Supporting Generalisation (Part 1): Challenges and Solutions

When we see our students for intervention, we all hope that they will make progress not just with the targets we set them during our sessions, but also that the linguistic skills and knowledge taught ultimately help them become better communicators in real-life contexts (Paul  and Norbury, 2012). Any less than this and our job could be said to be incomplete (Nelson, 1998). Yet, achieving “generalisation” or “carryover” has long been considered one of the harder parts of a Speech and Language Therapist’s (SLT’s) job.

Criticisms of the “pull-out” model

In recent years, the traditional “pull-out” model of therapy, where students are removed from their lessons and taken into a separate room, has been heavily critiqued for not supporting generalisation. Some have argued that in creating an artificial context, skills taught within sessions may not naturally transfer to real communicative situations and have limited relevance to students’ lives (Nelson, 1998). Indeed, research has found that children do not always generalise what they learn in this way to spontaneous conversation (Paul and Norbury,  2012).

As an SLT working privately in schools as part of a “buy in” service, much of the support that I provide is within this context. The criticism notwithstanding, there are many benefits to this way of working. A separate room can provide a quiet, nurturing, distraction-free space for those with  attention difficulties. It also allows the teaching of basic skills, tools and strategies, as well as more attention, talking time and scaffolding (Paul and Norbury, 2012), than would not be possible in a busy classroom environment. Older students may appreciate the freedom to practice new skills without the pressure of doing it in front of peers (Nelson, 1998).

However, it is important to remain conscious of the drawbacks, and to remember that generalisation is unlikely to occur automatically. According to Paul and Norbury (2012), it needs to be consciously built into any intervention program rather than considered an afterthought. They argue that it should form a part of every session; if not every activity, then at least every goal.

If your role involves any direct work with students, then you could try some of the following approaches to ensure that what you teach “carries over”: using more “natural”, contextually based activities and materials, involving teaching staff, supporting metacognitive skills, teaching strategies as well as using a mediated learning approach and bridging questions.

Using “natural”, contextually-relevant activities and authentic materials

Whilst the “pull-out” room has been criticized for being contrived and “unnatural”, the intervention activities we choose to introduce within that environment may be more or less so. Activities can be placed along a “continuum of naturalness” (Fey, 1986, cited in Paul and Norbury, 2012, p.68), according to how similar they are to those our students normally engage in. So, asking them to complete a specially designed connectives worksheet might be considered less natural, whilst supporting them to integrate said connectives into a real class assignment might be considered more so.

Interestingly, research has shown that certain approaches which fall toward the less natural end of the scale such as drilling can be highly efficient at getting students to learn new language forms. No natural approach can elicit so many correct responses in such a short space of time (Paul and Norbury, 2012). But, like the pull-out room, the problem comes with transferring these gains to real-life communication. Paul and Norbury (2012) recommend a combined approach of more “clinician-directed” activities during the early stages of treatment, with more naturalistic ones either simultaneously or afterwards to encourage “carry over”.

With younger children, this might take the form of child-led play, or informal conversation extended by an adult. With older students of secondary school age, we can incorporate materials they are likely to encounter at other times of day into our sessions. So, as hinted earlier, we might use their textbooks, lecture slides or curriculum vocabulary lists rather than specially designed activities (Costello, 1983 in Paul and Norbury, 1998, p.86). Nelson (1998) also recommended balancing “pulling-out” students from lessons with “pulling in” curriculum content and materials. By doing so, our students immediately see the relevance of what we are teaching to their daily lives.

One idea already commonly used by SLTs naturally supports generalisation: pre-teaching curriculum vocabulary. To truly know a word, we must encounter it multiple times across different contexts. When you teach your student curriculum terms such as “tectonic plate”, “constructive” and “destructive margins” you ensure that they will hear them in at least two different contexts: their Speech and Language Therapy (SALT) session, and their geography lesson.

Paul and Norbury (2012) also recommend “priming”, by previewing materials and exercises students will go on to encounter in their lessons. They argue that since our students already know the right answers, this allows them to “look smart” in front of other students, boosting their self-esteem.

However, Larson and Mckinley (2003) caution us not to fall into the trap of simply tutoring our students the curricular material, which they claim is only a short-term solution to underlying language difficulties. Instead, we should aim to use these materials as a vehicle to teach specific language-based skills and strategies. So, when teaching a strategy like morphological analysis, where I teach students how to break complex words apart to work out their meanings, I often include some curriculum vocab words from different topics to practice on.

Say I have taught the prefix “omni” and the root words “potent” and “bene”, then students may be able to work out the meanings of their new R.E. words “omnipotent” and “omnibenevolent” as well as perhaps extending their reasoning to more everyday words like “beneficial”, “impotent” and “potential”. Teaching the prefixes “endo” and “exo” as well as the root “thermo”, may help them to work out “endo-“ and “exothermic reaction”, “thermal imaging” and perhaps “exoskeleton”. Or, if we are working on figurative language, then I might draw similes and metaphors from their curriculum text,  A Christmas Carol, and have us work out the meanings together.

Of course, sometimes curriculum materials may initially be too complex for lower level students. In these cases, you may need to start with more basic, specifically designed activities. For instance, I sometimes use simple inference worksheets from websites like Super Teacher. But where possible, I try to combine them with more authentic, contextually relevant tasks, such as getting them to write short PEE paragraphs (Point, Evidence, Explanation) – an acronym commonly used in secondary school English lessons to support extended writing skills – to answer the inference questions about the scenarios.

This way, my students immediately see how their developing inference skills directly support their ability to analyse and write in English lessons. Once they succeed with these simpler tasks, they can progress to applying these skills to their actual curriculum texts. Alternatively, if you have spent time teaching students to write relative clauses, then you could try spotting them in newspapers together.

Involving teaching staff

Using a pull-out model does not preclude the possibility of working more collaboratively. When I worked for the NHS, we would invite our students’ Learning Support Assistants (LSAs) into our sessions as a matter of course. Doing so allows you to share targets, model facilitation strategies and pass on ideas, hopefully allowing your work to be reinforced outside of SALT sessions, thereby supporting generalisation.

It can also give LSAs an insight into a new student’s difficulties, or conversely, their abilities. In cases of learned helplessness, some students become so reliant on support staff, believing themselves to be incapable that they effectively give up and stop trying. I once worked with a student at a Further Education (FE) college whose LSA attended our sessions. After watching him work, she commented that she had no idea he was so capable.

Supporting your students to succeed in a quiet, low-pressure environment may help to change others’ perceptions of their limits. When an LSA sees what a student can achieve with the right scaffolding, it might lead to a shift in expectations in the classroom, potentially leading to gentle encouragement to aim higher.

What next?

If we want our students to become more independent, we have to move beyond teaching basic language skills. In Part 2 of this series of posts, I discuss using a learning strategies approach and supporting meta skills to aid generalisation and independence. Click here: https://secondaryschoolslt.com/2026/08/12/supporting-generalisation-part-2-teaching-strategies/

Notes

Larson, V.L. and McKinley, N.L. (2003) Communication solutions for older students. Thinking Pub. 

Nelson, N. W. (1998) Childhood Language Disorders in Context: Infancy Through Adolescence. 2nd edn. Boston, MA: Allyn & Bacon.

Paul, R. and Norbury, C. (2012) Language Disorders from Infancy Through Adolescence: Listening, Speaking, Reading, Writing, and Communicating. 4th edn. St. Louis, MO: Elsevier.