Does mobilisation increase family engagement with an early childhood intervention programme? A randomised controlled trial
Sarah Cotterill+, Peter John* and Alice Moseley°
+ Centre for Biostatistics, University of Manchester,
* School of Public Policy, University College
° Department of Politics, University of Exeter
Corresponding author: Sarah Cotterill, Centre for Biostatistics, University of Manchester, Research & Development Support Unit, Salford Royal NHS Foundation Trust, Stott Lane, Salford, M6 8HD, UK
Does mobilisation increasefamily engagement with an early childhood intervention programme? A randomised controlled trial
Abstract
We compare the effect of two methods of mobilisation – doorstep canvassing and postal appeals – on family attendance at early childhood Sure Start centresin England with a sampleof 3,444 families. Families who were not already using Sure Start services were randomly assigned to one of three conditions: a visit from an outreach worker providing information and encouragement; a leaflet about Sure Start; and a control group that received the usual service. We found no evidence of a significant difference in Sure Start attendance between the three groups: neither visits nor leaflets were effective in encouraging non-attenders to go to Sure Start.
Keywords
Mobilisation, Sure Start, Randomised controlled trial, Outreach
Introduction
Experimental research in political science suggests that doorstep canvassing is more effective than postal appeals in encouraging people to vote, and that both methods are more effective than no contact (Green Gerber, 2008; John Brannan, 2008; Karp, Banducci & Bowler, 2007; Pattie Johnston, 2003). Less is known about whether mobilisation methods can increase engagement with public services, although doorstep visits can encourage people to use a kerbside recycling service (Cotterill, John, Liu & Nomura, 2009). In particular, we do not know whether personal methods are a more effective way for public service providers to make contact with citizensthan impersonal modes.
In this study weexaminethe effect of two mobilisation methods,doorstep visitsand leaflets,on family attendance at local Sure Start services in England.Sure Start is a universal state run service offering parenting and family support. Like the US Early Head Start programme, Sure Start focuses on enhancing children’s development and strengthening families, emphasising the importance of setting foundations in the first few years of life (Welshman, 2010; Love et al., 2005). Some families already use Sure Start, and of those who do not, many do not need the services or opt to use alternative provision. However, it is likely that a number of families are not aware of their local Sure Start provision, do not appreciate the detail of what is on offer, or simply fail torealise that they are eligible. It is possible that direct contact with families to inform them of what Sure Start has to offer might encourage them to visit their local centre and take up relevant services and activities. Outreach and home visiting services are an integral part of the design of Sure Start, aiming to encourage ‘hard-to-reach’ families to engage with the service (Ball Niven 2006); but these services may benefit from more precise knowledge about how to target families, which is where experimental research may help.
This randomised controlled trial compares the impact on attendance between families who receive a doorstep visit by a Sure Start outreach workerin addition to the usual service, those who receive a leaflet in the postin additional to the usual service and a control group who get the usual service only. Conventional wisdom and research findings in political science suggest that more personal methods, like doorstep visits, are better at gettingcitizens to participate. However we find that neither doorstep visitsnor leaflets are effective ways of encouraging non-registered families to attend Sure Start.
Mobilisation
Doorstep visits are short impromptu personal calls by a worker or volunteer to a person’s home to promote some behaviour or activity. The visit combines provision of information, friendly encouragement andgentle persuasion. This type of brief doorstep visit can be effective in a number of settings. Non-partisan doorstep visits have been shown in numerous studies to be effective in increasing voter turnout by about seven percentage points in local and national elections (John Brannan, 2008). Gerber, Green and colleagues (Gerber Green 2000a, b, 2001; Gerber, Green & Green, 2003; Green, Gerber & Nickerson, 2003; Green, 2004, Green and Gerber, 2008) show that face-to-face mobilisation has a demonstrable effect on voter turnout and is far more effective than more impersonal methods, such as telephoning and direct mail (see also McNulty, 2005). A recent systematic review of field experiments that test whether mail increases voter turnout at US elections found that ordinary mail has a small, although rather lacklustre effect of 0.1 percentage points, but the impact is increased to 2.6 percentage points by the inclusion of social pressure (indicating that voter turnout will be monitored or promising to disclose whether the recipient voted) and seems to be unrelated to salience of the election, partisan tone of the message or production quality (Green, Aronow & McGrath, 2010). Doorstep visits are used in other settings, including in the promotion of public services like Sure Start, and we might expect them to be similarly effective in such contexts. For example, face-to-face doorstep contact can also successfully persuade households to recycle their waste in local recycling schemes (Cotterill, John, Liu Nomura,2009; Bryce, Day & Olney, 1997; Reams Ray, 1993). There is a body of work on community mobilisation that indicates that asking citizens is a powerful way to get them involved (Lowndes, Pratchett and Stoker 2006).
