Positive behaviour in the early years: research report

CHAPTER THREE METHODS

3.1 Introduction

A general description is needed to provide background to the findings reported. Detail of specific measures used may be found in the Annexes. The measures used focused on parent, staff and service provider perceptions of young children's behaviour towards the end of the first school term of the year 2005-2006. A further set of measures was taken on a sub-sample of case study settings four months later.

Two local authorities: Edinburgh City and North Lanarkshire had agreed with Scottish Executive Education Department to host this research. They were asked to facilitate approaches to 3 pilot settings and 20 main study settings, and the associated staff and parents. Plans to seek informed consent from parents, practitioners and service providers for all aspects of the investigation were prepared as part of the ethics approval for the project. Participants have been given the opportunity to 'opt in' to the research, and subsequently to 'opt out' prior to each aspect of the investigation. University ethical procedures were in place at all points in the research design.

The study design aimed to recruit a sample of 1000 children and their early educators and families in each of two local authority areas. It was anticipated that 20 settings in each area would be needed to provide the stratified sample of children sought- aged 0-3 (100), 3-4 (300), 4-5 (300), and a transition group (300) entering Primary 1.

A representative sample of practitioners, parents and service providers was sought from the same settings. Given the very different staff-child ratios that operate either side of entry to school, numbers of participating professionals would be greater in early years nursery settings than in Primary 1.

Aiming for a sample of this size is a way of ensuring sufficient returns to enable significant valid and reliable results, to enable generalisation of findings, and to ensure potential for future research. Through working closely with all involved it was hoped to generate at least an average survey response. Actual responses varied by questionnaire, from 62.47% for the first round of well-being, to 34.28% on the parental adult strategies questionnaire. None of the average overall returns per measure fell below this figure. Table 3.2 gives the percentage returns on all measures used.

3.2 Number of children

The target number families and children per local authority and age group/strata are shown in the tables below. A small number of the settings originally approached by local authorities withdrew because of other commitments. The figures below show the numbers identified for the sample by age strata and by local authority.

Table 3.1 - Numbers of children identified for participation

	Number of Children
	0-3	3-4	4-5	P1
City of Edinburgh	124	212	288	240	864
North Lanarkshire	122	324	325	334	1105
Total	246	536	613	574	1969

The sample size was larger in North Lanarkshire than Edinburgh (Table 3.1), but percentage parental returns from Edinburgh exceeded returns from North Lanarkshire (Table 3.2).

3.3 Number of participating parents

As with the number of children there is a difference between actual families recruited, returns and the final number of cases with complete data. Some measures can be reported by the actual total per measure, eg the Parenting Daily Hassles which achieved 724 returns - 603 of which matched up with other data from measures completed.

Table 3.2 - Number of participating parents

Local Authority	Number of Children	Number of Parental Returns	Percentage Parental Returns
City of Edinburgh	864	360	42%
North Lanarkshire	1105	369	33%
Overall Total	1969	729	37%
Average Return Rate per Setting	45%
Number of parent cases with complete data	Although there were 729 parental returns overall, our merged parent file on which our analyses have been based has 603 cases.

3.4 Participating Settings

Forty-one settings took part in the study - 23 in North Lanarkshire and 18 in Edinburgh. Two of the Edinburgh settings provided access to groups of 'hard to reach' parents, but were not otherwise involved in the study. In each local authority nursery settings included 0-3 provision as well as 3-5 classes, schools and centres, including partner providers. Eleven primary schools in North Lanarkshire were included, five did not have an associated nursery class. In Edinburgh five primary schools were included, one of them being an independent school. In this way a sample of settings that were typical of each area were included in the study.

3.5 Deprivation indices

Of the 41 settings, 25 were in areas of high social deprivation, eleven were in the medium range, and six settings were in areas of low social deprivation. Not all settings received children from the local area - particularly in the case of independent and partner provisions, children travel outside their home area to nursery and school. Data was drawn from the Scottish Index of Multiple Deprivation (2004) and form the Social Focus on Deprived Areas, Scottish Executive National Statistics publication (2005) . Full detail is shown in Annex 3.

