When schools use student data to predict test results

A school may combine attendance, lateness, behaviour incidents, wellbeing notes and past assessment results to estimate how a student will perform on NAPLAN or another standardised test. The prediction can appear administrative, yet it may influence intervention groups, teacher expectations, access to extension work or how often a child is monitored.

For Australian families, the first task is to find out what is actually happening. A spreadsheet used by a year-level team is different from an automated model supplied by an education technology company, and both differ from a formal decision recorded in a student file. Ask for plain-language information before assuming that a score is being generated by artificial intelligence.

Predictive indicators are not proof of ability. A child who misses school because of asthma, caring responsibilities, anxiety, transport problems or cultural commitments may be classified as “at risk” without the underlying circumstances being understood. Behaviour records can carry similar problems when normal developmental differences, disability-related conduct or cultural misunderstanding are treated as evidence of likely academic failure.

Ask what data is being used

Request a meeting with the principal, year coordinator or data protection contact, and ask which fields feed the prediction. Relevant categories may include unexplained absences, late arrivals, detentions, classroom removals, wellbeing referrals, reading levels and previous test scores.

Ask whether the data is entered manually, imported from a student management system or processed by an external vendor. Families should also ask how long the information is retained, who can see the results, whether a student can be profiled without parental consent and whether the prediction is shared with another school or government agency.

Separate support from labelling

A useful early-warning system should lead to practical support, such as tutoring, attendance assistance, disability adjustments or a conversation with the family. It should not quietly become a permanent label that lowers expectations or places a student into a less demanding curriculum.

Ask what human review occurs before action is taken. A prediction based on attendance and behaviour should be checked against the student’s health, disability, language background, family circumstances and teacher observations. Request that staff record the context rather than treating a probability score as an objective fact.

Check Australian privacy protections

The Privacy Act 1988 and the Australian Privacy Principles may apply when a school or supplier is covered by federal privacy law, while government schools often operate under state or territory privacy and records legislation. In New South Wales, Victoria, Queensland and other jurisdictions, the correct complaint or access pathway can therefore depend on the school sector and the body holding the information.

Ask for the school’s privacy policy and make a written request to access and correct personal information. The response may involve an internal privacy officer, a state education department, an FOI process or a state information commissioner. Keep copies of emails and note dates, names and the exact data you were told was used.

Look closely at testing decisions

NAPLAN results are often discussed alongside attendance and engagement data, but a predicted result should not determine whether a child receives a rich curriculum. In Sydney or Melbourne, where schools may use several commercial platforms alongside department systems, families should ask whether separate databases are being matched.

Ask whether the model was tested for accuracy across students with disability, Aboriginal and Torres Strait Islander students, multilingual learners and students from different socioeconomic backgrounds. A system that performs well for one group may produce more false alerts for another. The school should be able to explain what happens when the prediction is wrong.

Build a written record

After a meeting, send a short email summarising what was said: the data sources, the purpose of the analysis, the staff who can access it, the support offered and any correction requested. Written records are especially valuable when a child moves between schools or when a vendor’s contract changes.

Parents can work collectively through a P&C association, school council or local advocacy group. In Australia’s education market, schools may purchase attendance dashboards, behaviour platforms and assessment tools from different suppliers, so a group request can reveal whether families are receiving consistent information about data collection and retention.

Use advocacy channels carefully

If the school will not explain the system, escalate through the principal, regional office, department privacy contact or the relevant state commissioner. Ask for the policy governing automated decision-making, data sharing and student records rather than relying on informal assurances.

Families comparing international policy debates can also review parent advocacy resources from New Yorkers United for Kids, particularly its focus on standardised testing, student privacy and local control. The legal framework differs from Australia’s, but the central concern is familiar: schools should be accountable when data affects a child’s opportunities.

A complaint should focus on specific issues: inaccurate attendance records, unverified behaviour descriptions, undisclosed profiling, excessive data collection or a decision made without meaningful human review. Avoid framing the matter as opposition to every form of data use; a precise request is more likely to produce a useful answer and a correction.

Start by emailing the principal this week with five requests: identify the data fields, name the software or supplier, explain who sees the prediction, provide the privacy and retention policy, and correct any inaccurate record.