As a teacher, what can you learn about your students from a good assessment? How can a good assessment support your understanding of individual students and help you plan for the next steps in teaching and learning? Teacher readers often let us know they'd like more content on assessment. It’s also an area that is constantly evolving and innovating. To unpack these questions, we’ve called on our colleague at the Australian Council for Educational Research (ACER), Dr Dara Ramalingam, Research Director of Assessment Development.
Assessment is at the centre of any good educational system. At its heart, assessment is evidence: when used well, assessment data can help us understand what students know and can do, inform decisions about what they need to learn next, and provide evidence to inform education programs and policy. The psychologist David Ausubel (1968) famously argued that the most important factor influencing learning is what the learner already knows, and that teaching should proceed accordingly. It is assessment that provides this critically important information.
Precisely because assessment is so consequential, the quality of assessment materials is of the utmost importance: if we do not have high-quality test materials, anything else that follows – from reporting, to targeted instruction in the classroom, to policy initiatives based on the test results – is compromised.
But what does that mean for teachers? What can a good assessment actually tell us about a student’s learning and progress?
At its simplest, assessment provides an opportunity for students to demonstrate what they know and can do. But good assessment can tell us considerably more than how many questions a student answered correctly. It can provide evidence about the knowledge and skills they have developed, the understandings they are beginning to develop, and what they may be ready to learn next.
The role of test developers
Assessment evidence comes in many different forms, with different purposes: while data from standardised assessments provide valuable information about student achievement and progress, teachers also gather evidence every single day through classroom tests, questioning and informal observation. These different sources of evidence can be used to build a more complete picture of where a student is in their learning journey. Regardless of the purpose of the assessment, the fundamental question remains the same: how does this evidence provide a sound basis for understanding where students are, and informing next steps?
For the large teams of specialist test developers at ACER, 2 questions are at the heart of every single assessment we create. The first is, what is the construct (knowledge, skills and abilities) I am trying to measure? The second is, how will learner responses to this material act as evidence of proficiency in the construct I am measuring?
The first question may seem a simple one, but consider, for example, an assessment of early mathematics. If a young student is presented with detailed worded problems in such an assessment, we risk assessing not their mathematical ability, but their reading ability. Fortunately, test developers are skilled not only at using simple and accessible language, but also at making targeted use of technology, for example by providing audio support where reading is not the construct being assessed.
As for the second question, test developers must always consider what student responses actually tell us. Every single item should be written so that it assesses a skill of interest in the domain being measured – for example, an item within a science assessment might focus specifically on the skill of drawing conclusions from evidence – and the student’s response provides evidence of their proficiency in that aspect of the construct.
Going beyond correct and incorrect answers
The result of an assessment is often summarised in an overall score. While this is of use in providing a global sense of a student’s proficiency in a domain of interest, 2 students receiving the same overall score may show quite different patterns of responses. It is therefore valuable for teachers to analyse the results in a more granular way to see what patterns are evident.
Expert test developers deliberately construct assessments in a way that supports this purpose. They are often supported by a blueprint that provides a definition for the domain being assessed. Sometimes, the blueprint also identifies important strands or sub-strands within that domain (for example, number and algebra is a widely accepted strand within mathematics) and specifies how each should be represented in the test instrument.
For example, 2 students may receive the same score on a reading assessment, but one may demonstrate a relative strength in making inferences while another may show strength in critically evaluating texts. Examining the patterns of relative strength and weakness students show provides more information than a test score alone. For teachers, this distinction is critical, as it points to quite different learning needs.
A broader frame of reference for the evidence
Understanding what students know and can do is critically important. But this information becomes even more powerful when interpreted in the context of how learning develops within a domain.
Learning progressions are descriptions of growth, uncoupled from age- or grade-based expectations. ACER has developed several such descriptions, including for mathematics and reading, which can be found on our ACER's Learning Progression Explorer site. Using such descriptions of growth acknowledges an important reality known to every teacher: in every classroom, there are students with a wide range of abilities.
Rather than considering only whether a student is, or is not, meeting grade-based expectations, using a learning progression allows teachers to see which skills a learner has mastered, which they are consolidating, and which are still to come. It provides a mechanism for dealing with the fact that some students arrive in a grade-level classroom lacking the foundational skills to engage with the curriculum at that grade, while others may have already mastered it.
Using assessment evidence to identify where a student is in a learning progression allows for more targeted teaching at the individual level, but also the identification of groups of students who might benefit from working together on a particular learning activity.
From evidence to action
Test developers can construct the most elegant and targeted materials to assess a particular skill but, ultimately, assessment data is not useful if it is not used by teachers, schools, systems and policy makers.
For teachers, descriptions of growth can be deeply helpful here: a review of the skills that sit immediately above where a student is currently located can suggest a next focus for teaching and learning.
ACER’s purpose is to improve learning for all. At the heart of that aim is a simple but powerful idea: to help students move forward in their learning, we must first understand where they are. High-quality assessment materials make this possible.
References:
Ausubel, D. P. (1968). Educational Psychology: A Cognitive View. Holt, Rinehart & Winston
