What if Every Test Made Students Smarter?

How assessment shapes student behaviour before, during, and after the test

Ask most students why they’re taking a test, and they’ll say, “Because the teacher told me to.” But assessment should be more than a bureaucratic hoop to jump through. Every test sends signals—about what matters, what’s worth remembering, and what success looks like. Done right, every assessment is an opportunity to learn. The question is: are we sending the right signals and making the most of each opportunity?

This Substack - 100% Assessment - exists to help teachers and school leaders develop a coherent, whole-school approach to assessment—one that ensures every exam, quiz, or performance actively promotes learning.

Assessment in schools is complex, shaped by competing agendas and the decisions of multiple stakeholders. Given this, we need a unifying principle that everyone can rally around:

All assessments, whether large or small, should be designed to improve student learning.

If we take this principle seriously, we must consider the many ways assessment influences learning—not just during the test itself, but before it’s taken and after the results are used. Students, teachers, parents, and school leaders all respond to assessments in ways that shape behaviour and expectations.

Let’s explore these mechanisms in more detail.

Expectation of Inference: Assessments Promote Learning Before They Are Taken

The mere presence of an assessment changes student behaviour. Before a test even takes place, students anticipate that an inference will be made about their learning—and this expectation shapes their actions. How strongly it influences them depends on how much the assessment matters to them.

How Does the Expectation of Inference Work?

Anticipating a future assessment directs students' attention towards specific knowledge as they form beliefs about what might be included. Teachers influence this by choosing how much to reveal about the content. Attention is the gatekeeper of learning—we must attend to something in order to learn it.¹ Since attention is limited, students prioritise what they believe will be assessed. This expectation helps establish a hierarchy of knowledge, guiding students on where to invest their time and effort.

But directing attention alone is not enough. Students must also:

1. Care about doing well.

2. Believe they can succeed.

3. Know how to succeed.

1. Caring About Success

Students’ motivation to perform well varies widely. It can depend on whether the assessment counts towards a qualification, their relationship with the teacher, how others will perceive their performance, their enjoyment of the subject, or the consequences of failure. While many of these factors are beyond a teacher’s control, some—such as how the assessment is framed, who sees the results, and whether success is linked to future opportunities—can be shaped by teachers and school leaders.

2. Expecting Success

Even if students care about doing well, they may disengage if they believe success is out of reach. Prior failures, inaccessible question formats, or a high pass mark can all contribute to a sense of inevitable failure. Teachers can help counteract this by ensuring assessments feel challenging but achievable, using scaffolding, accessible formats, and opportunities for low-stakes success.

3. Knowing How to Succeed

Perhaps the most controllable factor is ensuring students know how to perform well. As business theorist Frederick Herzberg observed, you can’t motivate someone to play the piano if they don’t know how to play the piano.² Students must be taught effective study methods and understand that their effort directly influences success.

The Impact on Learning Behaviour

When all these factors align—attention directed, motivation high, belief in success intact, and a clear path to achievement—the expectation of inference triggers productive learning behaviours. Students are more likely to study, participate, concentrate, and persist. Over time, this builds self-efficacy, a positive academic self-concept, and a reinforcing expectation of future success.

However, the reverse is also true. If students expect failure, do not see the value of assessment, or lack the tools to succeed, the expectation of inference can demotivate rather than inspire.

Testing Effect: Students Learn During Assessments

Assessments are often seen as taking time away from instruction, but substantial evidence shows that testing itself is a powerful learning tool. These benefits include strengthening the storage of knowledge in long-term memory through retrieval, which reduces forgetting. The act of retrieving information triggers myelination, strengthening synaptic connections between neurons and enhancing future retrieval success. Without regular retrieval opportunities, neural pathways may weaken, making information harder to access. As psychologist Robert Bjork notes, “using our memory shapes our memories”—each time we recall knowledge, we reinforce neural connections, improving future retrieval.³

Why Does Testing Enhance Learning?

Learning during assessments involves more than just recall; it also strengthens encoding by forming new neural connections and reinforcing existing ones. This improves the accessibility of both tested and untested knowledge, making learning more flexible and transferable.

Assessments also prime students for future learning by bringing prior knowledge to the forefront. When students retrieve information, they integrate new content more effectively. Tests also promote metacognition, helping students identify gaps in understanding and refine their study strategies.

Not All Assessments Are Equal

The effectiveness of testing as a learning tool depends on the cognitive demands it places on students:

• Retrieving well-established knowledge strengthens neural pathways and boosts long-term retention. The benefits are only limited once retrieval becomes highly automatic (or if retrieval isn’t desired).

• Applying knowledge in new contexts enhances flexibility but may not reinforce memory as strongly.

• Novel problem-solving tasks challenge students in new ways but do less to support long-term encoding unless retrieval is involved.

To maximise learning, assessments should balance retrieval practice and meaningful application, ensuring that testing both consolidates knowledge and encourages deeper thinking.

Consequences of Inference: How Attainment Judgements Promote Learning

The act of inference, a mainstay of assessment literature, only comes into play after the assessment is complete. On its own, inference —the process of drawing conclusions from assessment data—does not affect learning. What matters is how the information is used: Who sees it? In what form? What decisions follow?

How Post-Assessment Decisions Affect Learning

Assessment results influence multiple stakeholders:

• Teachers may reteach content, target support, or identify struggling students.

• Parents may praise or chastise their child, supervise study more closely, or hire a tutor.

• School leaders may interpret results as a reflection on teaching quality or predict future performance, directing resources accordingly.

• Students may adjust their self-concept, motivation, or subject choices based on feedback.

Each of these responses can enhance or hinder learning, depending on how the information is presented and interpreted.

Why Validity and Reliability Matter

The usefulness of an assessment depends on how well it measures what it claims to measure (validity) and how consistently it does so (reliability). Teachers play a critical role in ensuring both by designing fair assessments, marking accurately, and presenting results in a way that leads to effective action.

However, validity and reliability are not fixed properties of an assessment—they depend on how the results are used. The same data can mean different things to different people:

• A teacher may use a Year 7 history test to identify gaps in understanding and adjust teaching.

• A parent may see only the final grade, reacting without context.

• A school leader may analyse data at the department level but not need detailed student-level information.

If assessment outcomes signal misleading information, effort and resources may be misallocated. Ensuring that inferences lead to meaningful, appropriate actions is essential for assessment to truly support learning.

Conclusion

Assessment is more than a tool for measuring learning—it actively shapes it. Before a test, the expectation of inference directs students' attention and motivation. During a test, retrieval strengthens memory and deepens understanding. After a test, the way results are used can reinforce or undermine future learning. When we design assessments with these mechanisms in mind, we move beyond mere measurement and create assessments that truly help students learn.

1. We prefer not to attribute the phrase ‘gatekeeper of learning’ to any one individual since similar phrases are common in education from the early 20^th century onwards, though its current popularity is perhaps due to two sources:
Lemov, D. (2015). Teach like a champion 2.0: 62 techniques that put students on the path to college. Jossey-Bass.
McCrea, P. (2019). Learning: What is it, and how might we catalyse it? Ambition Institute. https://s3.eu-west-2.amazonaws.com/ambition-institute/documents/Learning_what_is_it_and_how_might_we_catalyse_it_v1.4.pdf.

2. Herzberg, F., Mausner, B., & Snyderman, B. B. (1959). The Motivation to Work. New York: John Wiley & Sons.

3. Anderson, M. C., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20(5), 1063–1087.