more implicit style of narration, often producing brief, minimalistic stories (Minami, 2002) or stories with minimal goal structure (Matsuyama, 1983). African American children have been noted to use a topic associating approach to narration similar to that used by Native Americans (T. Champion et al., 1995; Gee, 1989; Hyon & Sulzby, 1994) as well as a call–response interaction (Smitherman, 1977), which entails the narrator producing a story with interspersed “calls” from the listener.

    Despite cultural variations, most stories still have protagonists seeking to solve challenges, and all children who reside in the United States, including those from culturally, ethnically, and linguistically diverse backgrounds, can benefit from learning the story grammar and relational features expected in U.S. schools. Narrative intervention can be applied to any approach to narrative structure (e.g., Applebee, 1978; Hudson & Shapiro, 1991; Labov & Waletzky, 1967), yet what has been developed thus far has mostly followed Stein and Glenn's (1979) model of story grammar. For example, each of the following assessments draw from Stein and Glenn's story grammar framework to some extent: the Assessment of Story Comprehension (Spencer & Goldstein, 2019), the Edmonton Narrative Norms Initiative (Schneider et al., 2005), the Monitoring Indicators of Scholarly Language (S. L. Gillam et al., 2016), the Narrative Assessment Protocol (Bowles et al., 2020), the Narrative Language Measures subtest of the CUBED assessment (Petersen & Spencer, 2012, 2016), the wordless picture books or other sampling procedures used in conjunction with the Systematic Analysis of Language Transcripts and the Narrative Scoring Scheme (Heilmann et al., 2010; Miller et al., 2015), and Westerveld and Gillon's (2010) language sampling protocol. Additionally, the following manualized intervention programs promote Stein and Glenn's story grammar structure for U.S. children: Let's Know! (Language and Reading Research Consortium et al., 2017), Story Champs, (Spencer & Petersen, 2016), Story Grammar Marker (Moreau & Fidrych, 1994), and Supporting Knowledge in Language and Literacy (SKILL; S. L. Gillam & Gillam, 2016). Although it is common in the United States to teach the story grammar framework, doing so does not imply that such an approach to narration should be considered a replacement for the student's own culturally unique storytelling approach. Teaching children to tell stories in a manner that aligns with curriculum expectations is a matter of teaching code-switching, not code-overwriting.

    Interwoven throughout the narrative structure is the language used to express it. Whereas story grammar is used to organize the gross discourse structure of a narrative, syntax, morphology, semantics, and other features of language are used to make up the narrative microstructure (Colozzo et al., 2011). When a narrative contains microstructural features that reflect literate, academic language, it is often markedly more complex than dialogic conversation (MacLachlan & Chapman, 1988). This literate language expressed in the context of a story, which we refer to as narrative language, often consists of multiple causal and temporal adverbial, relative, and nominal subordinate clauses, elaborated noun phrases, adjectives, adverbs, appositives, and dialogue (Benson, 2009; Greenhalgh & Strong, 2001; Peterson & McCabe, 1983; Roth & Spekman, 1986). To tell about an event (either orally or in written form) that the listener/reader did not also experience, and for the story to be comprehensible, the storyteller frequently uses sophisticated, complex, precise language to paint a detailed picture (Bruner, 1986; Johnston, 2008; C. E. Westby, 1985). Effective storytelling requires complex grammar to depict specific images and evoke purposeful emotions, whereas effective conversation benefits from suprasegmentals such as gestures and joint attention, facilitating comprehension despite the use of imprecise sentences (Hudson & Shapiro, 1991). Narratives also often contain many complex and infrequent vocabulary words that facilitate comprehension (Petersen et al., 2020). The spontaneous use of complex vocabulary or domain-specific words during storytelling demonstrates semantic complexity and is arguably a better indicator of one's lexical aptitude (Hadley & Dickinson, 2018).

    Narratives are commonly examined at macro and micro levels, but that is not to suggest they are restricted to story grammar or morphosyntactic analyses. In fact, narratives showcase a large number of core skills that serve as the foundation to a range of academic (C. E. Westby, 1985) and social abilities (Brinton & Fujiki, 2017). Narratives bring several inconspicuous repertoires together to form an integrated higher order composite that is clinically convenient and useful (Boudreau, 2008). To be an accomplished storyteller, a child must orchestrate many simultaneously converging processes including attention, cognition, memory, inferencing, and theory of mind (Curenton, 2011; Dodwell & Bavin, 2008; R. B. Gillam & Johnston, 1992; Hudson & Shapiro, 1991; Johnston, 2008; Lahey, 1990; Nippold & Schwartz, 1996). Considering narratives are a pervasive and complex form of discourse, they have considerable clinical utility.

    A notable literature documents the predictive nature of early narrative language skills for later academic achievement (Bishop & Edmondson, 1987; Dickinson & McCabe, 2001; Fazio et al., 1996; Feagans & Appelbaum, 1986). Children's narrative abilities are related to listening comprehension (Bourg et al., 1997), receptive vocabulary (Heilmann et al., 2010), and writing (Griffin et al., 2004; Kim et al., 2015; Scott & Windsor, 2000). The relation between early oral narrative skills and later reading comprehension is particularly strong (Catts et al., 2002; Dickinson & McCabe, 2001; Gilmore et al., 1999; Griffin et al., 2004). For instance, C. E. Snow et al. (2007) reported that the ability to produce a narrative in preschool predicts reading comprehension performance in the fourth, seventh, and 10th grades. Likewise, Suggate et al. (2018) demonstrated that narrative skill at school entry predicted reading comprehension performance 10 years later. Moreover, evidence exists of causal relationships between narrative-focused oral language interventions and reading comprehension (Clarke et al., 2010; Language and Reading Research Consortium et al., 2019) and writing outcomes (Kirby et al., 2020; Petersen et al., 2019; Spencer & Petersen, 2018).

