What is ‘underachievement’ at school?1

Abstract

This paper formed the basis for our oral and written presentations to the House of Commons Select Committee on Education and Skills on the topic of underachievement. 'Underachievement' is now a widely used term in education policy and practice. It is used routinely to refer to nations, home nations and regions, to types and sectors of schooling, to physiological, ethnic and social groups, and to individuals. It has been used to mean simply low achievement, also lower achievement relative to another of these groups, and also lower achievement than would be expected by an observer. The paper presents examples of each. These multiple uses lead to considerable confusion which, coupled with common errors in assessing the proportionate difference between groups, mean that significant public money has been spent attempting to overcome problems that may not, indeed, exist. Where underachievement is understood to mean a lower level of achievement by an individual (or group) than would be expected using a model based on the best available predictors, then there is nothing we can know about underachieving individuals (or groups) that they have in common. They can not be disproportionately working-class males, for example, because class and sex would then be part of the 'best available predictors'. Even if, instead, we reserve some predictors from our best model, there is no evidence that underachievers have much in common (and examples from such models are presented in the paper). In raw-score terms, we might say that a particular social group exhibits lower achievement (in the sense of publicly available figures relating to pencil and paper tests) than another, as in the case of some ethnic groups. Or we might say that there is differential attainment between groups, as in the case of males and females. This is very far from saying that the lower-attaining group could and should do better on that assessment, or that the surface dissimilarity (such as ethnicity or sex) is in any way the cause of the difference in attainment. Making explicit what we mean by underachievement is an important step towards accepting that, collectively, we do not really mean anything by it. 3 Summary• 'Achievement' at school generally describes levels of attainment in public examinations such as

Who is missing from higher education?

“…Any underrepresentation is already as much in evidence in terms of the possession of entry qualifications at NQF level 3, and these in turn are based almost entirely on stayingon rates in schools and colleges, in turn based almost entirely on NQF level 2 qualifications, and so on (see Gorard and Smith 2004). This, in turn, suggests that WP activities need to be directed at the earlier-life of potential students more than at the point of possible transfer to HE.…”

(Mis)Understanding underachievement: a response to Connolly

“…First, there are many things that we are aware of that are not mentioned in the Gorard and Smith (2004) paper. Lack of citation is not evidence of lack of awareness.…”

“…Therefore, we might correctly say that this particular group 'exhibits lower achievement' but would not say that the English speakers 'could and should do better on that assessment' (emphasis in original, see above). Our condemnation here, and in Gorard and Smith (2004), is quite clearly of the inappropriateness of the assessment for this student or group of students, rather than of some inherent lack of talent among the students involved. For more on this see, for example, Gorard (2004).…”

“…Gorard and Smith (2004) cite the sudden appearance of the current much larger gap in attainment between male and female at age 16 at the same time as the reforms of the late 1980s such as GCSE, abolition of norm-referencing, more coursework and so on as suggestive of an explanation. We reproduce one part of this analysis in Figure 1, representing the GCSE/GCE results of all school-leavers in England from 1974/75 to 1997/98, separated by male and female.…”

(Mis)Representing underachievement: a rejoinder to Gorard and Smith

“…With these errors in mind, I felt it worth looking again at some of the other work published by Gorard and his colleagues; especially their later work on underachievement (Smith 2003;Gorard and Smith 2004). Central to this was the article by Smith that sought to address two wide-ranging research questions: 'how can underachievement be defined and measured?…”