Enhanced academic performance using a novel classroom physical activity intervention to increase awareness,attention and self-control: Putting embodied cognition into practice

Abstract

When language is processed, brain activity occurs not only in the classic ‘language areas’ such as Broca’s area, but also in areas which control movement. Our systems of understanding, including higher level cognition, are rooted in bodily awareness which needs to be developed as a precursor to intellectual reasoning. Cognition is embodied, and this concept may offer a radical new way of improving school education by improving children’s systems of physical understanding. A new classroom physical intervention, called Move4words, based on embodied cognition, was developed for pupils aged 7–13 years and trialled with 348 typical pupils in 10 mainstream UK schools. Three pilot controlled trials showed significant improvements in academic performance, particularly for struggling pupils performing in the lowest 20 percent. Effect sizes were large for the lowest achievers: Hedges’ g = 0.86 for national examinations at age 11 (KS2 SATs) and g = 1.24 for progress through National Curriculum Levels in reading, writing and maths. Performance gains were maintained for at least 1 year after the end of the intervention.

Keywords

Concentration improving reading inclusive classroom intervention mindful physical activity primary school raising achievement

Introduction

Schools still largely regard the mind and body as two distinct things, with teaching designed to train only the mind. However, recent developments in cognitive science suggest that there is much more to thinking and learning than previously supposed.

Throughout much of the 20th century, prevailing ideas about human thinking and cognition were based on a model which can be likened to the operation of a modern computer. Knowledge was assumed to be held in a memory area, separate from the brain systems which sense the world and move our bodies. The way our brains use our existing knowledge to learn and to solve new problems was thought to be through a series of logical steps (Barsalou, Kyle Simmons, Barbey, & Wilson, 2003). Thinking skills were ‘overwhelmingly seen as conscious, deliberate and rational’ (Claxton, 2012, p. 79).

New evidence from neuroscience and psychology has given rise to a radically different model of how the human brain thinks and learns. In the ‘embodied cognition’ model, the body and brain work together in an inextricably linked ‘brain–body’ system. Thinking skills are now seen as dependent on the body in previously unimaginable ways (e.g. Boncoddo, Dixon, & Kelley, 2010). The origin of this brain–body link is thought to derive from the evolution of language from gesture (e.g. Gentilucci & Corballis, 2006), and the key role of language in the development of sophisticated thinking skills (e.g. Perlovsky & Ilin, 2013).

The sophisticated and abstract knowledge of the adult is built, brick by brick, on physical experiences in the real world. For example, when we think of a car, our brains do not only provide us with a logical list of its physical features, the car’s purpose, words associated with car and so on, but neurons also fire to re-enact our physical and sensory experience of driving: the muscle movements of gear changing, sensations we had of movement when travelling, and the sound of the engine and so on (Barsalou et al., 2003).

So our systems of understanding are rooted in bodily awareness; the human brain is primed for action, even when processing abstract concepts. This includes higher level elements of cognition such as conceptualisation, reasoning, planning and judgement, as well as planning of and verbalising physical actions.

Neuroscientific and psychological evidence now points to a very surprising conclusion, that experience gained from learning accurate muscle control in order to achieve physical tasks allows the child to better understand how to achieve other, more abstract goals (Hung & Labroo, 2011; Sommerville, Woodward, & Needham, 2005).

Educationalists are now familiar with one aspect of embodied cognition, namely where the body is actively involved in a very specific learning experience. An example of this is the type of early-years phonics approach where the child makes a wriggly snake-like movement when making the ‘ssss’ sibilant sound, and so on, to embed the learning. However, embodied cognition has implications which go far beyond this. If our brains do not solve problems unaided, and the brain’s control of the body plays an essential role in any form of thinking or problem-solving, then schools could benefit by including physical training of sensorimotor control (Ionescu & Vasc, 2014).

This article explores this new approach to school interventions, based on embodied cognition. Our hypothesis is that improving attention and self-control through physical and visual bodily activities and auditory tasks can lead to enhanced academic attainment. We have here only looked at results of nationally standardised examinations, which only measure one aspect of academic attainment. Other important, but less quantifiable, aspects are beyond the scope of this article.

We have developed and trialled a new whole-class 12-week physical action programme for pupils aged 7–13 years, assessing its impact on national performance measures. The intervention can be used with a whole class. It is easily transferable between different areas and contexts, and can be delivered by the class teacher without upheaval of current teaching strategies.

We discuss three studies in this article. One study was a matched-pairs controlled trial of 16 schools’ performance in national exams taken at age 11–12 (Key Stage 2 Statutory Assessment Tests (KS2 SATs)); pupils participated in the intervention in their Year 6 (aged 10–11 years) in 8 schools (50%). The other two studies took place in the same 3 additional schools; pupils were aged between 8 and 10 years when they did the intervention. One study was a time-series trial, where termly performance in reading, writing and maths (National Curriculum Levels) was tracked for up to 3 years before the start of the intervention and for 1 year after the end of the intervention. The final study was a two-group controlled trial of reading performance (National Curriculum Levels) where one group did the intervention and the other did not.

Embodied cognition

The human brain evolved to solve problems in the physical environment, using sensory input, perceptual processing and muscle control (the sensorimotor system). Reading, writing and maths utilise already-existing brain systems which evolved for the very different purpose of controlling action. So the idea that the brain can only develop sophisticated intellectual reasoning capacity once well-developed sensorimotor representations have been formed through learning to control physical actions (e.g. Gallese & Lakoff, 2005), may not be surprising.

Many neuroimaging studies have shown that when language is being processed, brain activity occurs in the motor and sensory areas corresponding to the meaning being transmitted (e.g. Boulenger et al., 2006). Some cognitive scientists argue that all cognition is based on knowledge that comes from the body (e.g. Gallese & Lakoff, 2005; Wilson, 2002). Gesture and miniature muscle movements in the hands are a fundamental part of language production and comprehension (e.g. Olmstead, Viswanathan, Aicher, & Fowler, 2009). Hung and Labroo (2011) showed that controlling the body by tensing muscles actually improved participants’ self-control and mental focus in unrelated cognitive tasks, and facilitated ‘the self-regulation essential for the attainment of long-term goals’ (p. 1046).

