Children's Cognitive Behaviour Report - August 2007

ChangesinChildren's CognitiveDevelopment at the

Start ofSchool in England2000­2006

Christine Merrell, Peter TymmsandPaul Jones

CEMCentre, DurhamUniversity, MountjoyBlock4, StocktonRd., Durham,

England.

Contact:Christine.Merrell@cem.dur.ac.uk

Abstract

Englandhas seen massivechanges in theEarly Yearsoverthelast fewyears. Thereis nowan official early childhoodcurriculum, free nursery education for three­year­olds, the Sure Start programme has started, the Neighbourhood Nurseries programme has been introduced for the most deprived communities and a national network of children's centres was launchedin2003. Therehave also been important demographic changes. During this period, the CEM Centre at Durham University has been collecting consistent data from many thousands of children when they start schoolat the age of four on a range of variables that have been chosen because they are good predictors of latersuccess. These include vocabulary, concepts about print, letter and word identification, phonological awareness, ideas about maths, digit identification andsimplearithmetic(without any formal notation). Theextentto which thousands of children's scores on these measures have changed from 2001 to 2006 inclusively are examinedandthelinkbetween themajor initiatives andthefindings arediscussed.

Introduction

Since 1997, the English government has introduced several national and local initiatives intended to enhance the lives of children in England and to reduce the impact of poverty and social deprivation. Goodquality childcare andeducationinthe early years havebeen ahighnational priority andeducation intheearly years was one

of themainprovisions of the2002Education Act. TheAct introduced theFoundation Stage of theNationalCurriculum for childrenfrom aged3 years to theendof the first year at school (the Reception year). The Foundation Stage has specific curriculum guidance across sixareas of learningandastatutory assessment, theFoundation Stage Profile, which uses practitioners' observations to monitor each child's progress. Although many children will attend a pre­school setting and the full Reception year, they are not legally required to start attending school until the start of the term after their fifth birthday, unless they are home educated,so, potentially, somechildren will experience the full Foundation Stage and others will miss much of it. In practive almost all children now start school at the age of 4 and very high proportion have been tonursery beforethat.

Further initiatives cover a wider age­range than the Foundation Stage. SureStart' is an ongoing, widely implemented initiative supported by the Government that aims to achieve better outcomes for children, parents and communities (Sure Start 2007). Socio­economic status is related to academic achievement (Bordieu and Passeron, 1977) and so Sure Start is predominantly aimed at deprived neighbourhoods. The Sure Start Local Programmes (SSLPs), whichwere set up to improve the well­being, attainments and life chances of all children aged 0 ­ 4 years old in each area and to support their families, include a wide variety of local programmes. By 2004, there were 524 SSLPs established aimed at helping almost half a million children living in disadvantaged areas. Evidence that these initiatives have been successful is limited since many of the programmes are still fairly new, however some evaluations have been completed (see for example Brooks et al. 2003) and the National Evaluation of Sure Start (2007) is ongoing. The National Evaluation has evaluated SSLPs in terms of their implementation and their effectiveness, and has found variation between schemes (Anninget.al., 2007, National Evaluation of SureStart, 2005).

Another large scale UK Government initiative within the last 10 years has been the establishment of Education Action Zones' (EAZs) aimed at raising educational standards in disadvantaged urban and rural areas. EAZs usually run for three years with the possibility of extended funding for a further two years. The first round of EAZs were set upin1998, withasecondroundintroducedin1999. Someo f theEAZ programmes includedinterventions for pre­school andschool children. An evaluation of thesecondroundof EAZs by OFSTED(2003) reportedvariablefindings. Amongst the conclusions, it reported that "Although some success was evident, overall, the headway madeby zones, andtheschools in them, was too variable.In the majority of zones there was not enough deliberate and sustained attention to tackling difficult common issues".

More recently in 2003, the Government published a green paper called Every Child Matters, which prompted a debate about services for children, young people and families. Following theconsultation, theGovernment publishedEvery Child Matters: theNext Steps, andpassedthe Children Act 2004, providinglegislationto support the development of integratedservices to meet theneeds of children from birth toage 19 years.

