The Unrealistic Educational Expectations of High School Pupils: Is America Exceptional?

Abstract

There is growing concern that many American teenagers hold unrealistic educational plans. This may indicate a detachment from reality, which could be detrimental to well-being in later life. But is this problem specific to certain countries like the United States, or is it common among young people from across the developed world? This article uses data from the Trends in Mathematics and Science Study (TIMSS) and Programme for International Student Assessment (PISA) to investigate this issue. It shows how expected and actual college graduation rates differ across a number of countries but also that this gap is particularly large in the United States. Additional analysis suggests that this is being driven, at least in part, by the large proportion of low-achieving American children who believe they will go on to obtain a bachelor's degree. The implications of these findings are discussed in reference to educational policy and contemporary sociological debates.

ACKNOWLEDGMENTS

Thanks are given to the ESRC for providing funding for this project. I would particularly like to thank Ingrid Schoon, John Micklewright, Patrick Sturgis, Sandra McNally, and Sylke Schnepf for their helpful comments and the hospitality of the Institute for Research on Poverty at the University of Wisconsin, Madison, on a visit while this article was produced.

NOTES

Notes

¹ Bruce and Yearley (2006:130) describe habitus as “a set of acquired principles of thought, behaviour and taste that generates social practices and is particularly associated with a certain social class. It is described as a person's own knowledge and understanding of the world.”

² Low-achieving children with parents from professional backgrounds expecting to complete college may also be a problem; although this group may be able to afford to attend college, they are just as unlikely to keep up with the academic demands. However, I choose to focus on low-achieving children from disadvantaged backgrounds because they face two significant barriers to entry (lack of adequate academic preparation and lack of finances) rather than just one.

³ Around one in ten children picked the “don't know” category in the United States. In this article, I do not consider this as missing data. Rather, I view it as a legitimate substantive answer for children to provide, indicating that they have not committed to the educational path they are going to take.

⁴ Reading and science were also covered in the PISA 2003 study but in a lot less detail than math. All of the results presented are robust to the use of reading or science test scores instead of math test scores. Results are available from the author upon request.

⁵ I experimented using the other plausible values and by running five separate models and averaging the estimated coefficients and standard errors. Results are very similar to those presented.

⁶ An absolute skill threshold refers to children scoring below a certain number of points on the test. In PISA, for instance, the survey organizers define any child scoring below 420 test points as in achievement “level one.” The implication of using such an absolute threshold is that the proportion of children defined as low achievers will vary across countries. Results using this absolute skill threshold are available from the author upon request.

⁷ In TIMSS, the only information available on family background are children's reports of parental education and only in a selection of countries. On the other hand, PISA contains detailed information on parental occupation in all participating nations.

⁸ As this information is based upon children's reports, measurement error may be a concern. However, Jerrim and Micklewright (2012) have investigated this issue with specific reference to the PISA data set. They find a strong correlation between child and parent reports of father's occupational status in the 10 countries with data available.

⁹ Approximately 10 percent of the U.S. sample is defined as working class, 60 percent as intermediate, and 30 percent as professional class. Figures for the OECD as a whole are approximately 15 percent working class, 65 percent intermediate, and 20 percent professional class. The social class distribution in England, Canada, and Australia is comparable to that in the United States, perhaps because of the similar labor market structures.

¹⁰ Maternal and paternal occupation may have independent influences upon children's educational expectations, and the relative importance of each parent may also depend upon the gender of the child. Here, I use the higher of the two to create a broad measure of socioeconomic background, though note that distinguishing between the impact of mothers and fathers is an important issue worthy of further research.

¹¹ There has been some debate as to whether comparing logit coefficients across groups is an appropriate empirical strategy (Allison 1999; Hauser and Andrew 2006). A few alternative approaches have been suggested, although none has been universally accepted (Williams 2009). Consequently, I have used a nonparametric approach to test the robustness of my results. The main conclusions drawn remain largely intact (results available from the author upon request).

¹² This measure is more attractive than alternatives such as the odds ratio and marginal effect (predicted probabilities), as it is linear, can take values at any point along the real number line, and is not sensitive to the point on the logistic distribution at which estimates are produced. Log-odds are therefore not influenced by differences between countries in the absolute proportion of children who expect to complete college.

¹³ It is also interesting to note that 50 percent of low-achieving children from the professional group expect to obtain a bachelor's degree in the United States—a greater proportion than in most other developed countries. Although these young people may be able to afford to attend college, they are still unlikely to have the academic ability to do so. This group may therefore also undertake inappropriate educational pathways (potentially leading to a waste of human and financial resources).