Published evidence for academic impact of physical activity interventions

There is a view in education that sensorimotor programmes are ineffective. Some high-profile commercial techniques such as Brain Gym have been publicised without adequate evidence, made exaggerated claims and used pseudoscientific explanations. The Dyslexia Dyspraxia Attention Treatment (DDAT) programme was taken up by the press as a ‘cure’ for dyslexia, following the work of Reynolds, Nicolson and Hambly (2003) which used unmatched control and experimental groups and had many other methodological flaws (e.g. Snowling & Hulme, 2003; Stein, 2003).

Hyatt, Stephenson, and Carter (2009) concluded that research on perceptual motor programmes fails to support their use. However, the main cornerstone of their argument rested on a 30-year-old and technically primitive statistical ‘meta-analysis’ by Kavale and Mattson (1983) which only used studies available in the early 1980s.

Now, however, there is a growing body of evidence that physical activity interventions do have positive impact on academic performance. Tomporowski, Lambourne, and Okumura (2011) carried out a large review of the evidence to date, and found significant impact of physical activity and exercise on children’s intellectual function, cognitive abilities and academic achievement. Booth et al. (2014) found long-term positive impact of medium-to-vigorous physical activity on academic attainment in a study of almost 5000 UK adolescents. Chaddock-Heyman et al. (2014) found that increased aerobic fitness improved neural connectivity within the brains of 9- to 10-year-old children, particularly between left and right brain hemispheres, which the authors suggest may explain the connection between greater fitness and higher cognitive performance in children.

Focussed awareness and mindfulness of physical actions may be more important than aerobic impact (e.g. Best, 2010), which may explain why some studies show a positive relationship between the amount of exercise and academic attainment and other studies do not (Rasberry et al., 2011). High-intensity physical activity may actually be less effective at improving cognitive performance than moderate exercise (e.g. Kashihara, Maruyama, Murota, & Nakahara, 2009) because of biochemical factors produced during intense exercise. There are also some indications that physical activity performed in the classroom may have a greater impact on academic achievement than classic exercise done on the sports field or in the gym (e.g. Donnelly & Lambourne, 2011; Hill et al., 2010).

Various research groups have explored the impact of exercise on attention spans, and find that short bursts of physical activity can have a surprisingly positive impact on improving concentration levels, at least in the short term (e.g. Budde, Voelcker-Rehage, Pietraßyk-Kendziorra, Ribeiro, & Tidow, 2008; Hill et al., 2010; Pontifex, Saliba, Raine, Picchietti, & Hillman, 2013). Diamond and Lee (2011) showed that executive function (the ability to predict, plan and act) in children aged 4–12 years could be developed through physical as well as cognitive activities, so long as the physical activities involved incremental steps which gradually increased the challenge and which included repeated practice.

More specific elements of physical activity have been researched. Children’s phonological awareness and reading can be improved by participating in rhythmic exercises (Moritz, Yampolsky, Papadelis, Thomson, & Wolf, 2013). McPhillips, Hepper, and Mulhern (2000) and Jordan-Black (2005) assessed the impact of a year-long thoughtful classroom physical activity programme, and reported significant outcomes of a randomised, double-blind, movement-placebo–controlled trial with combined effect sizes on reading age of d = 0.77 (our calculation from their published figures).

A controlled trial of the effect on reading ability of participation in a 6-week two-handed coordination programme (Uhrich & Swalm, 2007) showed a significant improvement in reading comprehension in the experimental group compared to controls (p < .05). A movement-placebo–controlled trial by Byl, Byl, and Rosenthal (1989) showed that balance exercises had significantly greater impact on reading grade levels 6 months after the intervention, compared to an aerobic exercise regime (p < .05).

The practice of relaxation exercises and mindfulness also appears to positively improve aspects of academic performance. Secondary school students who were taught self-relaxation exercises performed significantly better in orthography tests than controls (Krampen, 2010). Weare (2013) reviewed work on the positive impact of mindfulness on improving concentration in lessons.

This wide range of effective physical interventions suggests that working with the body offers a valuable tool to improve cognitive functioning in a very broad sense.

Method

The intervention

Pupils’ mindful control of visual, motor and auditory skills was trained by following videos of child actors modelling a set of over 200 individual activities. Activities started out very simply at the start of the 12-week programme, and built in complexity with each week. To make delivery easy for participating teachers, the Move4words intervention was highly prescriptive, with each element of the 60 daily activity sessions laid out in step-by-step format. Short video segments, 45 seconds to 2 minutes in length, gave the children clear verbal, musical and visual instruction how to perform each activity, so the children effectively taught themselves the necessary skills.

The intervention is based on physical training in incremental steps, with repeated practice, aiming to improve focussed attention and executive function, and has a number of elements as follows:

Visual attention and eye tracking skills;

Classroom mild aerobic exercise;

Concentrating attention on limb-movement patterns;

Cross-lateral body coordination;

Relaxation.

A few examples of how pupils’ attention is focussed are as follows. Their awareness is drawn to physical sensations in hands, arms and feet by rubbing or tapping the area before physically moving a limb. Animations teach how ribcage muscles work, including the diaphragm, during breathing to help the pupils to feel the muscles working as they do mindful breathing exercises. Animations show how eye muscles work during tracking and convergence, and pupils are encouraged to feel these muscles activating during eye tracking exercises.

Teachers and learning-support assistants who participated in recent trials received one 2.5-hour training session to enable them to deliver the intervention. Longer training is not necessary because they play a supporting, rather than a teaching, role.

Trial design and participants

Study 1 − two-group controlled trial of exam results at age 11

Study 1 compared the impact of the intervention in whole classes of 10- to 11-year-old children (Year 6) in 8 matched pairs of primary schools: 16 schools in total. The intervention was provided to 1 in each pair of schools in either 2011 or 2012, but the outcome measure was determined for both schools in each pair, in four consecutive years − 3 years before the intervention, and the intervention year. The total number of pupils in the study was 1955 over 4 years; 235 pupils from 8 schools participated in the intervention in the final year.

The outcome measure was the percentage of pupils in each school who achieved or exceeded the threshold of Level 4 in English and Maths in national examinations taken at the end of primary school (SATs). These figures are published each year by the Department for Education. Pupils need to achieve a minimum score in UK SATs tests to be able to access the secondary curriculum successfully; this threshold is Level 4 in English and Maths or 27 points. In 2011 and 2012, the Government-set floor standard was 60 percent of all pupils reaching their target of Level 4 in English and Maths. Schools were considered to be failing if they fell below these standards.