Another development in 2003 was the Government's rebranding of Early Excellence Centres, Sure Start projects that offer childcare and Neighbourhood Nurseries, as Children's Centres which aim to integrate daily care, education, family support and health services. Otherinitiatives aretoo numerous tomention.

If these recent programmes have enhanced the cognitive development of young children, there should be evidence of a reduction in the previously documented gap between children from affluent and deprived neighbourhoods. There may also be a general improvement in thecognitivedevelopmental levels of youngchildren.

Monitoring improvements over time requires reliable assessment data, whose content remains unchanged,collectedon awide scaleover anextended period. Although the intention of the statutory Foundation Stage Profile is to monitor children's progress, in practise it cannot beusedfor this purposebecause it does notdiscriminateamongst the above average children, it is dependent on judgements which may change over time, it has only partially investigated psychometric properties and it was not introduced until 2003. By contrast, the Performance Indicators in Primary Schools (PIPS) On­entry Baseline Assessment collects very reliable objective data which has well­establishedpredictivevalidity andhas remainedunchangedfor several years(see

for exampleCEM Centre1999,Tymms etal. 2000).

Measures andSample

The Performance Indicators in Primary Schools (PIPS) project is one of several projects run from the Curriculum, Evaluation and Management (CEM) Centre at Durham University (CEM Centre, 2007), which aim to provide schools with data on the attainment, progress and attitudes of their pupils. The CEM Centre provides assessments and monitoring systems for children aged 3 – 18 years. Schools (and sometimes whole education authorities) subscribe to the projects and CEM analyses data from more than one million children each year. The CEM Centre provides the necessary assessments andthen marks andanalyses thedatafor schools, andprovides them with standardised feedback on the attainment, attitudes and progress of their pupils. As aresult of theseservices, theCEM Centreholds largelongitudinal datasets thatcan befurtheranalysedfor research purposes such as this paper.

The PIPS On­entry Baseline Assessment (PIPS BLA) was developed by the CEM Centre and is administered within the first six weeks of a child starting compulsory education on an individual basis, taking approximately 20 minutes per child. Its content is basedon areas of children's development which have been shown to bethe best predictors of later success or difficulty at school (Tymms and Middleton, 1995 andTymms 1999). Theassessment is available intwoversions atext version, where the administrator works through a book with the child, or a computer­delivered version. This paper analysed data from children who were assessed with the computer­delivered version. The assessment includes measures of vocabulary, early reading, mathematics andphonological awareness sections, specifically:

• Writing–thechildis askedtowritehis/herown nameandthequality of

writingis scoredagainst examples.

• Vocabulary –thechildis askedto identify objects embeddedwithin apicture.
• Ideas about reading–assesses concepts about print.
• Repeating Words – the child hears a word and is asked to repeat it in this assessment of phonological awareness.
• RhymeDetection –thechildselects awordtorhymewith atarget word froma

choice of three options in this assessment of phonological awareness.

• Letter identification –afixedorder of mixedupper andlowercaseletters.
• Wordrecognition andreading. This starts with wordrecognition andmoves

on to simple sentences that the child is asked to read aloud. The words within these sentences are highfrequency and common to mostreading schemes. This is followed by a moredifficult comprehension exercises which requirethechildto readapassage and at certain points select onewordfrom a choiceof three that best fits that position in thesentence.

• Ideas about mathematics –assessment of understandingof thevocabulary

associatedwith mathematical concepts.

• CountingandNumerosity –thechildis askedto countfour objects. These

disappear from the screen and then the child is asked how many objects they saw. This is repeatedwith seven objects.

• Sums –addition andsubtraction problems presentedwithout symbols.
• Shape identification.
• Digit identification –single, two­digits andthree­digits.
• Maths problems –includingsums with symbols.

The sample consisted of 124 primary schools in England that carried out the PIPS BLA with all of their children entering the Reception class in September every year from 2001 to 2006 inclusively. The number of pupils assessed each year is shown in Table1.

Table1 Number of PupilsAssessed

Additional informationon the pupils' date of birth, sex, special needs, first language, entitlement to free school meals and the number of terms in pre­school was also recorded.

