Reviews for Educational Objectives and National Assessment

The average rating for Educational Objectives and National Assessment based on 2 reviews is 4 stars.

Review # 1 was written on 2009-08-02 00:00:00 David Bartlett A lot of this book isn't really so much what you might expect from the title - that is, proposing a new system of assessment for schools and teachers. Really, that only comes right at the end. Most of this book explains the limitations and complications of current systems of assessment used in schools and why these, despite their supporters and detractors, have clear limitations, as well as unintended consequences. The author covers a lot of important ground here - the difference between formative and summative assessment, or between norm and criterion based assessment tools. What is particularly interesting with much of this is that large assessment strategies (the kind of assessments that test the entire cohort of an education system to gauge system wide health) require a level of statistical sophistication that most teachers simply do not have. This means that they are required to trust that those who put the tests together not only know what they are doing, but that they are also not going to 'over egg the pudding'. Something Australians have learnt to their cost that people in these positions are likely to do. This doesn't mean that I'm opposed to cohort wide assessments - just that we need to be told what these assessments can and cannot tell us. The other problem identified here is that assessments are not neutral. How you plan to assess something will impact what is taught and how it is taught. As Bernstein said - education has three 'message systems': curriculum (from the Latin for running track, it is what you plan to teach), pedagogy (from the Greek for guiding boys, it is how you plan to teach it) and assessment (from the Latin for to sit beside, it is how you will know what you need to teach, what your student is ready to learn). These three message systems do not sit in isolation from each other - each impacts the others. At the moment, our obsession with assessment (based mostly on our inability to trust teachers - they are mostly women after all) curtails the other two. It is also the case that our choices of assessments is not neutral in terms of gender or ethnicity either. Everyone seems to know now that I.Q. tests were iffy because they privileged white, middle class, male knowledges over all other ways of knowing. But if that has become standard knowledge for a generation, the fact other forms of assessment of 'intelligence' are equally problematic hasn't seemed to register with people at all. But, if this wasn't the case, why would it be that you can rank people as easily by their postcode as by their high stakes test scores - and far more cheaply too, by the way. Of course, it denies you the ability to say 'merit' or 'meritocracy', but otherwise the result is identical. I would strongly suggest reading the last chapter of this book - lovely stuff. Here is a big long quote I found particularly interesting: "The most challenging task in developing a theory of educational assessment is that of reconceptualizing reliability. Underlying this challenge is a shift in our world view. The psychometric model carried with it a notion of objectivity'that ability or attainment is a property of the individual which can be reliably (accurately) measured and that the resulting 'score' is unaffected by context or the testing situation. We now know however that performance is very much context bound, affected by motivation and the assessment mode itself. It is also construed according to the perspectives and values of the assessor' whether it is the one who designs the assessment and marking scheme or the one who grades the open-ended performances. We do not therefore see assessment as a scientific, objective, activity, this we now understand to be spurious. "Assessment is not an exact science, and we must stop presenting it as such. This is of course part of the post-modern condition'a suspension of belief in the absolute status of 'scientific' knowledge (Gipps, 1993b; Torrance 1993b). The modernist stance suggests that it is possible to be a disinterested observer, while the post-modernist stance indicates that such detachment is not possible: we are social beings who construe the world according to our values and perceptions. The constructivist paradigm does not accept that reality is fixed and independent of the observer; rather reality is constructed by the observer, thus there are multiple constructions of reality. This paradigm would then deny the existence of such a thing as a 'true score'." …. "Evaluation within the constructivist and naturalistic paradigms rejects the traditional criteria of reliability, validity and generalizability and looks instead for qualities such as trustworthiness and authenticity (Guba and Lincoln, 1989). We need similarly to reconceptualize the concepts of reliability and generalizability in relation to assessment, and the educational evaluation literature is a potential source. "Trustworthiness according to Guba and Lincoln is based on credibility, transferability and dependability. Credibility comes from prolonged engagement and persistent observation, ie regular on-going assessment in the classroom, and including parents in the dialogue about pupil performance. Transferability could replace the notion of generalizability: since performance is context bound the assessor must specify the context in which a particular achievement was demonstrated. Others then judge whether this will be transferable to other contexts; these are referred to as the sending and receiving contexts respectively: the more the sending and receiving contexts are alike the more likely is transfer. It is this description which, through providing extensive information about the sending context, allows judgments about transferability to be made. Dependability replaces traditional reliability, it is related to the process of assessment and the judgments made which must be open to scrutiny. Guba and Lincoln suggest that this be achieved through an audit process'possibly like the quality control process in moderation. Authenticity is to do with the extent to which the relevant constructs (and this means all the stakeholders' constructs) are fairly and adequately covered in the assessment. The fairness aspect of authenticity suggests that all groups' constructs are included rather than just the test developer's." (pp.167-8)

Review # 2 was written on 2012-08-09 00:00:00 George Delashmutt Good points, but buried under self-important language. I enjoy a good academic theoretical discussion, but this guy makes Foucault look like Dr. Seuss.