All schools were in urban areas of relative deprivation in the Midlands and North West of England. The ability range and gender distribution of participating children were typical of regular classrooms in the United Kingdom, including approximately 20 percent of children with special educational needs (Department for Education, 2011b). Most participating schools had a diverse ethnic mix, with up to 60 different languages being spoken at home.

Eight comparison schools were chosen from similar environments to the intervention schools, and matched as closely as possible to each of the intervention schools on final year performance in each of the three previous years (using matched pairs, following the method of Mant, Wilson, & Coats, 2007). A further four factors were also taken into account in the matching process: urban environment, Year 6 pupil numbers, percentage of children with special educational needs and percentage of children receiving free school meals.

Study 2 − one-group time-series trial of reading, writing and maths performance

Study 2 followed individual pupil’s academic performance for up to 3 years before the start and 1 year after the end of the intervention. The intervention was delivered in the summer term of 2011 to whole classes. A total of 113 pupils aged 8–10 years participated from three schools in a total of five classes. One participating class was in Year 3 (age 7–8 years) and two from each of Year 4 (age 8–9 years) and Year 5 (age 9–10 years).

The performance measures were assessments of National Curriculum Levels in reading, writing and maths which the schools routinely made three times each year for each pupil. These were assessed by each class teacher as part of each school’s normal practice, using normalised tests which allow comparison with national performance averages. Results can be presented as levels (e.g. Level 3c) or the equivalent points score (e.g. 19 points).

The Government set benchmark standards at age 7 (at the time of this trial: Level 2b = 15 points), and at age 11 (Level 4b = 27 points). The Department for Education (2011a) published data on how pupils’ performance in reading, writing and maths advances through Key Stage 2. We have used these data to determine national average points scores for each term through KS2 to compare against pupils’ performance in the intervention schools.

Tracking started at age 7, so the time period of tracking before the intervention depended on the age of the participating children, ranging from 6 months for children in Year 3 (aged 8 years at the time of intervention) up to 2 years for children in Year 5 (aged 10 years at the time of intervention). All children’s performance was monitored for one complete year after the end of the intervention.

The three schools were in urban areas of relative deprivation in the Midlands and North West of England. One of these schools participated in study 1, because they also ran the intervention with Year 6 pupils; the other two schools did not participate in study 1.

Study 3 − two-group controlled trial of National Curriculum reading scores

The data from 51 pupils from one school in study 2 provide a two-group controlled trial of reading performance. Three year groups (Years 3, 4 and 5) participated in the intervention, and the school provided tracking data for all three groups. We compared group mean reading scores for pupils of the same age but from two different cohorts. Thus reading scores assessed in the spring of 2011 for Year 5 pupils (i.e. before they participated in the intervention in their Year 5) were compared with scores assessed in the spring of 2012 for a group of Year 5 pupils who were a year younger than the first group (i.e. 9 months after they had participated in the intervention in their Year 4). We repeated the procedure for reading assessments made for the two groups of pupils who were in Year 4 in 2011 and 2012.

Intervention delivery

The 20-minute intervention was delivered daily for 12 weeks to the whole class at the start of the school day during normal lessons, using the video-based programme on DVD provided to each school. Delivery was by trained class teachers or learning-support assistants.

Statistical analyses

Children were included in the data analysis in study 2 only if they had been present in school for sufficient tests to allow progress rates to be calculated before and after the intervention. A total of 21 pupils (18% of the total) were excluded from analysis because they had been absent for several tests, had left the school or joined part way through the relevant time period. In order to determine whether the observed changes in performance were statistically significant, we calculated effect sizes and carried out t-tests, and checked whether inequalities between groups might contribute to the observed effect using analysis of covariance (ANCOVA). T-test results are presented as the probability, p, that the difference between outcome measures has come about by chance. If p is less than .05, then we are more than 95 percent sure that the differences are real and significant; this is the widely accepted cut-off for significant effect.

Effect sizes were calculated from our controlled trials in order to determine the magnitude of the impact of the intervention. The effect size is the improvement in learner achievement brought about by the intervention divided by the variation in student performance. Effect sizes provide a standardised measure which allows fair comparison between similar studies.

We used the effect size calculator provided by the Centre for Evaluation and Monitoring at Durham University, to calculate the recommended effect size and its confidence interval (CI). This is the bias-corrected Hedges’ g, which compensates for small sample numbers. The Educational Endowment Foundation (Sutton Trust) consider that the educational impact of effect sizes can be categorised into ‘low’ (d = 0.02–0.18), ‘moderate’ (d = 0.19–0.44), ‘high’ (d = 0.45–0.69) and ‘very high’ (d equal to or greater than 0.7). We have used these categories in our interpretation of our data.

Results

Study 1: two-group controlled trial of exam performance at age 11 − Key Stage 2 SATs

In the 3 years before the intervention, the average percentage of pupils reaching the government’s target of Level 4 or more in English and Maths in national examinations (KS2 SATs) was 59 percent in both the experimental and comparison groups of schools. The national average score was 73 percent over this time period. Thus 12 of the schools were performing considerably below the average, and some were in danger of being put into special measures (Table 1).

Table 1.

Comparative study of academic performance at the end of primary school exams (KS2 SATs) taken in Y6 at age 11, in eight matched pairs of comparison and intervention schools. N indicates the number of pupils in Y6 in each school in the intervention year (total of 258 in comparison schools, 235 in intervention schools). Stars indicate that teachers boycotted the KS2 exams in 2011.