3.6 Instruments

In seeking ethical approval for the research approach, a range of possible instruments was identified to draw from as appropriate. Approval was sought, though not all instruments would necessarily be used, for the following range:

• questionnaires - practitioners, service providers and parents;
• semi-structures interviews - practitioners, service providers and parents;
• in-depth case-studies - practitioners, service providers and parents;
• observations in settings - occurrence, collaboration and management and promotion;
• self-evaluation performance indicators;
• tools for child monitoring and whole class screening (Laevers);
• information from councils, in relation to e.g. socio-economic levels;
• follow-up in-depth case studies;
• videos in nurseries - for tracking case study examples, and

documentary sources of information (including data from therapeutic services and social work services).

The measures used in Strand A were:

• Strengths & Difficulties Questionnaire (Goodman, 1997; 2005) completed by educators on all children;
• Leuven Well-being and Involvement Scales for Young Children. Educators were asked to do a whole class screening after training at the introductory conferences. A second round of this process-oriented child monitoring system was undertaken in April and May 2006;
• Hutchinson and Smith Screening Schedule completed at case study stage by practitioners in relation to a sample of children identified from the whole class screening as having consistently low well-being and involvement as well as a matched number of children with consistently high well-being and involvement;
• Adult Strategies Questionnaire: completed by educators;
• P1 children's transition records evaluation: Emotional, personal and social dimensions of practice (in one local authority);
• Transition Questionnaire completed by educators for children at all stages;
• Infant/Toddler Environment Rating Scale ( ITERS) and Early Childhood Environment Rating Scale ( ECERS) completed through observation by the project research team in all settings;
• Head of Centre/School Interviews - all settings, and
• Staff focus groups - Strand A research team supported by Childhood and Families Team.

The measures used in Strand B were:

• Daily Hassles Questionnaire (Crnic K A & Greenberg M T, 1990);
• Strengths and Difficulties Questionnaire completed by all parents;
• Adult strategies questionnaire completed by all parents;
• Transition Questionnaire completed by all parents;
• POMS - child profile in four domains completed by a small number of parents in each case study setting, and
• Parent focus groups - Childhood and Families team supported by Strand A team.

The purposes of research tools used in the project are shown in table 3.3 below.

Table 3.3 - Purposes of the research tools

Research Tools	Aim of Research Tool	Parents	Educators	Research Team
Strengths & Difficulties Questionnaire	Looks at a range of behaviours in areas of emotions, concentration and relationships and addresses to what extent these behaviours are of concern and impact on home and school life.	v	v
Adult Strategies Questionnaire	Identifies commonly used strategies for promoting pro-social behaviour and addresses the extent to which professionals and parents feel skilled and supported	v	v
Transition Questionnaire	Looks at transitions between different stages and types of provision and addresses behavioural aspects at these times of change. What practices can be identified by staff and parents as successful in relation to supporting transitions?	v	v
Parenting Daily Hassles	Explores behaviours and events that occur in daily family life that may make parenting difficult	v
Observation Screening (whole-class): Well-being and Involvement	Identifies children's levels of well-being and involvement as key indicators of quality in education, positive behaviour and development		v
Background Information Form	Provides contextual socio-economic information on families participating in project	v
Infant and Toddler Environment Rating Scale & Early Childhood Environment Rating Scale	Identifies environmental setting aspects of care and education provision - used to contextualize questionnaire findings			v
Leuven Well-being and Involvement Scales	Whole class screening to establish child levels on scale 1-5.			v

3.7 Process

Research materials were sent out to all settings in November / December 2005 and again in January 2006 for late settings - clear and concise step-by-step guidance was developed to accompany all materials. This was differentiated for parents, practitioners and service providers. Instructions for giving each practitioner and family a personal code were included. Prior to January and February team observation visits, each setting was sent a summary of the structured interview schedule which included a request for documentation to be available on the visit day. Numbers of settings requested further training in using the well-being and involvement scales- this was undertaken by our research assistant.

