Dr. Frank SullowayIn a recent study on birth order and intelligence, Norwegian researchers report that eldest children have higher I.Q.s than their younger siblings. Why would that be? There are as many guesses as there are scientists. “Virtually anyone who has a sibling is a birth-order theorist,” says Dr. Frank Sulloway, an expert on family dynamics at the University of California, Berkeley. Dr. Sulloway answered readers’ questions about the study and its implications; his answers appear below. [Editors’ note: questions are no longer being accepted.]
2007
1:00 am
What about differences between the second and third or fourth child? Does this three-point gap also apply to the subsequent children?
— Posted by V. Nhan
2007
2:00 am
One of the most interesting features of the new Norwegian study is that birth-order differences in I.Q. become smaller with increasing birth rank. The difference between a firstborn and a second-born in a family of two children is about 2.3 I.Q. points. The difference between a firstborn and a second-born in a family of three children is 2.1 I.Q. points. By contrast, the difference between a second-born and a third-born in the same family is only 1.1 I.Q. points. In a family of four children, the I.Q. difference between siblings is reduced to only 1 I.Q. point per birth rank, and the difference between the third-born and the last sibling drops to only 0.2 I.Q. points. In general, then, the largest birth-order differences are observed between firstborn siblings and second-born siblings in small families. Children with successively higher birth ranks suffer relatively smaller deficits in I.Q. because of their birth orders.
These intriguing results might be explained by niche partitioning within the family, although this is not the only possible explanation. In two-child families, the firstborn and the second-born may be partitioning their roles into that of the mature and studious achiever and that of the less mature, perhaps more athletic, younger sibling who pursues other types of interests in order to be different (a process called “deidentification”). As other siblings are added to the family system, they appear to position themselves somewhere between these two extremes. Hence the overall difference in I.Q. in a family of five children, between the eldest and the youngest offspring (a total of about 2.7 points), is not much greater than the difference we observe between a firstborn and a second-born in a family of two children (2.3 points).
Theories about the dilution of family resources might also explain why birth order makes less of a difference with higher birth ranks. As more children are added to the family, the relative disparity decreases between the intellectually rich environment of a firstborn, with exclusive access to parental attention before the arrival of other siblings, and the intellectual environment of a family that includes many offspring competing for parental attention.
— Posted by Dr. Frank Sulloway
2007
2:01 am
Why do some psychologists make such a fuss over a few I.Q. points? In this instance, I think we’re talking about a little more adult attention producing a slightly higher level of verbal and mathematical ability, which might help in school. A pinch of creativity dwarfs that in importance.
— Posted by P. Dorell
2007
4:00 am
As I noted in the Science Perspective that accompanies publication of the new study by Petter Kristensen and Tor Bjerkedal, the 2.3 I.Q. points that differentiate the average Norwegian firstborn from the average Norwegian second-born in a two-child family is equivalent to the firstborn having a 13 percent greater chance of getting into a better college. This difference is also equivalent to the firstborn having 1.3 times the odds of getting into a better college, compared with the second-born.
It is also worth noting that 2.3 extra I.Q. points (the advantage enjoyed by a firstborn over an immediately younger sibling) is approximately equivalent to scoring an extra 15 points on each SAT test, or a combined 45 points on the three current tests, which have a mean combined score of about 1,500 points. The cutoffs for acceptance to the best colleges, based on SAT scores, often hinge on where one stands within a range of just 40 to 50 points on the three tests combined.
Seen in this perspective, these documented differences in I.Q. by birth order are hardly negligible. However, as I said in a recent interview published in part by Nature, if I had the choice of having 2.3 extra I.Q. points or having the “enlarged curiosity” that Charles Darwin’s uncle, Josiah Wedgwood, recognized in his nephew on the eve of Mr. Darwin’s departure on the Beagle to circumnavigate the globe, I would unhesitatingly choose the latter.
So, yes, I.Q. is hardly everything, and much that makes people successful in life has to do with how people use their intelligence rather than with their intelligence per se. In addition, there is considerable evidence suggesting that siblings born later use their intelligence differently from the way firstborns use theirs. Indeed, later-born siblings would appear to have 2.3 extra points of one difficult-to-measure intellectual skill, associated with unconventional thinking, that firstborns sometimes lack.
— Posted by Dr. Frank Sulloway
2007
5:37 am
The problem I have with this study is that it was done in a very homogeneous population. Should not this hypothesis first be tested in other populations before drawing any firm conclusions?
— Posted by P Penko
2007
6:55 am
This reader raises a very interesting point. Yes, the Norwegian population is probably more homogeneous than, say, the American population. However, one must keep in mind that many well-designed studies of birth order and intelligence — studies that have controlled at minimum for differences in family size and socioeconomic status — have been conducted in America, the Netherlands, France and Israel, among other countries. These studies have consistently shown that firstborn siblings have higher I.Q. scores than second-born siblings, that second-born siblings have higher I.Q. scores than third-born siblings, and so forth.
This newest study from Norway uses a marvelous within-family study design, in which brothers were all compared with other brothers from the same families. It obtains essentially the same results that were found in another large Norwegian sample of individuals who grew up in different families. In short, the results strongly suggest that the findings from previous between-family studies were actually giving us an accurate portrait of birth-order differences in intelligence, as long as these previous studies adequately controlled for differences in family size and socioeconomic status. So we do have reasonably convincing data, from many different countries, that confirm the relationship between birth order and intelligence.
— Posted by Dr. Frank Sulloway
2007
7:00 am
Is age of the parents a factor?
