Who can help me with writing this Compare and 6 Research: on RESEARCH REASONING

This is a repository copy of Increasing adolescents' depth of understanding of cross-curriculum words: an intervention study..

White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/109137/

Version: Accepted Version

Article:

Spencer, S., Clegg, J., Lowe, H. et al. (1 more author) (2017) Increasing adolescents' depth of understanding of cross-curriculum words: an intervention study. International Journal of Language and Communication Disorders. ISSN 1368-2822

https://doi.org/10.1111/1460-6984.12309

[email protected] https://eprints.whiterose.ac.uk/

Reuse

Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item.

Takedown

If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request.

1

Increasing adolescents' depth of understanding of cross-curriculum words: an

intervention study.

Abstract:

Background: There is some evidence that vocabulary intervention is effective for children,

although further research is needed to confirm this within contexts of social disadvantage.

Very little is known about the effectiveness of interventions to increase adolescent knowledge

of cross-curriculum words.

Aims: This study evaluated the effectiveness of an intervention programme designed to

develop adolescents’ knowledge of cross-curriculum words.

Methods and Procedures: Participants were 35 adolescents aged between 12 and 14 years

who were at risk of educational underachievement with low scores on a range of assessments.

Participants received a ten-week intervention programme in small groups, targeting 10 cross-

curriculum words (e.g. ‘summarise’). This was evaluated using a bespoke outcome measure

(the Word Knowledge Profile). The study involved an AABA design, with a repeated

baseline, delayed intervention cohort and blind assessment. Intervention included both

semantic and phonological information about the target words and involved the adolescents

using the words in multiple contexts.

Outcomes and Results: Results were promising and participants’ knowledge of the targeted

words significantly increased following intervention. Progress was demonstrated on the Word

Knowledge Profile on the item requiring participants to define the word (for the summer

intervention group only). These increases in depth of knowledge were seen on taught words

but not on matched non-taught words.

2

Conclusions and Implications: Cross-curriculum words are not consistently understood by

adolescents at risk of low educational attainment within a low socioeconomic context. A 10

week intervention programme resulted in some increases to the depth of knowledge of

targeted cross-curriculum words.

INTRODUCTION

Importance of cross-curriculum vocabulary learning

Vocabulary skills are important to classroom learning because of a) the complex and abstract

words used in the curriculum content, b) the need to use words as tools to access learning and

facilitate ‘academic thinking’, c) the role of vocabulary in processing new disciplinary

concepts and ideas, and d) the vocabulary used by teachers during oral pedagogy in

conveying information (Alexander 2006; Nagy and Townsend 2012). There is evidence for

an association between vocabulary knowledge and academic attainment. Vocabulary

assessment scores accounted for variation in academic attainment in mathematics and English

upon leaving school (Spencer et al. 2016). Knowledge of cross-curriculum words accounted

for considerable amount of variation in academic attainment on tests of reading, mathematics,

social sciences and science in a cohort of 339 12-14 year old children (Townsend et al. 2012).

Insufficient knowledge of cross-curriculum vocabulary has been associated with the

attainment gap between different groups of students based on socioeconomic background

(Gardner and Davies 2014).

Beck, McKeown and Kucan (2002) outline the importance of vocabulary knowledge for

learning, in particular cross-curriculum vocabulary (also known as Tier 2 words or general

academic words), which are abstract and occur in multiple contexts, for example coincidence,

3

industrious, fortunate, introduce. While more low-frequency subject-specific words may be

specifically taught in the classroom (e.g. peninsula, isotope, stanza), these cross-curriculum

words are used across topics in schools but may not be taught explicitly (Beck, McKeown

and Kucan 2002; Justice et al. 2014). Vocabulary interventions should target these cross-

curriculum words, given their potential impact on success within the classroom (Gregor

2009; Justice et al. 2014).

Principles of teaching cross-curriculum vocabulary

Evidence for the effectiveness of vocabulary intervention is well established during the

preschool and primary school years (Best, Dockrell and Braisby 2006; Lubliner and Smetena

2005; Justice et al. 2014; Marulis and Neuman 2010; Steele and Mills 2011) with emerging

evidence for secondary school aged children, particularly using whole-class approaches

(Lesaux et al. 2010; Snow et al. 2009). These studies typically involve children acquiring

relevant information about different aspects of a word (including semantic, phonological,

morphological, grammatical and orthographic) in order to establish clear lexical

representations (Stackhouse and Wells 2001). The rationale for this is that when children do

not develop phonological (information about the sound structure of a word) and/or semantic

knowledge (information about meaning, function and relationships with other words) about

new words, inadequate representations of words are stored, resulting in impoverished

vocabulary or difficulties with accessing and retrieving words in the lexicon quickly and

accurately (McGregor et al., 2002).

Studies have also investigated principles for teaching academic vocabulary words including

cross-curriculum words (see Nagy and Townsend 2012 for a review). Principles of rich

vocabulary instruction (Beck, McKeown and Kucan 2002) are used to support in-depth

knowledge of highly functional words. Principles include providing direct and explicit

4

definitions and attributes of words to be learned; promoting depth of processing,

opportunities for repeated exposure and use of the target word, encouraging learning of words

across contexts, and supporting children to use words in varied sentences (Stahl and

Fairbanks 1986). New word learning must happen in relation to authentic contexts, with

multiple opportunities to explore links with other words (Nagy and Townsend 2012).

Teaching word learning strategies and word consciousness is also an important component of

rich vocabulary interventions (Graves 2006). Learning a new cross-curriculum word is an

incremental process, with new knowledge deepening with multiple exposures and multiple

opportunities to use a word in new contexts (Lesaux et al. 2010).

There is a growing recognition that vocabulary interventions should promote the deep

understanding of a relatively small number of words, their elements and semantically and

morphologically related words in rich contexts (Lesaux et al. 2010: 45; Graves 2006). Depth

of processing moves from: 1) association, where a new word is learned in relation to a

definition or single context, to 2) comprehension in which a child demonstrates

understanding in a sentence or utterance, or where a child puts definitional information to

use, for example by finding an antonym; to 3) generation, in which the child produces a novel

response to a word such as an original sentence, their own definition or applying the word to

a new context (Stahl and Fairbanks 1986).

Interventions for adolescent vocabulary skills in contexts of socioeconomic disadvantage

In areas associated with social disadvantage the prevalence of language difficulties is higher

in children attending pre-school and primary school (Department for Education (DfE) 2012;

Law, McBean and Rush 2011) and secondary school (Spencer, Clegg and Stackhouse 2012)

when compared to non-disadvantaged areas and vocabulary difficulties are a particular

feature of this (Farkas and Beron 2004). In response, emerging evidence shows that older

5

children in contexts of social disadvantage can successfully learn new words through direct

intervention (Joffe 2006; 2011). Snow, Lawrence and White’s (2009) Word Generation

Project in the USA embedded instruction of five words per week across the school

curriculum with 6th – 8th grade students (aged 11-14 years) in a 24 week whole-school

programme. Five schools implemented the programme, with pre- and post-test data available

for 697 students. These data were compared to that from 319 students from three schools that

did not implement the programme. The words targeted were ‘all-purpose academic words’

(Snow et al. 2009, p 326), defined as words that occur in a range of subjects and hence have

maximum functionality across the curriculum. Adolescents in participating schools made

accelerated progress in vocabulary knowledge with higher levels of educational attainment

when compared to those in non-participating schools (Snow et al. 2009).