    The value of narratives reflected by the current body of research is mirrored in prevailing academic standards. For example, the Common Core State Standards, adopted by the majority of U.S. states (National Governors Association Center for Best Practices & Council of Chief State School Officers, 2010), specifically details story grammar and language complexity outcomes for all primary, middle, and high school grade levels. Objectives in the Common Core State Standards related to narration can be found in the English Language Arts Standards, dispersed throughout specific reading literature, speaking and listening, and language standards.

    Narratives are socially important too. Parents of children with disabilities often express their desire for children to tell them about their day or what happened at school (Pituch et al., 2011). Teachers need information about events that took place on the playground or over the weekend, and peers want to be entertained. All of these are related to storytelling, and, in general, children with language disabilities do not excel at social communication (Colozzo et al., 2011; Dodwell & Bavin, 2008; Fey et al., 2004; Kaderavek & Sulzby, 2000). Children who are good storytellers are more accepted by their peers (Hart et al., 2004; P. C. McCabe & Marshall, 2006) and receive parent and teacher approval (Bliss & McCabe, 2012). Additionally, increased exposure to narrative conversations predicts children's future ability to solve social problems (Leyva et al., 2014).

    Storytelling appears to be a pivotal social skill in that children naturally extend what is learned in training environments (e.g., classroom or clinic room with SLP) to untrained social contexts with minimal effort. From a clinical perspective, what could be better? Narrative intervention uses an authentic, functional context, which entails the formation of a genuine connection between adult and child, where meaningful information is conveyed, and language processing, pragmatics, and social–emotional learning are integrated (Brinton & Fujiki, 2019). The natural environment (i.e., peer and adult attention) can maintain children's efforts to tell stories and, by doing so, provide additional opportunities for practicing the newly acquired language skills. Possible evidence of this generalization is seen in an important study comparing a written text comprehension intervention, an oral narrative intervention, and a combination intervention (Clarke et al., 2010). Researchers found that students in the oral language intervention-only condition, but not the other two conditions, continued to improve their reading comprehension performance, even after the intervention was withdrawn. Brinton and Fujiki (2019) examined whether a narrative-based language intervention promoted language processing, pragmatics, and social and emotional learning for children with developmental language disorder. Intervention sessions were focused on storytelling and story enactment. Children were taught how to write or dictate a brief journal entry focusing on new concepts learned from the story. Teacher ratings showed improvement in prosocial behavior, and the school-based SLP reported improvement in expressive narrative language and syntax.

    Improving oral narration can improve written narration. Spoken narratives play a fundamental role in the development of written stories and are more commonly addressed by SLPs. Because young children cannot write what they cannot say, children with language disabilities are at a particular disadvantage with respect to academic writing tasks (Koutsoftas & Gray, 2012; Pavelko et al., 2017). Fortunately, recent research indicates that an explicit and intensive focus on oral narrative language has a direct and meaningful impact on the quality of written narratives children produce, even for children with significant disabilities (Kirby et al., 2020; Petersen et al., 2019; Spencer & Petersen, 2018). Researchers suggest that this transfer from oral narration to written narration occurs because the story grammar and linguistic structures taught and practiced through oral storytelling evolve into a cognitive schema that generalizes to alternative modalities, including writing (Hayes & Flowers, 1980; Rumelhart, 1980; C. Snow, 1983; C. E. Westby, 1984). Based on schemata theory (Anderson & Pearson, 1984; Mandler, 1987), the quickest way to promote story writing is to promote oral storytelling that takes advantage of the shared story level and sentence level structures.

    One of the greatest advantages of addressing oral narration for the purpose of improving written narration is that it eliminates the interference during intervention from other potentially delayed or emerging skills related to transcription (i.e., handwriting mechanics, letter formation, spelling, and letter sounds), while garnering children's cognitive focus for generating and organizing their thinking and speaking (V. Berninger & Graham, 1998; McCutchen, 1996). Even when transcription and decoding skills are automatized, the quality of written narratives and reading comprehension is intimately dependent on the oral language foundation the child brings to the task (V. W. Berninger & Abbott, 2010; Catts et al., 2006; Fey et al., 2004). Hence, it is recommended that oral narratives receive considerable attention during therapy, especially for young students and those with language-based disabilities.

    Clinicians must be strategic and planful to be able to facilitate the development of narrative schemas that will readily transfer to academically and socially meaningful repertoires. Ultimately, the flexible and pragmatic use of acquired narrative content and form is the goal. Various conditions (Hughes et al., 1997) in which storytelling occurs form a continuum of narrative tasks (see Figure 1). For example, stories can be told or retold. Generating quality narratives can be more challenging than retelling a story (R. B. Gillam & Johnston, 1992; Kaderavek & Sulzby, 2000), yet there can be simple story generations and very complex story retells. Because retells reflect both listening comprehension and story production skills (Boudreau, 2008; Petersen & Spencer, 2012; Wagner et al., 1999), individuals who have weak receptive language may find generating a story easier than retelling one (Weddle et al., 2016). Telling about recently heard/read events is a common academic task and would be considered a retell of a model story. Stories about personal experiences are most often produced for social reasons such as gaining attention from peers or approval from adults (McCabe, 2017). The first telling of the experience would be considered a story generation, but subsequent repeated productions would be best classified as retells. Given social attention is the natural (and powerful) consequence of storytelling and it is easier to talk about real events (A. McCabe et al., 2008), personal stories are pervasive in children's spontaneous communication (P. C. McCabe & Marshall, 2006; Preece, 1987; C. Westby & Culatta, 2016). When a story is not true and has never been told before (either orally or written), we would characterize it as a fictional generation. Fictional stories can be more challenging than personal event narratives (Allen et al., 1994; A. McCabe et al., 2008), yet this is not always the case. A less common type of narrative is a script, which has been defined as a recount of typical events, based on a collection of personal experiences rather than a single event (Hayward et al., 2007; Hudson & Shapiro, 1991; Hughes et al., 1997). For instance, if a child tells about what he usually does on Sundays or how his family celebrates a specific holiday, he is generating a script narrative. This type of discourse is highly related to expository discourse because it does not focus on a specific event.

