Home | Third Culture | Digerati | Reality Club

Harris places a high degree of reliance on self-report studies, whereas I am more cautious about such measures. Over the last year I have conducted one of the largest birth-order studies ever undertaken in order to determine why self-report data are often less supportive of birth-order effects than are real-life studies. This new study involves more than 5,700 subjects who rated themselves, their siblings, friends, spouses, and, in some cases, their parents and offspring.

One of the most serious problems with self-report data is that measurement scales are typically unanchored. What does it mean for us to assign ourselves a "7" on a 9-step scale for empathy, given that we have no accurate idea about who belongs at the closest reference points (6 and 8)? In a study of 660 CEOs, I found only minimal birth-order effects based on self-ratings using unanchored scales. By contrast, when I had the same CEOs rate themselves relative to a sibling, birth-order effects were highly significant and a whopping 5 times larger than by the previous method of assessment. Using this same method of direct sibling comparisons with more than 4,800 additional subjects, aged 8 to 95, I have confirmed every general claim about personality that I made in Born to Rebel. In terms of their overall magnitude, birth-order effects for 30 different personality traits are somewhat smaller than those for sex but somewhat larger than those for age (Sulloway, in press). (8) Moreover, the similarity between the effect sizes for Big Five personality dimensions in this new study, and the confirmation rates for the same personality dimensions in my previous meta-analysis of the birth-order literature, is reflected by the substantial correlation, which is .89. In other words, meta-analysis provides a remarkably accurate measure of the birth-order trends that are recognized by siblings themselves.

Harris maintains that we leave our birth orders behind us when we depart the family for the wider world, a claim that is fundamental to her peer-based theory of personality development. In order to test this assertion, I asked 757 people to rate themselves and their spouses on 30 personality traits. Not only do ratings among spouses reveal significant birth-order differences, but these differences are almost as large as those found in direct sibling comparisons. In addition, the specific traits that exhibit these birth-order differences are the same traits that correlate significantly with birth order in direct sibling comparisons. (9) I have recently replicated this result in a sample of 135 roommates, which demonstrates that unrelated individuals recognize birth-order effects in the people with whom they live on a day-to-day basis. (10) In sum, birth-order effects are not limited to sibling relationships or confined, as Harris argues, to behavior within the family. On this crucial matter, her theory is clearly falsified. The theory is also falsified by evidence relating to other aspects of family niches, including age spacing between siblings and changes in functional birth order owing to the death of siblings or the acquisition of step-siblings.

For some traits, middle children score significantly higher or lower than either firstborns or lastborns. In my previous interview with John Brockman, I cited Catherine Salmon's recent researches bearing on this topic (Salmon 1998; Salmon, in press; and Salmon and Daly, 1998). By testing a series of Darwinian hypotheses about middle children in a variety of different behavioral contexts, Salmon found that middle children, compared with their siblings, are less closely tied to parents and more closely identified with their peers. In her commentary on my interview with John Brockman, Judith Harris ridiculed my discussion of Salmon's study, saying that "the fun part" in birth-order research comes in making up ad hoc explanations to fit such findings. It is worth noting that Harris had not read any of Salmon's studies; this is of a piece with her less than thorough study of the original birth-order literature. (11) To be sure, reading the original literature on any subject does not guarantee the accuracy of one's conclusions. But it is generally thought to help.


Birth-order effects are alive and well in the psychological literature, despite Judith Harris's claims to the contrary. Birth-order effects are just the tip of the iceberg in terms of how family environments resolve themselves into a series of microenvironments for each individual member. Apart from the effects of birth order, we know very little about the ways in which the nonshared environment influences human development, mainly because psychologists have not been successful in developing direct measures of this environment. The challenge for future researchers lies in devising ways to test competing hypotheses that bear on the nature and influence of this elusive environment. Peer groups are doubtless an important aspect of this source of environmental influences, but so are family microenvironments, as well as life experiences more generally. To claim, as Harris does, that peer groups explain almost everything about the environmental sources of personality, and that family microenvironments explain almost nothing, seems like a scientific parody of the currently known facts.



FRANK J. SULLOWAY is the author of Freud, Biologist of the Mind : Beyond the Psychoanalytic Legend, and Born to Rebel : Birth Order, Family Dynamics, and Creative Lives.

Previous | Page 1 2 3 4 5 | Next