Cognitive Daily

ResearchBlogging.orgHow do you raise “good kids”? It’s one of the questions that plagues parents even before their kids are born. Although everyone’s child can’t be above average, we all want our kids to be nice to others, to “get along” in the world. But kids don’t necessarily cooperate. Babies scream, pull hair, defecate and urinate where they’re not supposed to. Toddlers throw fits in the middle of supermarkets, and older children lie to us and steal from each other. How do we keep them from becoming delinquents, convicts, or worse?

Unfortunately a lot of the research suggests that parents don’t actually have much influence on their kids’ behavior — peers, other environmental factors, and genetics seem to have a larger impact. Yet as parents, we can’t simply throw up our hands and give up. We exert whatever small influence we do have, and hope it doesn’t backfire.

Some studies have suggested that mothers have a disproportionate influence on kids, and that an authoritative parenting style leads to the best results (more prosocial children, who get along with others better). But according to a research team led by Paul Hastings, many of these studies are flawed because they don’t measure masculine prosocial behaviors — the actions more likely to be seen in boys than girls. Girls, they say, tend to be more helpful, sympathetic, and passionate, while boys are more friendly, engaged, and assertive (without being aggressive). All these behaviors are really prosocial.

Hastings and his colleagues had the parents of 133 two- to five-year-olds read several stories about interactions between children and asked how they would respond to the situation if their child had been involved in the situation. The parents also filled in a survey to determine how authoritative their parenting style was.

From five to nine months later, the children themselves were observed playing in a group. After playing for 40 minutes, the experimenter gave them a new toy drum and observed how well they interacted with the new toy. Did they take turns? Did they make sure everyone had a chance to play? In addition, the kids’ preschool teachers answered a questionnaire about each child’s prosocial behavior.

Was their a relationship between the parents’ earlier responses and the children’s behavior later? Here are some of the results:


This graph shows the relationship between various parental strategies and children’s “giving” behavior in the laboratory 5 to 9 months later. The filled data points are the only correlations that are statistically significant — most strategies have no significant effects. Giving behaviors are those stereotypically considered feminine: giving a toy to another child, accepting an invitation to play, and offering a different activity. “Attributions” corresponds to parents stating that prosocial behaviors were part of their child’s disposition. For mothers, there was a significant positive correlation between attribution and girls’ later giving behavior. Mothers who discussed positive behavior with their sons had sons who later exhibiting giving. Finally, mothers’ indirect praise of their daughters (to teachers, for example) was also correlated significantly with later giving behavior. None of the fathers’ strategies correlated significantly with giving.

But what about more masculine behavior? The researchers also looked at how willing kids were to take turns, share, and be cooperative rather than aggressive. This “turn-taking behavior is charted against parenting strategies below:


In this graph, only one correlation was significant: when mothers attributed prosocial behavior to their sons, this correlated significantly with later turn-taking. When boys and girls behavior was combined, then an authoritative mother also correlated significantly with later turn-taking.

The teacher ratings of the children’s behavior followed a very similar pattern, confirming the behavior observed in the laboratory. So it seems that some strategies used by mothers have some small effect on children’s behavior, but few fathering strategies show any effect at all [and of course, since these are just correlations, the strategies themselves may not be causing the results]. When the researchers statistically corrected to consider the mother’s strategy, the fathering effect disappeared entirely.

The researchers say this could be due to the fact that mothers tend to spend more time with younger children, so fathers need not dismay entirely. Maybe a new study could consider the impact of stay-at-home fathers with working wives.

Hastings, P.D., McShane, K.E., Parker, R., Ladha, F. (2007). Ready to make nice: Parental socialization of young sons\’ and daughters\’ prosocial behaviors with peers. Journal of Genetic Psychology, 168(2), 177-200.


  1. #1 agnostic
    May 20, 2008

    It’s worth emphasizing that to the extent parents can influence their children, other than through the genes they pass on, it tends to be stronger when the children are at the parents’ mercy. When the kids grow up and you test them as adults, there is typically no parenting influence (the same is true of the shared-environment term for IQ differences).

    Basically, when kids are still in the transient stages of development, you can push the rank-order around a little bit by interventions, but developmental dynamics will have smoothed out these effects in steady-state.