Similarly, Lesaux et al. (2010) delivered an 18 week vocabulary intervention programme in

middle schools in the USA. Teachers delivered the intervention to 296 children aged 11-12

years in 13 treatment classrooms, and compared outcomes with 180 children in 8 control

classrooms. Only 24% of participants spoke only English at home. The intervention targeted

75 cross-curriculum words in whole-class intervention sessions which focused on building

depth of vocabulary knowledge via multiple meanings and morphological analysis. The study

found significant gains on a researcher-designed measure of comprehension of targeted

words and a morphological decomposition task for the intervention group but not the control

group. There were no changes to standardised measures of reading vocabulary

comprehension.

With both of these intervention studies, teachers delivered the content of the intervention in

the classroom. This has a number of advantages. As all adolescents within the class are

targeted, selection criteria and scheduling intervention sessions are not an issue. Furthermore,

the intervention embeds vocabulary instruction within the whole school curriculum, allowing

6

for repeated and consistent exposure to the target words. However, some schools may be

unable or unwilling to implement whole-school vocabulary programmes, particularly in low

socioeconomic contexts where there is increased risk of low quality schooling (Lupton 2005).

Evaluating cross-curriculum vocabulary interventions

Despite an emphasis on rich vocabulary instruction and supporting deep knowledge of taught

words, many intervention studies use outcome measures which access a relatively surface

level of word processing. For example, changes to knowledge of targeted words are

frequently assessed using multiple choice tasks, by asking participants to link targeted words

to definitions (Snow et al. 2009) or synonyms (Lesaux et al. 2010). Such tasks are suited to

being administered to large numbers of participants in short periods of time but they are

limited in that performance is likely to be mediated by literacy skills and changes to depth of

word knowledge cannot be measured over time. In contrast, Townsend and Collins (2009)

used a bespoke outcome measure in their vocabulary intervention study with 37 children who

were English language learners. Participants were asked if they had seen or heard the word

before, and then asked if they thought they knew what the word meant and to provide a

definition and an example of the word used in a sentence. This type of self-evaluation of

word knowledge has been put forward as a means of examining the incremental nature of

building word knowledge (Nagy and Scott 2000; McKeown and Beck 2004). Many

vocabulary intervention studies use bespoke assessments featuring targeted words as outcome

measures rather than standardised vocabulary assessments. This is because published

vocabulary assessments measure knowledge of general words and so are unlikely to change

following intervention focused on specific targeted words. Such assessments are also unable

to show changes to depth of knowledge of words over time. In addition, standardised

vocabulary assessment is most open to item bias on testing due to vocabulary reflecting

7

individual experiences, the language of the home and familiarity with school curricula

(Stockman 2000; Townsend and Collins 2009).

Research Aims

This study aimed to evaluate the effectiveness of an intervention programme designed to

increase adolescents’ depth of knowledge of cross-curriculum words. It addressed the

following research question: Is an intervention programme effective in increasing

adolescents’ knowledge of cross-curriculum words within the context of high risk of

educational underachievement and social disadvantage?

METHOD

Design

This study was a single-blind cross-over trial with random allocation to groups and blind

assessment to group status. The study was carried out during one school year at a mainstream

secondary school. In collaboration with the school, it was agreed that small-group

intervention led by the research team (rather than teaching staff) was best suited for

developing the intervention programme and for the priorities of the school at the time. This

was due to a perceived lack of staff time, resulting from necessary responses to the recent

Ofsted report (see below).

The study used a delayed intervention design, comprising an intervention group (Spring

Group) and a waiting control group who went on to receive intervention (Summer Group). A

repeated baseline of word knowledge was taken pre-intervention for both the Spring and

Summer Groups. This allowed examination of the stability of knowledge of targeted and

matched non-taught words over time, without intervention. A post-intervention assessment

was carried out for both groups, and a delayed follow-up assessment of word knowledge was

8

also carried out approximately three months later for the Spring Group. Thus, there were

four assessment time points across the study (see Table 1), with all participants being

assessed pre-intervention, immediately before intervention, and immediately after

intervention. The Spring Group was also assessed approximately three months after

intervention ended to examine maintenance of progress.

Insert Table 1 around here.

Participants

Participants attended one mainstream secondary school in an urban area of social

disadvantage. The school’s catchment area was ranked in the bottom 5% of England’s wards

in terms of socioeconomic status, using the Indices of Deprivation (McLennan et al. 2011).

These indices rank England’s 32,482 super-ordinate areas in terms of seven domains

(income, employment, health and disability, education training and skills, barriers to housing

and services, crime, and environment). In 2013, 46% of students in this participating school

achieved five or more A* to C grades including English and mathematics in their General

Certificate of Secondary Education (GCSE) exams upon leaving secondary school. Many

options after compulsory schooling have entry requirements including obtaining five or more

GCSE grades at A*-C level including mathematics and English. This is an important

benchmark for school evaluation data as schools with fewer than 40% of pupils achieving at

this level are considered to be underperforming and in need of improvement. The school was

judged as performing inadequately by the Office for Standards in Education, Children’s

Services and Skills (Ofsted) and was in the 'special measures' category during the project,

meaning it was undergoing an intensive programme of school improvement.

Participant criteria were agreed in collaboration with staff at the school, as follows:

9

 In school years 8 or 9 (aged between 12 and 14 years).

 In low ability classes when curriculum subjects were ability-streamed.

 Not making expected progress in these low ability classes as determined by teacher

informal assessment and judgement.

 No statements of special educational needs, indicating no identified significant

learning needs.

 English is a first language or were bilingual but had attended all schooling in an

English speaking context.

 Not receiving existing interventions or other support within the school.

 Not currently on the caseload of speech and language therapy services.

Parental and pupil consent was obtained for 44 participants. The Spring Group initially

comprised 21 participants and the Summer Group 23 participants (this difference in number

was due to timetabling differences across Year 8 and Year 9 participants). Nine participants

left the study following the first assessment: three participants left the school, four

participants were unable to take part because of timetabling clashes, and two were school

non-attenders. This left 19 participants in the Spring Group, and 16 in the Summer Group. No

participants chose to withdraw from the study partway through the intervention.

Demographic information for all participants is summarised in Table 2, along with the mean

number of intervention sessions attended.

Insert Table 2 around here.

Independent t-tests showed that the Spring and Summer groups did not differ significantly on

age (t(42)=-1.398, p=.169) or the number of intervention sessions attended (t(33) = 2.023,

p=.051). Table 2 also shows how many participants in each group had disruptive behaviour,

as measured by the number of participants who needed referral to the school’s behaviour

10

support team during at least one of the sessions (9 in the Spring Group and 4 in the Summer

Group).

Participants completed an initial assessment session to profile their overall language skills

(see Appendix 1). In addition, the research team had access to the results of the nationally

used educational assessment Cognitive Abilities Test (CAT) which participants completed on

entry into secondary school at 11 years of age. The CAT is used as a baseline from which to

predict pupil performance in national examinations and to inform target setting. The verbal

subtest (CATV) includes tests of receptive vocabulary, sentence completion, verbal

classification, and verbal analogies which are presented in a written form. The quantitative

subtest (CATQ) assesses reasoning ability with numbers, and the nonverbal subtest

(CATNV) assesses the ability to think and reason with non-verbal material such as shapes

and designs. Standardized scores are calculated with a mean of 100 and SD of 15. Table 3

shows that participants scored significantly below the normative mean on all language and

CAT subtests. Spring and Summer intervention groups were matched on all measures of

language; independent t-tests showed that the Spring Group and the Summer Group did not

differ significantly on any of the assessment information.

Insert Table 3 around here.

Materials and measures

Cross-curriculum vocabulary selected for intervention

The intervention sessions targeted knowledge of 10 cross-curriculum words. These are

abstract words that occur in multiple contexts, referred to as Tier 2 words according to the

11

tiered system of Beck, McKeown and Kucan (2002). The Academic Word List (Gardner and

Davies 2014) was used to select words for inclusion in the study, in collaboration with

teaching staff. The resulting list consisted of 10 highly functional, cross-curriculum words

(all verbs) in use across the academic subjects taught in schools.