A recent line of work suggests that impersonal methods of mobilising voters might do as well as more personal ones.These methods have the advantage of being low cost – both for the conveyer of the message and the recipient – and provide a way of sending messages in a timelymanner.They also have the added advantage that the general public has greater familiarity with these forms of contacting, such as direct mail and electronic mailings. One UK studyfinds comparable effects on voter turnoutbetween postal mailings as compared to telephoning (Fieldhouse et al, 2010, 2011). Direct mail is an effective way to raise charitable donations (Huck & Rasul, 2011; Cotterill, John Richardson 2010). In this study we extend this line of investigation to the context of local public services to determine whether doorstep visits or mail are effective in encouraging families to attend a local children’s centre.
There is some evidence that the effectiveness of mobilisation campaigns varies according to whothe canvassers are, suggesting that ‘the messenger matters’ (Green Gerber 2008: 39).Ethnic minority households are more likely to be reached by canvassersfrom the same ethnic group (Michelson, 2006). Canvassing seems to be more effective when carried out by people from the local area rather than outsiders (Sinclair, McConnell, Michelson & Bedolla, 2007). The mechanism here is unclear, but effective mobilisation depends on: a canvasser who is motivated to contact households and deliver the message persuasively; a canvasser who is comfortable in the environment they are working in; and householders who are willing to open the door to the canvasser and to be receptive to the message they deliver (Michelson, 2006: 1195). These studies have implications for the design of interventions aiming to recruit participation in social welfare schemes in that staff may be uncomfortable working in environments that are unfamiliar to them, householders may be reluctant to open the door to strangers, and some householders may be hesitant in dealing with official visitors.
Engaging families with an early childhood intervention programme
The Sure Start programme, first launched by the UK Government in 1998 drawing on lessons from the US Early Head Start programme (Welshman, 2010),aims to improve the life chances of young children, particularly those growing up in economically deprived areas, by providing support to both parents and children. The programmeis open to all families from pre-birth to their children reaching the age of five and includesthe core elements of integrated childcare and early learning; antenatal care;outreach and family services; links to the national employment and training service; and support for childminders and for children and parents with special needs (NAO, 2006; Clarke, 2010). Implementation varies locally, influenced by a complicated array of local priorities, implementer’s beliefs and unintentional consequences (Carter, 2012).Sure Start Centres are provided in a variety of community locations from schools to colleges, churches and standalone buildings and are offered to all families in their catchment area. Originallyimplemented in the most deprived communities in England, centres have now been established in every local authority in England[1] and local authorities have a duty to ensure sufficient number ofcentres to cover the needs of their local population.
Evidence from a national evaluation which compared five year outcomes for children and families in Sure Start designated areas as compared to those in areas that were, at the time of the study,without Sure Start provision but matched demographically, indicates better overall wellbeing of children in Sure Start areas (Melhuish, Belsky & Leyland, 2010). Families in Sure Start areas attained better scores on eight measures including physical health, obesity rates amongst children, parenting and family functioning,and life satisfaction of mothers. On eleven measures related to cognitive, emotional and social development there were no differences between Sure Start and non-Sure Start areas, and on two measures of maternal depression and contact with schools Sure Start areas fared slightly worse. Sure Start is valued by parentsfor its effect on self-esteem, parenting confidence and reduction in feelings of isolation (Williams Churchill,2006), providing a justification for researchto investigate the effectiveness of methodsfor recruiting families.
Parents can engage with Sure Start on four levels: taking up the services that are offered; involvement in service delivery as volunteers or paid workers; parental self help; and community action to achieve a controlling influence over services (Clarke, 2010). The interventions trialled in this study only seek to improve engagement at the first level, encouraging access to and take up of services. A variety of methods are already in practice for recruiting families to Sure Start. One method of advertising the programme is through door-to-door visits from outreach workers. Official guidance states that every family with a new baby should be visited by the Sure Start Centre within two months of the child’s birth as a way of raising awareness of the services on offer (DfES, 2006).Community health workers, social workers, general practitioners and other professionals also play an important role in signposting the service to parents (DfES, 2006). In some areas where service uptake is low,large-scale leaflet distribution has been carried out(Williams Churchill, 2006). Another important method of parents finding out aboutSure Start Centre services is through word of mouth (TNS Social, 2009). Research indicates that enrolment on parenting courses is highest among parents who can clearly see a benefit in participating (Spoth, Redmond & Shin, 2000), and where parents perceive that their children have behavioural problems (Seefeldt, Heinrichs & Eggert, 2008; Winslow et al., 2009). The content and quality of the service offered to parents affect the level of engagement (Chaffin et al., 2009). Monetary incentives areunsuccessful in increasing enrolment or frequency of attendanceat a parenting programme (Dumas, Begle, French & Pearl, 2010). While there are a variety of methods for promoting Sure Start services, there is limited evidence about which methods are most effective for increasing uptake.