Comparison schools								Intervention schools
School code	Percentage of pupils achieving Level 4+ in English and Maths in KS2 SATs						Change in %	School code	Percentage of pupils achieving Level 4+ in English and Maths in KS2 SATs						Change in %
	Year 1	Year 2	Year 3	Average of Years 1–3	Year 4	N			Year 1	Year 2	Year 3	Average of Years 1–3	Year 4	N
2011								2011
C1	34	48	*	41	57	46	+16	Ex1	27	50	*	39	67	14	+29
C2	75	88	*	82	81	20	−1	Ex2	77	88	*	83	100	25	+18
C3	95	90	*	93	82	44	−11	Ex3	97	90	*	94	100	31	+7
2012								2012
C4	52	*	34	43	41	29	−2	Ex4	49	*	37	43	79	29	+36
C5	48	*	42	45	47	34	+2	Ex5	44	*	43	44	69	48	+26
C6	58	33	59	50	68	22	+18	Ex6	43	43	58	48	68	31	+20
C7	58	44	61	54	68	22	+ 14	Ex7	59	44	63	55	79	24	+ 24
C8	48	71	73	64	80	41	+16	Ex8	49	77	71	66	70	33	+4
Average (SD)	58.50 (18.78)	62.33 (24.09)	53.80 (15.64)	58.92 (19.02)	65.50 (15.81)		+6.58 (10.65)	Average (SD)	55.63 (21.98)	65.33 (22.12)	54.40 (14.10)	58.75 (20.13)	79.00 (13.7)		+20.25 (10.73)
National average				73	77		+4	National average				73	77		+4

KS2 SATs: Key Stage 2 Statutory Assessment Tests; SD: standard deviation.

After using Move4words, 79 percent of pupils in the intervention schools reached their target, a 20 percent rise compared to the previous 3 years (Figure 1). This increase was statistically significant (t = 5.340; p = .001), so is unlikely to have come about by chance.

Figure 1.

Study 1 – percentage of children reaching target level (Level 4+) in English and Maths in final-year Primary School KS2 SATs over a 4-year period in eight trial schools (total N = 995) and eight comparison schools (total N = 960). White bars indicate data from comparison schools; grey bar indicates data from intervention schools in three previous years when intervention was not used; black bar indicates data from intervention schools where pupils did the intervention in their Year 6 (N = 235).

Only 66 percent of pupils from the comparison schools reached their target in the intervention year, a 7 percent increase compared to the previous 3 years, only marginally better than the national average improvement of 4 percent over the same time period. This was not a statistically significant improvement (t = 1.748; p = .124). The intervention schools’ performance in the trial year was statistically significantly better than that in the comparison schools (N = 16; t = 2.630; p = .034; effect size g = 0.86, CI = [−0.16, 1.89]).

The increase in the percentage of children achieving their target in KS2 SATs tests in the intervention schools was three times greater than the improvement seen in the comparison schools over the same time period. The intervention schools’ performance in the trial year was statistically significantly better than that in the comparison schools (N = 16; t = 2.630; p = .034; effect size g = 0.86, CI = [−0.16, 1.89]).

Because the range of school performance was large, the CI was large. Six schools in each group were performing below the average, and the comparison between these schools may provide a more reliable estimate of the effect size (g = 1.51, CI = [0.23, 2.80]). ANCOVA analysis showed that the differences between schools in previous years’ performance do not explain the differences between control and experimental schools’ performance in the trial year.

Participating heads and teachers reported many improvements. Formerly difficult year groups began doing well, reading ability increased, classrooms were more focussed and children were in a better learning mood.

Study 2: one-group time-series trial of progress through National Curriculum Levels in reading, writing and maths

The data comprised National Curriculum Levels (converted to points scores) in reading, writing and maths before and after the intervention from all pupils from five classes from three participating schools, and the national average scores at each age. Figure 2 shows examples of group average reading scores from two classes from one participating school. The rate of progress in reading was considerably faster throughout the year after completing the Move4words programme, than before the intervention.

Figure 2.

Studies 2 and 3 – two examples of reading progress through National Curriculum Levels for whole classes in one school, for (a) pupils who did the intervention in their Year 4 and (b) pupils who did the intervention in their Year 5. White bars indicate national averages at each age; grey bars indicate the group average scores before intervention; black bars indicate group average scores after intervention. The hatched area shows the 12-week intervention period. The lines show average rates of progress before and after the intervention. The black stars indicate data compared in study 3 and Figure 4.

The study-average progress rates for the time periods before and after the intervention were calculated for reading, writing and maths scores (Table 2). The slopes of the best-fit lines in Figure 2 are examples of these rates. We have also subdivided the data into performance measures for pupils performing in the top 50 percent, those between 50 percent and 20 percent and those performing in the bottom 20 percent (Table 2) and have calculated progress rates for these sub-groups.

Table 2.

Comparing Key Stage National Curriculum Level progress rate in points per year during a time period of 1–2 years before intervention (Pre), and 1 year after the end of the intervention (Post) for three ability bands. Data combined from five classes in three schools. T-tests were used to compare pre and post rates; t is the test statistic and p is the probability level; g is Hedges’ bias-corrected effect size.

Ability range	N	Progress rates through National Curriculum sub-levels converted to points scores per year
		Reading (SD)		Writing (SD)		Maths (SD)		Overall combined score (SD)
		Pre	Post	Pre	Post	Pre	Post	Pre	Post
All	96	2.94 (1.94)	4.80 (3.01)	3.37 (2.19)	4.02 (2.29)	2.36 (1.68)	4.44 (2.21)	2.82 (1.34)	4.32 (1.91)
		t = 4.450; p < .001;		t = 1.832; p = .070;		t = 7.187; p < .001;		t = 5.970; p < .001;
		g = 0.73		g = 0.29		g = 1.06		g = 0.91
Below 20th percentile	11	1.93 (2.67)	7.82 (4.87)	3.85 (2.88)	5.0 (2.45)	2.38 (2.5)	5.82 (2.89)	2.72 (2.06)	6.21 (3.23)
		t = 3.052; p = .012;		t = 1.511; p = .151;		t = 3.294; p = .002;		t = 2.738; p = .021;
		g = 1.44		g = 0.41		g = 1.22		g = 1.24
Between 20th and 50th percentile	43	2.40 (1.81)	5.29 (2.82)	3.05 (2.08)	4.05 (2.74)	2.02 (1.24)	4.16 (1.95)	2.42 (1.22)	4.40 (1.76)
		t = 5.369; p < .001;		t = 1.722; p = .093;		t = 6.046; p < .001;		t = 5.755; p < .001;
		g = 1.21		g = 0.41		g = 1.30		g = 1.30
Above 50th percentile	42	3.78 (1.52)	3.50 (1.59)	3.57 (2.10)	3.37 (1.63)	2.71 (1.79)	4.37 (2.19)	3.24 (1.13)	3.74 (1.21)
		t = −0.775; p = .443;		t = –0.387; p = .701;		t = 3.348; p = .005;		t = 2.173; p = .036;
		g = −0.18		g = −0.11		g = 0.82		g = 0.42

SD: standard deviation.