    Figure 1.

    Figure 1. General continuum from easier to more difficult storytelling.

    Often, story retelling is the go-to context for initial teaching. Once children have acquired a proper story grammar framework, it can be extended to the generation of personal stories and eventually to the generation of fictional stories and writing. Retelling a short, familiar book or a story that was specifically designed for intervention allows for graduated, controlled, and targeted practice because the clinician can select stories based on an individual child's language skills.

    Photos or illustrations that provide clues about story content can be used to modify the demands of storytelling, depending on the goal of narrative intervention and the child's abilities. Although visual materials can scaffold retells (e.g., Meyer, 1969) and fictional stories (S. L. Gillam & Gillam, 2016), not all classroom academic tasks, to which narrative skills should generalize, involve visuals (e.g., listening and reading comprehension), especially as the grades increase. Likewise, pictures and icons are generally not available in social contexts. Some research suggests that young children and children with language impairment retell more complete episodes and include more information units when they listen to stories without pictures as opposed to orally and pictorially presented stories (Schneider, 1996; Schneider & Dubé, 2005). Photos and illustrations can support storytelling, but they may result in a low estimate of a child's abilities. Therefore, it is important to use visuals strategically, prevent dependence on them, and promote the generalization of storytelling skills to more natural and sophisticated contexts.

    Principles to Practice

    There is not one right way to implement narrative intervention. Nonetheless, the extant research evidence and our clinical experiences have revealed 10 principles of narrative intervention that can help guide practice. Notwithstanding the collection of principle-driven guidelines described below and listed in Table 1, it is important to remember that children's Individualized Education Program goals and instructional objectives should drive the selection of specific intervention parameters. Relevant assessment data should inform the selection of individual goals and objectives, ensuring a direct link from assessment to intervention. A chief purpose of valid, reliable, and meaningful assessment is to motivate efficient and effective intervention. Therefore, SLPs are responsible for using purposeful assessments, converting assessment data to goals, and making wise choices about whether to work on story grammar or complex language, retells or generations, personal or fictional stories, and what type and timing of scaffolds to use. There are a number of commercially available narrative intervention programs (e.g., SKILL, Story Grammar Marker, Story Champs) that can reduce the time and energy needed for preparation, but clinicians are still responsible for making evidence-based choices regarding the selection of programs and for using them effectively with their clients.

    Table 1. Principle-driven narrative intervention guidelines.

    1. Build story structure before vocabulary and complex language
    2. Use multiple exemplars to promote metalinguistics and generalization
    3. Promote active participation
    4. Contextualize, unpack, and reconstruct stories
    5. Use visuals to make abstract concepts concrete
    6. Deliver immediate corrective feedback
    7. Use efficient and effective prompts
    8. Differentiate, individualize, and extend
    9. Arrange for generalization opportunities
    10. Make it fun
    Build Story Structure Before Vocabulary and Complex Language

    If a child cannot yet tell or retell a story using a complete episode, that is where we recommend clinicians start. To work efficiently on syntax, grammar, or academic vocabulary (or anything else) during storytelling, it is helpful for children to be able to tell and retell basic stories. Once a child can produce the basic structure, storytelling becomes the context for teaching a wide variety of other skills (Justice et al., 2008; Swanson et al., 2005). Ensuring children develop a cognitive schema related to story grammar early can help reduce frustration of children with extremely limited language skills, by gradually increasing the cognitive and linguistic demands. Spencer et al. (2013) provide an example of how this might unfold. They examined an individual narrative intervention with 3- and 4-year-old children with developmental disabilities, focusing exclusively on story grammar for 12 sessions. Once storytelling established and each child could retell a minimally complete episode, language complexity targets such as temporal and causal subordinating ties were systematically folded into the sessions.

    Interventions with older children may not need to begin at the level of story grammar construction. Research indicates that children with language disabilities tend to struggle with both story grammar and grammar/syntax, but their use of complex language is often affected to a greater degree (Colozzo et al., 2011; Fey et al., 2004; R. B. Gillam & Johnston, 1992). If that is the case, then the goal can be to provide a child with extra, incidental practice telling stories while specifically modeling, prompting, and encouraging the use of correct syntactical forms. The same principle applies to vocabulary. If definitions of words and the use of more sophisticated words are the desired targets, children will learn them more efficiently if their cognitive energy is not exhausted on constructing the narrative schema. Gardner and Spencer (2016) investigated explicit vocabulary instruction embedded in a narrative intervention delivered to small groups of at-risk first graders. They observed that, as retells improved, children became better at learning the target vocabulary words. They also became better at figuring out the meaning of novel, unknown words from a story context. S. L. Gillam et al. (2014) also embedded vocabulary targets in their classroom narrative intervention. Although, as a group, the first graders who received the intervention learned the vocabulary specifically taught to them, children who were at risk acquired fewer new word meanings.

