|
Article Excerpt
This paper examines the relationship between birth order and later educational outcomes in a birth cohort of more than 1,000 New Zealand young adults studied to the age of twenty-five. Being later born was associated with gaining fewer educational qualifications at secondary level and beyond. The use of nested models to control for the confounding effects of family size on birth order revealed that birth order effects on educational attainment were not disguised by family size effects. Following adjustment for potentially confounding factors, there remained a statistically significant association between being later born and a lower likelihood of obtaining educational qualifications. It was concluded that the intra-family dynamics initiated by birth order may have a lasting effect on the individual in terms of later educational and achievement outcomes.
Keywords
birth order
achievement
longitudinal studies
family background
education
qualifications
Introduction
A large body of research has examined linkages between birth order and later psychological outcomes including mental health, intelligence, academic achievement, and personality (for reviews see Elliott, 1992; Heer, 1985; Steelman, Powell & Werum, 2002; Stewart & Stewart, 1995; Zajonc, 2001; Zajonc & Mullally, 1997). One aspect of this research has been an examination of the linkages between birth order and educational achievement. In general, this research has led to the conclusions that later birth position is associated with lower educational achievement and career outcomes. For example, Black, Devereux and Salvanes (2005) examined the effects of birth order on educational achievement, using a sample consisting of the population of Norway, and found that earlier birth position was associated with a significantly greater number of years of education. Similarly, Herrera and colleagues (Herrera, Zajonc, Wieczorkowska & Cichomski, 2003), using data from a large representative cross-sectional sample in Poland, found that first- and earlier-born respondents reported a significantly greater number of years of schooling and higher levels of occupational prestige. Also, Travis and Kohli (1995) reported that earlier birth order was related to the total number of years of education, particularly for those individuals from middle-class families. These results were replicated using a prospective design in an Australian sample by Marjoribanks (1997). The evidence also suggests that this effect can be observed whilst children are still in school. For example, Sputa and Paulson (1995) found a relationship between earlier birth order and school achievement (grades and test scores) in a sample of adolescents. Also, Cherian (1990) reported a similar finding for adolescents in South Africa.
Supporting these findings are studies that have examined the birth order status of individuals in leadership positions. For example, Newman and colleagues (Newman, Pettinger & Evan, 1995; Newman & Taylor, 1994) found that amongst state governors and town supervisors in the United States, and amongst Australian prime ministers, there were more individuals who were first born than would be predicted by chance. Andeweg and Van Den Berg (2003) reported similar findings for a full list of incumbent politicians in the Netherlands. These findings also extend to more general measures of achievement, including the ratings of other family members. For example, Paulhus, Trapnell and Chen (1999) found in a series of four studies that individuals consistently reported first- and earlier-born children from their own families as being the highest achievers, even after controlling for participants' beliefs about birth order.
These linkages between educational achievement and birth order may be explained in a number of ways. One explanation for these findings may be referred to as the 'family resources' explanation (Hertwig, Davis, & Sulloway, 2002; Marjoribanks, 2001; Travis & Kohli, 1995). Under this explanation, any family has a given set of resources, and with increasing family size, families have decreasing resources to provide for the education and development of children. Later-born children are therefore necessarily accessing a dwindling pool of resources, taking slices of an ever-shrinking pie. A second explanation may be referred to as the 'family niche' explanation, in which successive siblings tend to differentiate themselves from other siblings in an attempt to maximise outcomes and create a unique identity within the family (Feinberg et al., 2003; Schachter & Stone, 1985; Sulloway, 1996; Super & Harkness, 1986; Wachs, 1996). In this view, if older siblings tend to favour scholastic and educational activities, then younger siblings would be more likely to favour an alternate life course.
Despite the existing evidence for the links between birth order and educational achievement and outcomes, there are a number of limitations to the existing literature. One major issue concerns the potentially confounding role of family size on the effects of birth order. In particular, family size sets the upper limit of an individual's birth order, with the result that later-born children will tend to come from larger families than earlier-born children. In turn this raises the possibility that the apparent associations between birth order and educational achievement may be a disguised association between family size and educational achievement. The methods by which family size has been controlled in studies of birth order have varied. Many studies have failed to take family size into account (Andeweg & Van Den Berg, 2003; Herrera et al., 2003; Marjoribanks, 1997; Newman et al., 1995; Newman & Taylor, 1994; Travis & Kohli, 1995). In some studies family size has been introduced as correlated covariate factor (Black et al., 2005; Cherian, 1990; Sputa & Paulson, 1995). Finally, Paulhus et al. described the use of a within-family design which compared ratings of achievement for individuals from the same family using data on all family members (Paulhus et al., 1999). The current study aimed to illustrate how control for the effects of family size can be achieved using nested regression models (see Method).
Further limitations in the literature include the fact that a number of studies have employed a cross-sectional design (Cherian, 1990; Herrera et al., 2003; Paulhus et al., 1999; Sputa & Paulson, 1995), and have used retrospective recall to estimate the influence of family and other risk factors in determining educational outcomes (Marjoribanks, 1997; Travis & Kohli, 1995). Other studies have not controlled for a range of personal and family-related factors that may affect children as they develop into adulthood and that may be confounded with birth order (Black et al., 2005; Herrera et al., 2003; Marjoribanks, 1997; Paulhus et al., 1999). Further studies have employed small sample sizes or examined achievement in selected samples or specialised populations (Andeweg &Van Den Berg, 2003; Marjoribanks, 1997; Newman et al., 1995; Newman & Taylor, 1994).
Taken together, these considerations suggest that the ideal study design to examine linkages between birth order and educational outcomes is a longitudinal design in which a representative cohort is studied from infancy to adulthood. In such a study measures of birth order and family, social, and emotional context can be collected at various points along the life span, and a wide range of measures can be employed to effectively control for covariates. Furthermore, the use of a large representative sample would allow for the use of nested models in which the effects of birth order on educational outcomes can be estimated for each stratum of family size in order to effectively control for the confounding of birth order and family size.
The present research uses data gathered over the course of a twenty-five-year longitudinal study to examine the linkages between birth order and subsequent educational achievement over the fife course. The aims of this study were:
1 To examine linkages between birth order and educational achievement over...
|
