I believe tests are useful. Administered in the right way, they can help a student see what they know and what they still need to improve at. This can aid in raising standards, which is generally the goal of any test.
The problem with this is that many people identify standardised testing with raising standards. This isn't the case. Borrowing from the preachings of Ken Robinson et al., standardised testing serves to create conformity among it's recipients. It makes sure everyone can so the same thing at the same stage. This has been referred to 'fast food education' - a lot of the same thing. The argument against this type of conformity is that humans are not machines. We don't all perform at the same level at the same time for the same rate. We are individuals who all develop at different rates and, if you believe in multiple intelligences, learn in different ways.
We should do the best we can to create systems of education that cultivate conditions to let our students flourish. By trying to ensure that everyone can do the same things, at the same stage of their lives, we are sending a message to some students that they are not good enough. What does this do their psyche if they begin hearing it at the age of 8?
It concerns me that education in Australia seems to heading down the same path as the US is finally beginning to shy away from. The No Child Left Behind Act failed and the NAPLAN assessments introduced in Australia in 2008 similarly has the potential to lower standards, instead of raise them.
The school I am currently working at is at the pointy end of the process of developing learning outcomes for each grade level at our school. This is helping them achieve their goal of creating a clear and connected curriculum for K-12. Being an International School that is a private institution, they don't have the shackles of a national curriculum to adhere to. The idea is that stake-holders will be able to see what students will be able to know, understand and do at each year of their education. I think it is important for a school to have a clear and strong curriculum. How else can we drive daily instruction in relation to the school's values? During our process, the usual concerns of OBE have been raised: Are we measuring everyone to a minimum standard? What happens if students don't achieve the required level? What happens if they do it before they're 'meant' to? How will these affect assessment and reporting processes?
I advocate certain aspects of OBE. It is based on constructivism theories, it doesn't compare students to other students and assumes nothing in terms of student's capabilities. What concerns me is that if outcomes are too shallow (ie. 'a mile long and an inch deep' - as Linda Darling-Hammond puts it) or restrictive and narrow, then we run the risk of measuring our students to the same standards and this can also lends itself to an over-emphasis on standardised testing procedures. We need to ensure that we develop continuum based outcomes that allow students to achieve at a rate that is developmentally appropriate for them.