Results ChangesinBackgroundVariables

Firstly, the stability of the number of pupils with English as an Additional Language (EAL), the number of pupils entitled to Free School Meals (FSM), the average numberof terms of attendance at nurseryand sex of thechildrenwas investigated. If the income of a family is below a certain threshold, the children in that family are entitled to free school meals, therefore FSM is an indicator of deprivation level. If these sample characteristics changed significantly over the period of investigation, they should beconsideredwhen interpretingthetrends found inthePIPS BLA scores overtime. Table2 shows thetotal numberof pupils for whom backgrounddatawere available each year and the percentage of EAL, FSM, the average number of terms spent in nursery andthepercentages of boys andgirls.

Table2 InformationaboutBackgroundVariables

Therewas no significant difference between the number males and females each year or over time. These figures for FSM, EAL and nursery attendance are illustrated in Figure 1 below. Attendance at nursery is expressed as the percentage of children attendingratherthan themean length of time.

Figure1 BackgroundVariables

100

90

80

70 % Englishas

60 AdditionalLanguage

% FreeSchool

50

Meals

40 % Nursery

30 Attendance 20

10

0

2001 2002 2003 2004 2005 2006

Although graphically, thedifferences in EAL and FSMovertime looksmall, all three variables didchangesignificantly (p<.05) overtheperiodof timeof interest.

Nationally, thepercentage of children entitledto free school meals inEnglish nursery and primary schools was estimated to be 16.9% in 2005 and 16% in 2006 (DfES, 2006). The percentage of pupils with English as an Additional Language in English primary schools was 11.5% in 2005 and 12.5% in 2006 (DfES, 2006). The percentage free school meal entitlement of the samples analysed in this paper was a little higher than the national average in 2005 but slightly lower in 2006. The percentage of pupils for whom English was an additional language was noticeably higher in the samples analysed in this paper than the national figure in2005and2006 perhaps becauseit deals with youngchildren ratherthan thefull primary age range.

Attainment on entry to school was compared to the national average. To do this, the mean standardised total PIPS BLA score and standard deviation for the schools included in the analysis for this paper were compared to the national sample. These areexpressedas T scores with amean of 50andstandarddeviation of 10. Thesample only differed significantly from the national average in 2003 (sample mean = 49.73) and 2006 (sample mean = 49.68). Although statistically lower than the national average,less than 0.05of astandarddeviationis very small ineducational terms.

ChangesinPIPSBLAScoresOverTime

Table 3 shows the mean raw scores and standard deviations from the PIPS BLA for early reading, phonological awareness, vocabulary, mathematics and the total score each year between 2001and2006inclusive. TheEarly Readingscalewas made upof Writing, Ideas about Reading, Letter Identification, Word Recognition and Reading. The Phonological Awareness scale was made up of Repeating Words and Rhyme Detection. TheEarly Mathematics scalewas madeupof Ideas about Maths, Counting and Numerosity, Sums, Digit Identification, Shape Identification and Number Problems. the Total raw score, which is the total of the Early Reading, Early Maths, Phonological Awareness andPictureVocabulary scores.

Table3 Raw Scores fromPIPSBLA

Themean scores areillustrated in Figure2.

Figure2 Raw Scores fromPIPSBLA

70

60

50

EarlyReading

40 EarlyMaths

PhonologicalAwareness 30 Picture Vocabulary

Total

20

10 0

2001 2002 2003 2004 2005 2006 Academic Year

Looking at Figure 2, it is difficult to discern any changes over time. The total score did not change significantly over time however a one way analysis of variance did indicate statistically significant differences over time for the four separate areas. Although significant, these differences were very small, for example comparing the mean scores from 2001 with those of 2006, the Effect Sizes were ­0.02 for early reading, 0.11 for phonological awareness, 0.04 for mathematics and ­0.03 for vocabulary.

The analysis of the background variables of FSM, EAL and nursery attendance over time indicatedthat thesechanged significantlyand thereforemight havean impact on theBLAscores. Age at timeof assessmentwas stable overtheyears and thechildren

in the sample started school in the autumn term. Table 4 and Figure 3 show the predicted mean scores for early reading, phonological awareness, mathematics and vocabulary aftercontrollingfor FSM,EAL, age at timeof assessment andsex, but not for numberof terms in nursery. Table4also shows thestandarddeviations.