In addition, documentary sources were gathered from local authorities and individual settings ( e.g. school handbooks, planning proformas, newsletters, transition records, behaviour policies, socio-economic deprivation indices etc.). The research team also undertook semi-structured interviews with service providers/heads of centres and schools.

To gain further contextual information, research team members observed in all settings using the Infant and Toddler Environment Rating Scale and Early Childhood Environment Rating Scale after having been trained in house on usage of the tools. These are widely used early years provision quality assessment instruments and consist of respectively 39 and 43 items in the following overarching categories.

Table 3.4 - ECERS and ITERS item categories

Infant & Toddler Environment Rating Scale (Strata: 0-3)	Early Childhood Environment Rating Scale (Strata: 3-4, 4-5 and P1)
Space & Furnishings	Space & Furnishings
Personal Care Routines	Personal Care Routines
Listening & Talking	Language-Reasoning
Activities	Activities
Interaction	Interaction
Program Structure	Program Structure
Parents & Staff	Parents & Staff

In parallel with the case study phase which completed the project, practitioners in all settings were asked to undertake further action-research by doing a second well-being and involvement screening. In the case study settings this sampling targeted a small number of children in each of the low, medium and high categories for well-being and involvement, with the aim of looking a little more closely at their behaviour in the context of the setting attended.

3.8 Response rate

Response rates are shown in full in the tables in Annex 1(ii). The average percentage returns against the total number of each measure distributed were 36.21% of the Background Information Forms, 36.77% of Parenting Daily Hassles Questionnaire, 62.47% of Well-being Observations, 61.35% of Involvement Observations, 36.21% of Parental Strengths and Difficulties Questionnaire, 63.64% of Staff Strengths and Difficulties Questionnaire, 34.28% of Parental Adult Strategies Questionnaire, and 35.65% of Parental Transitions Questionnaire. A percentage return cannot be given for either the Staff Adult Strategies Questionnaire (n=168) or the Staff Transitions Questionnaires (n=128) since there was no fixed target figure for distribution of these measures. The second round of Well-being and Involvement measures yielded 41.29% and 41.95% respectively.

3.9 Data gathered

These percentages represent 713 Background Information Forms, 724 Parenting Daily Hassles Questionnaires, 1,230 Well-being Observations, 1,208 Involvement Observations, 713 Parental Strengths and Difficulties Questionnaires, 1,253 Staff Strengths and Difficulties Questionnaires, 675 Parental Adult Strategies Questionnaires, and 702 Parental Transitions Questionnaires, 168 Staff Adult Strategies Questionnaires, 128 Staff Transitions Questionnaires, 813 second round Well-being and 862 second round Involvement measures.

3.10 Analysis

Results from the questionnaires provide both qualitative and quantitative data. Both quantitative and qualitative information was coded for analysis using SPSS, to provide both descriptive and correlational statistical analysis. Textual analysis of comments from parents and teachers was also undertaken using a grounded approach to developing the coding system, resulting in a series of systematic coding categories. A team of six completed all the coding and undertook reliability checks - a high level of congruence was achieved. Coding systems are available - a sample accompanies this report ( Annex 4). Application and analysis of the measures in the first phase of the project enabled the team to identify approaches for further in-depth case study which included interviews and focus groups with parents and professionals to elaborate the information provided in questionnaire returns.

3.11 Answering the four main questions

In these ways it is intended to be able to answer the four main questions posed in this project, which are:

1 What is the extent and nature of behaviour difficulties among children in early years and early primary settings?

2 What strategies do parents, practitioners and service providers use to manage behaviour and promote pro - social behaviour?

3 What practices can be identified by staff and parents as successful in relation to supporting transitions from nursery/pre-school to school?

4 What effective approaches to training and support can be identified for staff in early years settings?

3.12 Presentation of findings

The report has been organised to present the data by key questions and on the basis of two sets of case studies into settings, and into themes of special interest or concern. Results are presented by age strata and across the sample as a whole.