These 10 taught words were matched with 10 non-taught words, which were not taught

during the intervention programme (See Appendix 2). Matching was based on:

• Number of syllables, stress pattern, and number of phonemes, using the Medical Research

Council (MRC) Psycholinguistic Database (Wilson 1998). This database generates lists of

words according to criteria input by the user.

• Measures of frequency from The British National Corpus (BNC) (2007). The BNC is a

100 million-word collection of samples of written and spoken language from a wide range

of sources, designed to represent a wide cross-section of current British English.

Word Knowledge Profile

The bespoke outcome measure, the Word Knowledge Profile, examined depth of word

knowledge of the ten taught words and the ten non-taught words. This was developed to

capture change to depth of word knowledge over time, measuring participants’ phonological

and semantic knowledge about the target and control words. For each of the 20 words,

participants were asked to: repeat the word, generate a rhyming word or non-word, rate their

own knowledge of the word, provide a verbal definition of the word (resulting in the Word

Meaning Score, divided into taught words and non-taught words), to use the word in a

sentence, and relate the word to their own experiences by first saying where or when they

might hear the word, and then being asked when they would perform the word, for example

‘when might you consult something?’. Participants were also asked to spell six of the words:

12

interpret, convince, consult, generate, summarise and establish (due to time constraints it was

not possible to ask participants to spell all of the words). Thus eight components were

assessed for each word, in the order outlined above (see Appendix 4). For each word, each of

the eight components were scored correct/incorrect. A copy of the Word Knowledge Profile

sheet and examples of correct and incorrect responses are provided on the project website

(Spencer 2016): https://adolescentvocabulary.wordpress.com/resources/how-to-evaluate-new-

word-learning/.

Correlations between the standardised language assessments and the Word Knowledge

Profile suggested that the measure had validity as a measure of vocabulary ability. For

example, Pearson’s product-moment correlation coefficient showed significant correlations

between the baseline Total Word Meaning Score on the Word Knowledge Profile the BPVS,

(r = 0.571, p= <0.001), CELF4UK Formulated Sentences (r = 0.493, p= 0.001) and CATV (r

= 0.511, p= <0.001) but not with the spoonerisms task from the PhAB (r = 0.209, p= 0.173),

the CATNV or CATQ (r = 0.241, p= 0.128; r = 0.239, p= 0.119) nor the CELF4UK

Recalling Sentences (r = 0.263, p= 0.084).

Intervention programme

A ten-week programme was designed, consisting of once weekly one hour-long word

learning intervention session carried out in small groups of 3-5 participants within the school

setting. Each session targeted the knowledge and use of one word. The intervention

programme incorporated principles of word-learning that have been successfully applied in

other studies, as outlined in Appendix 3 (Beck, McKeown and Kucan 2002; Graves 2006;

Justice et al. 2014; Lesaux et al. 2010; Marulis and Neuman 2010; Nagy and Scott 2000;

Snow, Lawrence and White 2009; Stahl and Fairbanks 1986). The intervention was guided by

principles of rich instruction to increase depth of knowledge of the targeted words, with an

emphasis of providing multiple opportunities to use the words in multiple contexts, engaging

13

the participants in personalising word meaning and in applying strategies for learning new

words, including morphological analysis. All session plans are freely available on a project

website (Spencer 2016): https://adolescentvocabulary.wordpress.com/example-word-

learning-session-plans/.

Each session followed a broadly similar structure, providing multiple opportunities to activate

components of the speech processing system (Stackhouse and Wells 2001). In every session

each participant was supported to think about what the target word means, what it sounds

like, to say the word aloud, to use it in a sentence and to write it down during a concept

mapping activity. Participants were supported to use suffixes and prefixes to build derivatives

of the word. Each session included an experiential activity (e.g. where they evaluated,

summarised, interpreted etc) and a main topic activity which gave them lots of opportunities

to use the word in multiple contexts. Each week, participants were given a challenge card

which reminded them to look out for the target word across the week and to use it if possible.

The next session began with participants reporting back on this challenge.

Procedure

The study received ethical approval from the Research Ethics Committee, Human

Communication Sciences, at the University of Sheffield.

1. Assessment Procedure

Each participant was seen individually for assessment in a quiet room within the school.

Assessments were digitally recorded for later analysis and transcription. Two researchers

(both qualified speech and language therapists) administered assessments at the first

assessment point. The first assessment session lasted one hour and included the battery of

language assessments outlined in Appendix 1 as well as the Word Knowledge profile.

14

Participants could request breaks at any time during the assessment session. The Word

Knowledge Profile took between 10 and 20 minutes to complete for each participant. Post-

intervention Word Knowledge Profiles were carried out by qualified speech and language

therapists (not the researchers) and trained speech and language therapy students who were

blind to both the group status of the participants (i.e. those who had received intervention

versus the waiting control) and to the treatment status of the words (taught words versus non-

taught words). At each assessment point, two members of the research team marked 25% of

the Word Knowledge Profiles using written transcripts resulting in a 92-96% rate of

agreement overall. Any disagreements were discussed and a decision was made on the correct

response.

2. Intervention Procedure

Intervention sessions were delivered to small groups of 3-5 participants within the school

setting. Due to timetabling, participants received intervention with peers from the same

school year group. Random allocation was achieved by dividing participants into school year

groups, each participant was assigned a random number and then the anonymised numbers

were divided into two groups using Windows Excel. Within intervention groups, participants

were assigned to a small peer group for intervention sessions according to their school

weekly timetable, so that core subjects were not missed.

The first session took the form of a group discussion in order to engage all participants in the

learning process by exploring the concept of word-learning and its importance. Following the

first discussion session, the Spring Group received nine weekly one-hour intervention

sessions. Due to timetabling constraints, the Summer Group received eight intervention

sessions, thus receiving intervention for eight words. The mean number of sessions which

participants actually attended was 7.06 (SD 1.2, range 4-9); Spring Group attended a mean of

15

7.42 sessions (SD 1.1, range 5-9) while the Summer Group attended a mean of 6.63 sessions

(SD 1.2, range 4-8).

Intervention was delivered by two of the research team (SS and HL), both qualified and

experienced speech and language therapists. Treatment fidelity was ensured by designing a

lesson plan for each of the ten sessions. The researchers developed the intervention

programme together and met weekly to discuss implementation and ensure consistency

across groups.

Analysis

All quantitative data was inputted into SPSS 19 (IBM 2010) for analysis. A series of paired-

sample t-tests were conducted to evaluate the impact of intervention on different items of the

Word Knowledge Profile, as has been done in previous language intervention studies with

adolescents (e.g. Ebbels et al. 2014). Bonferroni corrections were used to avoid type I error.

Repeated measures ANOVA was not suitable because the intervention was delivered in small

groups and therefore participants’ progress may have been affected by peers’ progress.

RESULTS

Knowledge of cross-curriculum words at baseline

For the whole cohort, at baseline 1) repetition of words was a relative strength, with a mean

of 19/20 (95%) correct (range 14-20); 2) participants were able to describe the meaning of a

mean of 5.2/20 (26%) words (range 0-12), as indicated by the Total Word Meaning Score:

(combining both taught and un-taught words), as well as by participants’ own evaluation of

their word knowledge as ‘green’ (4.3/20); 3) participants were able to provide a rhyme for a

16

mean of 9.64/20 words (48%, range 0-20), with explicit instruction and the acceptance of

non-words.