Average reading progress rates for the whole group were statistically significantly faster after the end of the intervention (p < .001; Table 2). Reading progress rate increased by 63 percent to 4.80 points per year (a considerably greater rate than the national average rate of 3 points per year for this age range). This is a large effect (g = 0.73). A similarly large impact was seen for maths with an 88 percent increase in progress rate (g = 1.06). The intervention had only a small impact on writing, with a 19 percent increase in progress rate (g = 0.29).

Schools are also interested in overall performance scores, with reading, writing and maths scores combined, so these have also been calculated, and tabulated in Table 2. Figure 3 and Table 2 show the progress rates before and after intervention for the combined reading, writing and maths scores. Study-average progress rates were significantly enhanced after the intervention (Figure 3, Table 2), with very high impact (g = 0.91). The greatest impact was experienced by the bottom 20 percent, who experienced a 128 percent increase in progress rate of reading, writing and maths combined (g = 1.24). Children between the 20th and 50th percentile improved by 81 percent (g = 1.3), while those in the top 50 percent improved by 15 percent (g = 0.42, a moderate effect). All these effects are statistically significant.

Figure 3.

Study 2 – progress rates in combined reading, writing and maths scores in points per year before the start of and after the end of the intervention. A total of 92 children from five classes in three schools; intervention took place in Years 3–5. Data from National Curriculum sub-levels converted to points scores for three achievement bands. The white bars show national average progress rates for the age ranges studied here.

Figure 3 also shows national average progress rates calculated from published figures for the three achievement bands. Before the intervention, the participating pupils’ average reading progress was similar to the national average. After the intervention, the combined progress rates in reading, writing and maths were surprisingly large (Figure 3), particularly for pupils in the bottom 20 percent (exceeding 6 points per year, more than double the national average rate of 2.8 points per year). These high progress rates have transferred into substantial improvements in KS2 scores, narrowing the gap between lower and higher achievers.

Study 3: two-group controlled trial of National Curriculum reading scores

Figure 2 illustrates how trial 3 was constructed using data from Year 5 pupils. Black stars indicate class-average reading scores in spring Year 5 for two groups of pupils of the same age but from two different cohorts. One test was carried out in spring 2011 (Group A) before these pupils participated in the intervention in summer term of their Year 5 (Figure 2(b)). The other test was carried out in spring 2012 (Group B) on a different cohort, 9 months after this cohort had completed the intervention in their Year 4 (Figure 2(a)). We used the same procedure to compare performance in spring Year 4 for pupils who were 1 year younger than described above.

Figure 4 shows group mean reading scores for all four groups in the controlled trial. Post-intervention scores were 3.4 points higher than pre-intervention scores for the Year 4 trial (a statistically significant improvement; p = .006). The intervention had very high impact with Year 4 (g = 0.94). The intervention also had a large effect for Year 5 pupils (an increase of 2.8 points; g = 0.64), which was significant at the 90 percent confidence level (p = .069). Expected progress rates were 3 points per year (two complete KS Levels in 4 years), so results of this trial showed that, after using Move4words, pupils were 1 year ahead, compared to pupils of the same age in the previous year, for both Year 4 and Year 5 pupils.

Figure 4.

Study 3 – controlled trials of the impact of the intervention on group average National Curriculum points scores for reading with pupils in Year 4 and Year 5. White bars indicate National Average scores at this age. Grey bars indicate reading performance in spring 2011 for pupils who had not yet experienced the intervention, compared to reading scores for a different cohort at the same age (black bars), assessed in spring 2012, 9 months after the end of the intervention.

Discussion

All three trials showed that pupils who participated in Move4words had significantly enhanced academic performance levels compared to pupils who did not use the intervention. After using the intervention in low-achieving schools, 13 percent more pupils had reached the Government’s target in national examinations (KS2 SATs) at age 11 compared to matched comparison groups. Nine months after the intervention finished, pupils aged 9 and 10 years were 1 year ahead in National Curriculum Levels in reading, writing and maths, compared to how pupils of the same age had performed in the previous year. Overall, progress through National Curriculum Levels in reading, writing and maths was 50 percent faster in the year after the intervention compared to pre-intervention rates. Pupils in the bottom 20 percent experienced the greatest improvement in reading, progressing almost three times faster than before.

The new movement intervention appears to have had significant impact and high effect (effect sizes exceeding 0.64) on academic performance in reading and maths with whole classes. The impact was greater for pupils performing below average. Hattie (2009) reviewed a substantial body of educational interventions and recommended that new interventions should be taken up by education only if they produced effect sizes greater than 0.4. He found that effect sizes of approximately 0.5 were equivalent to increasing General Certificate of Secondary Education (GCSE) grades by one complete grade.

The high effect sizes are comparable to the largest impact for pupils in Year 3 or older reported by Brooks (2013) in his review of what works for pupils with literacy difficulties. Brooks’ greatest reported impact was for the one-to-one Catch Up Literacy pilot (Year 3; d = 1.0). The new intervention is designed to take 15 minutes per day for 3 months with the whole class. It requires considerably less teacher time than one-to-one interventions, offering a cost-effective addition to current provision. It does not replace literacy support, but adds to it, and does not require high levels of teacher expertise to implement.

Long-term impact

Successful literacy interventions of many types are often found to have a ‘wash out’ effect (e.g. Hurry & Sylva, 2007) where performance gains are not maintained and literacy performance slowly trends back towards the original baseline. But our data suggest that gains in reading, writing and maths performance were maintained for at least 1 year after the end of the intervention.

Comparison with impact of aerobic exercise on academic performance

Effect sizes for the impact of exercise on academic performance are about 0.25 (see the review by Etnier, Nowell, Landers, & Sibley, 2006), which is considerably smaller than the effect sizes calculated in these pilot studies of the new intervention which exceed 0.8 for low-ability children and 0.4 for above-average ability children.

