School type and academic culture: evidence for the differentiation–polarization theory

Abstract

Several decades ago it was shown that the differentiation of pupils into tracks and streams led to a polarization into ‘anti‐school’ and ‘pro‐school’ cultures. Support for this differentiation–polarization theory is mainly based on case studies. This paper presents findings of a quantitative study in Belgium (Flanders). Attention is given to the conceptualization of the polarization component of the differentiation–polarization theory. The findings suggest that the culture of pupils is less study‐oriented in technical/vocational schools than in general (grammar) schools. The differentiation–polarization theory also applies to school staffs: the staff culture is less academically‐oriented in technical/vocational schools than in general schools. Moreover, staffs' attitudes towards pupils—their judgements on the teachability of pupils and the trust they place in pupils—are different.

Keywords:

• ability‐grouping
• pupil culture
• teacher culture
• tracking systems
• vocational education

Acknowledgement

This paper was originally presented at the XV^th International Sociological Association World Congress of Sociology, Brisbane, Australia, 2002.

Notes

1. There are exceptions, e.g. Berends (1995).

2. Frank (1989, 1997) has argued—with respect to consumption—that an individual's subjective sense of well‐being depends strongly on his or her particular frame of reference. This frame of reference does not necessarily correspond with the most advantageous frame of reference from society's point of view.

3. E.g. ‘knowledge, beliefs, values, codes, tastes, and prejudices that are traditional in social groups and that are acquired by participation in such groups’ (Cohen 1955: 12), or ‘transmitted and created content and patterns of values, ideas, and other symbolic‐meaningful systems as factors in the shaping of human behaviour and the artifacts produced through behaviour’ (Kroeber and Parsons 1958: 583). Summarizing, Rousseau (1990: 154) states that culture is usually seen as ‘a set of cognitions shared by members of a social unit’.

4. As such, this measure should not be confused with measures of approaches to learning or studying (Biggs 1987, Entwistle et al. 1989) which measure the extent to which pupils endorse different approaches to the act of learning as well as the motives and strategies around these approaches. In this project, I investigate pupils' general feelings about studying, not their learning styles.

5. Responses were imputed for missing values by way of item correlation substitution: a missing value for one item is replaced by the value of the item correlating most highly with that item (Huisman 1999). A confirmatory factor analysis (LISREL 8.12; Jöreskog and Sörbom 1993) produces a significant loading higher than 0.40 for each item, so all items were retained to form one scale (χ ² = 27.99; df = 21, p = 0.14, RMSEA = 0.0094, AGFI = 1.00 (after adding the error correlations for several items, which can be seen as an adjustment to reality—see Bollen (1989)), Cronbach's α = 0.82).

6. Measured by means of the occupational prestige of father and mother (Erikson et al. 1979)—the higher value is used as an indicator of the SES of the family.

7. To do this, the aggregation has to be reliable and represent something really shared at the group or organization level (Glick 1985). A useful measure is the mean rater reliability (Glick 1985), calculated by means of the Spearman‐Brown formula based on the intra‐class correlation of a one‐way analysis of variance (Shrout and Fleiss 1979, Glick 1985). The result must be above 0.60 to sustain aggregation to the group or organization level (Shrout and Fleiss 1979, Glick 1985). Aggregation of study‐involvement to the school level by way of the mean per school is reliable (0.86). Study involvement is clearly something shared by the pupils of the same school, so it is legitimate to speak of a study‐culture.

8. Later these scores were reversed; the most important goal got the highest score.

9. This aggregation was reliable (0.73).

10. On the basis of a confirmatory factor analysis by LISREL 8.12 (Jöreskog and Sörbom 1993), only those items with significant loadings higher than 0.30 were retained.

11. Aggregation of this measure to the school level by calculating school‐level means is reliable (0.93).

Additional information

Notes on contributors

Mieke Van Houtte

Her research covers diverse topics within the sociology of education, particularly the effects of structural and compositional school features on outcomes for pupils and teachers.