Table 4 Predicted Raw Scores from PIPS BLA after controlling for FSM, EAL, age andsex

Figure3 Raw Scores fromPIPSBLA after controllingfor FSM, EAL, ageand sex

30

25

20 Early Reading

Early Maths

15

Phonological Awareness 10 PictureVocabulary

5

0

2001 2002 2003 2004 2005 2006 AcademicYear

After controlling for these variables, Figure 3 essentially shows flat lines. The only significant change over time was for picture vocabulary but this was a very slight decline, as shown in Table4.

As seen earlier in Table 2 and Figure 1, the percentage of children attending nursery increased over time, particularly between 2001 and 2002, and the mean number of terms also increased. It might have been expected that this increase in nursery attendance would be reflected in the BLA scores but, as shown in Figures 2 and 3, thattrendwas notfound.

TrendsforChildrenEntitledtoFreeSchoolMeals

Themainaim of many of theGovernment's initiatives inthe last 10yearshas been to enhance the lives of children from deprived backgrounds. Figures 5 to 11 compare the mean baseline scores of children entitled to FSM with those who were not. The error bars represent the 95% confidence intervals. These plots were restricted to children with English as their first language.

Figure5 EarlyReading andFSM

FSM No

Yes

30

25

20

2001 2002 2003 2004 2005 2006

AcademicYear

Figure6 EarlyMathematics and FSM

28

FSM No

Yes

26

24

22

20

18

2001 2002 2003 2004 2005 2006

AcademicYear

Figure7 PhonologicalAwareness andFSM

FSM 10

No Yes

8

6

2001 2002 2003 2004 2005 2006

AcademicYear

Figure8 VocabularyandFSM

FSM 16

No Yes

14

12

2001 2002 2003 2004 2005 2006

AcademicYear

As expected the mean BLA scores of the group of children entitled to FSM were significantly lower than the group from more affluent backgrounds for each of the four areas of development. The trends seen in all four figures indicate that the gap between children from deprived backgrounds and children from more affluent backgrounds neitherdecreasednor increasedstatistically overtime.

Investigating the sub­sections of the areas in more detail, there were some interesting differences between children entitledto FSMandchildren who werenot.

Figure9shows Ideas about reading, which assesses concepts about print includingan appreciation of thedifference betweenawriterandareader, thedifference between a picture and print, knowledge of where to begin reading a passage and being able to identify where a sentence begins and ends. The difference between those children entitled to FSM and those who were not increased significantly (p=0.009) over time although this was only becauseof largedifference in the 2006cohort andperhaps the followingyear this difference wouldhavedecreasedagain.

Figure9 Ideas about Reading andFSM

FSM No Yes

4.8

4.5

4.2

3.9

3.6

3.3

2001 2002 2003 2004 2005 2006

Academic Year

Thedifference between mean scores for Repeating Words of thetwo groups was very small each year and not significantly different in 2005. For this part of the assessment, children listened to a list of words of increasing complexity of which most were either unfamiliar or nonsense. They were asked to repeat each word immediately after hearing it. By contrast, Figure 10 shows a significant difference in Rhyme Detection each year between the two groups however the difference between the groups did not increase or decrease significantly over time. This part of the assessment also assessed phonological awareness. Children were presented with an item (for example hat') and asked to select an item from a choice of three which rhymed (for example cat', head', ring'). This skill requires a higher level of understandingof language andsounds than simplyrepeatingwords.

Figure10 RhymeDetection andFSM

FSM 5

No Yes

4.5

4

3.5

3

2.5

2001 2002 2003 2004 2005 2006

Academic Year

In the Early Mathematics scale, the ability to do simple sums, presented to the children without formal notation, showed a decline over time, with the exception of 2006 for the group of children not entitled to FSM. This is illustrated in Figure 11 wherea steady decline intheability of children who areentitledto FSMto do simple sums is apparentandstatistically significant.

