The research question was "What factors seem to be related to whether or not a child is classified as SEN" This author hypothesised that the main factor related to whether or not a child is classified as SEN would be test scores (APS). A search of relevant scholarly literature found that other factors, particularly low social economic status, and cultural-linguistic factors has been shown to increase the proportion of students classified as SEN.
The model used was a GLM log one. The two main variables from the hypothesis - APS and SEN - was completed in R 2.7.0 to model the expected relationship. Further analysis including charts and tables was done to examine the correlation between SES factors (such as being eligible for free school meals), and whether or not a child is classified as SEN. Additional work was completed to discover the relationship between whether or not a child is classified as SEN and socio-linguistic factors (Mother tonge and ethnicity).
An important issue for this statistical analysis is understanding the variables and the effect of other factors (both available in the dataset and ones not contained within the dataset). Blindly plugging variables into a copied statistical model and performing an analysis based on those is of little to no practical or even theoretical use. The following two sections (2.1 and 2.2) provided some context and background research on factors affecting the mis-identification of students needing special education services. Such factors help explain why a correlation between one or more explainatory variables (i.e. test scores or school grades) and the response variable (in receipt of special education services) is not as strong as one would expect.
2.1 Low Achievement vs Learning Disabilities.
One of the frequently occurring reasons why students are referred for special education services is for having a learning disability. "Learning disability (LD) is a condition characterized by the occurrence within an otherwise adequate level of intellectual functioning of one or more specific deficits in the cognitive processes that undergird learning." (Blair, C. And K. G. Scott, 2002). The diagnosis of having a learning disability has increased in the past few decades. In the United States the rate of identification and placement in special education services, has doubled between 1982 and 2002 (Lewit & Baker, 1996).Much of this increase can be attributed to the increased awareness and acceptance of this invisible disability as a real special need requiring special education services. (Formally, students were told "If only you'd apply yourself and try harder."). However, as mentioned above some of the increase in referral and identification is likely due to other, non-LD factors causing a mis-identification. Statistical analysis published in 2002 by Clancy Blair and Keith G. Scott "indicated that 30% of LD placements among boys and 39% of LD placements among girls were attributable to what can be considered low-SES markers." This section discusses the situation where low achievement (LA) from resulting from low social economic status (SES), results in students being misidentified as having a learning