IMPROVING FAILING SCHOOLS: TOWARDS A RESEARCH BASED MODEL Stating the problem – what are failing schools?

Since the emergence of the school effectiveness movement in the 1970’s (Teddlie & Reynolds, (2000), increasing attention has been paid to the differential effectiveness of schools. While it is true that the majority of variance in pupil achievement can be seen to result from non-school factors, such as ability and socio-economic background, school level factors account for between 10 % and 40 % of the variance in pupil outcomes depending on such factors as the specific study reported on, national context (e.g. school level variance is higher in countries with lower levels of central control and higher levels of school choice), population served (school level variance is greater for pupils from low SES backgrounds than for those from more advantaged contexts), and subject (school level variance is greatest for those subjects that are least reliant on previous out-of-school knowledge, and is thus greater for English than maths, for example). While school effectiveness research has pointed to the existence of between school variance and some of the factors associated, school improvement research has pointed to the possibility of improving the performance of schools in different contexts, and has made many practical contributions to helping schools better serve their communities and pupils, as well as building up a knowledge base on some of the processe