At baseline, the greatest number of participants correctly reported some meaning for the word

‘exclude’ (81.8%) followed by ‘convince’ (79.5%). No participants gave a correctly scored

response to the words ‘infer’, ‘conglomerate’, ‘incur’, ‘distinguish’ nor ‘liberate’. No answers

included multiple contexts in the definitions. Furthermore, often one context was provided

across the cohorts’ responses. For example, ‘exclude’ was defined only in relation to school

exclusions for behavioural incidents and submit was defined in relation to computer tasks.

Tables 4 and 5 show the mean scores on each component of the Word Knowledge Profile,

across four assessment times for the Spring and Summer groups.

Insert Tables 4 and 5 around here.

Evaluation of the intervention

Repeated Baseline Phase – stability of outcome measure

The results of the Word Knowledge Profile for both taught and non-taught words across the

four time-points in the study are outlined in Table 4 (Spring Group) and Table 5 (Summer

Group).

The results across the repeated baseline in the study showed that there were no significant

differences at Time 1 or Time 2 baseline measures for either intervention groups on any item

on the Word Knowledge Profile. For example, t-tests on the Word Meaning component for

taught words show both the Spring Group (t (18) = -1.61, p = .13) and Summer Group (t (12)

= .49, p = .64) made no progress during the autumn term over the repeated baseline.

Spring group

17

A series of paired samples t-test compared results on the Word Knowledge Profile at Time 2

and Time 3. This showed that the Spring Group made progress on the Word Meaning score

for taught words over the intervention phase (t (18) = -2.28, p = .035). Although this was

approaching significance, it was not significant when Bonferroni corrections were applied.

The Summer group, the waiting control group who did not receive intervention, made no

progress on the Word Meaning score for taught words (t (15) = .25, p = .81). Figure 1

compares the progress made on the Word Meaning score for taught words over the spring

term for the Spring and Summer groups.

Insert Figure 1 about here

However, the Spring Group made progress on the two other items of the Word Knowledge

Profile over the intervention phase:

1) Their ability to give a context when they might use the word increased from a mean of

4.95/20 to 10.74/20 (t (18) = -5.48, p = <.001). The eta squared statistic (2.50) indicated a

large effect size. The score for taught words increased from 2.53 (SD 1.74) at the start of the

intervention phase to 6.37 (SD 3.37) post-intervention (t (18) = -5.83, p = <.001). The score

for non-taught words did not significantly increase over the same period.

The Summer group also made progress on their ability to give a context when they might use

the word (t (15) = -4.36, p = .002), though this did not reach significance when Bonferroni

corrections were applied. However, the Spring Group did not make more progress on this

item than the Summer Group (t (30) = -1.74, p = .092), suggesting that progress may not have

resulted from attending the intervention.

2) Their ability to give an example of when they enacted the word (e.g. tell me about a time

when you evaluated something) increased from a mean of 4.47/20 to 7.32/20 (t (18) = -4.53,

18

p = <.001). The eta squared statistic (0.53) indicated a large effect size. The score for taught

words increased from 2.11 (SD 1.94) at the start of the intervention phase to 4.21 (SD 3.05)

post-intervention (t (18) = -4.41, p = <.001). The score for non-taught words did not

significantly increase over the same period.

The Summer group also made progress on their ability to an example of when they enacted

the word (e.g. tell me about a time when you evaluated something) but these did not reach

significance once Bonferroni corrections were applied (t (15) = -3.30, p = .008). However,

the Spring Group did not make more progress on this item than the Summer Group (t (30) =

1.90, p = <.001), suggesting that progress may not have resulted from attending the

intervention.

Summer Group

A series of paired samples t-test compared results on the Word Knowledge Profile at Time 3

and Time 4. A paired samples t-test showed that the Summer Group made significant

progress on the Words Meaning score for taught words over their intervention phase with the

Words Meaning score for taught words increasing from 2.69 (SD 1.96) at the start of the

intervention phase to 4.38 (SD 2.13) post-intervention (t (15) = -5.40, p = <.001). The eta

squared statistic (0.66) indicated a large effect size. The Spring Group, who did not receive

intervention in this period, maintained progress following their spring intervention but did not

make any further progress (t (15) = .37, p = .72). The Summer group also made more

significant progress than the Spring group during this intervention period (t (30) = -4.92, p =

<.001), and suggesting that progress on this item resulted from attending the intervention

programme. Figure 2 compares the progress made on the Word Meaning score for taught

words over the summer intervention period for the Spring and Summer intervention groups.

Insert Figure 2 about here

19

The significance of progress on the Word Meaning item when both intervention groups are

combined

To analyse the overall effectiveness of the intervention programme, one-sample t-tests were

conducted to see if progress was significantly higher than 0 across the intervention period.

This proved to be the case for the Word Meanings for taught words following intervention

(Mean 1.17; SD 1.40; t (34) = 4.94, p = <.001), with an eta squared effect size of 0.42, but

not for the Word Meaning score for non-taught words (Mean 0.03; SD 1.04; t (34) = .16, p

=.87). Figure 3 compares the progress made on taught versus non-taught words for both

intervention groups.

Insert Figure 3 about here

Figure 4 shows the distribution of gains made in participants’ knowledge of taught words

following intervention. This chart presents the increase to the Word Meaning score for taught

words, comparing pre- and post-intervention results. It shows that five participants were able

to describe the meaning of fewer words following intervention (of note, all five participants

needed referral to the school’s behaviour support team during the intervention sessions).

Most participants were able to describe the meaning of two extra words following

intervention (n=12).

Insert Figure 4 about here

DISCUSSION

This study designed, delivered and evaluated a ten week intervention programme to increase

adolescents’ knowledge of cross-curriculum words, delivered in small groups. Results were

promising and showed that participants significantly improved their knowledge of target

20

words, as indicated by their increased ability to describe the targeted words’ meaning during

the Summer intervention period.

Adolescents' knowledge of cross-curriculum words

Adolescents in this area of social disadvantage had limited knowledge of commonly used

cross-curriculum words – scoring correctly on only 25% of these words pre-intervention.

Where participants did have some knowledge of words, answers were related to one

functional context (e.g. ‘exclude’ was related to school exclusion and ‘submit’ to online

games), rather than demonstrating rich word knowledge across contexts. This is important

given the association between poor vocabulary knowledge and academic outcomes

(Townsend et al. 2012). A key priority for teaching and educational policy is therefore to

raise awareness of the need to explicitly teach cross-curriculum words, particularly with

students at risk of low educational attainment in contexts of social disadvantage.

Knowledge of cross-curriculum words was measured using the Word Knowledge Profile,

which may be useful in other educational and clinical contexts when profiling adolescent

language skills as part of a wider battery of measures. The Word Knowledge Profile’s Word

Meaning score (based on participants being asked to describe the meaning of the target word)

was positively correlated with the BPVS, the Formulated Sentences subtest of the CELF-4

and the CATV, suggesting a shared underlying construct with standardised measures of

language ability (it was not correlated with the nonverbal measure - CATNV). Further study

is needed to examine the reliability and validity of the Word Knowledge Profile in order to

establish its utility as a clinical tool.

Evaluating the intervention programme on knowledge of functional cross-curriculum words

21

An intervention programme was designed to build on evidenced principles of effective

vocabulary instruction. Cross-curriculum words were chosen due to their potential to impact

on learning across the secondary school curriculum. Participants were given many

opportunities to practice and personalise word meanings (Townsend and Collins 2009) and

were provided with explicit instruction in how to use the words in authentic speaking and

writing contexts (Nagy and Townsend 2012). The ten-week programme targeted one main

word per week, although each lesson included morphological analysis and adding suffixes or

prefixes to build related words.

