Introduction

Assessment is the area of language teaching where theory and classroom practice often collide. We all want to assess fairly, meaningfully, and efficiently—but tight timetables, systemic pressures, and time constraints frequently pull us in other directions. For many of us, assessment becomes a stressful afterthought—rushed, inconsistent, and often disconnected from the learning we so carefully plan.

This article offers a set of ten key principles to guide classroom-based assessment in a way that both supports learning and aligns with research. The aim is not to provide a rigid checklist, but a set of flexible reminders: principles we can strive for when designing, adapting, or reflecting on assessment practices.

Importantly, these principles are aspirational. Many teachers, including myself, often lack the time and resources to implement them all fully. But knowing what “better” looks like can help us make informed trade-offs, tweak existing practices, and build assessment routines that actually support learning.

If I’ve learned anything about language testing, it’s thanks to the late Professor Cyril Weir, whose clarity of thought and deep understanding of assessment shaped my thinking when I studied under him over 30 years ago.

1. Validity – Test What You Actually Taught

What it means

Validity means your assessment measures what it claims to. If you say you’re testing listening skills, but the real challenge lies in understanding unfamiliar vocabulary, cultural references, or decoding poor-quality audio, the task is not valid.

In language classrooms, invalid assessments happen all the time:

• A listening task in the textbook asks learners to infer whether a character is happy or sad—but students struggle because the audio includes many unfamiliar structures (e.g., passé composé forms not yet taught
• A writing task asks students to “describe a past holiday” before they’ve learned the passé composé with être or irregular verbs like faire. Most learners will either write in the present or produce error-ridden texts.
• You mark “grammar accuracy” but test includes structures never modelled in class.

According to Messick (1989), a valid assessment must measure the intended construct without interference from unrelated skills or knowledge. In L2 contexts, Fulcher and Davidson (2007) also warn against “construct-irrelevant variance”—where results are skewed by factors that shouldn’t matter (e.g., background knowledge, decoding ability, handwriting).

Possible solution

Make sure assessment tasks reflect what learners have actually practised. Recycle taught language, limit new vocabulary, and avoid assessing grammar that hasn’t been modelled repeatedly. Use familiar task formats to reduce cognitive load. When teaching using sentence builders or substitution tables (as in the EPI approach), assess those same patterns—not “creative writing” from scratch.

2. Reliability – Be Consistent and Fair

What it means

Reliability refers to how consistent assessment results are across time, teachers, or contexts. If the same piece of work would earn different marks from different teachers—or from you on different days—your system isn’t reliable.

This is a well-documented problem. Research by Alderson et al. (2000) and Barkaoui (2007) shows that in language assessments, scoring variation is high, especially in writing and speaking.

In practice, we see this in:

• Teachers applying mark schemes inconsistently in GCSE-style writing tasks.
• Oral assessments where confidence or accent sways scores more than structure use.
• Comments like “Nice effort” for one learner and “Needs more complexity” for another—despite similar output.

Inconsistency undermines trust in assessment. Learners can’t improve if they don’t know what counts as “good” and why.

Possible solution

Use transparent success criteria (e.g., “Includes at least three time phrases,” “Connects ideas using et, mais, parce que”) instead of vague rubrics. Moderate samples with colleagues, or cross-check with exemplar responses (e.g., AQA/Edexcel sample responses for the GCSE). Use whole-class marking codes to streamline and reduce subjectivity.

3. Authenticity – Make Tasks Feel Real

What it means

Authentic tasks mirror real-life language use. They feel purposeful, relevant, and engaging. When assessments are too artificial or contrived, learners don’t see the point—and often perform poorly because the situation feels unfamiliar.

Examples of poor authenticity:

• Writing a postcard from a theme park—when learners have never seen or written a postcard.
• Listening to a contrived conversation between “two cousins visiting the Eiffel Tower” in exaggerated accents and stilted dialogue.
• Reading an article about “Why children should avoid too much screen time” in formal register—impossible to relate to for Year 8 students.

Gilmore (2007) highlights that authentic tasks increase motivation and better reflect communicative competence. When learners can imagine themselves using the language, their engagement and performance improve.

Possible solution

Use purpose-driven tasks that reflect real communication: voice messages, roleplays for booking or complaining, WhatsApp-style exchanges, social media posts. Avoid tasks that only serve the test. Adapt textbook prompts if needed (e.g., turn “write about your school” into “write a review for an exchange partner”).

