Scholarlink

<p>Understanding the structural drivers of teaching cost is a persistent challenge in higher education systems internationally, including in Australia. To extend research on cost heterogeneity and structural cost drivers, we develop and apply a Structural Cost Profiling (SCP) approach, based on Latent Class Analysis (LCA), to analyse 189,811 ‘units of study’ records from 12 Australian universities. Centring on cost per enrolled student, we identify eight distinct cost profiles that capture systematic, multidimensional variation in staffing, enrolment scale, delivery mode and campus location. While the data show significant variation across the dimensions, our analysis demonstrates that cost heterogeneity is not random but structured, with higher average costs per student typically associated with configurations such as (1) regional delivery with small cohorts and labour-intensive postgraduate provision, whereas (2) lower costs emerge from scale-based online delivery models. A separability analysis shows that staffing allocation, salary expenditure and teaching hours are the indicators most strongly aligned with the latent classes, while categorisations such as discipline (Field of Education) or delivery mode play a more limited and contextual role in differentiating the cost classes, instead helping to indicate where particular cost profiles occur, rather than defining those profiles on their own. These results suggest the need for clarity around the intentions and purposes of funding policies, which risk over- or under-funding where cost structures are assumed without an evidence base.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analysing the costs of teaching in universities through structural cost profiling

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月19日

查看全文
  • Gwilym Croucher,
  • Mark Rahimi

摘要

Understanding the structural drivers of teaching cost is a persistent challenge in higher education systems internationally, including in Australia. To extend research on cost heterogeneity and structural cost drivers, we develop and apply a Structural Cost Profiling (SCP) approach, based on Latent Class Analysis (LCA), to analyse 189,811 ‘units of study’ records from 12 Australian universities. Centring on cost per enrolled student, we identify eight distinct cost profiles that capture systematic, multidimensional variation in staffing, enrolment scale, delivery mode and campus location. While the data show significant variation across the dimensions, our analysis demonstrates that cost heterogeneity is not random but structured, with higher average costs per student typically associated with configurations such as (1) regional delivery with small cohorts and labour-intensive postgraduate provision, whereas (2) lower costs emerge from scale-based online delivery models. A separability analysis shows that staffing allocation, salary expenditure and teaching hours are the indicators most strongly aligned with the latent classes, while categorisations such as discipline (Field of Education) or delivery mode play a more limited and contextual role in differentiating the cost classes, instead helping to indicate where particular cost profiles occur, rather than defining those profiles on their own. These results suggest the need for clarity around the intentions and purposes of funding policies, which risk over- or under-funding where cost structures are assumed without an evidence base.