The results evaluating the impact of the ten-week intervention programme on adolescents’

knowledge of cross-curriculum words were promising. Both intervention groups increased

their knowledge of the taught words but showed no progress in their knowledge of matched

non-taught words. The Spring group made progress during the spring term on two items of

the Word Knowledge Profile (when asked to describe when they would use the targeted word

and when asked when they would ‘perform’ the word, e.g. ‘when would you summarise

something?’). However, the Summer group (waiting control) also made some progress on

these two items, so the progress may not be attributable to the intervention progress.

During the summer term, the Summer group made significant progress on the Word Meaning

item, indicating that they were more able to describe and give definitions of the targeted

word. Furthermore, the Summer group made significantly more progress than the Spring

group during this intervention period. This, plus the lack of progress during the repeated

baseline period, strengthens the assertion that change in word knowledge was a result of the

intervention.

It is interesting to note that the Summer group made more progress than the Spring group on

this Word Meaning score. It is unclear why this was the case and a number of factors could

22

be relevant. Although based on established research into new word learning, this translational

project involved delivering a new programme of intervention in a real-world context for the

first time. It could be that the speech and language therapists delivering the programmes were

in some way more effective during the second implementation of the programme in the

summer term. The Spring group had more participants who required referral to the school’s

behavioural management system during the sessions (9 participants in the Spring group and 4

in the Summer group). This may have impacted on these participants’ own learning and that

of their peers during the sessions. Results may also be due to the relatively small cohort sizes

as the Spring group did make progress which was significant (p = .035) before Bonferroni

corrections were applied.

While there is an established evidence base for supporting vocabulary intervention with

young children (Lubliner and Smetena 2005; Best, Dockrell and Braisby 2006) and those

with diagnosed language impairment (Steele and Mills 2011), there is currently very little

evidence for the effectiveness of small-group vocabulary interventions aimed at adolescents,

particularly within contexts of social disadvantage. Notable exceptions include a small-group

intervention for younger adolescents targeting vocabulary and narrative skills (Joffe 2006).

Whole-class vocabulary interventions have also been evaluated: Word Generation in the

USA that demonstrated the impact of whole-school vocabulary programmes with older

children (Snow et al. 2009). This study therefore contributes to the existing small evidence-

base in this area, which is important given that developing and implementing evidence-based

practice is of increasing importance in the changing landscape of service commissioning

arrangements (DoH 2010) and an increasing need for joint working across the health and

education sectors (DfE 2012).

Although participants made a significant improvement on the Word Knowledge Profile

outcome measure, the mean number of words learnt was 1.17 during an average of 6 sessions

23

attended by participants. Progress also did not generalise to untaught matched words. Such a

small and specific change is not unusual, for example in the Word Generation, project

participants improved by an average of four out of the 40 targeted words on a multiple choice

comprehension test over a yearlong programme (Snow et al. 2009). As highlighted by Nagy

and Townsend (2012: 101), rich instruction resulting in ownership of new words is time

intensive and intervention studies therefore typically target small numbers of words. Non-

intensive instruction does not reliably increase comprehension of new words (Stahl and

Fairbanks 1986). Further research is required to investigate why this may be, for example it

could be due to the complex and abstract nature of taught words, or that verbs are more

difficult to acquire than nouns (Nash and Snowling 2006).

This study also used an outcome measure that required an advanced depth of processing of

the targeted words. Participants were asked to generate a novel response to a word,

explaining the meaning in their own words and relating the word to personal contexts (Stahl

and Fairbanks 1986). It may be that an outcome measure that was able to measure increases

in associative knowledge of the word (where the targeted words are matched with a specific

definition or context) or increased comprehension of the word (for example, by asking

participants to use definitional information to find an antonym) would be more sensitive to

progress following intervention. The intervention sessions also included significant work on

morphological analysis and participants needed support to use suffixes and prefixes to build

derivatives of the word. While the sessions targeted one word (e.g. evaluate), the session

included information on related words (e.g. positive and negative) as well as derivatives (e.g.

evaluation, evaluated) and gains in such related words were not captured by the outcome

measure.

The current study is important as it shows the investment required to remedy adolescents’

poor knowledge of academic vocabulary. This may have implications for schools and speech

24

and language therapists planning related policies and interventions. Future research is needed

to investigate the clinical significance of the findings and whether the programme is a viable,

cost effective option for schools.

Evaluation of the study and future directions

The study had a robust evaluation design and was implemented across one school year to

enable: a) multiple baseline measures to be taken, b) delayed intervention allowing

comparison of intervention group with a control group, c) the use of taught and matched non-

taught words, d) the development of a non-standardised Word Knowledge Profile to assess

the specific impact of the intervention, and e) blind assessment. However, the study design

had important limitations. The evaluation conducted was small in scale and based in one

school which means that further research is needed to investigate generalisation to other

groups of adolescents at risk of educational underachievement. Future work could investigate

whether similar or more positive results are found with other age groups, or when the

programme is delivered by other professionals within the school. A larger cohort would also

enable a more thorough examination of potential influencing factors on the outcomes of the

intervention, such as disruptive behaviour, group effects, or initial knowledge of words.

Further research could also follow-up adolescents in the classroom to see if they use their

new vocabulary knowledge outside of the intervention sessions. Indeed, an important

potential confound in the study could be that they were exposed to the target words in their

subject lessons. The situation in the participating school at the time of the study did not allow

for investigation of exposure to target words or generalisation of word knowledge in relation

to the curriculum. However, the lack of progress during the repeated baseline period does

suggest that the targeted words were not being learned in the classroom.

25

The intense nature of this intervention programme is challenging. While no participants chose

to withdraw from the programme, challenges such as timetabling sessions to minimise the

impact on participants’ core curriculum classes was difficult and school absenteeism was also

a challenge with participants missing an average of 3/9 of sessions (33%). Such challenges

meant that group sizes varied between 3 and 5, which may have impacted on equal

opportunities to engage with the programme across groups. Alternative service delivery

models are needed, and whole-school interventions or supporting school staff to deliver

intervention sessions may be more appropriate (Joffe 2006) and multidisciplinary

collaboration may increase the impact of intervention (Steele and Mills 2011). However, the

current study was based in a very challenging secondary school, which was implementing a

rigorous scheme of school improvement measures as stipulated by school regulators. When

designing the project in collaboration with this school, teachers strongly favoured a small-

group intervention for these particular adolescents over a whole-school or school-led

intervention programme. This was due to both the schools’ limited staff time and resources

and the nature of the adolescents’ difficulties requiring more targeted and specialist support

than could be given in a whole-school approach. This study therefore raises questions about

the best way to support language skills in such challenging contexts, particularly where the

resources of both secondary schools and speech and language therapy services are limited.

Notwithstanding these concerns, this study has provided evidence that a busy secondary

school was able to host and support the intervention programme, resulting in the participants

making progress.

Conclusion

Adolescents at risk of low educational attainment in a context of social disadvantage had

limited knowledge of cross-curriculum words which are commonly used in the classroom

(e.g. ‘summarise’). Adolescents were not making progress in relation to their knowledge of

26

these words without specific intervention. This is important, given the potential consequences

of poor vocabulary skills for literacy skills and educational attainment. Therefore explicit

teaching of such vocabulary is needed, particularly in secondary schools in areas associated

with social disadvantage. Delivering a small-group vocabulary intervention programme in a

busy secondary school was challenging. Despite this, the study provides evidence that the

programme was effective in increasing adolescents’ knowledge of cross-curriculum words,

adding to the evidence base for successful language intervention for adolescents in

mainstream secondary schools.

Acknowledgements:

The study was funded by a seed funding scheme provided by the Faculty of Medicine,

Dentistry and Health at the University of Sheffield. Thank you to Carla Rohde who assisted

with some data input and assessments. Thank you to the trainee speech and language

therapists who assisted with data collection, including: Helen Cooper, Emma Tomlinson,

Frieya Capes, Emma Adams, Bianca Hopkinson, Chris Davies, Philip Gailey, and Chloe

Beech. Many thanks to the participating school and of course we are very grateful to the

participating pupils.

