The four main innovation methods Nesta brought to the trial discovery projects were data science, design, behavioural science and experimentation. Following initial workshops with each of the local areas, the project teams identified how these innovation methods might help to explore some of the challenges identified.

The project team in each local area then developed a bespoke project plan, drawing on combinations of these innovation methods, alongside qualitative research and desk research, to produce new actionable insights. Some examples of how we used these innovation methods in the discovery projects are set out below.

In all of the local areas, Nesta worked with local professionals to analyse existing early years datasets and produce new insights. Some of the data we looked at included:21

• Children’s outcomes in the Early Years Foundation Stage Profile (a measure of child development completed in reception year for all children in English schools).
• Children’s ASQ (Ages and Stages Questionnaire) scores.
• Take-up of the Health Review for two to two and a half year olds.
• Take-up of the free entitlement to early education at age two.

In York the project aimed to overcome barriers to parents taking up services for two year olds, particularly focusing on the universal Health Review and targeted early years childcare offers for two year olds. We analysed data on current take-up of the two year old health review and learned that York performs well on children’s outcomes compared with national and regional figures; around 94 per cent of children who receive a Health Review are at or above the expected level of development at age two. However, only around 73 per cent of parents are taking up the Health Review. This is a relatively low proportion compared with nearly 80 per cent nationally and 83 per cent regionally (in Yorkshire & the Humber). When we consider which children we know to be at the expected level of development (for example by analysing data for the full sample of families, including those who did not take up a Health Review), the proportion of children meeting this target reduces to 68 per cent.

We also analysed data on take-up of the Health Review in York both geographically and according to children’s ethnicity (see below). This shows considerable variation in take-up across the city and for different demographic groups.

This analysis led to further work to explore how take-up of the two year old health review could be improved among those groups not currently engaging (see more detail on this below).

We also began work on creating a live data dashboard to enable quicker and easier analysis of service data by early years professionals in York. The data dashboard is a tool into which real-time data on the two year old health review can be fed, to enable service delivery to become more responsive to communities. The dashboard allows us to monitor service usage by showing take- up across different geographical areas and demographic groups over time. In the future it could be used to monitor risk factors, use data to target resources more efficiently, and systematically aggregate and monitor feedback from families about their service experience.

Figure 5 below shows a mock-up of how the data dashboard could look when it is fully operational.

The project team in Leeds had identified the aims of improving children’s speech, language and communication outcomes and of understanding how to better meet the needs of families who are currently not accessing services. The starting point of this work was analysing data on children’s communication and language outcomes in reception year, and breaking this down geographically by children’s centre reach area (see map below). This provided a visual representation of how children’s outcomes vary across the city, so that we could identify geographical areas where children were currently less likely to meet expected levels of development. We also analysed the proportion of families eligible for free early education at age two whose children were enrolled in childcare.

Through this analysis we identified Harehills as an area where there were fewer children meeting expected levels of development in communication and language, and where there was also lower take-up of funded early education. We made Harehills the focus of qualitative research with parents and design work to generate ideas for how we could improve support for children’s communication and language development.

In Stockport, the project team sought to establish the impact of existing family support services across the local authority.

Initially we looked at the paths children followed in the ASQ-SE (Ages and Stages Questionnaire – Social and Emotional Development), a screening tool that is used during health visitor appointments at several points in a child’s early years. Looking at each ASQ-SE assessment, we found that 50 per cent of children who did not pass their first assessment then passed their second assessment. However, 17 per cent of children who did pass their first assessment then fell into ‘did not meet the expected level of development’ or were borderline for their second assessment.

Although this is a provisional analysis, this concerned Stockport practitioners as it indicates they may be missing some opportunities to identify children who would benefit from support. Following on from this analysis, the Stockport team are considering the possibility of introducing an interim assessment in an attempt to catch those children who are not continuing to meet their milestones.

Through our data analysis and qualitative research with parents, we learned that in Stockport there are more family support services concentrated in the first few months of children’s lives. Our analysis also found that improvements in children’s outcomes tended to occur around the same time that interventions were generally in place. This can be seen in results related to children’s development of fine motor skills. We saw a decline in fine motor skills as children get older, happening at the point at which fewer interventions are being provided. The team in Stockport have identified this as an avenue for potential further investigation, to explore the potential to extend their support offer for older age groups of children.

Following the initial data analysis and some qualitative research with practitioners in Leeds, we refined our original research question, so that the next stage of design work addressed the question:

Our main aims at this stage of the project were to deepen our understanding of the experiences of families in the Harehills area, and learn about parents’ behaviours and challenges relating to their children’s speech, language and communication development. We would use these findings to develop ideas for new solutions in collaboration with local parents and practitioners.

We recruited 13 local parents and interviewed them about their experiences of bringing up children in Harehills. Once we had interviewed all 13 parents, we carried out thematic analysis and also developed user journey maps (see Map 1) and relationship maps (see Map 2), to help visualise and understand their personal experiences, as well as their challenges with accessing support from local services. These maps were used to frame the conversations in two co-design workshops we ran with local parents and practitioners, to encourage the participants to empathise with the pain points experienced by parents and to think about solutions from their perspectives.

