Challenges of conducting systematic research in Australia's Northern Territory

Abstract

This paper explores the challenges of conducting systematic research, using our experiences of conducting a study to evaluate the effectiveness of ABRACADABRA, an online tool for early childhood literacy instruction as the contextual framework. By discussing how the research team resolved such perennial issues as high teacher turnover, low or erratic Indigenous student attendance, difficulties with collecting reliable data on student outcomes, and the time and funding required to travel long distances, we show how rigorous research might still be conducted, to counter the usual proffering of such challenges as reasons why experimental research should not be attempted. Without minimizing the dimension of the logistical and funding challenges facing the conducting of experimental research in regional and remote settings, we end with an appeal that such work be prioritized, lest already disadvantaged education settings suffer further neglect in terms of national research priorities.

Keywords::

• early childhood literacy
• Indigenous research
• Northern Australia

Acknowledgements

This research is supported by the Australian Research Council (ARC) Linkages Project (LP0990171) in financial partnership with the Telstra Foundation and with in-kind support from the Northern Territory Department of Education and Training and the Centre for the Study of Learning and Performance, Canada. Additional research support was provided by the Fred Hollows Foundation, the Australian Institute of Aboriginal and Torres Strait Islander Studies; the Collier Foundation; and the Channel 7 Children's Research Foundation.

Notes

^1. The Northern Territory ABRA study was conducted in partnership with colleagues from Concordia University, McGill University, the University of Western Australia, the University of Sydney and the Menzies School of Health Research.

^2. To our knowledge, this is the first randomized controlled trial of literacy software targeting early childhood learners with mixed Indigenous and non-Indigenous student populations to be conducted in this region and certainly the first education project to be listed with the Australian and New Zealand Clinical Trials Registry (ACTRN) 12609001008213, which is set up to ensure that published trials both reflect their original aims and do not repress any unfavourable outcomes.

^3. In 2009, 26% of teachers in the NT left their positions, but this figure does not distinguish between teachers who left the system entirely, those who moved to another teaching position, and those who remained employed in the system, but left their classrooms to take up a non-teaching position. Further, this figure does not distinguish between urban and remote settings, and it can be assumed that the turnover rate in remote schools is much higher (Northern Territory Government Department of Education & Training, 2009).

^4. Schools in the Northern Territory are grouped into three “geolocations” for the purposes of reporting student data. These are “Provincial” (Darwin and Palmerston schools), “Remote” (Alice Springs and Katherine schools), and “Very Remote” (all other NT schools).

^5. Town Camps are quasi-permanent Indigenous settlements sitting on the outskirts of towns; in the Northern Territory these include Darwin and Alice Springs, as well as smaller towns. Despite their proximity to local services, town camps tend to have high levels of social disadvantage and poor infrastructure.

^6. Each year following an initial training to use ABRA, teachers then received continued support from a person who was an expert in early childhood literacy and could act as a literacy coach. This person was also a researcher who while gathering data on implementation fidelity also provided pedagogical and technical support (Helmer, Bartlett, Wolgemuth, & Lea, 2011).

^7. Using the power analysis equation appropriate for when students are individually randomly assigned to treatment and control groups within sites (schools) and that treated schools as fixed effects (Schochet, 2008), it was calculated that six schools with 50 students each equally allocated to treatment and control groups would have more than an 80% power to detect a difference of 33% of a standard deviation.

^8. The project secured an initial grant from the Telstra Foundation in late 2007 which allowed us to begin the first phase of the proposed 3 year study, but the funding necessary to upscale and complete the study required a continuous round of grant applications.