    An important point to remember is that, although teaching children to tell stories may be necessary, narrative intervention does not need to stop there. Numerous other therapeutic targets can be contextualized in storytelling once story structure is established to a minimal degree (S. L. Gillam, Gillam, & Reece, 2012). If stories are strategically and carefully designed for intervention, it may be easier to facilitate instruction on the embedded targets (see Spencer & Petersen, 2016) than if clinicians are dependent on what is available in published storybooks. Stories can be designed (or selected) for instructional purposes that include the promotion of grammar, syntax, cohesion, elaborated noun phrases, mental state verbs, dialogue, inference, theory of mind, vocabulary, and social problem solving, to name a few. The focus of narrative intervention is on oral narratives at this moment, but the path from strong oral narratives to written narratives is unconstrained (Petersen et al., 2019; Spencer & Petersen, 2018). Other critical academic skills that clinicians can address from a firm oral narrative foundation include listening comprehension (i.e., answering questions about what one hears; e.g., Spencer et al., 2013), reading comprehension (i.e., answering questions about what one reads; e.g., Clarke et al., 2010), and expository discourse (i.e., telling or retelling information; e.g., Lee, 2020; Petersen et al., 2019).

    Use Multiple Exemplars to Promote Metalinguistics and Generalization

    Multiple exemplar training is a systematic procedure to promote the generalization of higher order concepts (LeFrance & Tarbox, 2020). For narrative intervention, this means that clinicians should use several different stories (as opposed to one or two) with the same story grammar elements to teach story grammar. If the goal of the intervention was for a child to tell or retell a specific story, then practicing the same story over and over would work fine. However, the child may not learn how to tell stories generally and may not know much about story grammar as a higher order concept that could be transferred to other stories or contexts. To avoid promoting the memorization of specific stories, several exemplars (i.e., different stories with the same story grammar components) should be used as models in consecutive intervention sessions. Rapid acquisition of story grammar concepts can be accomplished if the same stories are not repeated because, in narrative intervention, the multiple exemplars promote children's abstraction of the pattern. Thus, by presenting children with different exemplars of the concept across sessions (Layng, 2019), story grammar schema emerges. Children quickly learn that story content is not the salient feature, and what is important is the schematic pattern. Knowledge of this schema and the ability to describe the pattern of stories (i.e., naming character, problem, feeling, action, and ending) is metalinguistic. A child's successful generation or retelling of a novel story resulting from narrative intervention is evidence of generalization and metalinguistic development.

    Although several narrative interventions have employed storybooks (Adlof et al., 2014; Brinton & Fujiki, 2017; Catts et al., 2016; Petersen et al., 2010), some researchers (including us) have strategically designed stories for initial intensive and targeted narrative practice (Brown et al., 2014; Favot et al., 2018; Hayward & Schneider, 2000; Petersen et al., 2014; Spencer et al., 2020, Spencer, Petersen, & Adams, 2015; Spencer & Slocum, 2010; Swanson et al., 2005). Some advantages of using prepared stories (over storybooks) for narrative intervention include being able to control what structures, vocabulary, and complex language children are exposed to, easily teaching metalinguistics through multiple exemplar training and rapid practice cycles within session because stories are brief, and building narrative comprehension and production simultaneously. Most approaches to narrative intervention eventually move away from a focus on program-specific stories to the integration of authentic children stories and curriculum materials.

    Promote Active Participation

    Decades of research indicate that active participation leads to greater learning opportunities (Archer & Hughes, 2011; Ellis & Worthington, 1994; Engelmann & Carnine, 1982). Maintaining a brisk teaching pace increases children's opportunities to respond and reduces opportunities for behavior problems (Forsyth & Archer, 1997). When students are actively engaged and responding, they are learning (Greenwood et al., 1984; Pratton & Hales, 1986). Interestingly, Knapp and Desrochers (2009) found that students prefer interactive instruction in which they play a prominent role.

    Children need repeated opportunities to practice meaningful language within an activity or session to ensure long-term maintenance of the targeted skills (Ukrainetz, 2006). Focused practice, supported by modeling and corrective feedback, will enhance children's ability to use the skills spontaneously in nontraining environments. With respect to narratives, this means that children must have ample opportunities to tell or retell stories. Listening to stories may improve comprehension of stories and have marginal impacts on narrative expression, but to hasten children's ability to tell/retell stories, they must practice it. More practice is better.

    To maintain high numbers of opportunities to respond in the context of narrative intervention, the size of the intervention group needs to be considered. If intervention is delivered individually, it is easy to support the child to do most of the talking. However, when other students are in the group, time should be evenly distributed across the students so that each student is responding frequently and everyone has an opportunity to tell/retell at least one complete story, beginning to end. It is common for children to have individual turns during small group narrative interventions, but it is important to keep the groups small enough so that each child receives numerous opportunities to practice storytelling. Some small group narrative interventions encourage the children who are listening to play games that demonstrate they are actively listening and participating (Spencer & Petersen, 2018; Spencer & Slocum, 2010; Story Grammar Marker, 2020; Weddle et al., 2016). Although all the children may not be talking at the same time, they are doing something that requires their attention to story grammar, which benefits comprehension and serves as scaffolding for when it is their turn to tell the story. Through active responding games, children also learn to be better listeners and can even be encouraged to play the role of co-teacher.

