Research & methods · Reviewed 30 October 2026
Read the evidence.
Clear sources. Careful comparisons. A campaign whose requests can be examined.
This independent campaign calls for a new Royal Commission into child protection in South Australia. It is not an official inquiry. This page explains the evidence used, what it cannot establish, and the principles behind presenting a difficult subject clearly.
Evidence, interpretation and a proposal
- Evidence: a number, statement or finding attributed to an identifiable report, with its date and definition.
- Interpretation: an explanation of what that evidence may mean. An interpretation must preserve uncertainty and consider other explanations.
- Campaign proposal: a request for further independent examination, accountability and protection. Citing a report does not mean its author endorses this campaign or recommends another Royal Commission.
The question of whether to establish a new inquiry is a public policy judgment. Official statistics alone cannot settle it.
What the care-concern figures count
The Guardian’s R20 process concerns referrals about alleged sexual abuse or exploitation, including circumstances where a carer allegedly failed to protect a child. These are administrative records of matters reaching oversight. They are not a count of proven offences or a survey of all abuse.
| Reporting year | Referrals | Children and young people involved |
|---|---|---|
| 2023–24 | 162 | 280 |
| 2024–25 | 130 new referrals | 222 |
Sources: 2023–24 annual report, printed page 26, Table 8; 2024–25 annual report, page 43.
A referral can involve more than one child. The 2024–25 report separately identifies eight matters involving nine children carried over from the previous year; those are not added to the 222 figure above. The report cautions that the figures do not represent prevalence: not every instance becomes known or is reported through the relevant channels.
Comparison limit: fewer recorded referrals cannot, by itself, demonstrate less abuse or a safer system. Reporting, disclosure, referral eligibility and case handling all matter. These annual figures also cannot be added to produce a verified total of distinct people across years: the published totals do not establish that there is no overlap.
Why historical and annual figures stay separate
Mullighan’s 2008 State care inquiry identified 242 people whose records established that they were in State care at the time of alleged abuse. They made 826 allegations concerning the 1940s–2004. These are different units, collected through a historical inquiry over a very different period. Mullighan report, printed page 24.
Neither 242 people nor 826 allegations can be plotted against the Guardian’s annual 222 as a trend in abuse. Care status established from records is also different from a criminal finding. The separate 2008 APY Lands inquiry is not included in these State care totals.
From recommendations to everyday practice
Mullighan’s recommendation 14 called for an allocated social worker and at least monthly face-to-face contact for every child in care. Nyland’s recommendation 77 returned to allocation and monthly contact. The official 2022 Trust in Culture review used this history to illustrate the difficulty of judging implementation: a policy response and evidence of delivery are different things. It acknowledged professional judgment about individual needs and said a full assessment would require an audit of children’s files. Trust in Culture, printed pages 55–56.
The Department for Child Protection’s 2025–26 annual report provides three separate measures:
- Allocation: all children and young people in care had a Primary Assigned Case Worker at 30 June 2026.
- Contact: 52.5% had a face-to-face meeting with their caseworker in the month ending 30 June 2026.
- Recommendation status: 251 of Nyland’s 260 recommendations were reported completed; five remained in progress and four were not accepted.
Allocation is not the same measure as contact. June’s contact figure does not describe every month of the year or explain individual circumstances. Completion status is a government implementation measure, not a direct measure of every child’s experience or safety.
Campaign interpretation: these distinctions support asking how reforms are checked in practice. Campaign proposal: an independent examination should consider delivery, oversight and children’s experiences alongside policies and completion reports. This is a proposed focus, not an agreed inquiry’s terms of reference.
Claims that need more evidence
A circulating “300+” claim has not been adopted as a campaign statistic. Its original source, reporting period, definitions and method for identifying distinct children have not been verified. Repetition cannot replace those checks.
Accounts from children, care leavers, families and workers can identify issues that administrative records miss. Personal testimony must be represented with permission and context; it should not be converted into a population estimate. This website does not invite public disclosure of identifying details about children.
Five choices for clearer information
Research on comprehension can inform design. It does not prove that a particular campaign page changes opinions, improves safety or prevents distress. The following are design principles, with their limits made explicit.