Possible causative mechanisms

A number of possible mechanisms have been proposed which might explain the positive impact of physical activity on academic performance. These proposed mechanisms include increased embodied cognition, improved focussing of attention (e.g. Budde et al., 2008), improved cognitive arousal (e.g. Lambourne & Tomporowski, 2010), enhanced executive function (e.g. Best, 2010; Davis et al., 2011), promotion of growth of new neurons and connections between existing neurons through stimulation of the molecular machinery of the brain (e.g. Cotman & Berchtold, 2002), increased inter-hemispheric neural connectivity (Chaddock-Heyman et al., 2014), and improved brain function in magnocellular (Solan, Shelley-Tremblay, Hansen, & Larson, 2007) or cerebellar (Krafnick, Flowers, Napoliello, & Eden, 2011) systems.

The trials described in this article were not designed to determine the underlying mechanisms by which change might be brought about. We propose two possible mechanisms which might specifically contribute to the impact of the programme, but do not exclude the contribution of others listed above. Further study would be required to resolve this question.

Attention

Improving the brain’s attentional systems is the first step in utilising embodied cognition to improve learning capacity. The process of attention requires the filtering out of most of the sensory input to the higher centres of the brain (Lennert & Martinez-Trujillo, 2011), leaving only the information which the brain expects to be important (based on previous experience) available for cognitive processing. Physical, visual and auditory attention is therefore key to how the brain deals with the sensory information which provides our only interaction with the outside world, and to the functioning of the integrated brain/body problem-solving systems.

The coordinated movement programme is partly designed to focus children’s attention to their limb and eye movements, so this may contribute to its observed impact. To follow the 200+ activities in the programme, pupils have to carefully concentrate and plan their physical actions, promoting focussed attention. It was not within the scope of this study to directly assess the impact on attention, although anecdotal evidence suggests that teachers observed increased alertness, increased ability to pay attention, to work for longer periods without distraction and better listening skills.

Inter-hemispheric coordination

Part of the programme involves a graduated set of limb-control exercises building up to slow and controlled cross-body coordination, touching the hand to the opposite knee. The coordination of movement on opposite sides of the body requires information transfer between left and right brain hemispheres, via the corpus callosum, a network of nerve fibres which joins the brain hemispheres.

The corpus callosum plays an important role in reading, allowing inter-hemispheric coordination of activity in hemispherically specialised elements of the reading network into a fluent orthographic process (e.g. Henderson, Barca, & Ellis, 2007). Carreiras et al. (2009) showed that the process of learning to read in initially illiterate adults demonstrably changes the structure of the brain, in particular by increasing the size of the corpus callosum. Shillcock and McDonald (2005) observed impaired inter-hemispheric coordination of orthographic information in reading in dyslexic subjects.

The corpus callosum develops throughout childhood, normally reaching approximately 90 percent of full maturity by age 11 (Paul, 2011). Perhaps repetitive conscious planning of contralateral body movements stimulates the maturation of the corpus callosum (e.g. Geffen, Jones, & Geffen, 1994), which in turn supports the inter-hemispheric elements of reading (Carreiras, Armstrong, Perea, & Frost, 2014; Chaddock-Heyman et al., 2014).

Possible confounding factors

Hawthorne effect

In trials where the outcomes of delivering a new intervention are compared against a ‘no-treatment’ condition, as in these studies, the pupils might be responding positively to increased attention from their teachers, and the intervention actually has no real impact. In other words, the changes may be due solely to the ‘Hawthorne effect’.

We argue that it is unlikely that there is a significant Hawthorne effect contributing to the positive outcomes of this study. Progress rates through National Curriculum Levels more than doubled after the intervention period for the 11 pupils who were performing in the bottom 20 percent before the intervention period in study 2. These pupils had been receiving a considerable amount of one-to-one support and teacher attention in the period before the intervention. If the Hawthorne effect could improve academic performance so dramatically by increased teacher attention, then previous special needs support should have caused the same response.

Regression to the mean

A possible explanation for the greater improvement observed for below-average pupils is statistical regression to the mean. This occurs when a low- or high-achieving group is identified on the basis of how they perform on a pre-test. The impact of regression to the mean depends on the accuracy of the testing procedure. If the test is not very accurate, then some pupils who appeared to be below average in the pre-test should not have been included, and some other poor readers will have been given spuriously high test scores but should have been included. At the re-test, after the intervention, then most of the incorrectly included higher-ability pupils will now provide higher test results, and the group average will increase, even if the intervention has no effect. The same effect occurs in the opposite direction, for any group chosen on the basis of above-average scores.

It is unlikely that regression to the mean is an important phenomenon in this study, because the time-series trial in study 2 involved up to 10 separate termly assessments, which would reduce noise and substantially reduce any effect of regression to the mean.

Helping the ‘tail’ of low achievers in England

Recent international comparative research has shown that in England, more than twice the number of pupils leave school without basic skills in literacy and numeracy than in comparable countries with similar levels of affluence such as Australia, Canada and Japan (data reviewed by Amadeo & Marshall, 2013) despite record levels of spending on education. This leads to there being a ‘long tail’ of 40 percent of English children who fail to achieve the national standard of 5 good GCSEs at age 16 (Amadeo and Marshall). The lowest-achieving 20 percent of UK children leave school poorly educated and singularly ill-equipped for a successful or productive adult life.

The new intervention discussed in this article may have a considerably greater impact on reading and academic performance for the bottom 20 percent than for higher achievers. Thus progress through National Curriculum Levels for pupils below the 20th percentile increased by 128 percent during the year after the end of the 3-month intervention (effect size d = 1.698; p = .021), in contrast to a 15 percent increase for those above the 50th percentile (effect size d = 0.440; p = .036).

There is little data of this kind available to allow comparison of the new intervention with other literacy interventions. Tanner, Brown, and Day (2011) assessed the impact of the Every Child a Reader programme on different ability cohorts, and found that reading progress for children aged 5–7 years and performing below the 10th percentile was larger, but not significantly greater than for the year group as a whole. Our new intervention may therefore offer a powerful new approach to add to methods currently available to help improve prospects for low-achieving children.

Conclusion

These pilot trials were very successful. The greatest impact was seen for pupils performing below the 20th percentile (with large effect sizes in excess of 1.0). Higher-achieving pupils achieved smaller but still significant improvements, justifying the inclusive use of the Move4words intervention. These improvements were long-lasting. The new intervention has a very simple delivery model which requires very little adaptation of a school’s current teaching strategy, and its inclusive approach allows all children to participate.