  2. #2 s
    May 20, 2008

    “Hastings and his colleagues had the parents of 133 two- to five-year-olds read several stories about interactions between children and asked how they would respond to the situation if their child had been involved in the situation.”

    If this is an accurate description of the methodology, it’s hard to assess causation. Parents presumably tailor their strategies to the perceived disposition of their child. For instance, the strongest correlation is with attributions. Unless parents are entirely blind to their children’s personalities, surely they are more likely to state that their child has a prosocial disposition if she normally acts prosocial. Why assume that the parent’s attributions caused the child’s behavior, rather than that both are influenced by the child’s actual disposition?

  3. #3 Dave Munger
    May 20, 2008

    Why assume that the parent’s attributions caused the child’s behavior, rather than that both are influenced by the child’s actual disposition?

    I didn’t assume that, and I don’t think the researchers did either. To me that’s more evidence that parents don’t have much influence at all, at least over the short term.

    [update: I realize now that some of my conclusions in the post may have been unclear, so I’ve updated them to reflect the fact that we’re talking about correlations here, not a controlled study]

  4. #4 E
    May 20, 2008

    Can you define what is meant by: Authoritative, Attributions, Discuss, Indirect, and Affection.

  5. #5 jim
    May 20, 2008

    “Hastings and his colleagues had the parents of 133 two- to five-year-olds read several stories about interactions between children and asked how they would respond to the situation if their child had been involved in the situation.”

    What is the correlation between what parents say is their strategy, and what their strategy is in practice? I doubt that there would be a very high correlation at all, and you’re measuring more contrast against their local parenting environment (my parents were “soft” so I think I’m “hard” and so forth).

  6. #6 Dave Munger
    May 20, 2008

    E: It’s all pretty much there in the post, but here you go, in handy summary format:

    Authoritative: Parents who score highly on a scale of parenting philosophy extending from permissive to authoritative.

    Attributions: Parents stating that prosocial behaviors were part of their child’s disposition

    Discuss: Parents who discuss positive behavior with their children

    Indirect: Parents who indirectly praise their kids’ behavior (to teachers, for example)

    Affection: Parents who show affection for their kids when they behave prosocially

  7. #7 Jim Thomerson
    May 20, 2008

    I think neither my wife nor I were willing to be inconvenienced by bad behavior. On the other hand we tried to do good things for the kids. I think a lot of parenting is by example. Kids see how their parents deal with situations and what they think important vs. trivial. They think, that is (isn’t) how I want to deal with that. Anytime you can discuss with your kid rather than argue, you are ahead of the game. When you do argue, let the kid win the minor arguments, but the parent win the major ones.

    My daughter, a teen-age handful who turned out fine, commented, “It scares me how much of you guys I see in myself.”

  8. #8 Steve
    May 20, 2008

    I keep reading articles and studies that talk about how little influence parents have on their children’s behavior. Most of the methodologies have a fairly short time vector. I don’t think these test measure meaningful parental influence at all, only short term obedience, which can be achieved through behavioral conditioning if the parent is willing to be dramatic enough and consistent enough in their efforts ( I can vouch for that from firsthand childhood experience ). Note that I’m not *advocating* that approach, and do not employ that approach with my own daughter – and she’s not nearly as obedient as I was. I suspect if I was willing to address her with the violence I was ( as were many of my generation, and most of the generation prior to my own ), she would be more obedient. I suspect if this study had been performed in the 1930s, it would have shown a different result.

    But that doesn’t mean that parents don’t influence their children immensely. Simple meta studies make that clear – unless you believe being Democrat or Catholic is genetic in origin. I think it’s very difficult to study the the level of influence exercised by the parent’s behavior rather than their intentional parenting strategies, but if you figure out a way, I’m wagering you’ll find that “Lead by example” is the very best advice you can give a parent.

  9. #9 Anonymouse
    May 20, 2008

    Did they correlate the amount of time spent parenting per week with the affect on the child? I would think that would have more direct correlation.

  10. #10 R E G
    May 20, 2008

    Parents don’t count, but the peer group does.

    Smart parents are choosing the peer group. Starting with the other parent, then the number of siblings, the day care, the school, the neighbourhood ….

    My kids had “friends” I just never had time to invite over, other friends who were always welcome.

