NAPLAN, Conflict of Interest and Research Ethics

Rejoinder to Timna Jacks – Company marking NAPLAN accused of conflict of interest

Timna Jacks’ article on Pearson Australia is a welcome reminder of Jean-François Lyotard’s (1979) seminal observations about data banks and the commercialization of knowledge.  The issues Lyotard identified have been emerging for decades and can be addressed through the tradition of research ethics.

Koomen_Poster_DataUses

Research ethics seeks to protect the vulnerable in data collection, in this case students.  The traditional ethical concerns of informed consent and conflict of interest are central to the issue brought to light by Jacks and are concerns that have been somewhat disregarded in the data frenzy currently capturing the education sector. On informed consent there are four key ethical issues: 1) do students have a choice about participation, 2) do students trust the data collection process, 3) are students confident that their results will be used fairly, and 4) are the interests of students, data agencies and third parties balanced.  The nature of students’ informed consent is unclear, there is a social compulsion to participate on the basis that there is a legislative mandate for students to attend school, but the mandate to attend school translates into a social expectation, and not compulsion, for students to participate in NAPLAN.  The key issue in Jacks’ article centers on the balanced interests of parties. Are the interests of students sufficiently balanced with the interests of others? The article suggests no.

By way of background, education is currently experiencing a clash of data collection traditions.  Traditionally educational assessment focused on providing a reliable indicator that teachers could use to report to parents and that systems could report to students for use in the broader education and job markets.    A second distinct data tradition relates to school evaluation and accountability such as the evaluative reports made available through the MySchool website – www.myschool.edu.au.  That NAPLAN provides a reliable indicator to parents and systems generates broad public support for the program – even if its curriculum coverage is somewhat limited. But there are now three other data traditions operating across education that may also be infiltrating NAPLAN and which may not be so transparent: scientific education research, quality management, and market research.

Public support for scientific education research tends to be high but this is a little more fraught. Scientific research is littered with disturbing episodes (e.g. Albert Neisser, Willowbrook, Tuskegee) but has been largely tamed through initiatives such as the Declaration of Helsinki and the work of university ethics committees. The extent to which NAPLAN data is used for scientific research and the ethical frameworks surrounding this research is unclear.  Therefore there is some justification for public concern on these matters.

Walter Shewhart in quality management provides yet another data tradition.  This tradition became prevalent during the industrial age to ensure quality and reproducibility of manufacturing. These techniques are now widely applied in the service sector and are increasingly being applied in education. In education, this tradition is used by educational administrators to influence the work of schools and teachers.

Finally, it is the tradition of market research that is the most pernicious in education and the issue at the heart of Jacks’ article; the possibility that data collected on the basis of creating a common understanding is being used for concealed strategic action.  While it is unlikely that this may be happening within such a large organisation, it is the possibility that it might be happening that is of concern, and it questions the social expectation that we as adults place on children to participate in NAPLAN.

The existing regulatory framework around data collection in Victoria, for example, is quite fragmented and patchwork. Children are mandated to attend school through the Education and Training Reform Act 2006 which is silent on participation in testing. There’s also the Privacy and Data Protection Act 2014, the Health Records Act 2001 and Public Records Act 1973. It is uncertain if this legislative and regulatory framework appropriately addresses the underlying issues first identified by Lyotard and which Jacks’ alerts us to in her article.

Jacks alerts us to a significant issue in education and we need to be thankful for her efforts. But it’s potentially only the tip of the iceberg in terms of ethical issues.  Two actions are required of government. First, a comprehensive review of the legislative and regulatory frameworks around data collection in education.  Should any shortcomings be identified these need to be addressed and new standards promulgated to bureaucrats, contractors, parents, teachers and students. The second action relates to conflict of interest. Government needs to centralize data collection in a new statutory agency independent from education administration and commercial education services.  That is, data collection, indicator production and reporting should reside in an authority independent from the Department responsible for the management of schools and teachers, and reside in an authority with no other responsibility but data, indicators and reporting. This would also mitigate the kind of ethical issues the Victorian department has recently experienced.