Low-cost, large-group interventions which improve performance and ability for low-achieving pupils, which are easily transferable between local areas and contexts, and which can be delivered by the class teacher without upheaval of current teaching strategies, particularly in areas of relative deprivation, would be a very cost-effective addition to current provision. E.L. Hill (2010) wrote ‘Classroom interventions that do not single out specific children, and appear to benefit all children, will be crucial in improving outcome for all’ (p. 888).

E.L. Hill (2010) contended that the development of motor skills (i.e. the development of good physical muscle control) is closely related to cognitive achievement, and emphasised how the contribution of skill (or difficulty) in one domain can have a positive (or negative) impact on development in other apparently unrelated areas. Our data support this view.

The use of physical intervention programmes has caused considerable controversy. However, the new model of ‘embodied cognition’ provides an exciting possibility that improving brain/body communication and body control could indeed improve cognitive performance, by improving the effectiveness of the whole interactive brain/body problem-solving system.

This study suggests that well-targeted choices of physical, visual and auditory intervention elements which have a firm grounding in science (e.g. the key role of rhythm in the development of phonological awareness) have led to a viable educational intervention approach. Larger scale and independent trials are required to determine whether this may be a new paradigm for education.

Footnotes

Acknowledgements

We thank the participating schools, teachers and children for their invaluable contributions to these pilot studies. Professor Tim Brighouse is thanked for his support in providing contacts and recommendations so that the trials could be set up. Move4words is available to schools via a Community Interest Company, formally regulated by the UK Government Regulator to ensure the company works for the benefit of the community and is not for profit.

Funding

This study received no grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Amadeo

Marshall

(2013). Low educational attainment in England: An audit. In Marshall

(Ed.), The tail: How England’s schools fail one child in five – And what can be done (pp. 135–145). London, England: Profile.

Barsalou

L. W.

Kyle Simmons

Barbey

A. K.

Wilson

C. D.

(2003). Grounding conceptual knowledge in modality-specific systems. Trends in Cognitive Sciences, 7, 84–91.

Best

J. R.

(2010). Effects of physical activity on children’s executive function: Contributions of experimental research on aerobic exercise. Developmental Review, 30, 331–351.

Booth

J. N.

Leary

S. D.

Joinson

Ness

A. R.

Tomporowski

P. D.

Boyle

J. M.

Reilly

J. J.

(2014). Associations between objectively measured physical activity and academic attainment in adolescents from a UK cohort. British Journal of Sports Medicine, 48, 265–270.

Boncoddo

Dixon

J. A.

Kelley

(2010). The emergence of a novel representation from action: Evidence from preschoolers. Developmental Science, 13, 370–377.

Boulenger

Roy

A. C.

Paulignan

Deprez

Jeannerod

Nazir

T. A.

(2006). Cross-talk between language processes and overt motor behavior in the first 200 msec of processing. Journal of Cognitive Neuroscience, 18, 1607–1615.

Brooks

(2013). What works for children and young people with literacy difficulties? London, England: The Dyslexia-SpLD Trust.

Budde

Voelcker-Rehage

Pietraßyk-Kendziorra

Ribeiro

Tidow

(2008). Acute coordinative exercise improves attentional performance in adolescents. Neuroscience Letters, 441, 219–223.

Byl

N. N.

Byl

F. M.

Rosenthal

J. H.

(1989). Interaction of spatial perception, vestibular function, and exercise in young school age boys with learning disabilities. Perceptual & Motor Skills, 68, 727–738.

10.

Carreiras

Seghier

M. L.

Baquero

Estévez

Lozano

Devlin

J. T.

Price

C. J.

(2009). An anatomical signature for literacy. Nature, 461, 983–986.

11.

Carreiras

Armstrong

B. C.

Perea

Frost

(2014). The what, when, where, and how of visual word recognition. Trends in Cognitive Sciences, 18, 90–98.

12.

Chaddock-Heyman

Erickson

K. I.

Holtrop

J. L.

Voss

M. W.

Pontifex

M. B.

Raine

L. B.

Kramer

A. F.

(2014). Aerobic fitness is associated with greater white matter integrity in children. Frontiers in Human Neuroscience, 8, 584.

13.

Claxton

(2012). Turning thinking on its head: How bodies make up their minds. Thinking Skills and Creativity, 7, 78–84.

14.

Cotman

C. W.

Berchtold

N. C.

(2002). Exercise: A behavioral intervention to enhance brain health and plasticity. Trends in Neurosciences, 25, 295–301.

15.

Davis

Barquero

Compton

D. L.

Fuchs

L. S.

Fuchs

Gore

J. C.

Anderson

A. W.

(2011). Functional correlates of children’s responsiveness to intervention. Developmental Neuropsychology, 36, 288–301.

16.

Department for Education. (2011a). How do pupils progress during Key Stages 2 and 3? (Research Report, DFE-RR096). Retrieved from https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/182413/DFE-RR096.pdf.

17.

Department for Education. (2011b, January). Special educational needs in England (Statistics). Retrieved from https://www.gov.uk/government/publications/special-educational-needs-in-england-january-2011.

18.

Diamond

Lee

(2011). Interventions shown to aid executive function development in children 4 to 12 years old. Science, 333, 959–964.

19.

Donnelly

J. E.

Lambourne

(2011). Classroom-based physical activity, cognition, and academic achievement. Preventive Medicine, 52(Suppl.), S36–S42.

20.

Etnier

J. L.

Nowell

P. M.

Landers

D. M.

Sibley

B. A.

(2006). A meta-regression to examine the relationship between aerobic fitness and cognitive performance. Brain Research Reviews, 52, 119–130.

21.

Gallese

Lakoff

(2005). The brain’s concepts: The role of the sensory-motor system in conceptual knowledge. Cognitive Neuropsychology, 22, 455–479.

22.

Geffen

G. M.

Jones

D. L.

Geffen

L. B.

(1994). Interhemispheric control of manual motor activity. Behavioural Brain Research, 64, 131–140.

23.

Gentilucci

Corballis

M. C.

(2006). From manual gesture to speech: A gradual transition. Neuroscience & Biobehavioral Reviews, 30, 949–960.

24.

Hattie

(2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Abingdon, UK: Routledge.

25.

Henderson

Barca

Ellis

A. W.

(2007). Interhemispheric cooperation and non-cooperation during word recognition: Evidence for callosal transfer dysfunction in dyslexic adults. Brain and Language, 103, 276–291.