— Posted by R. Doherty
2007
8:00 am
Studies have shown the mother’s age at the birth of her first offspring to be a significant predictor of the child’s I.Q. More educated mothers and mothers with higher I.Q. scores tend to have smaller families, and these mothers also tend to bear their first children at later ages than other mothers. As shown in the Norwegian study, differences in I.Q. between firstborns and second-borns were largest among the most highly educated mothers. Hence we would expect birth-order differences in I.Q. to be larger among mothers who have their first child at a later age, since such mothers are likely to be more educated than other mothers.
— Posted by Dr. Frank Sulloway
2007
9:00 am
Is the difference in test points between brothers constant across the range of scores? For example, would the difference between the eldest and youngest still be about 3 points if the eldest’s I.Q. were 120? Would the difference increase or decrease if the I.Q. of the eldest were significantly greater?
I am also surprised not to hear of any testing of interactions with respect to the impact of the presence of one or more sisters between the brothers.
— Posted by L. P. Bucklin
2007
10:00 am
The first is a really interesting question, and one to which we do yet not know the answer. One intriguing result of a study just published in 2007 by Dr. Bjerkedal et al. in Intelligence was that the differences in I.Q. scores between firstborns and second-borns were significantly larger if mothers were highly educated, compared to other mothers. This finding is consistent with the notion that firstborns and second-borns frequently partition their family niches into an adult-like, responsible role, emulating parental values and standards, or the role of a younger, more adventurous child who seeks out alternative and sometimes more unconventional ways of achieving.
To the extent that mothers (and parents in general) value education, this circumstance would tend to contribute to the degree to which niche partitioning hinges on differences in intellectual and educational values. In short, if parents are particularly educated and intelligent, and if they have particularly intelligent offspring, the disparity in I.Q. between a firstborn child and a second-born child might be even greater than the 2.3 points observed in the Norwegian sample as a whole. This is a conjecture, however, that can only be adequately answered by further empirical investigation.
Although this new and impressive Norwegian study tested only the I.Q. scores of males, the sample included many families in which subjects had sisters who were not tested. If, for example, we examine the scores of firstborn and second-born males in families with only two children, and if we compare these findings with other members of the sample from sibships of two (about half of whom would have had a sister rather than a brother in one of the two sibling positions), there does not appear to be much difference in the overall results. This suggests that the sex of one’s closest sibling is not much of a factor in the differences reported by birth order. A more formal statistical analysis, however, would have to be conducted on the original data to confirm this general impression.
— Posted by Dr. Frank Sulloway
2007
11:00 am
What is the finding for only children? Are their I.Q. scores most similar to the firstborn or oldest children in multichild families?
— Posted by B. Simpson
2007
12:00 pm
In this Norwegian study, the I.Q. scores of only children were significantly lower (by about 1.2 points) than those of firstborns with one younger sibling. Only children, however, are more likely than firstborns with siblings to come from single-parent homes. When I.Q. scores were adjusted to account for differences in maternal education, parental income and parental marital status, the difference in I.Q. scores between firstborns and only children was reduced to just 0.6 points.
This remaining difference does suggest that Robert Zajonc’s theory about older siblings teaching younger siblings may help to explain why firstborns have higher I.Q.s than only children, since only children have no one to teach. However, this I.Q. difference is also consistent with the theory of niche partitioning within the family. Only children do not compete with younger siblings for parental favor. Hence they are presumably less motivated than firstborns with younger siblings to nail down the niche of the family “achiever.”
— Posted by Dr. Frank Sulloway
2007
1:00 pm
Many families now consist of siblings who are often 10 or more years apart in age. Are large differences in age between siblings addressed in this study?
— Posted by D. Kincade
2007
2:00 pm
The Norwegian study did examine the role of age spacing between siblings. They found that the largest birth-order differences in intelligence were between pairs of brothers who were less than a year apart in age. Smaller differences were found for sibling pairs separated by a gap of two to six years. Differences in I.Q. were even smaller for pairs separated by seven or more years. These findings are consistent with resource dilution theories about birth order and intelligence. Such findings are also consistent with the possible role of niche partitioning among siblings, as offspring compete for parental favor. Larger gaps in age between siblings tend to allow each sibling to obtain greater resources from parents, and such large age gaps also tend to reduce sibling rivalry.
— Posted by Dr. Frank Sulloway
2007
3:00 pm
This article states that the study was conducted on men, but that the results are expected to be the same for women. If that is the case, why is the study gender selected in the first place? Both the use of males in studies, and the reporting of the results as if they apply equally to both genders, appears to be common in behavioral studies. Is this a historical artifact, and if so, why does it persist?
— Posted by H. Stevens
2007
4:00 pm
The inclusion of only males in the Norwegian study was determined by the circumstance that 18- and 19-year-old Norwegian males must all register for military service and are given a standard test of intelligence. These data are unfortunately not available for Norwegian females. However, numerous well-controlled, between-family studies conducted in many different countries have included almost equal numbers of males and females, and those studies have shown that the effects of birth order do not differ in any appreciable way by the sex of subjects included in these studies.
— Posted by Dr. Frank Sulloway
2007
5:00 pm
Does this research shed any light on the oldest sibling in families that have adopted children?
— Posted by B. Mosca
2007
6:00 pm
Although the Norwegian study does not consider the issue of whether siblings were adopted or not, it does show that I.Q. is influenced by how children were raised, not by how they were born. That is to say, a third-born child who lost both older siblings before these siblings reached the age of 1, and who grew up as a functional firstborn, scored the same on the Norwegian intelligence test as other biological firstborns. From these results, we may infer that the I.Q. scores of adopted children would accord, insofar as birth order is concerned, with how these adopted children have been raised rather than born.
Conversely, in families with both adopted and biological offspring, the biological offspring ought to have I.Q. scores corresponding with how they were raised, after considering the relative age and functional birth order of any adopted children.
— Posted by Dr. Frank Sulloway