Sleep Training, Breastfeeding and Daycare: What the Research Says

The approach

Many heated early-parenting debates rest on studies that cannot separate cause from correlation (families who make one choice differ in many other ways). Oster separates the better studies (randomized trials, careful comparisons) from the weaker ones, and argues that for many choices the honest answer is "the differences are small - choose what works for your family".

Examples she discusses (her reading of the evidence)

• Breastfeeding: real short-term benefits are well supported (for example fewer gastrointestinal infections, and some evidence for fewer ear infections and less eczema in infancy), while many long-term claims (IQ, obesity) look weaker once family differences are accounted for. Fed is what matters most; feeding decisions should not be driven by guilt.
• Sleep training: the studies she reviews do not find lasting harm to attachment or behavior and generally find it improves sleep; not sleep training is also a fine choice.
• Safe sleep: back-sleeping is strongly supported. On bed-sharing she discusses how risk concentrates in particular circumstances (sofas, smoking, alcohol, very young babies); the AAP still recommends room-sharing without bed-sharing, and Cradle follows the AAP on safe sleep.
• Daycare vs. home, screen time, potty training, early learning: effects in good studies are generally small; quality of care and family fit matter more than the category.
• Vaccines: she is clear the evidence supports vaccinating on the recommended schedule.

Her framework for decisions (from The Family Firm)

In The Family Firm she suggests running family decisions like a thoughtful organization: frame the question, fact-find, make a final decision, and follow up - and to agree on the family's "big picture" priorities first, so individual decisions have something to be judged against.

How to use this well

Useful for relieving guilt and making informed trade-offs. It is population-level evidence, not a statement about one child, and it does not override the pediatrician on medical questions.