The ARR is based on those who are undertaking study at the year 10 or year 12 level as at August in the reference year and they may not go on to complete year 10 or year 12. The NSSC data items used to construct the ARRs are consistent and comparable over time, and support assessment of annual change. The ARR measures change over a period of time. The numerator and denominator are sourced from different annual cycles of the NSSC, to follow the same age-cohort. Given the long analysis period, student transitions, such as migration or re-entry to the school system, have an effect on the accuracy of this calculation. In addition, the denominator is sourced from two different NSSC years due to different starting years for secondary school. For example, for the ARR from year 7/8 to year 10 in 2009, the denominator for NSW, Vic, Tas and ACT is sourced from NSSC 2006 (year 7) and for Qld, WA, SA and NT is sourced from 2007 (year 8). There is some variability in the reporting of Indigenous status, particularly in relation to not stated responses. This may result in some under reporting of Indigenous status, see Appendix 2: Collection of Indigenous Status of Students (Cat. no. 4221.0) 2009. Increases in the number of Indigenous students due to improvements in the reporting of Indigenous status may lead to increases in ARRs for Indigenous students independently of changes in actual retention. Before sending data to the ABS, each State education department cleans the data and removes duplicate records so that students are only counted once. Due to the different enrolment systems, the ability to remove duplicates varies among jurisdictions and this may result in over-reporting of school students in some states. Decreases in the number of students due to improvements in the identification and removal of duplicate enrolments may lead to decreases in ARRs independently of changes in actual retention. |