1. Give every number a unit and a date
Use “referrals”, “children involved” and the reporting period beside a number. Provide readable text or a table alongside any chart. Research found substantial differences in adults’ ability to interpret health graphs; a review of numerical communication recommends clear language, consistent denominators and time periods. These findings come mainly from health communication, so their application here is a design inference. Galesic and Garcia-Retamero, 2011; Fagerlin, Zikmund-Fisher and Ubel, 2011.
2. Put essential limits beside the claim
A visitor should see “allegations”, the period covered and the main comparison limit without opening a footnote. Experiments on communicating uncertainty suggest that acknowledging numerical uncertainty need not substantially undermine trust. They did not study this campaign or child-protection allegations. Uncertainty should be described honestly, never replaced with an invented margin of error. van der Bles and colleagues, 2020.
3. Offer depth without hiding meaning
Start with a short explanation, then offer sources and definitions. Essential caveats remain visible; optional detail can expand. Multimedia learning research supports managing unnecessary mental effort and presenting material in manageable segments, while usability guidance describes progressive disclosure. Neither establishes an ideal page length or a conversion benefit for this subject. Mayer and Moreno, 2003; Nielsen Norman Group, progressive disclosure.
4. Make interaction help understanding
An optional question such as “Can these annual totals be added?” can invite reflection, followed immediately by an explanation. Research on retrieval practice found benefits for delayed recall in learning experiments. That does not establish that clicking a reveal button produces learning. Avoid scoring visitors, trivialising abuse or making interaction a barrier to the evidence. Roediger and Karpicke, 2006.
5. Give visitors control over a difficult subject
Use factual language, optional detail and an accessible support route. Avoid graphic imagery and autoplay. SAMHSA’s framework emphasises safety, transparency, collaboration and voice; applying those principles to a website is not a claim of clinical efficacy. Research on content warnings does not support promising that a warning prevents distress. A brief content description can support an informed choice about reading. SAMHSA, 2014; Bridgland, Jones and Bellet, 2024.
What comparable public sites offer
The following observations come from parsed first-party page content reviewed on 30 October 2026, not hands-on testing of their interfaces. They describe features, not evidence of effectiveness or endorsement.
- Scotland’s Promise Progress Framework connects child-centred themes to outcomes and evidence. It informs the campaign’s question-first structure: start with what a child should experience, then examine what a measure actually tells us.
- The Children’s Commissioner’s Big Ambition maps pair topic and area choices with evidence and table alternatives. The transferable principle is to keep data definitions and an accessible reading route alongside an interactive display.
- Home Stretch connects a clear care-to-21 policy request with state-by-state information and lived-experience stories. Its useful lesson is to distinguish the requested change from each jurisdiction’s current position.
- NSPCC Talk PANTS pairs a small set of plain-language ideas with conversation examples, FAQs and accessible or translated resources. Its UK child-safety material serves a different audience from this South Australian campaign.
- Bravehearts’ Ditto programme presents age-specific learning formats and resources. Its broader separation of education, research and support is useful; a children’s game or mascot is not a template for presenting allegations to adults.
In the media
Our news collection brings together selected reporting on child protection, with publication dates, short summaries, material responses and links to the original articles. It includes current coverage and historical reporting on the Mullighan and Nyland inquiries.
Reporting provides context. It does not imply endorsement of this campaign, and findings about particular matters should not be generalised to every child or placement.
Source inventory and review approach
Core official records
- 2008 Mullighan report: historical evidence, counting definitions and recommendation 14. The relevant passages are linked above.
- 2016 Nyland Royal Commission and government response: the completed inquiry and its recommendations, distinct from this proposal for a new inquiry.
- 2022 Trust in Culture: assessment of implementation and practice, including the monthly-contact example.
- Guardian annual-report collection: original reports and surrounding explanations. The comparison above uses 2023–24 and 2024–25.
- DCP 2025–26 annual report: the department’s current published allocation, June contact and recommendation-status figures cited above.
Sources are selected for relevance and traceability, with original reports preferred to repeated headlines. Publication dates, reporting periods and definitions are checked separately. Page references use printed numbering where stated; PDF links may use a different page count.
A useful comprehension check is whether a reader can explain the campaign’s request, distinguish a referral from a proven offence, recognise a comparison’s limits and find their MP. Clicks, time on a page or an emotional response do not establish understanding. These are evaluation criteria, not claimed results from testing this website.