4. Washback – Make Assessment Drive Good Learning

What it means

Washback refers to the impact assessment has on teaching and learning. Hughes (2003) and Bailey (1996) show that learners focus on what they think will be tested. So if your assessments reward memorisation and neatness over fluency and risk-taking, that’s what students will prioritise.

Negative washback examples:

• Marking only grammatical accuracy in writing—but encouraging creative use of chunks in lessons.
• Testing only reading and writing—so students disengage from speaking tasks.
• Giving “fill-the-gap” tests every term—so learners memorise phrases rather than learn how to manipulate language.

Possible solution

Design assessments that reflect and reward the habits you value: recall, improvisation, clarity, range. If you’re using retrieval practice in lessons, assess it. If fluency-building is a goal, include spontaneous speaking. Use assessment as a continuation of teaching, not a switch in focus.

5. Transparency – Be Clear About What Counts

What it means

Transparency means that learners understand what they’re being assessed on, and what success looks like. Without this clarity, they can’t prepare effectively or act on feedback.

Research by Sadler (1989) and Black & Wiliam (1998) shows that learner understanding of criteria is essential for progress.

Lack of transparency often looks like:

• Ambiguous phrases in rubrics: “some complex language,” “generally accurate,” “shows understanding.”
• No explanation of how tasks are graded—learners focus only on the number.
• Tasks set without clear models, so students are shooting in the dark.

Possible solution

Before assessments, show worked examples at different levels and ask students to annotate what works and why. Use visual rubrics or traffic-light criteria (“must include a time phrase, an opinion, a justified reason”). After assessments, give feedback aligned to the criteria—not just a number.

6. Practicality – Keep It Manageable

What it means

Practicality is about whether your assessment system is sustainable. If you’re drowning in marking, or learners are overwhelmed, something’s got to give.

Poor practicality examples:

• Termly assessments that take weeks to mark but give little insight.
• Tasks that require long periods to explain or complete—leaving little teaching time.
• A speaking test where only one pair speaks while the rest of the class waits.

Research by Brindley (2001) and Green (2014) confirms that assessments must fit operationally into teaching. Otherwise, they’ll be done infrequently—or poorly.

Possible solution

Assess little and often. Use mini-tasks (e.g., one-minute oral summaries, sentence corrections, 40-word writing bursts). Use retrieval-based assessment (e.g., do-now tasks, hinge questions) for a snapshot of learning. Share the load with peer- or self-assessment.

Also, avoid high-stakes “mega-tests.” Use a portfolio of smaller, focused assessments over time. That’s how real learning is best captured.

7. Inclusivity – Remove Unnecessary Barriers

What it means

Inclusivity means all learners—regardless of SEN, EAL background, or cognitive processing style—can access and succeed in assessments if they know the material.

In practice, exclusion often comes from:

• Listening once at full speed with no visual support.
• Writing prompts that require imaginative thinking but don’t scaffold basic sentence formation.
• Tasks that rely on background knowledge learners may not have.

Research by Kormos & Smith (2012) and Mitchell (2014) stresses that inclusive assessment isn’t about lowering expectations—it’s about designing assessments that test the language, not processing speed or inferencing.

Possible solution

• Slow down audio or allow two listens.
• Use images or brief context to support listening/reading.
• Provide writing scaffolds (sentence starters, vocabulary boxes).
• Offer tiered task options (e.g., basic/extended) to allow learners to challenge themselves.

8. Formative Usefulness – Make Assessment Feed Learning

What it means

Formative assessment is assessment for learning, not just of learning. If the feedback you give never gets used, it’s wasted effort. One common piftall of feedback on assessment is that the students are not actively engaged in it.

Sadler (1989) and Wiliam (2011) show that feedback only works when learners are trained to understand and respond to it.

Common classroom fails:

• Red-pen marking with no time to respond.
• “You’re working at a 6” comments—without next steps.
• Feedback that’s generic (“use more detail”) or unclear.

Possible solution

Build in DIRT time (Dedicated Improvement and Reflection Time) in the feedback-handling phase of the assessment process with a very narrow focus (focus on 2 or 3 key issues only). Ask students to act on feedback immediately:

• “Fix these 3 sentences.”
• “Add a second reason to this opinion.”
• “Re-record your oral answer including a time phrase.”