    If intervention will be delivered to large groups of children, there are a number of other available strategies that can support active engagement. In S. L. Gillam et al.'s (2014) classroom-based narrative intervention, students used graphic organizers. When children retold stories individually, a teacher walked around the room to check on them. Choral responding (Heward et al., 1989), which means that all children respond in unison, has been used in several large group narrative intervention studies (Petersen et al., 2019; Spencer, Petersen, & Adams, 2015; Spencer, Petersen, Slocum, et al., 2015). This strategy is helpful when there are too many students for the teacher or interventionist to give each one individual support. In these studies, children chorally named the parts of the story (i.e., “character, problem, feeling, action, ending”) as well as the story content, one sentence at a time (e.g., “John was sad because his knee hurt.”). After children had told the story in parts with the group, they were paired with a friend in a class-wide peer tutoring arrangement to retell the model story individually (Bowman-Perrott et al., 2013). Choral responding ensures that all children in the class/group talk, not just the students who are confident enough to raise their hands, and peer tutoring provides an opportunity for each child to retell a story while their peer is attentive and playing the role of teacher.

    Contextualize, Unpack, and Reconstruct Stories

    When teaching storytelling, it is important to make the activity meaningful and relevant to children (R. B. Gillam & Ukrainetz, 2006). Storytelling is used for purposeful communication with one or more listeners. This intentional, communicative purpose should not be lost during the intervention process. It may be tempting to eliminate story content that children are not yet able to retell independently. Although that may be a good reason to select stories designed for narrative intervention over storybooks, there are several reasons to retain all the main episodic elements (e.g., problem, attempt, and consequence) of a story. First, comprehension of a narrative is facilitated through the relations between the story grammar elements (Curenton, 2011; Schick & Melzi, 2010) and, when components are missing, children cannot detect the causal and temporal relations between them. Stories are much more than the sum of their parts. Having children practice stating the character and setting of a story without cohering it to the rest of the story eliminates the purposeful, communicative nature of storytelling and renders the activity a sentence repetition or picture description task. Second, children do not necessarily develop story grammar in the sequence they are told and often taught (e.g., character and setting first) because the meaningful, related parts are in the middle. Although young children can talk about characters and settings, storytelling does not become storytelling until a child associates an initiating event with an attempt and a consequence (cause and effect). Development of storytelling can be thought of occurring from the inside out, beginning with the feeling that some past event triggered. As children develop, they begin to add layers of information about the problem that caused the feeling or what was done to resolve it, and eventually add the more distant contextual details such as the setting and resolution. Third, when the goal of narrative intervention is to teach the cognitive schema related to narration, it is necessary for every story to adhere to that pattern, otherwise, meta-skills will not be achieved. It is difficult to promote the development of narrative schema using multiple exemplars unless there are clear patterns across those exemplars.

    We recommend that stories used in intervention contain, at a minimum, a basic, minimally complete episode (problem, attempt, consequence) and that, before discussing the individual components in the intervention session, the entire story is presented in its complete and complex form first. This provides contextualizing information and ensures children see it as a whole, thereby enhancing purpose and motivation for the task (R. B. Gillam & Ukrainetz, 2006). Once the whole is modeled, the story can be broken down into more manageable chunks, ensuring multiple opportunities for children to practice each part. Yet, even when each part is practiced individually, it is important to contextualize the child's utterance and attach meaning to it. For example, asking questions related to story grammar (e.g., “What was John's problem in this story?” or “What did he do to fix his problem?”) or having children practice retelling the parts, one part at a time (e.g., “John was riding his bike down the street.” or “He asked his mom for a bandage.”), is an appropriate scaffolding technique, allowing the interventionists to provide several models and feedback during guided practice, yet we also encourage children to place those parts back into either the entire narrative or a larger section of the narrative. For example, after prompting a child to include the character's feeling, we often ask the child to go back one element in the story (e.g., to the problem) and to retell both the problem and the feeling so that there is cohesion and purpose in the task. After a period of supported telling/retelling of the parts within the same session, the story should be reconstructed into its whole. Children should be encouraged (and supported) to tell/retell the story in its entirety and in order, and to the extent possible, use cohesive devices to causally and temporally connect the episodic elements. Thus, within a single intervention session, a story is presented as a whole unit to contextualize it, then unpacked as parts for guided practice, and reconstructed as a whole in independent practice (R. B. Gillam & Ukrainetz, 2006). In this manner, children are frequently reminded of the purpose of each story grammar element and the importance of their cohesion for generating a more meaningful whole. This whole–part–whole framework driven by purposeful, explicit instruction has been shown to be effective in a number of narrative intervention studies (e.g., Brinton & Fujiki, 2017, 2019; S. L. Gillam, Gillam, & Reece, 2012; all Story Champs studies).

    Use Visual Support to Make Abstract Concepts Concrete

    Because storytelling requires the confluence of numerous higher order processes, it is a demanding task for students with language disabilities. We have already established that the structure of stories is a concept (or schema) learned through exposure to multiple exemplars. However, teaching children about schemas can be a slow process without the help of visual support (e.g., pictures, drawings, icons, graphic organizers). Although clinicians should be cautious about allowing children to always use visuals when telling or retelling stories, they can be an extremely effective teaching tool, especially when introducing a new story. Specifically, visual supports can be used to make abstract concepts such as story grammar and linguistic features (e.g., causal subordinate conjunctions) more tangible and concrete for children.

