Recently we have seen growing momentum on the complex topic of "evidence" in our field. Evaluation results are sometimes disappointing and the evidence base for policy and services is not strong. But it is reassuring to see more discussion on evidence, including the need to engage with "what is not working" in an open and transparent way.

The best research may not always have a good story to tell. In fact, rigorous research often reveals what went wrong with a program or policy. It is hard to know what to do when results are negative, and there can be anxiety about making these results public.

At AIFS, we are aware that our work needs to go beyond research findings if it is to help solve the problems faced by practitioners and policy-makers. We are committed to building useful evidence and getting it into action on the ground.

A culture that values experimentation and innovation needs to pay close attention to "failures" and to use this knowledge for continuous improvement. We need to be agile and adaptive. Sometimes called "radical incrementalism", a positive experimental approach to system and service reform also requires the maturity to have an honest dialogue across sectors (Halpern, 2015).

To improve outcomes for children and families across Australia in a sustainable way, we need research on the scalability of promising services, and to understand the principles underlying positive outcomes.

At AIFS, we are committed to discovering what works for families and building useful evidence to assist with policy development and service design and delivery.

We are putting all our efforts into this through our research program, our roundtables, webinars and conferences, and our policy and practice summaries. And as Family Matters approaches its historic 100th issue later in 2017, we are looking at ways that it, and AIFS more broadly, can continue to facilitate cross-sector and interdisciplinary dialogue and learning, to accelerate positive outcomes for children, families and communities. We ask you to complete the Family Matters survey in this issue (or online) to help us in that mission.

This issue of Family Matters opens debate about evidence, gathering as many perspectives as possible to see what it means to various people across the sectors. The issue is focused on how evidence is being generated and used in practice. It also highlights that there is a strong aspiration from government and service deliverers for evidence-based and data-driven services, but that it is important to identify and address the barriers to making this happen.

For example, it is common to focus on programmatic evidence at the expense of the expertise of frontline social service practitioners, or of understanding what matters most to service users and non-users. As a result, it is possible that we don't have a deep understanding of the complexity of what evidence is, where it can be generated from and how it can be used to improve services.

In addition, as our practitioner profiles in this edition show, we shouldn't underestimate the significant cultural change and capacity building that is required - within service delivery and policy agencies - to make an evidence-based approach work.

This issue

Building on the 2016 AIFS Conference theme of "research to results", issue 99 of Family Matters discusses current debates in Australia about the use of evidence to improve policy and practice in the child and family welfare sectors.

This Family Matters examines issues in the child and family welfare sector from different perspectives, providing a voice to all those affected by the pressing debates. In addition to articles from researchers and experts, the edition contains a series of "practitioner perspectives" on the use of evidence, and departmental updates on significant child and families programs.

AIFS' Acting Deputy Director (Research) Kelly Hand begins the edition overviewing how evidence is being used to inform practice in the current policy and program context. In the debate about using evidence in practice, Hand believes there is more agreement than disagreement between stakeholders.

Adapted from their keynote speeches delivered at the 2016 AIFS Conference, the lead articles from Professors Greg Duncan and John Lynch and from Justice Jennifer Coate discuss the importance of quality research to improving outcomes in child school achievement, early childhood programs and support for victims of child abuse. Despite their different fields, at the heart of each article is a commitment to relying on robust evidence to inform their work.

Professor Duncan discusses findings indicating the potential of combining child and parent-based programs in closing the gap between children of high and low socio-economic status in school achievement. In contrast, Professor Lynch takes a broader view on the issue of evidence. From the perspective of an epidemiologist, Lynch scrutinises the evidence for prominent early childhood programs, before arguing powerfully for researchers to "lift their game" to better assess programs. As a Commissioner on the Royal Commission into Institutional Responses to Child Sexual Abuse, Justice Coate outlines how the Royal Commission is being supported by a significant interdisciplinary research program (headed by Associate Professor Leah Bromfield) to advance understanding of the alarming issues raised by the Commission's investigation.

Remaining focused on "research to results", our shorter practitioner perspectives snapshot how evidence is being used on the ground. Working at both small and large community service organisations, these practitioners all echo the importance of having a good evidence base to understanding their clients and communities, and to selecting the right programs and interventions. Evidence is key to knowing whether they are getting the results they intended. However, another important message is repeated in these articles - that is, that practitioners need to be able to adapt single issue, evidence-based programs to meet the multiple issues and risks that many children and families struggle with daily.

Ros Baxter and Elly Robinson update on two significant initiatives, funded by the Department of Social Services, aimed at improving social services. Baxter overviews how "big data" is starting to be used to better understand how people are using social services in Australia. Robinson's update on the progress of the Expert Panel project outlines the benefits and challenges of an initiative that is deeply embedded in the "research-policy-practice interface". The emerging findings from both authors will undoubtedly interest practitioners seeking to improve services.

A pair of AIFS research reports complete the issue, illuminating the research in two important but under-explored areas - Indigenous program evaluation and supported playgroups for vulnerable young families.

Finally, it is with sadness that we say goodbye to two esteemed and longstanding members of the AIFS family, who have moved on from AIFS to take up exciting new challenges in child and family research. Tributes for Associate Professors Daryl Higgins and Ben Edwards appear in this issue, and we thank them for the great contributions they have made to AIFS and their commitment to working to improve the lives of families.