Official guidance specifies that Sure Start Centres must make special efforts to target ‘hard to reach’ groups that are often excluded from mainstream services (DfES, 2006). Parental engagement with local Sure Start services varies across localities and across demographic groups. Case study research suggests that fathers and some ethnic minority groupsare under-represented interms of Sure Start service use(Tunstill, Allnock, Akhurst & Garbers, 2005; Williams Churchill, 2006; Potter Carpenter, 2008). Catering for a complex set of conflicting interests within Sure Start catchment areas can be difficult to negotiate and may lead to some groups feeling excluded (Clarke, 2010).Obstaclesto engagement include ‘logistic barriers’ and ‘interpersonal barriers’ (Winslow et al., 2009). Logistic barriers include lack of transport, problems with fitting attendance atSure Start activities around othercommitments and the financial cost of some of the Sure Start activities offered(Spoth, Redmond & Shin, 2000; Tunstill, Allnock, Akhurst & Garbers, 2005; Williams Churchill, 2006). Interpersonal barriers includeconfusion about eligibility,language barriers,culturally inappropriate services, mistrust of providers, prior negative experiences and fear of stigmatisation (Keller and McDade, 2000, Tunstill, Allnock, Akhurst & Garbers, 2005; Williams Churchill, 2006).There can be a perception that Sure Start ‘is not for me’, with parents rejecting unwanted labels that they are inadequate parents or live in poverty (Clarke 2010:110). Service providers can play an important role in ‘problem solving’ to help parents confront and overcome these barriers to participation (Winslow et al., 2009)
In the Englishcity where the current study was undertaken, 59 percent of eligible families were already registered with Sure Start. A comparison of registered and non-registered families in this local authority suggests that there is little difference between these two groups on key demographic variables, apart from an over-representation among the registered families of those from more deprived areas of the city. This may be a reflection of greater need for Sure Start Services by families in deprived areas or an indication of specific targeting of these areas by the local authority concerned. Registration with Sure Start and engagement with the service are not synonymous: families can register with Sure Start and then never attend, and can attend occasional sessions without being registered. This study sought to determine whether a mobilisation campaign could successfully attract those who were not already registered with Sure Start.
Research Design
Research Objectives and Hypotheses
The research objective was to evaluate the effectiveness of home visits and leaflets on the take up of Sure Start services. Theresearch question was: Are visits from an outreach worker providing verbal and written information about Sure Start more effective than written information alone or the usual service in encouraging attendance at Sure Start among non-registered families?Using a randomised controlled trial we tested whether a brief doorstep visit or a leaflet delivered by post was more effective than the usual service in promoting the Sure Start service.
Firstly, we expectedthat the posted leaflet wouldincrease attendance at Sure Start, because families receiving the leafletwould be better informed about Sure Start activities than those in the control group. Secondly, we expected that a short visit by an outreach worker would make it even more likely that families would attend Sure Start, because they would be better informed about Sure Start activities than those in the control group, but also becauseof the extra personal encouragementprovided by the doorstep visitor. A flow diagram of the experiment is shown in Figure 1.We follow the CONSORT guidelines for reporting randomised controlled trials (Schulz, Altman & Moher, 2010).
[Figure 1 about here]
Data and Randomisation
In the autumn of 2010, in collaboration with the local government, we organised a mobilisation campaign in an urban area in the North West of England to encourage families to attend local Sure Start centres. The target population was all families with a child aged between two months and eighteen months that were not already using Sure Start services. This age range was targeted because of the importance to Sure Start of establishing contact with families early in the life of the child, based on evidence that early childhood intervention programmes have the potential to alter the trajectories of vulnerable children (Brooks-Gunn, 2003). We omitted three electoral wards from the list: the first was an early childhood pilot area where a number of new ways of working were already being tested; the second was the site of a council by-election and UK electoral law prohibits the local authority from undertaking campaigns during elections; in the third area a Sure Start centre had not yet opened.The sampling unit was the child. Our local government partner identified children born during the previous 18 months from the Register of Births. All UK parents are legally required to register the birth of any child with a local Registrar within 42 days of the birth.Local government researchers checked the names and addresses of the births against a Sure Start databaseand removed families who were already registered to receive Sure Start services. They removed families of children who appeared on the local register of looked after children (children in local authority care) or who were listed on the register of deaths, to avoid causing any unnecessary distress. They undertook a check against the electoral register to identify families who had moved after the birth of the child, but discrepancies between the two databases meant that we could not undertake a comprehensive check. A summary of the procedure for identifying the study population is provided in Table 1.