How it works
A transparent method.
estimATAR turns a student's NAPLAN domain percentiles into a likely ATAR range using a transparent, auditable pipeline: percentile conversion, literacy/numeracy baselines, and a small set of contextual adjustments for school sector, ICSEA, domain disparity, and Y7→Y9 trend.
Sample and limitations
v1 is trained on a sample of ~1,200 Queensland students attending schools with ICSEA values ranging from 1015 to 1170. Individual students within those schools can sit at any ICSEA. This means:
- Estimates are calibrated to the QLD cohort. State ATAR scales are now considered interchangeable, so out-of-state results are comparable — but the training sample is QLD, so treat interstate estimates with a little extra caution.
- Government-sector students are under-represented; point estimates for gov-sector may be less reliable.
- Schools with ICSEA outside the 1015–1170 training band are extrapolations; the SE widens and adjustments may not generalise.
- Students without a Year 9 sitting have meaningfully wider confidence intervals.
What it isn't
estimATAR is a prediction tool, not a prophecy. A student's ATAR depends on their Year 11–12 subjects, effort, and the cohort around them. Use this as a calibration check, not a ceiling.