Evidence LedgerFamily science, in context
Research in sequence

Ideas change.
The record shows how.

The field did not move in a straight line from one theory to one answer. It widened from separation and a primary caregiver toward observable relationships, context, multiple attachment networks, synthesis, and explicit limits on court use.

Eight turning points

How the research frame changed

1951 to the present evidence base

1951–1958

Separation and continuity enter public policy

Bowlby’s work foregrounded the developmental importance of a child’s enduring caregiving relationships, but was often simplified into a one-mother, one-child rule.

1960s–1970s

Attachment becomes observable

Ainsworth and colleagues developed relationship-specific observational methods. Schaffer, Emerson, Lamb, and others widened attention to multiple caregivers and fathers.

1970s–1990s

Context replaces simple deprivation claims

Rutter and developmental psychopathology research distinguished separation from deprivation and emphasized resilience, timing, context, and protective relationships.

1997–2005

Meta-analysis tests the theory

Quantitative syntheses estimated sensitivity, security, and disorganization associations, strengthening some claims while showing that effects were probabilistic and multifactorial.

1999–2013

Infant overnights become an empirical dispute

Solomon and George, McIntosh and colleagues, and Tornello and colleagues reported cautionary associations. Designs, measures, and causal interpretation became heavily debated.

2000–2019

More shared-care outcome studies become available

Reviews by Kelly and Lamb, Bauserman, Baude, Nielsen, and Steinbach increasingly reported equal or better average outcomes. Researchers continued to debate how much of the difference came from shared care itself and how much came from the kinds of families using it.

2014

Two integrations seek common ground

Warshak opposed blanket restrictions. Pruett, McIntosh, and Kelly joined attachment and involvement perspectives while retaining conditional caution about conflict and unestablished care.

2022–2024

Networks and limits become clearer

International consensus endorsed familiar caregivers, continuity, and attachment networks. New participant-level synthesis tested how relationships with both mothers and fathers operate together.

Focused researcher timeline

What McIntosh published after 2014

2010 to 2024, with later papers shown

McIntosh's publication history is often described as either permanent opposition to infant overnights or a complete reversal. Neither description is accurate. She continued to defend the caution in her earlier work, while her later publications placed much more emphasis on context, caregiving quality, safety, conflict, and protecting a child's relationships with both parents.

Read the organised infant overnight guideOpen Jennifer McIntosh’s publication archive
2010

The first report raised concern about some frequent overnight groups

The Australian report found poorer results on some parent-reported measures for selected groups of very young children who frequently stayed overnight. McIntosh and colleagues nevertheless argued for caution, particularly for infants and in difficult family circumstances.

The study did not prove that overnights caused the reported differences.

Read the report summary and evidence limits
2014

Attachment and an ongoing relationship with both parents are presented as compatible aims

Their framework still recommended closer attention where a parent was unfamiliar, conflict was frequent, safety was uncertain, or a demanding schedule did not suit the child.

McIntosh, Pruett, and Kelly rejected a general rule against all overnight care during the first three years. They said early security and continuing relationships with both parents could both be protected.

Read Part I: what the research was said to supportRead Part II: how the framework was applied
2015

McIntosh did not withdraw the original caution

McIntosh, Smyth, and Kelaher replied directly to Warshak. They defended the purpose, methods, findings, and stated limits of their overnight study. They also rejected the claim that their study supported primary maternal care or discouraged shared parenting by fathers.

Their conclusion was that science had not yet resolved the infant overnight debate.

Open the 2015 reply to Warshak
2015 to 2016

The focus widens from time alone to what children experience in each home

Sadowski and McIntosh interviewed children aged 8 to 12 who lived in shared-time arrangements. It did not study infants or prove that a particular percentage of care is best.

The work examined what helped children feel secure and content, what they missed while away from a parent, and which parental behaviours made shared care easier or harder.

Read On laughter and lossRead the related security and contentment study
2016

Shared care is examined as a balance of possible benefits and risks

Smyth, McIntosh, Emery, and Higgs reviewed shared-time parenting, defined as children spending about 30 to 50 percent of their time with each parent. The chapter considered both benefits and risks rather than treating shared care as automatically good or automatically harmful.

Its practical message was that outcomes depend on the child, the parents, their relationships, conflict, safety, and the way the arrangement operates.

Open the ANU record and citation
2017

The conditional approach becomes a parent education program

McIntosh and Tan reported a pilot program for separated parents of young children. This was intervention research, not proof that one overnight schedule produces better outcomes.

The program was designed to help parents consider their individual infant, their family circumstances, and a developmentally sensitive care arrangement.

Read the Young Children in Divorce and Separation study
2018 to 2021

Later attachment research places family risks and caregiving quality ahead of parent labels

McIntosh and colleagues studied couple relationship quality, intimate partner violence, risks present before and around birth, changes in attachment, and reports of post-separation safety concerns. Most of these papers did not test custody schedules, so they should be used to explain relevant risk and protective factors, not to prescribe a number of nights.

This body of work makes the surrounding caregiving environment more important to the analysis.

2018: couple relationship quality and attachment2019: family violence and attachment security2020: attachment stability and change2021: mother and father reports of safety risks
2021

Updated infant guidance states both sides of the practical risk

McIntosh and Booth said the evidence for babies and preschool children remained limited and that no one arrangement fits every family. They retained concern about violence, conflict, poor parenting, unsettled schedules, and repeated separation from an attachment figure.

They also stated that the long-term benefit of safe involvement by both parents was not in question and acknowledged the concern that too little overnight time may weaken the second parent-child relationship.

Read the 2021 revised infant guidance
2024

New population research shows that attachment can change with caregiving

McIntosh and colleagues followed an Australian population cohort from infancy to preschool. Secure attachment was observed in 59 percent of the sample at age one and 71 percent at age four. Gaining or losing secure status was associated with changes in maternal caregiving sensitivity. This was attachment research, not a study of post-separation overnight schedules.

The researchers found no difference in caregiving sensitivity scores for mothers and fathers of preschool children.

Read the abstract, figures, and full citation
2022 onward

The wider field draws a clearer boundary around attachment evidence

This is not a McIntosh paper. It is included so readers can compare her later conditional framework with the broader direction of the field.

International attachment consensus describes children as forming relationships with several familiar, non-abusive caregivers. It says attachment theory cannot determine an exact parenting-time schedule.

Read the international consensus and remaining disagreements

What changed in McIntosh’s published emphasis

Her later work does not treat attachment as a reason to exclude one parent. It recognizes that children can benefit from safe, continuing relationships with both parents. It increasingly asks how caregiving quality, safety, conflict, transitions, relationship history, and the child's own responses affect an arrangement.

What did not change

McIntosh did not retract the original infant study or endorse equal overnight care for every young child. She continued to argue that very young children can be vulnerable to unsafe, conflicted, unfamiliar, or poorly supported arrangements. Her later position is best described as conditional, not as a complete reversal.

What the later papers can support in court

They support asking for evidence about the actual care each parent provides. They support examining safety, cooperation, conflict, the child's relationship with each parent, and whether transitions are manageable. They do not support assuming that a parent is primary or secondary because of gender alone.

Research collection

Publication timeline

28 linked research records

2009
Family law research · Government evaluation using multiple national datasets

Evaluation of the 2006 Family Law Reforms

Rae Kaspiew, Matthew Gray, Ruth Weston, Lawrie Moloney, Kelly Hand, Lixia Qu & the Family Law Evaluation Team