What this paper adds

27

What is already known on this subject

Knowledge of cross-curriculum vocabulary is important for literacy skills and for learning

in the classroom. Interventions can be effective in increasing vocabulary knowledge for

younger children and for children with diagnosed language difficulties. Research has

identified strategies and principles for intervention to increase depth of word knowledge.

Much less is known about interventions to facilitate depth of knowledge of cross-

curriculum words during adolescence.

What this study adds

A ten week intervention programme was delivered by a speech and language therapist to

small groups of adolescents in a mainstream secondary school. At the start of the

intervention, participants lacked knowledge of academic, cross-curriculum vocabulary,

potentially critical to educational success. The intervention was evaluated using a robust

study design (an AABA design, with a repeated baseline, delayed intervention cohort and

blind assessment). The intervention resulted in positive changes to participants’ depth of

knowledge of ten cross-curriculum words, indicated by increased scores on a bespoke

outcome measure.

Clinical implications

This study shows that adolescents identified by teachers as being at risk of educational

under-achievement lacked knowledge of cross-curriculum words such as evaluate,

summarise and discriminate. It provides evidence that intervention was effective in

supporting these adolescents to increase the depth of their understanding of cross-

curriculum vocabulary. Clinicians and educators may use this study: a) to justify support

for adolescents’ word learning; b) for information about interventions to increase

28

adolescents’ knowledge of cross-curriculum vocabulary; and c) to design robust

evaluations of intervention. The study’s session plans and outcome measures are also

freely available on a project website for use by practitioners.

29

REFERENCES

ALEXANDER, R. J., 2006, Towards dialogic teaching: Rethinking classroom talk.

Cambridge: Dialogos.

BECK, I., MCKEOWN, M. and KUCAN, L., 2002, Bringing Words to Life: Robust

Vocabulary Instruction. (New York: Guilford Publishers).

BEST, R., DOCKRELL, J. and BRAISBY, N., 2006, Lexical Acquisition in Elementary

Science Classes. Journal of Educational Psychology, 98, 824–838.

BRITISH NATIONAL CORPUS, VERSION 3 (XML Edition), 2007, Distributed by

Oxford University Computing Services on behalf of the BNC Consortium. URL:

http://www.natcorp.ox.ac.uk/

DEPARTMENT FOR EDUCATION, 2010, Investigating the Role of Language in

Children's Early Educational Outcomes. Research Report DFE-RR134 (London:

Department for Education).

DEPARTMENT FOR EDUCATION, 2012, The Better Communication Research

Programme: Improving provision for children and young people with speech, language

and communication needs. Research Report RR247 (London: Department for Education).

DEPARTMENT OF HEALTH, 2010, Equity and Excellence: Liberating the NHS.

(London: Department of Health). Retrieved on 25th January 2016 from:

https://www.gov.uk/government/publications/liberating-the-nhs-white-paper

DUNN, L., DUNN, D., SEWELL, J., and Styles, B., 2009, British Picture Vocabulary

Scale 3rd Ed. (Chesterfield: Winslow Resources).

EBBELS, S. H., MARIĆ, N., MURPHY, A., and TURNER, G., 2014, Improving

comprehension in adolescents with severe receptive language impairments: a randomized

30

control trial of intervention for coordinating conjunctions. International Journal of

Language & Communication Disorders, 49, 30-48.

FARKAS, G. and BERON, K., 2004, The detailed age trajectory of oral vocabulary

knowledge: differences by class and race. Social Science Research, 33, 464-497.

FREDERICKSON, N., FRITH, U., and REASON, R., 1997, Phonological Awareness

Battery. (Chesterfield: Winslow Resources).

GARDNER, D., and DAVIES, M., 2014, A new academic vocabulary list. Applied

Linguistics, 35, 305-327.

GRAVES, M. F., 2006, The vocabulary book: Learning & instruction (Columbia:

Teachers College Press).

HIRSCH Jr, E. D., 2003, Reading comprehension requires knowledge—of words and the

world. American Educator, 27, 10-13.

IBM CORP, 2010, IBM SPSS Statistics for Windows, Version 19.0. (New York: IBM

Corp).

JOFFE, V. L. 2006, Enhancing language and communication in language-impaired

secondary school-aged children. In: Clegg J., and Ginsborg J. (Eds) Language and social

disadvantage: theory into practice. Chichester, West Sussex: John Wiley.

JOFFE, V. L., 2011, Vocabulary enrichment programme. Milton Keynes, UK:

Speechmark.

JUSTICE, L. M., SCHMITT, M. B., MURPHY, K. A., PRATT, A., and BIANCONE, T.,

2014, The ‘robustness’ of vocabulary intervention in the public schools: targets and

techniques employed in speech–language therapy. International Journal of Language &

Communication Disorders, 49, 288-303.

31

LAW, J., MCBEAN, K., and RUSH, R., 2011, Communication skills in a population of

primary school-aged children raised in an area of pronounced social disadvantage.

International Journal of Language & Communication Disorders, 46, 657-664.

LESAUX, N. K., KIEFFER, M. J., FALLER S. E., and KELLEY, J. G., 2010, The

effectiveness and ease of implementation of an academic vocabulary intervention for

linguistically diverse students in urban middle schools. Reading Research Quarterly, 45,

196-228.

LUBLINER, S. and SMETANA, L., 2005, The effects of comprehensive vocabulary

instruction on title I students' metacognitive word-learning skills and reading

comprehension. Journal of Literacy Research, 37, 163-200.

LUPTON, R., 2005, Social justice and school improvement: improving the quality of

schooling in the poorest neighbourhoods. British Educational Research Journal, 31, 589-

604.

MARULIS, L. M. and NEUMAN, S. B., 2010, The effects of vocabulary intervention on

young children's word learning: A meta-analysis. Review of Educational Research, 80,

300-335.

MCLENNAN, D., BARNES, H., NOBLE, M., DAVIES, J., and GARRETT, E., 2011,

The English Indices of Deprivation 2010 (London: Department for Communities and

Local Government).

NAGY, W. E., and SCOTT, J. A., 2000, Vocabulary Processes. In M.L. Kamil, P. D.

Pearson, R. Barr and P.B. Mosenthal (Eds.) Handbook of Reading Research, volume 3.

New Jersey, USA: Lawrence Erlbaim Associates.

NAGY, W., and TOWNSEND, D., 2012, Words as tools: Learning academic vocabulary

as language acquisition. Reading Research Quarterly, 47, 91-108.

32

NASH, H., and SNOWLING, M., 2006, Teaching new words to children with poor

existing vocabulary knowledge: A controlled evaluation of the definition and context

methods. International Journal of Language & Communication Disorders, 41, 335-354.

RICHES, N.G., TOMASELLO, M. and CONTI-RAMSDEN, G., 2005, Verb learning in

children with SLI: Frequency and spacing effects. Journal of Speech Language and

Hearing Research, 48(6), 1397-1411.

SEMEL, E., WIIG E., and SECORD, W. A., 2006, Clinical evaluation of language

fundamentals UK 4th edition (CELF4 UK) (London: Pearson).

SNOW, C. E., PORCHE, M. V., TABORS P. O., and HARRIS, S.R., 2007, Is literacy

enough? Pathways to academic success for adolescents (Baltimore: Brookes Publishing

Company).

SNOW, C. E., LAWRENCE, J., & WHITE, C., 2009, Generating knowledge of academic

language among urban middle school students. Journal of Research on Educational

Effectiveness, 2, 325-344.

