[thank you to all those supporting me to date in much greater numbers than I had expected. It’s a bit difficult to stick with my longish blogs. I’m sharing my pre-confirmation PhD thinking for today so apologies for the dryness and density, but I feel that we need to go here to engage the neoliberal agenda, lighter material to come later]
which metaphor for an assessment principle – constellation or continuum?
It is easy to agree with Geoff Masters (2013, p. 1) when he observes educational assessment as a field divided and in disarray. Educational assessment began by providing simple reliable indicators to parents as well as to students for currency in the job and education markets. Assessment has now grown to encompass school and system evaluation as well as scientific research, with elements of quality management and market research creeping in. Data collection is moving from research as event to embedded and ongoing research through ubiquitous and unobtrusive data collection (Behrens & DiCerbo, 2014). This transition is blurring the demarcation between educational assessment and other forms of data collection.
While it’s easy to agree on the disarray, Masters’ unifying principle to address the chaos is problematic. Masters proposes that the fundamental purpose of assessment is to establish where learners are in their learning at the time of assessment (2013, p. 5), but this principle seems too attached to the objective measurement school and its philosophical stance.
The problem that Masters is sensibly trying to address is the divided approaches and paradigms in contemporary assessment practices such as quantitative, qualitative, formative, summative and the like. Masters addresses this problem by suggesting a universal transcendent principle to underwrite all assessment practices. But for his principle to be unifying, universal and useful it must be better than competing alternate ways of formulating a principle. So is Masters’ principle something we could all agree to over other contenders for universal principles? While I do not propose to proffer an alternative at this stage, let’s explore Masters’ principle a little further.
Masters’ unifying principle is presaged by a learning space, either unidimensional (continuum) or multidimensional (continua), in which a student can be located at a particular point in time. The language of the principle is about mathematical space and location, and by incorporating this metaphor into a principle he seeks to subsume all assessment practices. His principle assumes that there is a true location at which each learner can be located at a point in time, and that once that location is determined that information can be used to fulfil all possible educational information purposes. So there are two issues, is the location metaphor the best way to describe contemporary assessment practices, and is a location – should it be able to be determined – once determined be sufficient to meet all educational information needs.
The foundation for Masters’ principle appears to be the objective school of measurement with its Rasch-based and IRT-based models (e.g. see Embretson & Reise, 2000; Masters, 1982; Rasch, 1980). It is this school of measurement with its concerns for true score and measurement error that lends itself to the ‘where is the student’ metaphor. However, there are increasing calls for the use of other measurement models for which the ‘who is the student’ metaphor is probably more appropriate. Notable examples of this work includes that of Mislevy as well as that of Leighton and Gierl (Almond, Mislevy, Steinberg, Yan, & Williamson, 2015; Leighton, Gierl, & Hunka, 2004; Leighton & Gierl, 2007, 2011). These alternative models, by moving away from the singular location metaphor, challenge the usefulness of Masters’ unifying principle.
There are several ways of describing and locating Masters’ unifying principle. One that comes to mind is that Masters takes a Kantian approach with its focus on objective transcendence presupposing learning as moving from location to location. From the objective measurement school this is couched as ‘the idea of the variable must transcend any particular set of observations and the measure on the variable must transcend the observed responses on which it is based’ (Wright & Stone, 1979, p. 141), where what is learning is seen as an a priori concept measured by the subject through empirical observation; along with appropriate application of measurement error. By casting Masters’ approach as Kantian allows us to quickly sketch out a landscape of alternative foundations for a unifying principle.
Unlike Kant, Hegel took history into account. Where Kant thought he could say on purely philosophical grounds what human nature is and always must be, Hegel accepted that the Human condition could change from one historical era to another (Singer, 2001, p. 13). The Hegelian notion of a dynamic history challenges the stability of Masters’ notion of ‘establish where learners are’, because this location is dependent on historical context. It’s then a fairly short leap to a Marxist critique of the principle, that any measure used to implement the principle could be biased against certain groups which of course could be mitigated by techniques such as DIF. It is at this point that I find we can discard Masters’ principle from being universal, and that it’s at best a useful heuristic. This brief analysis points to the danger of basing principles on an instrumental technique, in this case the Rasch model. A principle should probably come before selecting a technical implementation.
