Editorial

2020 is likely to go down in the history books as a memorable year, if not an easy or relaxing one for those involved in education. Between the fires, floods and the COVID-19 pandemic, there have been many interruptions to learning and cancellations of gathering to celebrate milestones such as birthdays and anniversaries. This Special Issue represents our socially distanced celebration of the coming together of two educational milestones – the 90th birthday of the Australian Council for Educational Research (ACER) and the publication of the latest results from the Programme for International Student Assessment (PISA) representing more than 20 years of participation by Australian students and schools in this international study.

PISA measures how well 15 year olds, who are nearing the end of their compulsory schooling in most participating education systems, are prepared to use their knowledge and skills in particular areas to meet real-life opportunities and challenges. Beginning in 2000, every three years, a sample of Australian students completes an assessment in reading, mathematical and scientific literacy, along with other areas of interest, to gauge how well students are able to apply their learning. In addition, PISA collects attitudinal data on students’ interest and self-efficacy, along with reports of students’ experiences of various aspects of school life, such as interactions with teachers and other students.

This Special Issue brings together both the ACER and PISA in Australia milestones, presenting six articles that focus on various aspects of Australian (and New Zealand) students in different cycles of PISA, including three by authors with affiliations with ACER. Three articles investigate changes in average performance in reading, mathematics and or science over time (using multiple cycles of PISA data), while the other three make use of data from a single cycle of PISA.

John Ainley, Dan Cloney and Jess Thompson focus on the declines in average PISA reading, mathematics and science scores that have been recorded among Australian students. They explore whether some of this decline may have resulted from shifts in the age-grade distributions of students in the samples brought about by policy changes regarding age of entry to school. PISA uses an age-based sample, focusing on 15-year-old students, rather than sampling intact classes at a single grade level. Australian students who participate may be in Grade 9 through Grade 12 and thus have had differing amounts of schooling at the time of the assessment. Using multiple regression methods, the authors model the effects of grade (or which year level students were in) on performance while controlling for student characteristics and jurisdiction. On average, students in higher grades score better in reading and mathematics than students in lower grades. The authors then explore how changes in the performance of students in different grades compare to changes in the average performance of the samples, and how changes in the distribution of grades in the PISA samples may have contributed to the overall decline noted in Australian students’ performance in PISA over the past decades.

A second article investigated changes over the seven cycles of PISA, this time focusing on changes in Australian students’ attitudes towards reading, mathematics and science, and the relationships between changes in attitudes and changes in student performance in these areas. Over two decades of assessment, there have been alterations in the way questions about students’ interest in and enjoyment of reading, mathematics and sciences have been worded, and I Gusti Darmawan first had to ensure that what was being measured as ‘student attitudes’ had not changed as a result. Multigroup Confirmatory Factor Analysis (M-CFA) was used to examine the measurement invariance of the construct of attitudes over the cycles of PISA. Once satisfied that there was consistency in what was being measured over the seven cycles, the author used Structural Equation Modelling to explore mediating effects of attitudes on performance in reading, mathematics and science over time. As reported elsewhere in this Special Issue, but also in the series of Australian National PISA reports, the average performance of Australian students has declined in each of the three areas (see https://www.acer.org/au/pisa/publications-and-data). At the same time, levels of enjoyment of mathematics and science increased over time, while levels of enjoyment of reading remained fairly stable, albeit relatively low. Around half of 15-year-old Australian students read only to get information they need or only if they had to, and one-third of the students reported that reading was a waste of time in 2000. Levels of instrumental motivation were relatively high and had slight upward trends for both mathematics and science. In terms of the relationships between attitudes and performance over time, enjoyment of reading had a relatively strong effect on reading performance in 2000, stronger again in 2009 (when reading was again the main focus of PISA) but slightly weaker, although still positive, in 2018. While Australian students may be becoming more aware of the importance and relevance of mathematics and science for their future studies and careers, this is not reflected in their performance, and the association between such positive attitudes towards mathematics and science and student performance in these key areas appears to be weakening over time.

Next, Megan Chamberlain and Emma Medina compare the performance of Australian and New Zealand students in PISA reading literacy over the past two decades. In PISA’s first cycle in 2000, when reading literacy was the major focus of the assessment, both countries’ students generally performed well above the OECD average. Nearly 20 years later, the picture is less positive, with the latest cycle of PISA in 2018, where reading literacy was again the main focus, showing a decline of students’ average reading performance over time in both countries. To understand this decline, the authors explore the performance of various social groups overtime, including socio-economically advantaged and disadvantaged students, indigenous and non-indigenous students, and genders, as well as key educational challenges facing both countries, such as poor disciplinary climates, declining attitudes towards reading, and increased bullying. Multiple linear regression is then used to examine relationships between student background – using those contextual and demographic variables examined earlier – and reading literacy performance for Australian and New Zealand students. In both countries, confidence and reading for enjoyment have the largest effects, while a higher sense of belonging has a negative effect on reading with all other variables held constant. This signals that if all other conditions for academic excellence are not met, having a strong social network at school may not be enough to assist students achieve. While creating a positive learning environment and developing positive attitudes toward reading are important, these alone will not erase the gaps in performance between sub-groups. The authors conclude that the strong similarities in the educational contexts of Australian and New Zealand provide policy-makers and educators in both countries with the opportunity to learn from the successes (or failures) of each other and to collaborate to improve outcomes, including reading literacy, for future cohorts of students.