SPENCER, S., 2016. Supporting new word learning in secondary schools.

https://adolescentvocabulary.wordpress.com/ (accessed 26th February 2016).

SPENCER, S., CLEGG, J., and STACKHOUSE, J., 2012, Language and disadvantage: a

comparison of the language abilities of adolescents from two different socioeconomic

areas. International Journal of Language and Communication Disorders 47, 274-284.

STACKHOUSE, J., and WELLS, B. (Eds.). (2001). Children's speech and literacy

difficulties: Identification and intervention. London: Whurr Publishers.

STAHL, S. A., and FAIRBANKS, M. M., 1986, The effects of vocabulary instruction: A

model-based meta-analysis. Review of Educational Research, 56, 72-110.

33

STEELE, S. C., and MILLS, M. T., 2011, Vocabulary intervention for school-age

children with language impairment: A review of evidence and good practice. Child

Language Teaching and Therapy, 27(3), 354-370.

STOCKMAN, I. J., 2000, The new Peabody Picture Vocabulary Test-III: an illusion of

unbiased assessment? Language, Speech, and Hearing Services in Schools, 31, 340–53.

TOWNSEND, D., FILIPPINI, A., COLLINS, P., and BIANCAROSA, G., 2012,

Evidence for the importance of academic word knowledge for the academic achievement

of diverse middle school students. The Elementary School Journal, 112, 497-518.

TOWNSEND, D., and COLLINS, P., 2009, Academic vocabulary and middle school

English learners: An intervention study. Reading and Writing, 22, 993-1019.

WILSON, M. D., 1988, The MRC Psycholinguistic Database: Machine Readable

Dictionary, Version 2. Behavioural Research Methods, Instruments and Computers 20(1):

6-11. Retrieved on 25th January 2016 from:

http://websites.psychology.uwa.edu.au/school/MRCDatabase/uwa_mrc.htm

34

APPENDIX 1.Languaeg assessments administered

The following battery of assessments was administered: British Picture Vocabulary Scale 3rd

Ed (BPVS) (Dunn, Dunn, Sewell and Styles 2009) assessed receptive vocabulary; Recalling

Sentences subtest of the Clinical Evaluation of Language Fundamentals, fourth edition UK

(CELF4 UK) (Semel, Wiig and Secord 2006) assessed expressive language ability;

Formulating Sentences subtest of the CELF4 UK (Semel et al. 2006) assessed participants’

expressive language ability including their ability to formulate sentences of increasing length

and complexity including simple, compound, and complex sentences; Spoonerisms subtest of

the Phonological Awareness Battery (PhAB) (Frederickson, Frith and Reason 1997) assessed

phonological awareness.

35

APPENDIX 2. Information about words targeted during intervention and matched non-taught

words

Taught words

Frequency according to BNC

Control words

Frequency according to BNC

Number of Syllables

Phoneme structure

interpret 1313 establish 5213 3 VCCVCCVC

evaluate 1087 eliminate 1100 4 VCVCVCVC

exclude 1319 exploit 1176 2 VCCCCVC

sustain 1228 submit 1224 2 CVCCVC

discriminate 446 conglomerate 331 (as a noun)

4 CVCCCVCVCVC

infer 319 incur 324 2 VCCV

contribute 2643 distinguish 1967 3 CVCCCVCCVC

generate 1967 dominate 992 3 CVCVCVC

summarise 199 liberate 164 3 CVCVCVC

consult 5350 convince 1200 2 CVCCVCC

36

APPENDIX 3. Principles of word learning applied during intervention

 Supporting deep understanding of a relatively small number of words.

 Multiple repetitions or exposures to new words in a variety of contexts.

 Both explicit discussions of word meaning and opportunities for participants to derive

meaning from implicit information.

 Collaborative learning, working with participants to actively construct a meaning based on

their experiences.

 Multiple opportunities to use the word in different contexts.

 Focus on different aspects of a word (including semantic, phonological, morphological,

grammatical and orthographic) in order to establish clear lexical representations

(Stackhouse and Wells 2001).

 Strategies and activities to increase motivation and engagement.

 Activities in which the participants experienced carrying out the action of the verb in order

to enable mapping the word onto the action, based on previous research into verb-learning

with younger children which involve miming an action while modelling the word (Riches,

Tomasello, and Conti-Ramsden 2005).

 Development of transferable word-learning strategies, for example dictionary skills,

inferring word-meaning from context, and encouraging participants to ask when unsure of

a word.

 Facilitation of generalisation to the class or home situation, through the use of personal

challenges at the end of each session.

37

Appendix 4. Further information about items on the Word Knowledge Profile

Item Description Word repetition Participants were asked to repeat each word after one adult model. Rhyme production

Participants were asked to generate a rhyming word for each word. If they were unable to think of a real word, they were encouraged to think of a non-word. Both real word and non-word rhymes were accepted as correct.

Spelling Participants were asked to spell six of the 20 words. Phonetically plausible attempts were accepted as correct.

Knowledge rating

Participants were asked to rate their own knowledge of the word according to a red- amber-green system. They were given the following instructions (adapted from Beck, McKeown and Kucan 2002): Red: I do not know this word. I have not heard it before and cannot use it. Amber: I know something about this word. I have heard it before but am not sure how to use it. Green: I know this word. I know what it means and can use the word. If the participant rated the word as ‘red’, the participant was usually unable to complete the rest of the Profile for this word.

Ability to use the word in a sentence

Participants were asked to use the word in a spoken sentence. This was scored as correct if the participants produced the word as a verb in a sentence which was syntactically and semantically correct. Sentences without an appropriate semantic component such as I exploited something were not scored as correct.

Word meaning task

Participants were asked to describe the meaning of each word. This was scored correctly if the student indicated that they knew something about the meaning of the word. Detailed definitions or multiple contexts were not required for a mark but were noted on the Profiling sheet. Very few responses included a detailed definition; therefore this liberal scoring procedure was adopted.

When or where would you use this word?

Participants were asked to think of an example of when or where they might use the word. This was used as an indication of the participants’ engagement with the word and the likelihood of the word being used. If the participant was able to give an example of when they would use the word, their response was scored correctly (e.g. ‘I would use eliminate during a computer game’).

Personal context Participants were asked to give an example of a personal context in which they had executed the action of the word for example for evaluate ‘Can you give me an example of when you have evaluated something?’. If the participant was able to give an example of when they had completed the action, their response was scored correctly (e.g. ‘I evaluated the word learning lessons in the interview’)

38

Tables and Figures

Table 1. Study design including delayed intervention and waiting control

Autumn Term

Spring Term Summer Term

Time 1

Time 2

Time 3

Time 4

Baseline

Repeated Baseline

- 6 weeks after Time 1

Spring Intervention (10 weeks)

Assessment – 10 weeks after

Time 2

Summer Intervention (10 weeks)

- 10 weeks

after Time 3

S p ri

n g

G ro

u p :

In te

rv e

n tio

n

Standardised assessment and Word

Knowledge Profile

Word Knowledge

Profile

Yes Word Knowledge

Profile, post- intervention

No Word Knowledge

Profile, follow up

S u m

m e

r G

ro u

p :

D e

la ye

d

in te

rv e

n tio

n g

ro u

p Standardised

assessment and Word

Knowledge Profile

Word Knowledge

Profile

No

Word Knowledge Profile, pre- intervention

Yes Word Knowledge Profile, post intervention

39

Table 2. Summary of participant characteristics

Spring group N=19

Summer group N=16

Age in years and months at start of study 12:8 (range 12:0 – 13:11)

12:11 (range 12:1 – 13:11)

Year group Year 8 15 9 Year 9 4 7

In receipt of free school meals 11 4 Gender male 11 11

female 8 5 Language Status: Pupils speaking English as an additional language

3 0

SEN status* No SEN 14 8 SA 1 1 SA+ 4 7

Mean number of intervention sessions attended 7.42 (1.12) 6.63 (1.20) Number of participants with disruptive behaviour 9 4

* SA = School Action level of support (additional need identified and supported by school), SA+ = School Action Plus level of support (additional support provided by outside agency such as a mental health team or educational psychology).

