Verified by 
There seems to be widespread misunderstanding of science experiments and what they can tell us. Everyone should be aware of the uncertainty in experiments and the lack of generalizability to every single person or situation.
The scientific experiment is useful for understanding general trends for the type of people studied, but it usually doesn't apply to everyone (unless you are studying the effects of highly toxic substances that kill all). Hypothesis testing, the usual method for determining significant effects, tests group difference. All that is needed is to find a significant difference between the mean (average effect) of the treated group and the mean for the untreated group (the "treated" group receives the experimental intervention—e.g., a new medication). Only a subset of the treated group—a few more than in the untreated group—has to be affected for significance to be found. Typically, unless it's about poisons, a considerable number of people are not affected by any particular medicine or treatment.
But even more important for parents of young children is to realize that there really aren't any human experiments that can be done to inform you how to parent at any given moment. So, for example, experiments that show the "success" of cry-it-out parenting might be interesting but they have several flaws.
Science cannot recommend particular parenting practices at particular times for a particular child in a particular context. Why not? Because parenting is like white-water rafting (but much harder)—there is too much unpredictability and changing circumstances.
Each parenting situation is unique because of shifting conditions—what happened earlier in the day, the uniqueness of the child's personality, the point of growth in various systems in the child, the relationship of the parent to the child, the context, the history of behavior in that context, the moods of the parent and child, and so on and on.
Further, studies related to parenting often don't look at all the variables of importance. For example, studies often look at outcomes over a short time period, when the real effects occur later (as with tobacco use or cell phone use—effects come decades later). This is what happens with studies of infant formula and for cry-it-out parenting. How many studies look at the effects on child learning in school, or on psychological wellbeing or health a decade or more later?
Science cannot do randomized controlled trials—the type of experiments preferred today—with children because they are unethical (e.g., randomly assign individual babies to groups, one group is touched in some regular fashion and the other group is never touched; then measure effects on some predicted outcome from no touch—this can't be done; but accidental studies of non-touch with babies and intentional studies with monkeys show detrimental effects of minimal touch). When scientists are able to do randomized, controlled trials with humans, it is for a short time period or on a small behavior that doesn't account for all the individual differences in families, parents, kids and situations. And so the results cannot tell you what to do in a particular situation or what the effects will be in 20 years.
Even if scientists could ethically set up an experiment with parents and young kids, it would not work, because it's your unique relationship that matters and that's uncontrollable by a researcher—and should be. Good parenting is about a good fit between the needs of the young child and the parent's responsivness at any given moment. You generally want to be responsive to your child in each moment, and if you are, every time is slightly different—just like momentary shifts in white-water rafting. So there goes the 'control' that the scientific method generally lauds.
(I should say that I'm focusing on parenting young children here whose growth is phenomenal and requires good nurturance. When children are older or adolecents, there are particular practices that can be followed like rules, like eating dinner together, that help children do better. But again, there are always exceptions to what works best for a particular child and family.)
Thus parents should be cautious about the conclusions particular experiments draw. Just like you can find a Bible passage to support virtually any belief, you probably can find a scientific study to do the same.
So what is the alternative for guiding parenting practices and decisions? I suggest two baseline principles.
The first baseline principle is to follow evolved parenting practices. I mention these regularly in my posts. They evolved more than 30 million years ago initally and continued to co-evolve with the needs of the infant and young child. Science is showing us, little by little, the mechanisms of why they are so needed by young children for optimal development:
- Breastfeeding 2-5 years, with at least 6 if not 12 months exclusively
- Nearly constant touch in the first years of life
- Responsiveness to the needs of the baby before extensive distress occurs
- Playful, companionship interactions with the child (and the child with multiple-aged others)
- Positive social support for mother and child
- Natural childbirth (no interference with timing, no pain, no separation of mother and newborn)
- Multiple adult caregivers to help with all this
Following ancestral practices as closely as possible means you are providing the care your baby evolved to need. We have been conducting a great experiment on our children by eliminating many of these practices and the results are not going well for mental, physical and social health. (See my other posts, for example: Where are the happy babies?; Decline of children and the moral sense; Believing children are resilient may be a fantasy)
The other baseline principle is a precautionary principle: avoid things with known potential harm. This is informed by what studies have shown can happen. For example, not smoking is a precautionary action because although not everyone who smokes gets cancer, we know that the more you smoke the more likely you are to get cancer. So avoiding it altogether is a precautionary action, even though you might be the one person in a 1000 (or whatever the number is) who doesn't develop cancer from smoking.
Science can make general recommendations about best practices, ones that are correlated with good outcomes. It's wise to be informed about those.
Here are a few things we know in terms of applying the precautionary principle to parenting practices:
- We know that children tend to thrive more when they have parents who are mutually responsive, who influence and are influenced by the child in meaningful ways.
This is also supported by the first principle—it is the type of parenting observed in ancestral contexts. So it has double support.
- We know that breastmilk is invaluable for health, intelligence and many other good outcomes for kids. It has thousands of ingredients to build a healthy body and mind.
This is also supported by the first principle, so double support.
- We know that being held and cuddled is correlated with all sort of positive outcomes too and that touch helps mammals grow.
This is also supported by the first principle, so double support.
We could go on with other examples that match up the two principles (I do some of that in other posts). But how do parents decide about new activities that were not part of the ancestral context?
Like media use: Should young children watch television?
That is where there are a few studies to help out and we can apply principle #2 alone. A few years ago a longitudinal study indicated that television watching before age 2 is linked to ADHD (update described here). The study was criticized for not using the gold standard design, randomized control experiment, but remember you usually cannot do that with humans for ethical reasons.
The fact that there was a correlation with mental health problems brings into play the precautionary principle. So I would avoid letting children under age two watch television. It could be that my child would be one of those who were unaffected, but why take that chance? Just like with tobacco—why take the chance when you know there is a risk?
Also, when in doubt, go back to principle 1: How well does the activity match up with what the evolved brain expects for optimal growth? The child's brain expects lots of full-body play and movement, not sitting in front of a screen.
There is a third principle that can be used. Find a family who has raised remarkable children and imitate what they do. This is risky though because you don't have all the information on what happened to the children or how they will turn out. Some times early trauma is not evident till adolescence or adulthood.
Bottom line: For parenting, science can help inform us about trends (in experiments). Science can provide descriptive information about evolution. Science can analyze particular mechanisms (mostly through animal studies), like how positive touch affects gene expression. But science cannot give precise advice as to what to do in a particular situation with a particular child.That takes parental wisdom.
Generally, the best approach to parenting is being caring and responsive (not detached and not intrusive)—being a companion to your children, a gentle guide and consistent coach to their optimal development. This builds a good brain (growing from 25% at birth to 90% of adult size in 5 years) and a pleasant personality. Following principle #1 will also bring about strength and high intelligence (and I mean more than IQ and achievement scores).