When considering assessment from a Marxist perspective, and within the context of Lyotard’s (1984) analysis of knowledge , three further approaches become apparent. The first one is neo-liberalism and its concern for performativity (Ball, 2003) which Lyotard (1984, p. 54) describes as being defined by an input/output ratio. Masters’ Rasch model provides a particular advantage here over other models such as Bayesian networks . As Masters has earlier stated, in order to enable quantitative comparisons, or make ratios, we need a linear scale that makes differences between persons the same wether through hard or easy items(Wright & Masters, 1982, p. 8). That is, the Rasch model’s ability to create linear scales dovetails neatly into neoliberalism’s need for ratios. Masters may therefore be inadvertently buttressing a neoliberal agenda with his unifying principle.
Returning to Lyotard(1984), the Marxist agenda bifurcated around the time his book was published into what I characterise as post-structuralists and neo-modernists. On assessment, the post-structuralist due to their incredulity of grand-narratives (in particular those that involve numbers) continue to take a suspicious stance towards systems and system assessment. This stance has continued to grow since early days of the Frankfurt school in particular Marcuse and his notion of the Great Refusal (Marcuse, 1974, 2012). Post-structuralists therefore can find it difficult to engage with system assessment in a positive sense, but they have a lot to say about the lives of individuals within the lifeworld which continues to be valuable for system assessment. Neo-modernists on the other hand, in the tradition of Habermas (1985, 1987), are simpatico with the petit narratives of the post-structuralists but engage more constructively with systems. Neo-modernists consider the system to have emancipatory potential while having a tendency to colonize the lifeworld of communities that needs to watched and mitigated through transparency and deliberate democratic processes. From a neo-modernist perspective, a principle should be based around what is sought to be achieved, what needs to be understood, or what needs to be coordinated across the system. A neo-modernist will continue to embrace the objective measurement school strongly however, because of objective measurement has a strong ability to determine DIF, bias, and fairness. But objective measurement would not presage a universal principle on assessment.
This author will continue to work in a modernist tradition towards one or more universal principles for assessment to provide alternative to Masters which I consider too close to the neoliberal agenda.
Almond, R. G., Mislevy, R. J., Steinberg, L., Yan, D., & Williamson, D. (2015). Bayesian Networks in Educational Assessment. Tallahassee: Springer.
Ball, S. J. (2003). The teacher’s soul and the terrors of performativity. Journal of Education Policy, 18(2), 215–228.
Behrens, J. T., & DiCerbo, K. E. (2014). Harnessing the Currents of the Digital Ocean. In J. A. Larusson & B. White (Eds.), Learning Analytics:From Research to Practice (pp. 39–60). New York: Springer.
Embretson, S. E., & Reise, S. P. (2000). Item Response Theory for Psychologists. L. Erlbaum Associates.
Habermas, J. (1985). The Theory of Communicative Action: Reason and the rationalization of society. (T. McCarthy, Trans.). Boston: Beacon Press.
Habermas, J. (1987). Lifeworld and system: a critique of functionalist reason. (T. McCarthy, Trans.). Boston: Beacon Press.
Leighton, J. P., & Gierl, M. J. (2007). Cognitive Diagnostic Assessment for Education: Theory and Applications. New York: Cambridge University Press.
Leighton, J. P., & Gierl, M. J. (2011). The Learning Sciences in Educational Assessment: The Role of Cognitive Models. Cambridge University Press.
Leighton, J. P., Gierl, M. J., & Hunka, S. M. (2004). The Attribute Hierarchy Method for Cognitive Assessment: A Variation on Tatsuoka’s Rule-Space Approach. Journal of Educational Measurement, 41(3), 205–237. doi:10.1111/j.1745-3984.2004.tb01163.x
Lyotard, J.-F. (1984). The Postmodern Condition: A Report on Knowledge. Minneapolis: University of Minnesota Press.
Marcuse, H. (1974). Eros and Civilization: A Philosophical Inquiry Into Freud. Beacon.
Marcuse, H. (2012). One-Dimensional Man: Studies in the Ideology of Advanced Industrial Society (Vol. 8). Beacon Press.
Masters, G. N. (1982). A rasch model for partial credit scoring. Psychometrika, 47(2), 149–174. doi:10.1007/BF02296272
Masters, G. N. (2013). Reforming Educational Assessment: Imperatives, principles and challenges. Australian Education Review. Retrieved from http://research.acer.edu.au/aer/12
Rasch, G. (1980). Probabilistic Models for Some Intelligence and Attainment Tests. Chicago: MESA PRESS.
Singer, P. (2001). Hegel: A Very Short Introduction. Oxford: OUP Oxford.
Wright, B. D., & Masters, G. N. (1982). Rating Scale Analysis. Chicago: MESA PRESS.
Wright, B. D., & Stone, M. H. (1979). Best Test Design. Chicago: MESA PRESS.