40

Table 3. Participants’ scores on the language profiling assessments and the Cognitive

Abilities Test and comparison of intervention groups

Whole cohort (n=35) Intervention group means

Comparison of intervention

groups Standard

Score Mean (SD)

Standard Score Range

Comparison with the normative mean

Spring (N=19)

Summer (N=16)

t-test for equality of

means One

sample t- test

Cohen’s d

t df Sig.

CAT verbal 84.20 (7.23)

66 – 96 t= -12.92 P=<0.001**

-1.05 83.79 84.69 -.36 33 .72

CAT nonverbal

83.00 (7.16)

70 - 104 t= -13.63 P=<0.001**

-1.13 82.72 83.33 -.24 31 .81

CAT quantitative

81.69 (8.23)

68 - 112 t= -13.16 P=<0.001**

-1.22 80.58 83.00 -.86 33 .39

BPVS 81.69 (9.51)

69 - 109 t= -11.39 P=<0.001**

-1.22 81.42 82.00 -.18 33 .86

CELF4UK Recalling Sentences subtest

6.23 (2.64)

1 – 11 t= -8.43 P=<0.001**

-1.26 6.32 6.13 .21 33 .84

CELF4UK Formulated Sentences subtest

8.26 (3.71)

1 – 13 t= -2.78 P=0.009

-0.58 7.89 8.69 -.63 33 .54

PhAB Spoonerisms subtest

89.86 (8.40)

69 - 106 t= -7.14 P=<0.001**

-0.68 90.16 89.50 .23 33 .82

CAT, BPVS, PhAB standard scores = mean of 100, standard deviation of 15 CELF4UK scaled score = mean of 10, standard deviation of 3. **. Correlation is significant at the <0.001 level (2-tailed).

Applying a Bonferonni correction for multiple comparisons sets the significance level at .007.

41

Table 4. Spring Group Word Knowledge Profile performance for Time 1 – 4.

Time 1 Time 2 Time 3 Time 4 Baseline 1 Repeated

baseline Post

intervention measure

Follow-up

Mean (SD) Mean (SD) Mean (SD) Mean (SD) Word repetition (/20) 19.06 (1.52) 19.37 (1.1) 19.42 (1.9) 19.63 (1.0) Rhyme production (/20) 10.44 (5.7) 9.37 (7.4) 9.89 (6.6) 11.47 (6.6) Phonetically plausible spelling (/6) 4.63 (1.8) 4.47 (2.0) 4.68 (1.8) 5.11 (1.5) Words self- evaluated:

Red 8.89 (3.7) 8.58 (5.0) 8.16 (4.8) 8.87 (5.2) Amber 7.16 (3.0) 5.89 (3.2) 6.26 (3.0) 6.44 (3.5) Green 3.95 (3.5) 5.53 (3.8) 5.58 (3.8) 4.75 (3.2)

Use of word in a sentence (/20) 6.50 (3.2) 5.11 (3.6) 5.32 (3.7) 6.53 (3.7) Word Meaning Taught words

(/10) 2.42 (1.9) 2.95 (2.0) 3.68 (2.2) 3.38 (2.1)

Non-taught (/10)

2.63 (1.4) 2.47 (1.5) 2.58 (1.4) 2.81 (1.5)

Total (/20) 5.05 (2.9) 5.42 (3.0) 6.26 (3.3) 6.19 (3.4) Example of where participant would use this word

Taught words (/10)

3.26 (2.6) 2.53 (1.7) 6.37 (3.4)** 5.69 (3.1)

Non-taught (/10)

3.05 (1.8) 2.37 (1.6) 4.37 (3.0) 3.93 (2.2)

Total (/20) 6.32 (4.2) 4.89 (3.1) 10.74 (6.1)** 10.06 (5.5) Personal context Taught words

(/10) 2.63 (2.5) 2.11 (1.9) 4.21 (3.0)** 3.69 (2.6)

Non-taught (/10)

2.68 (1.4) 2.32 (1.5) 3.16 (2.1) 2.50 (1.8)

Total (/20) 5.32 (3.5) 4.32 (3.0) 7.37 (4.9)** 6.35 (4.0)

** A significant change at the <0.001 level (two-tailed) when compared to the last assessment time point (e.g. Time 1 compared to Time 2, Time 2 compared to Time 3, Time 3 compared to Time 4),

Applying a Bonferonni correction for multiple comparisons sets the significance level at .001.

42

Table 5. Summer Group Word Knowledge Profile performance for Time 1 – 4.

Time 1 Time 2 Time 3 Time 4 Baseline 1 Repeated

baseline Waiting control

Post intervention

measure Mean (SD) Mean (SD) Mean (SD) Mean (SD)

Word repetition (/20) 19.06 (1.52) 19.08 (1.8) 19.13 (1.5) 19.38 (1.1) Rhyme production (/20) 10.44 (5.7) 10.38 (5.8) 11.00 (6.0) 10.69 (6.6) Phonetically plausible spelling (/6) 4.63 (1.8) 4.69 (1.6) 4.94 (1.3) 5.27 (0.9) Words self- evaluated:

Red 7.13 (3.8) 8.38 (4.4) 7.38 (3.2) 7.25 (3.3) Amber 8.06 (3.4) 6.81 (3.6) 7.31 (3.1) 6.81 (3.4) Green 4.81 (2.9) 4.81 (3.5) 5.31 (4.2) 6.06 (4.2)

Use of word in a sentence (/20) 6.50 (3.2) 5.13 (3.8) 6.19 (2.7) 8.00 (4.0) Word Meaning Taught words

(/10) 2.88 (1.9) 2.75 (1.7) 2.69 (2.0) 4.38** (2.2)

Non-taught (/10)

2.94 (1.4) 2.69 (1.7) 3.19 (1.6) 3.13 (1.9)

Total (/20) 5.69 (3.1) 5.44 (3.2) 5.81 (3.4) 7.50 (3.6) Example given of where participant would use this word

Taught words (/10)

4.36 (2.4) 1.77 (1.2) 4.94 (2.5) 5.75 (2.1)

Non-taught (/10)

3.79 (2.6) 2.08 (1.4) 4.37 (2.9) 3.56 (2.1)

Total (/20) 8.87 (4.6) 4.94 (4.0)* 9.31(5.1) 9.31 (3.9) Personal context Taught words

(/10) 2.07 (1.9) 1.46 (1.7) 2.38 (1.9) 3.31 (2.0)

Non-taught (/10)

2.79 (2.0) 1.31 (1.4) 2.63 (2.3) 2.13 (1.8)

Total (/20) 4.94 (3.6) 3.38 (3.5) 5.00 (3.9) 5.44 (3.6) ** A significant change at the <0.001 level (two-tailed) when compared to the last assessment time point (e.g. Time 1 compared to Time 2, Time 2 compared to Time 3, Time 3 compared to Time 4),

Applying a Bonferonni correction for multiple comparisons sets the significance level at .001.

43

Figure 1: Progress on Word Meaning score for taught words across Spring term

44

Figure 2: Progress on the Word Meaning Taught Words Score of both intervention groups

across Summer term.

45

Figure 3: Progress on Word Meaning score on taught versus non-taught words across the

intervention phase.

46

Figure 4: Graph to show how many words participants learned during the intervention

(progress on the Word Meaning score for taught words).