Moving on to articles that focus on a single cycle of PISA data for Australian students, Claire Scoular, Sofia Eleftheriadou, Dara Ramalingam and Dan Cloney use data for the innovative domain for PISA 2015 – collaborative problem solving. Along with the three main domains of assessment – reading, mathematics and science – PISA has included an assessment of an additional, optional area, which has come to be called an innovative domain. In PISA 2015, this innovative domain was collaborative problem solving. The online assessment made use of computer-simulated agents, that is, sets of pre-programmed responses which were intended to act as collaborative partners for the students in place of other, real-life students in a range of problem-solving scenarios. The authors compare Australian students’ performance on this assessment with Australian students’ performance from two other, more recent assessments of collaborative problem solving to explore the extent to which the three assessments measure the same construct. They also explore how future assessments of collaborative problem solving might make improved use of technology and online presentation for a more in-depth assessment of collaboration and problem solving in a digital environment – an area that is certain to be of interest in the current remote learning and working circumstances and into the post-COVID world.

In another article focusing on a single cycle, Florence Gabriel, Sarah Buckley and Abhinava Barthakur analyse PISA 2012 data to examine how mathematics anxiety operates to influence self-regulated learning and mathematical literacy. Based on prior research and theory, they hypothesise a model in which self-concept and instrumental motivation affect mathematics anxiety, which, in turn, operates through perseverance and self-efficacy to ultimately influence mathematics literacy. Results of the structural equation model analyses show the strongest effect between self-concept and mathematics anxiety whereby lower self-concept is linked to higher mathematics anxiety. As a central concept in the model, mathematics anxiety is shown to relate negatively to perseverance with students expressing greater anxiety showing less inclination to persevere with a task. As the final element of the model, self-concept also affects self-efficacy which, in turn, has the second largest effect in the model whereby students expressing higher abilities in relation to completing specific tasks also perform at a higher level in mathematics. Together, these results show that working to increase students’ self-concepts has independent and concurrent desirable effects, lowering mathematics anxiety and raising self-efficacy, which ultimately have positive effects on mathematics literacy.

One unique aspect of PISA in Australia is the way in which successive PISA student samples have been integrated into longitudinal surveys of young people’s pathways through study and work and young adulthood. In this way, performance in reading, mathematics and scientific literacy in PISA may be related to later outcomes in terms of education, employment and wellbeing. In Australia, the students who took part in the PISA assessments in 2003, 2006, 2009 and 2015 were invited to participate in follow-up surveys and become respondents in the Longitudinal Surveys of Australian Youth cohorts for those years. 2020 also marked LSAY’s 25th anniversary, marking one quarter of a century of research into young Australian’s pathways and outcomes (https://www.lsay.edu.au/publications/search-for-lsay-publications/25-years-of-lsay-infographic).

Maien Sachisthal, Brenda Jansen, Jonas Dalege and Maartje Raijmakers combine the Australian PISA 2006 data with the LSAY data from 2008 to investigate relationships between students’ attitudes towards and performance in the PISA assessment of science and their decisions to enrol in a science course in Year 12. Their analyses use a relatively new technique, in which students’ interest in science is conceptualised as a network model. First, the model is mapped, using relationships between all of the indicators – items pertaining to students’ interest in and enjoyment of science, their intentions to pursue science-related studies or careers, knowledge of science, valuing of science and self-concept relating to science – to identify those that are more central to the network (have more, and stronger relationships with other indicators) and those that are more peripheral (have fewer, weaker relationships). The authors then investigate whether the indicators that are more central to the science interest network models (SINM) were associated with students’ responses in the 2008 LSAY survey about whether they had actually enrolled in a science course – specifically, physics, chemistry or biology – in Year 12. The results indicate that students’ intentions to pursue science-related study or careers and their enjoyment of science are the most central indicators for all three science courses, while actual knowledge of science, as measured by the PISA assessment, does not appear to play such a large role in the network. The findings have implications both for theorists working with network models and interest, as well as those investigating how to encourage more students to pursue studies and careers in science.

The articles in this Special Issue illustrate how further analyses of the PISA data can be used to address various questions about the education of our young people and highlight why Australia’s participation in such extensive and repeated data collections can be of value to educators, researchers and policy-makers.

Stay safe and happy reading!