    The most popular visual support strategies include using pictography, icons, symbols, gestures, or manipulatives that represent the story grammar elements or linguistic targets. Icons, symbols, and manipulatives can be designed to have correspondence to specific components (S. L. Gillam et al., 2018; Spencer & Petersen, 2016; Story Grammar Marker, 2020). In some studies, interventionists have taught each component with its symbol or icon in isolation and spent time explaining its purpose in stories (e.g., Brown et al., 2014; S. L. Gillam et al., 2014), but in our research, we teach children the names of the story grammar parts while modeling a new story. Through the repeated visual presentation of the icons in a specific order and repeated aural association with the names of the parts, children quickly acquire an understanding of the schema. Icons have also been used to represent less common vocabulary (i.e., Tier 2 words; Beck et al., 2013) and complex language features such as causal and temporal ties (Gardner & Spencer, 2016; Petersen et al., 2019). Pictography, which involves the use of simple line drawings to represent ideas and events in a story, can support story retelling, story generation, and the production of complex language (R. B. Gillam & Ukrainetz 2006; McFadden, 1998). We find pictography particularly useful when children are asked to generate their own story. By quickly drawing the content of the child's story, children can use the visual representation to support subsequent retellings (Petersen et al., 2014; Spencer et al., 2013). Graphic organizers and concept maps are additional examples of visuals that have been used to help children learn the pattern of stories (Boulineau et al., 2004; S. L. Gillam et al., 2015).

    When visuals are used during intervention, clinicians should be strategic about fading them. Sometimes, pictures or illustrations correspond to the specially designed stories, which reduce the demands for reticent children. If using storybooks, interventionists need to be cognizant that the illustrations can provide more information about the story than is needed. In some cases, relying on that information is beneficial, but sometimes, too many visual supports can interfere in the acquisition of the cognitive schema that allows them to use the structures in a generalized manner. Icons, symbols, pictography, and manipulatives serve as reminders of the narrative structure but contain less information about a specific story than detailed pictures and illustrations do. Therefore, we recommend fading pictures and illustrations before icons so that children can benefit from a smooth and gradual transition from easy to difficult tasks within a session. For example, in recent studies featuring small group narrative intervention with Spanish-speaking preschoolers learning English (Spencer et al., 2019, 2020), illustrations and icons were displayed while the story was modeled and when the first child retold the story independently. When the second child retold the story independently, only the icons were available, and when the third child retold the story, no visuals were available.

    Deliver Immediate Corrective Feedback

    Feedback is a powerful teaching tool (Hattie & Timperley, 2007). If a child is saying things you want them to say again or if they are using language you want them to repeat, the best positive feedback includes active listening (e.g., nodding head, eye contact) and recasting (Cleave et al., 2015). If the child is not including specific aspects of story structure or language that you desire, a correction is needed. Effective corrective feedback has three main characteristics, all of which apply to delivering feedback in narrative interventions. First, effective feedback focuses on what the child should do and minimizes attention given to the incorrect response. Students need information about their performance and what they can do to improve it. Ultimately, a successful correction leads to the child producing the correct response (Watkins & Slocum, 2004). A possible correction could simply be a model of what the child should say (e.g., “Say it like this. John was sad.”) with a subsequent opportunity for the child to try it again.

    Second, effective feedback is immediate (Watkins & Slocum, 2004). Practiced errors are likely to continue. Therefore, as soon as an error is made, a correction should follow. For example, if a child is retelling a story and skips a critical element (e.g., action), this would be considered an error. The interventionist should stop the story immediately and make an appropriate correction. It is important not to wait until the end of the story to provide the correction because feedback that is delayed from the performance is less effective than feedback that is delivered immediately and contingent up student responses (Archer & Hughes, 2011). This immediate feedback can be delivered in a naturalistic way, where the interventionist can express genuine interest or confusion about missing information in the story.

    Third, effective feedback is specific (Watkins & Slocum, 2004). It is tempting to give a reminder or hint at what the child should have said or provide some general information that a mistake was made (“Oops. You forgot something.”). Unfortunately, less intrusive corrections such as these can confuse or frustrate children, further depressing their motivation to try challenging things. If the response was firmly in a child's repertoire (in other words, it was easy), they would have made it. It is best to assume that the response is still too hard and to give that child-specific information about what to do instead. This can be resolved by explicitly telling the child what was missed (e.g., “Oops. You forgot to tell us his feeling. How did he feel?”) or modeling what the child should say and then giving the child the opportunity to imitate it.

    Telling or retelling story grammar elements in a consistent sequence can hasten the acquisition of the schema and prevent excluding critical information accidentally. In a way, the story grammar components operate like the links in a chain that function properly when arranged in a certain order. For that reason, when a child skips a story grammar element, it can be helpful to have them practice the correct sequence again immediately following the correction. In Petersen et al. (2017), the interventionists made immediate corrections and then asked the student to continue retelling the story starting with the component before the one that was missed. This type of overcorrection procedure honors the special cohesion between the elements and helps students to depend on the ordinal unfolding of events as they tell or retell stories.

    Use Efficient and Effective Prompts

    There is a large array of options when it comes to prompting (Cengher et al., 2018). One method of prompting (and fading) was described above as it relates to the use of visual materials. When the gradual withdrawal of visual materials (e.g., illustrations and icons) is embedded in the intervention's procedural steps, prompting is controlled by the curriculum. In other words, they are crucial active ingredients that should not be altered by the interventionist. Aside from these built-in prompts, clinicians are in control of the rest. It is the interventionist's responsibility to select and use efficient and effective prompts.