Also, teach students how to interpret and use feedback. Otherwise, we’re just commenting into the void.

9. Constructive Alignment – Match Teaching and Testing

What it means

Constructive alignment means your assessments reflect how and what you’ve taught. If students prepare for tasks one way, and the test demands another, that’s not fair.

Examples of misalignment:

• Teaching sentence builders, then removing them completely in writing assessments.
• Practising oral fluency, but testing only reading comprehension.
• Focusing on vocabulary sets, then testing unseen grammar.

Biggs (2003) emphasises that aligned assessment helps learners transfer classroom knowledge into successful performance.

Possible solution

Assess the language routines learners have been taught: sentence frames, structures, collocations. If you’ve used retrieval activities and repetition in class, design your tests accordingly.

Include familiar scaffolds in assessments and gradually remove them as learners grow in confidence—not all at once.

10. Balanced Assessment – Don’t Let One Skill Dominate

What it means

Assessment should reflect the full skillset of language learning—not just writing or grammar.

In many English MFL classrooms, speaking and listening are neglected in favour of writing. This creates a distorted view of progress.

Examples:

• Only written work is formally assessed each term.
• Speaking is practised but never assessed.
• Listening is sidelined because “they find it too hard.”

As noted by Macaro (2003) and Turnbull & Arnett (2002), unbalanced assessment demotivates students and provides an incomplete picture of what they can do.

Possible solution

Track which skills you assess. Aim for at least one task per skill each half-term, even informally. Record speaking via Flip, use live listening tasks, and mark comprehension through gist or detail tasks.

Summary Table: Ten Key Principles

Principle	What it means (in plain English)
Validity	Test what you taught—don’t sneak in surprises
Reliability	Mark consistently and fairly, not on gut instinct
Authenticity	Use tasks that feel real, not made-up or schooly
Washback	Make sure tests encourage the habits you want
Transparency	Tell students what counts and what to expect
Practicality	Don’t overcomplicate—make it doable for all
Inclusivity	Give everyone a fair shot at showing what they know
Formative Usefulness	Make sure feedback leads to change, not just a tick
Constructive Alignment	Match what you test to how you’ve taught it
Balanced Assessment	Assess all four skills—don’t let writing rule the roost

Conclusion

Assessment, when done right, drives learning, improves motivation, and gives you an accurate picture of what your students can do. Done wrong, it frustrates, confuses, and demoralises.

These ten principles don’t require perfect conditions. Just better thinking, smarter choices, and small tweaks over time. Even adjusting one of them in your next assessment can make a meaningful difference.

We owe it to our learners to make assessment not just something that happens to them—but something that happens for them.

References

• Alderson, J. C., Clapham, C., & Wall, D. (2000). Language Test Construction and Evaluation. Cambridge University Press.
• Bailey, K. M. (1996). Working for washback: A review of the washback concept. Language Testing, 13(3), 257–279.
• Barkaoui, K. (2007). Rating scale impact on ESL essay scores. Language Testing, 24(1), 51–72.
• Biggs, J. (2003). Teaching for Quality Learning at University. Open University Press.
• Black, P., & Wiliam, D. (1998). Inside the black box. Phi Delta Kappan, 80(2), 139–148.
• Brindley, G. (2001). Language assessment and professional development. Cambridge.
• Fulcher, G., & Davidson, F. (2007). Language Testing and Assessment. Routledge.
• Gilmore, A. (2007). Authentic materials and authenticity in foreign language learning. Language Teaching, 40(2), 97–118.
• Green, A. (2014). Exploring language assessment and testing. Routledge.
• Hughes, A. (2003). Testing for Language Teachers (2nd ed.). Cambridge.
• Kormos, J., & Smith, A. M. (2012). Teaching languages to students with specific learning differences. Multilingual Matters.
• Macaro, E. (2003). Teaching and Learning a Second Language. Continuum.
• Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational Measurement.
• Sadler, D. R. (1989). Formative assessment and the design of instructional systems. Instructional Science, 18, 119–144.
• Turnbull, M., & Arnett, K. (2002). Teachers’ uses of the target and first languages in second and foreign language classrooms. Annual Review of Applied Linguistics, 22, 204–218.
• Wiliam, D. (2011). Embedded Formative Assessment. Solution Tree Press.