Figure11 Sums andFSM

FSM No

Yes

3.5

3

2.5

2

2001 2002 2003 2004 2005 2006

AcademicYear

SummaryandConclusion

This paper analysed the development and skills of children starting compulsory education in England between the years of 2001 and 2006 inclusive. The sample comprised a group of 124 primary schools that had carried out the same baseline assessment over the period in question. Nearly 35,000 children were assessed. The sample was generally nationally representative with very slightly lower means in the 2003 and 2006 intakes. An analysis of background variables suggested that the characteristics of the sample were fairly stable over the years with the exception of attendance at nursery, which increased dramatically in 2002 but remained reasonably constant thereafter.

Analysis of thedatasuggestedthat:

• The overall development and skills of children in the sample of schools analysed (i.e. the total score on the PIPS BLA) remained stable between 2001

and2006inclusive. There was some change in the separate areas of early reading, phonological awareness, early reading and picture vocabulary but although statistically

significant, educationally thesechanges werevery small. After controlling for deprivation, language, age and sex, the only significant changeovertimewas avery slight declinein picturevocabulary. Therewas no indication that thegapbetween thoseentitledand not entitledto free school meals was decreasing. In one area of development (Ideas about Reading) it was increasing significantly, however this was a consequence of thedata from the2006cohortonly.

Why should the scores on the BLA have been so stable? On the one hand one might expect that the major initiatives would have resulted in measurable changes. On the otherhandit is important not to jumptoo rapidly toevaluativeconclusions. Thestudy had no measures of the involvement of the children in any of the national initiatives. Further, we do not have data on the interactions within families and how that had changedduring theperiod inquestion. But it is well knownthat attempts to influence the development of young children through early intervention are not easy. Whilst there are many positive examples of successful interventions there are also many examples of less successful initiatives (see for exampleRamey andRamey, 1998).

Ouraim is to provideasingleperspectiveon thechangingprofiles of children staring school in England during a time of rapid change. As further evidence builds from other studies a clear picture should start to emerge which will either confirm the pictureof stability foundin thepaperor challengeit with better information.

References

Anning, A. andtheNational Evaluation of SureStart (2007) Understanding

Variations inEffectiveness amongstSureStartLocalProgrammes: Lessons for SureStartChildren's Centres. National Evaluation Summary. Nottingham:DfESPublications.

Bordieu, P. andPasseron, J.C. (1977) Reproduction: ineducation, societyand culture.

London:Sage.

Brooks, G., P. Cole,M. Hines, M. Lewis , T. Ohn, Pollock, A., Ritchie, L. and

Vincent, C. (2003). AchievementInAdversity: RotherhamRawmarshSure Startin2002. Sheffield:University of Sheffield.

CEM Centre.(1999). PerformanceIndicators inPrimarySchools:Baseline

Assessment: TechnicalReport: CD­ROMVersion. Durham:CEM Centre, University of Durham.

CEM Centre(2007) www.cemcentre.org Websiteaccessed23rd July 2007.

Department for Education andSkills (2006) Schools andpupils inEngland, January

2006(Final). National Statistics First Release. London:DfES. http://www.dfes.gov.uk/rsgateway/DB/SFR/s000682/SFR38­2006.pdf Websiteaccessed23rd July 2007.

National Evaluation of SureStart(2005) EarlyImpacts ofSureStartLocal

Programmes onChildrenandFamilies. National Evaluation Report Number 13. Nottingham:DfESPublications.

National Evaluation of SureStart(2007) http://www.ness.bbk.ac.uk/ Websiteaccessed

on 21st July 2007.

OFSTED(2003) Education Action Zones:tackling difficult issues in round2zones.

Document referencenumber:HMI 1711www.ofsted.gov.uk. Website accessed12th August 2007.

Ramey, C.T. andRamey, S.L. (1998). Early Intervention andEarly Experience.

AmericanPsychologist 53(2):109­120.

SureStart(2007) www.surestart.gov.uk Websiteaccessedon 22nd July 2007.

Tymms, P. B. (1999) Baselineassessment, value­addedandtheprediction of reading.

JournalofResearchinReading, 22(1), 27­36.

Tymms, P. andMiddleton, P.(1995), Value­Added5– 7, ReportonProgress,

Curriculum Evaluation andManagement Centre, University of Durham.

Tymms, P., Merrell, C. andHenderson, B. (2000). BaselineAssessment andProgress

duringtheFirst Three Years atSchool. EducationalResearchandEvaluation 6(2):105 ­ 129.