    Over the years, we have come to appreciate a simple prompting technique that can reduce the amount of time spent prompting while ensuring that the child is successful in two easy steps. We call this procedure two-step prompting. For any clinician-controlled response prompt, we recommend first asking a question that includes specific information about what the child missed or should have mentioned. Although furthering prompts like “Then what happened?” or “What happened next?” can be used, we do not consider these response prompts because they do not provide information about what response is expected from the child. Furthering prompts are typically used to encourage the child to keep going or redirect them to the task, but assume they do not need help forming the expected response. In two-step prompting, we suggest that clinicians use questions such as “How did John feel about his problem?” or “What did he do to fix the problem?” because they direct the child's attention to the specific part of the story that needs to be told/retold. They also enhance children's ability to answer comprehension questions about a story without explicitly teaching children to answer wh-questions (Spencer et al., 2013). The second part of two-step prompting is to model it for the child. If the child is unable to tell/retell the expected target within a few seconds after asking the question, the clinician should quickly follow it up with a model sentence such as “John was sad because he cut his knee. Now, you say that.” or “He asked his mom for a bandage. Your turn.”

    Two-step prompting has a number of advantages. First, by using wh-questions as a prompting strategy, children's ability to answer wh-questions improves without specifically teaching it. In Spencer, Petersen, Slocum, and Allen (2015), preschoolers who received large group narrative intervention that addressed retelling using wh-questions as prompts made substantial gains over the control group on a story comprehension measure (d = 54). Second, the two-step prompting procedure helps children be successful. Children do not have to fail multiple times to get the amount of help they need. It is important that children experience success and receive approval and praise for their efforts, which ultimately determines if they will try again. Third, two-step prompting is efficient. It can take less than 10 s to complete both steps if they are needed. This reduces the amount of time a child experiences difficulty and reduces the time other children are passively observing. Fourth, the two-step prompting procedure is easy for the clinician to remember. Multistep, systematic prompting can become laborious for the interventionist to remember the steps, especially if those steps are individualized for different children. Fifth, two-step prompting is generally applicable. This means that it works for most children and most targets. It simplifies how an interventionist differentiates within a group, depending on the needs of the children. For example, let us say one child is still working on retelling a story with all the basic story grammar elements and another child is working on using complex vocabulary. For the first child, the interventionist can ask a question about the story grammar element missed, and for the second child, the interventionist can ask, “What is the special word that goes in this sentence?” and then model the sentence if needed (e.g., Weddle et al., 2016). Finally, two-step prompting leads to independence while reducing prompt dependency. By starting the prompting with the question, which is often a natural part of storytelling between conversational partners, the clinician does not provide more help than is needed. This allows the child an opportunity to produce a mostly independent response when they can do it.

    While two-step prompting works with most children, question prompts are not equally effective with all children. In a recent narrative intervention study with children with autism (Garcia et al., 2019), we discovered that participants rarely responded to the first step of two-step prompting. Presumably, they did not understand the questions. Modeling sentences and delaying the timing of models were more effective than the story grammar–focused wh-questions. It became apparent that, if a child has to be taught how to respond to the intended prompts, then they cannot be used effectively to “prompt” behavior. In our experience, two-step prompting is a good procedure to start with, but clinicians may need to adapt their prompting procedures to align with individual children's strengths and weaknesses like we had to do in the study with children with autism. Regardless of adaptation of prompting, clinicians should continue to follow the general recommendation of using the most efficient and effective prompts possible. For children with autism and other developmental disabilities, this may involve providing longer wait times, limiting verbal prompts, and strategically fading visual prompts (Favot et al., 2018; Garcia et al., 2019; S. L. Gillam et al., 2015; Petersen et al., 2014). If interventions are delivered individually, more gradual least-to-most prompting (i.e., more than two steps) is easier to manage and may be needed for some children with autism to acquire storytelling skills (Whalon et al., 2019).

    Differentiate, Individualize, and Extend

    Small group narrative intervention is extremely versatile and can facilitate differentiation and individualization (Petersen et al., 2019; Spencer et al., 2018; Weddle et al., 2016) because so many different targets can be addressed within the context of a single story and in a single session. If the story is purposefully selected, clinicians can focus intervention for some children on basic targets such as using the story grammar elements while encouraging other children to practice advanced targets such as using more complex sentences or complex vocabulary. It is even possible to include children in the group who use augmentative and alternative communication (Soto et al., 2007) or whose targets involve receptive language. Likewise, the promotion of perspective taking (Dodd et al., 2011) and inferencing (Gardner & Spencer, 2016) can be targeted within a narrative intervention. It is important to set the demands of the intervention task (e.g., retelling, answering questions, manipulating pictures) according to each child's language targets (Westerveld & Gillon, 2010) and to use two-step prompting flexibly. School-based SLPs are frequently pressured to do more with less time. Small group narrative intervention is one way to accomplish that.

    Depending on the age and development of the children, clinicians can extend their interventions to promote grade-appropriate academic and social skills. This means that intervention should not end when children can tell or retell a story using relevant story structure. This is just the starting place that serves as a solid foundation for teaching other critical skills that children with language disabilities often lack (e.g., personal and fictional generation, writing, using complex vocabulary). Even the discourse and sentence level structures of expository language can be more easily addressed once the narrative language foundation has been established (Petersen et al., 2019). If students become bored during intervention sessions, this is a red flag that the clinician has not moved them along to more challenging tasks.

    Sometimes, clinicians can struggle with what to do next. Fortunately, there are a number of published programs that can be helpful in this regard. Story Champs (Spencer & Petersen, 2016) includes 57 different lesson plans with eight add-on lessons so that teachers and clinicians have a large variety of targets to choose from, including simple to complex oral and written narratives, oral and written exposition, vocabulary, and complex language. The Story Grammar Marker program has a number of variations that facilitate social emotional learning, text complexity and close reading, and writing (Moreau & Fidrych, 2008). SKILL has explicit procedures that move children from producing simple story grammar elements to more complex stories with academic language (S. L. Gillam et al., 2018). If a clinician decides not to use a published program, we recommend consulting the general education curriculum and academic standards used locally or visiting for helpful materials and grade-level academic sequences.