    In fact the only conspiracy I have ever been a part of involved the mothers of my daughter’s dearest friends. When they were all about twelve we held a meeting over a bottle of wine and decided curfews, dress codes, driving rules and spending money for the next six years.

    Don’t tell me Cindi can stay out past midnight … I know better.

    Honestly, every time I read this stuff I am wondering if some researcher somewhere just wants to abdicate responsibility for their own kid.

  11. #11 TheNerd
    May 21, 2008

    As I am a working mother with a stay-at-home husband, I would be very interested in the most effective parenting approach for our situation.

  12. #12 Cog
    May 21, 2008

    It seems to me that all these studies work from far too narrow definitions of concepts like learning and behaviour. If parents really had so little influence on their children, we’d all be defecating on the living room floor, eating with our hands and speaking gibberish. The more likely conclusion to draw is that many, perhaps most, parents have inadequate strategies for training their children’s habits in the studied areas of behaviour.

  13. #13 granny
    May 21, 2008

    what about all the unconscious influences?

  14. #14 psychitis
    May 21, 2008

    Couldn’t the results suggest parenting styles are ineffective? Extraneous & confounding variables…

  15. #15 Jim Thomerson
    May 21, 2008

    How does this relate to Eric Berne’s theory that one’s psychology is made up of a parent = taught concepts of life, adult = thought concept of life, and child = felt concept of life? Are we saying that the parent is more the peers than the actual parent(s)? I encountered this theory in “Winning, The Psychology of Competition” by Stuart H. Walker. Let the parent control the preparation, let the adult do the mental toughness, then turn the child loose to have fun and win. My oversimplification, of course.

  16. #16 Tasha
    May 21, 2008

    Something should be said about the fact that the study compares the parents’ self-reported behavior with the children’s observed behavior. Ideally it would have been better to compare parents’ observed behavior than ask for self-report.

  17. #17 Tony Jeremiah
    May 22, 2008

    I agree with Tasha about the problem of using self-reported rather than actual parental behavior as a variable, especially concerning research that indicates attitudes and behaviors are often inconsistent; more commonly known by the expression, “do as I say, not as I do”. The correlation is likely to be higher between a parents’ actual giving behavior and the child’s giving behavior, especially since prosocial behavior is likely to be learned more through observational learning.

    I’m also curious as to why the researchers did not do the analysis based on Diana Baumrind’s (1971) parenting categories (authoritative, authoritarian, permissive-indulgent, neglectful). Based on Baumrind’s work, a particular parenting style is based on three factors: amount of acceptance and involvement with a child, amount of demands a parent makes for a child, and amount of freedom a parent allows a child. So Baumrind’s parenting styles are multi-dimensional. The parenting styles reported here seem to suggest that parents only use one particular strategy when dealing with a child; which seems questionable in the context of Baumrind’s work.

    In that regard, it seems a better analysis would have involved an additive-factors approach, whereby the combined frequency of strategies used by each parent (e.g., Discuss + Indirect + Affections + Attribution), is the predictor criterion for a child’s giving behavior rather than each individual strategy.

  18. #18 Bob Calder
    May 22, 2008

    Were there any feral children in the study?

  19. #19 Montathar Faraon
    May 28, 2008

    Nice charts, may I ask politely which software is used to create them?

  20. #20 Dave Munger
    May 28, 2008

    Yes, I used iWork’s Numbers application for the Mac. Works pretty well, but no error bars, trendlines, or true scatter plots.

  21. #21 mrG
    May 29, 2008

    it would seem to me that before any of these correlations can be said to say anything at all, we would need to have a group of parents willing to be trained to change their strategies, both for the better and for the worse, and then measure the effect of that change.

    and I don’t think that is likely to happen.

  22. #22 Bonnie Zahl
    June 1, 2008

    I’d be curious to know how the three significant correlations in giving compare with each other (i.e., are they the same or significantly different)? Were the effect sizes equal?

  23. #23 Neil Whitaker
    June 6, 2008

    As has been pointed out, causation here does seem very fuzzy. However, it is interesting.

    I haven’t seen any mention of margin of error. How wide is it? It might be a good idea to somehow show this in the graphs.

New comments have been disabled.