Innovating Pedagogy – Dynamic Assessment
This is the third in a series of annual reports on innovations in teaching, learning and assessment. The Innovating Pedagogy reports from Open University(UK) are intended for teachers, policy makers, academics and anyone interested in how education may change over the next ten years.
Giving the learner personalized assessment to support learning
Potential impact: medium
Timescale: long (4+ years)
The basic premise of dynamic assessment is that it is important to assess students’ potential to learn rather than measure what they have just done. Testing acts as a diagnostic tool that enables a teacher or the computer to offer guidance to the student during the assessment process. Thus, it differs from conventional testing in both assessing and guiding the progress of the student. The relationship between the assessor and student is not neutral because the whole purpose is to find suitable ways to promote student learning. A role of the assessor is to identify barriers to the student’s success and then apply an appropriate strategy to overcome the difficulties. Assessment and intervention combine in the process of dynamic assessment.
As well as being a way to offer direct support to the learner, dynamic assessment can inform the teacher about topics and skills that many students are finding difficult and so help the process of re-designing and improving the teaching. It can also motivate learners to reflect on their learning journeys and decide on which skills they need to improve. It is particularly valuable for developing 21st-century skills of reasoning, problem-solving, decision making, leadership, creativity and literacy.
Zone of Proximal Development
The dual function of assessing and improving a student’s level of attainment is based on the notion of a zone of proximal development (ZPD). This is the difference between what a student can do unaided and what is possible with assistance. For example, a student who struggles to think of ways to measure the height of a building might be guided ﬁrst to consider how the timer or the tilt sensor on a mobile phone could be used in solving the problem, and then how this could be done without the need for algebra (by sighting a 45-degree angle to the top of the building, at which point the distance from the building to the student is the same as the distance from the base to the top of the building).
The combined measures of the student’s actual development to date plus the zone of proximal development (what they could achieve with appropriate help), are more powerful indicators of attainment than a single score of current development, such as that given by an IQ test.
The methods of dynamic assessment are designed to allow the assessor (who may be a human teacher, or a computer-based tutoring program) to measure the effect of an intervention or lesson on the student’s performance. Typically, this is done by setting a pre-test of the student’s knowledge or skill, followed by a teaching intervention, and then a post-test to see how the knowledge or skill has changed. The ‘pre-test, intervention, post-test’ method resembles a traditional research design, but the aim is not to carry out a research study on a group of students, but to use this information to guide the learner and inform the teacher.
In the ‘sandwich format’ students receive coaching between the pre-test and the post-test. Alternatively, in the ‘cake’ format, students receive hints during the testing session. These two methods could be combined so that, in the example above, the student might be given hints to use a tilt sensor or to find an angle that would remove the need for algebra, then all the students could be taken out into the playground to measure the height of the school building, then they could be individually re-rested to see if they are able to apply their knowledge to a similar problem.
There is also an interactionist methodology that takes more account of the student’s zone of proximal development. In this case, teacher and student work together to solve a problem, with an emphasis on the teacher providing continual appropriate guidance.
The differences between dynamic assessment and conventional or static assessment are that:
- the focus of dynamic assessment is on guiding future development, whereas static assessment measures past achievement;
- the assessor and student relationship is different, since the assessor intervenes during the process;
- with dynamic assessment there is feedback to the student during the assessment process.
An example of dynamic assessment at university level was the assignment of students to an undergraduate Spanish course by finding a good match between the level of development for each student and appropriate Spanish teaching. In this case, the assessor prompted the students to revise their answers if they had made mistakes. Some students were able to improve with prompting, while others were not. Those who did revise their answers correctly were placed in an advanced language course.
A second example comes from a school science lesson with children aged eight and nine years who were learning about magnetism. The teacher discussed the topic with the children and was sensitive to each child’s ZPD, helping the class to move from everyday language, such as ‘hold’ and ‘push’, to the use of scientific terms such as ‘attract’ and ‘repel’.
A third study was designed to support English as a Foreign Language with young adults who were considered to be ‘at-risk immigrants’ entering Israel. The students were given a reading and comprehension pre-test. Assessors then went through the test with the students, building strategies with them to address each test item and, more importantly, showing them how these strategies could be transferred from one example to another. A learning potential score was devised by the assessors and the students were then designated as high, medium or low performers. This dynamic assessment facilitated more accurate instructional recommendations that helped these students to improve their English language skills.
Providing dynamic assessment of each student and giving timely guidance is a demanding task for a classroom teacher, so attempts are being made to develop software to perform this process. For example, the OpenEssayist system gives immediate feedback to university students on their draft essays. It analyses the essay and produces a summary of the key sentences the student has used, enabling learners to decide whether these sentences are what they intended to emphasise in their assignment and then go back and refine the main argument of their essay. A more general system of dynamic assessment is provided by the suite of Cognitive Tutors® from Carnegie Learning that provide individualized assistance to students as they work through problems in computer programming, algebra or geometry.
Dynamic assessment has been criticised on the grounds of its reliability. In order to construct a reliable test, the test items need to be stable, but the dynamic assessment procedure is deliberately associated with change, not stability. However, the main value of any assessment lies in the inferences that can be made from it about how well the student is progressing, and with dynamic assessment these inferences are sound, since they are tightly connected to administration of its test procedure. Although dynamic assessment puts new demands on teachers, test instruments have been constructed for use in the classroom. It should be considered as part of a range of assessment tools that can support individual students to reach their full learning potential.
Sharples, M., Adams, A., Ferguson, R., Gaved, M., McAndrew, P., Rienties, B., Weller, M., & Whitelock, D. (2014). Innovating Pedagogy 2014: Open University Innovation Report 3. Milton Keynes: The Open University
Classroom Aid editor note:
If well designed and planned, Experience API (xAPI) is an excellent vehicle to collect evidences along the learning path. The records should be automatically visualized to learners and educators to build the feedback and communication loops. Combined with “Learning to Learn”, data-driven learning strategy also can help learners own their learning.