    Arrange for Generalization Opportunities

    Narratives are a great choice for language intervention because they have direct relevance to highly valued academic and social skills. To maximize the effect of intervention, we recommend intentionally planning for generalization opportunities. One of our favorite ways to ensure this happens is to involve classroom teachers in the intervention. For instance, SLPs can support teachers' implementation of narrative intervention, in large or small group arrangements. A class can be divided into multiple small groups, and the teacher and SLP can each focus on a group to work with at the same time. In one study examining a multitiered implementation of narrative intervention in preschools (Spencer et al., 2018), teachers and teaching assistants were both involved in delivering the intervention. Cimburek and Petersen (2017) examined a similar multitiered implementation for preschool and kindergarten students. General education teachers delivered large group narrative intervention while SLPs delivered small group narrative intervention to at-risk students. By engaging teachers in narrative intervention, they learn the complex language that they should be encouraging their students to use and are more likely to extend practice opportunities to other academic domains (e.g., social studies, writing). If teachers choose not to participate in the intervention directly, the SLP might try to provide intervention in the classroom (e.g., S. L. Gillam et al., 2014; Weddle et al., 2016) so that the teacher at least observes and hears what is being taught.

    Another option is for SLPs to create activities for students to practice narrative-related targets in other contexts and consult with teachers about how to facilitate their use. For example, a simple lesson extension can be written out that directs the teacher to have children “turn-n-talk” about what they did over the weekend. SLPs can also supply teachers with graphic organizers or concept maps displaying the icons or symbols from narrative intervention, making it easier for children to see how the skills taught in intervention can be used during classroom activities such as reading comprehension or writing. If special stories are used for initial training, then storybook reading is an excellent context for generalization, whether the books are read by an SLP, a teacher, or a parent. Family engagement activities can be sent home so that children have support retelling stories to family members. In a recent study, we supplied teachers with posters of each of the vocabulary words targeted during intervention (Spencer et al., 2020). The purpose was to help the teacher and all the students be on the lookout for those words and their synonyms, which increases word consciousness (McKeown et al., 2017). Resourceful SLPs can use their creativity to craft activities that promote the generalized and spontaneous use of narrative-based language targets. The important thing is that generalization is not left to chance. Planning for it will hasten the acquisition and maintenance of trained skills, as well as promote their use in meaningful contexts (Alessi, 1987; Fey, 1988).

    Make It Fun!

    Simply put, people talk because they get attention from others when they do. Other benefits may be associated with specific requests (e.g., asking for a drink), but as humans, we are primarily motivated by social consequences. When engaged in narrative intervention, SLPs should keep in mind that the natural consequence of telling a story is attention and approval from a listener. As clinicians strategically use their attention to reinforce storytelling and make the interaction enjoyable, children will be willing to try challenging things. It is helpful to choose published narrative intervention programs that include games, materials, or activities that appeal to children, but it is dangerous to rely too heavily on them. If children are reticent, bored, or complain, then it is the SLP's responsibility to facilitate engagement. Because children naturally want to talk about themselves and their experiences, narratives are the perfect context for practicing language without it feeling like work.

    We have a number of suggestions for how to make narrative intervention fun. First, SLPs can increase the value of their attention by being animated when listening. Use overexaggerated facial expressions and comments (e.g., “No way!”) to show children you are listening carefully. Recasting is an effective way to show approval while having the added benefit of modeling complex language and expanding the child's utterance (Cleave et al., 2015). Second, teach other children to be active, engaged listeners and praise them when they show respect for the storyteller. Peers can be explicitly taught how to paraphrase a peer's story briefly. SLPs can model paraphrasing and encourage the other children to do it after listening to a story. Third, play games and use competition but do not eliminate attention and approval. It is okay to use consequences other than your attention (e.g., points, getting to color in a square on a chart, or moving a game piece), but although they may be needed temporarily for training purposes, attention will be necessary to maintain storytelling skills outside the intervention context. Fourth, take advantage of peer mediation. Putting children together in teams or pairs can increase motivation for the storytelling tasks. Children often work to impress their friends and may accept help from peers better than from adults. Finally, engage children in role plays, using props or puppets, or movement related to the stories. Recent research suggests that combining movement with storytelling enhances children's language outcomes (Brinton & Fujiki, 2017; Culatta et al., 2010; Duncan et al., 2019).


    Our purpose was to provide an overview of the academic and social value of narratives and to offer principle-driven guidance based on research and our own clinical experience implementing narrative intervention. Narratives are an extremely versatile conduit for learning a large number of academic and socially meaningful skills. Narrative intervention can be used with children as young as 3 years old, all the way through high school. Research has documented its effectiveness for typically developing children, those who are at risk, children with language disabilities, children with autism, and English language learners. While there are a handful of published narrative intervention programs available for clinical use, narrative interventions can be developed using materials that most SLPs have readily available. When narrative intervention is implemented following the principles of intervention outlined in this article, SLPs can have confidence that their approach is supported by research and extensive clinical experience. It is our hope that every school-based SLP can harness the power of stories in their clinical practice.


    Author Notes

    Disclosure: The authors receive financial benefits from the sale of Story Champs, which is one of the published narrative interventions mentioned in this article.

    Correspondence to Trina D. Spencer:

    Editor-in-Chief: Holly L. Storkel

    Editor: Katie Squires

    Additional Resources