26.

Hill

E. L.

(2010). The importance of motor skill in general development. Developmental Medicine & Child Neurology, 52, 888.

27.

Hill

Williams

J. H.

Aucott

Milne

Thomson

Greig

Mon-Williams

M. A. R. K.

(2010). Exercising attention within the classroom. Developmental Medicine & Child Neurology, 52, 929–934.

28.

Hung

I. W.

Labroo

A. A.

(2011). From firm muscles to firm willpower: Understanding the role of embodied cognition in self-regulation. Journal of Consumer Research, 37, 1046–1064.

29.

Hurry

Sylva

(2007). Long-term outcomes of early reading intervention. Journal of Research in Reading, 30, 227–248.

30.

Hyatt

K. J.

Stephenson

Carter

(2009). A review of three controversial educational practices: Perceptual motor programs, sensory integration, and tinted lenses. Education and Treatment of Children, 32, 313–342.

31.

Ionescu

Vasc

(2014). Embodied Cognition: Challenges for Psychology and Education. Procedia-Social and Behavioral Sciences, 128, 275–280.

32.

Jordan-Black

J. A.

(2005). The effects of the Primary Movement programme on the academic performance of children attending ordinary primary school. Journal of Research in Special Educational Needs, 5, 101–111.

33.

Kashihara

Maruyama

Murota

Nakahara

(2009). Positive effects of acute and moderate physical exercise on cognitive function. Journal of Physiological Anthropology, 28, 155–164.

34.

Kavale

Mattson

P. D.

(1983). ‘One jumped off the balance beam’: Meta-analysis of perceptual-motor training. Journal of Learning Disabilities, 16, 165–173.

35.

Krafnick

A. J.

Flowers

D. L.

Napoliello

E. M.

Eden

G. F.

(2011). Grey matter volume changes following reading intervention in dyslexic children. Neuroimage, 57, 733–741.

36.

Krampen

(2010). Improvement of orthography test performance by relaxation exercises: results of a controlled field experiment in basic secondary education. Educational Psychology, 30, 533–546.

37.

Lambourne

Tomporowski

(2010). The effect of exercise-induced arousal on cognitive task performance: A meta-regression analysis. Brain Research, 1341, 12–24.

38.

Lennert

Martinez-Trujillo

(2011). Strength of response suppression to distracter stimuli determines attentional-filtering performance in primate prefrontal neurons. Neuron, 70, 141–152.

39.

Mant

Wilson

Coates

(2007). The effect of increasing conceptual challenge in primary science lessons on pupils’ achievement and engagement. International Journal of Science Education, 29, 1707–1719.

40.

McPhillips

Hepper

P. G.

Mulhern

(2000). Effects of replicating primary-reflex movements on specific reading difficulties in children: A randomised, double-blind, controlled trial. The Lancet, 355, 537–541.

41.

Moritz

Yampolsky

Papadelis

Thomson

Wolf

(2013). Links between early rhythm skills, musical training, and phonological awareness. Reading and Writing, 26, 739–769.

42.

Olmstead

A. J.

Viswanathan

Aicher

K. A.

Fowler

C. A.

(2009). Sentence comprehension affects the dynamics of bimanual coordination: Implications for embodied cognition. The Quarterly Journal of Experimental Psychology, 62, 2409–2417.

43.

Paul

L. K.

(2011). Developmental malformation of the corpus callosum: A review of typical callosal development and examples of developmental disorders with callosal involvement. Journal of Neurodevelopmental Disorders, 3(1), 3–27.

44.

Perlovsky

L. I.

Ilin

(2013). Mirror neurons, language, and embodied cognition. Neural Networks, 41, 15–22.

45.

Pontifex

M. B.

Saliba

B. J.

Raine

L. B.

Picchietti

D. L.

Hillman

C. H.

(2013). Exercise improves behavioral, neurocognitive, and scholastic performance in children with attention-deficit/hyperactivity disorder. The Journal of Pediatrics, 162, 543–551.

46.

Rasberry

C. N.

Lee

S. M.

Robin

Laris

B. A.

Russell

L. A.

Coyle

K. K.

Nihiser

A. J.

(2011). The association between school-based physical activity, including physical education, and academic performance: A systematic review of the literature. Preventive Medicine, 52(Suppl.), S10–S20.

47.

Reynolds

Nicolson

R. I.

Hambly

(2003). Evaluation of an exercise-based treatment for children with reading difficulties. Dyslexia, 9, 48–71.

48.

Shillcock

R. C.

McDonald

S. A.

(2005). Hemispheric division of labour in reading. Journal of Research in Reading, 28, 244–257.

49.

Snowling

M. J.

Hulme

(2003). A critique of claims from Reynolds, Nicolson & Hambly (2003) that DDAT is an effective treatment for children with reading difficulties – ‘Lies, damned lies and (inappropriate) statistics?’ Dyslexia, 9, 127–133.

50.

Solan

H. A.

Shelley-Tremblay

J. F.

Hansen

P. C.

Larson

(2007). Is there a common linkage among reading comprehension, visual attention, and magnocellular processing? Journal of Learning Disabilities, 40, 270–278.

51.

Sommerville

J. A.

Woodward

A. L.

Needham

(2005). Action experience alters 3-month-old infants’ perception of others’ actions. Cognition, 96, B1–B11.

52.

Stein

(2003). Evaluation of an exercise based treatment for children with reading difficulties. Dyslexia, 9, 124–126.

53.

Tanner

Brown

Day

(2011). Evaluation of Every Child a Reader (ECaR) (Research Report DFE-RR114). London: Department for Education.

54.

Tomporowski

P. D.

Lambourne

Okumura

M. S.

(2011). Physical activity interventions and children’s mental function: An introduction and overview. Preventive Medicine, 52(Suppl.), S3–S9.

55.

Uhrich

T. A.

Swalm

R. L.

(2007). A pilot study of a possible effect from a motor task on reading performance. Perceptual and Motor Skills, 104, 1035–1041.

56.

Weare

(2013). Developing mindfulness with children and young people: a review of the evidence and policy context. Journal of Children’s Services, 8(2), 141–153.

57.

Wilson

(2002). Six views of embodied cognition. Psychonomic Bulletin & Review, 9, 625–636.