This map shows a family’s experience of seeking help with their young child’s language development. Some details were changed to protect the parent’s anonymity. This map was used in the two design workshops to help introduce the participants to some of the challenges that parents experience in accessing support and the range of services they interact with.

This map shows the network of relationships that one parent had in her community in Leeds. This is useful to help understand which organisations in her community this parent already interacts with and how she may wish to give and receive support from others. Alongside the map, some key information from the parent’s story is summarised. Some details were changed to protect the parent’s anonymity.

This map was used in the two design workshops to help participants think about the role that a range of services, organisations and community members might play in supporting children’s communication and language development.

We invited early years practitioners to the first design workshop to hear some of the findings from the interviews with parents. Practitioners thought about the challenges and barriers that parents were experiencing, as well as the parents’ interests and behaviours, such as their high use of outdoor play areas. They then generated a range of ideas for supporting young children’s speech and language development.

We invited parents and carers from Harehills to the second design workshop. Some of these parents and carers had participated in interviews previously and some were new to the project. At this workshop we also shared findings from the interviews and then invited the parents to generate their own ideas.

We also asked parents to give feedback on the practitioners’ ideas. We found that there was an encouraging amount of overlap, but parents also highlighted the additional need for accessibility in services (including interpreters for non-English speakers), flexibility (as parents would need to fit any additional activities into busy lives) and continued access to specialist services, such as speech and language therapy, for those who needed them. These insights will inform future service improvement work in Leeds.

In Stockport, we brought together a group co-design new sources of support that could of parents we had spoken to during the meet some of the parental needs identified interview stage for two online workshops, to in our work. The parents developed two prototypes:

The project team in York had identified the goal of increasing take-up of Health Reviews for two year olds. They decided to use insights from behavioural science to develop and test a new approach to inviting parents and carers to attend. The team drew on the EAST framework developed by the Behavioural Insights Team (BIT, 2014).23

Behavioural science tells us that people are more likely to do something if it is Easy,

Attractive, Social and Timely (EAST). These principles were applied to this intervention in the following ways:

• Easy: minimising the hassle or ‘friction cost’ to parents of making a Health Review appointment.
• Attractive: personalising the letter to the child.
• Social: making use of social norms – most people prefer not to break a commitment by not attending an appointment.
• Timely: encouraging parents to make a specific plan to attend the Health Review.

In the pilot, we changed the default for the Health Review appointment from an ‘opt- in’ model to an ‘opt-out’ model. Previously parents would have received a letter inviting

them to ‘opt-in’ to an appointment, and they would then have to call and arrange a time. The new ‘opt-out’ letter would invite parents to a pre-booked appointment. If they didn’t want to attend, they would need to call the Healthy Child Service to rearrange or cancel. By removing the administrative burden of signing up, the Healthy Child Service could potentially reach more parents who might otherwise forget to book an appointment. Both letter A (the previously used ‘opt-in’ letter) and letter B (the new ‘opt-out’ letter) are included in the appendix.

The new letter was then piloted in the York North children’s centre reach area. Initial findings were positive; during the month of the pilot, when the new letter was being used, nearly 70 per cent of children invited to an appointment in York North were brought to their appointment, compared with 43 per cent of children invited in York West and 18 per cent of children invited in York East, where the old version of the letter was being used.

As there were only small samples of parents involved, with between 22-45 children being invited to their Health Review in each children’s centre reach area, it was not possible to run a randomised controlled trial as part of this experiment (the sample size would have been too small to generate statistically significant findings). While these results are promising, it cannot be assumed that it was the new ‘opt-out’ letter that led to improved take-up of appointments in York North compared with York East and York West. However, given that the initial findings of the pilot project indicated some positive changes in appointment attendance, both Nesta and York would be interested in continuing this work and potentially evaluating it with a randomised controlled trial in the future, to create stronger evidence of its impact.

The project team in York also piloted text messaging surveys as a new method of gathering rapid feedback from parents and building an evidence base to inform future communications. Forty six parents responded to a 12-question survey, during which we explored some of the topics parents would most like advice on from health visitors and their motivations for attending or not attending the two-year health check. The team was then able to feed some of the findings from this survey into developing future communications with parents.

We learned through this survey that parents preferred the messages that emphasised the child health benefits of health reviews, as well as the qualifications of the nurses involved, rather than the messages that emphasised social norms.

Through trying out this new engagement method, we learned that communicating with parents via text message is a promising avenue for future work. Over half of the parents involved agreed to be contacted again in future and three-quarters said they would prefer to receive future updates about health appointments via text.

Due to the rapid nature of this project we did not obtain demographic data about the parents who took part in the text messaging surveys. It is therefore possible that they included parents who had already taken up a Health Review. In the future it would be valuable to do further work exploring the value of text messaging specifically with parents who had not taken up their child’s Health Review. This would help us to better understand this group of parents and which messages they respond to most positively.