Most people, when asked whether parental involvement benefits children academically, would say, "of course it does". But evidence from our research suggests otherwise. In fact, most forms of parental involvement, such as observing a child's class, contacting a school about a child's behaviour, helping to decide a child's high school courses, or helping a child with homework do not improve student achievement. In some cases, they actually hinder it.

Over the past few years, we conducted an extensive study of whether the depth of parental engagement in children's academic lives improved their test scores and grades. We pursued this question because we noticed that while policymakers were convinced parental involvement positively affected children's schooling outcomes, academic studies were much more inconclusive.

We analysed longitudinal surveys of American families that spanned three decades (from the 1980s to the 2000s) and obtained demographic information on race and ethnicity, socioeconomic status and the academic outcomes of children in elementary, middle and high school, as well as information about the level of parental engagement in 63 different forms.

What did we find? One group of parents, including blacks and Hispanics, as well as some Asians (like Cambodians, Vietnamese and Pacific Islanders), appeared quite similar to a second group, made up of white parents and other Asians (like Chinese, South Koreans and Indians) in the frequency of their involvement.

A common reason given for why the children of the first group performed worse academically on average was that their parents did not value education to the same extent. But our research shows that these parents tried to help their children just as much as the parents in the second group.

Even the notion that kids do better in school when their parents are involved does not stack up. After comparing the average achievement of children whose parents regularly engage in each form of parental involvement with that of their counterparts whose parents do not, we found that most forms of parental involvement yielded no benefit to children's grades, regardless of racial or ethnic background or socioeconomic standing.

In fact, there were more instances in which children had higher levels of achievement when their parents were less involved than there were among those whose parents were more involved. Even more counterintuitively: When involvement does seem to matter, the consequences for children's achievement are more often negative than positive.

When involvement did benefit kids academically, it depended on which behaviour parents were engaging in, which academic outcome was examined, the grade level of the child, the racial and ethnic background of the family and its socioeconomic standing.

For example, regularly discussing school experiences with your child seems, to positively affect the reading and mathematics test scores of Hispanic children, to negatively affect test scores in reading for black children, and to negatively affect test scores in both reading and mathematics for white children (but only during elementary school).

Policymakers should not advocate a one-size-fits-all model of parental involvement.

What about when parents work directly with their children on learning activities at home?

Regardless of a family's social class, racial or ethnic background, or a child's grade level, consistent homework help almost never improved test scores or grades. Even more surprising to us was that when parents regularly helped with homework, kids usually performed worse.

Our findings also suggest that the idea that parental involvement will address one of the most salient and intractable issues in education, racial and ethnic achievement gaps is not supported by the evidence. As it turns out, the list of parental involvement that generally works is short: expecting your child to go to college, discussing activities children engage in at school and requesting a particular teacher for your child.

Do our findings suggest that parents are not important for children's academic success? Our answer is no. We believe that parents are critical for how well children perform in school, just not in the conventional ways that our society has been promoting. The essential ingredient is for parents to communicate the value of schooling, a message that parents should be sending early in their children's lives and that needs to be reinforced over time. But this message does not need to be communicated through conventional behaviour.

When the US government issues mandates on the implementation of programmes that increase parental involvement, schools often encourage parents to spend more time volunteering, to attend school events, to help their children with homework and so forth. There is a strong sentiment in this country that parents matter in every respect relating to their children's academic success, but we need to pay attention to what the evidence is telling us.

Future research should investigate how parental involvement can be made more effective, but until then, parents who have been less involved should not be stigmatised. What should parents do? They should set the stage and then leave it.

NEW YORK TIMES

Dr Keith Robinson, an assistant professor of sociology at the University of Texas, Austin, and Dr Angel L. Harris, a professor of sociology and African and African-American studies at Duke, are the authors of The Broken Compass: Parental Involvement With Children's Education.