Shared Standards: Making Quality Reusable

Every assessment creates evidence about quality.

It may be a marked response, a portfolio, a performance, an annotated script, a standardisation exercise or a comparative-judgement session. In each case, assessors develop a clearer understanding of what stronger and weaker work looks like.

But that understanding is often temporary.

It may live in a marking system, a standardisation pack, a report or the experience of a small group of experts. When the assessment event ends, the evidence is archived and much of the shared understanding has to be rebuilt when the next cohort, task or assessor group arrives.

Shared Standards is an approach in RM Compare designed to change that.

It helps organisations retain the evidence behind their standards, govern it with confidence and reuse it wherever people need to make consistent judgements about new work.

From assessment event to shared standard

A shared standard is more than a description of performance. It is a visible, trusted reference point that helps people understand what quality looks like in practice.

A rubric can describe the characteristics of a strong response. An authentic example can make those characteristics tangible. A carefully governed collection of examples, evidence and professional judgement can become something more powerful still: a reusable standard.

RM Compare | 🧭Hub provides the home for this approach. It sits between RM Compare | 💻Studio, where comparative-judgement standards can be created, and RM Compare | 📳Live, where approved standards can be applied to new work on demand. Together, the modules help organisations move from one-off assessment activity to standards that can be shared and used over time.

The model is simple:

The Standards Repository holds authentic work and the context that makes it meaningful. Ranked evidence establishes how selected work relates in quality. A ruler turns suitable, approved ranked evidence into a reusable standard. Live then enables authorised users to place new work against that standard when needed.

Try it NOW!

Standards can begin in more than one place

Not every standard starts with an RM Compare judging session.

Many organisations already hold authoritative evidence created through established marking and assessment processes. Candidate work may have been marked against a rubric, awarded a grade, annotated by an examiner or approved as a standardisation exemplar. When these outcomes are used to place items in order, they create valuable mark-derived ranked evidence.

That evidence belongs in the Standards Repository too.

An organisation can retain the original item alongside the information that explains its status: the assessment task, mark scheme, mark or grade, annotations, specification, approval history and permitted use. This turns a set of historic scripts into a governed reference set that can support training, standardisation, moderation and meaningful professional discussion.

A mark-derived rank and a comparative-judgement rank are both useful—but they are supported by different evidence.

A mark-derived rank shows how items were ordered through an established rubric-based marking process. A comparative-judgement rank shows how a set of items relates in quality following repeated expert comparisons in Studio.

The Standards Repository brings these sources of evidence together without confusing them.

Where a collection of marked evidence is suitable for holistic comparison, an organisation may choose to calibrate a selected set through RM Compare. That can add a stable comparative rank and create a reusable ruler. Where this is not appropriate, the mark-derived reference set remains valuable in its own right.

Shared Standards does not require organisations to abandon the evidence they already trust. It gives that evidence a governed home and a route to greater reuse.

Ranks make quality visible

A rank shows how a selected set of items relates in quality.

In Studio, judges compare pairs of items and decide which demonstrates stronger overall quality against an agreed comparison statement. Across multiple comparisons, RM Compare creates a quality continuum: a rank order.

This is especially useful for complex work, such as extended writing, portfolios, projects, design work, practical performance and multimedia evidence. In these contexts, quality can be difficult to understand through isolated marks alone. Seeing work in relation to other work can make a shared standard more tangible.

Ranks help organisations answer:

“How does each item in this set relate to the others in quality?”

The rank may be created directly through comparative judgement. Or it may begin as a mark-derived order from an organisation’s existing assessment evidence. In both cases, the important principle is to retain the context and evidence that explain the order.

A rank becomes genuinely useful when it can outlive the assessment event that created it.

Rulers make standards reusable

A ruler is a shared reference scale built from approved ranked evidence.

Rather than running a new assessment exercise from the beginning every time fresh work appears, assessors can compare new work against an established standard.

The question changes from:

“Which response is better in this new group?”

to:

“Where does this new response sit on the standard we already trust?”

For an RM Compare ruler created through comparative judgement, the original rank is reviewed and approved before becoming available for reuse. It can then be governed and shared through Hub, with authorised users applying it in Live.compare.rm+1

This makes a standard more practical.

A team can use a ruler to support examiner induction. A moderator can use it as a point of reference when reviewing work. A school group can use it to discuss shared expectations. An assessment organisation can make an approved quality scale available across teams, centres and assessment moments.

The ruler does not replace professional judgement. It gives that judgement a common, evidence-rich frame of reference.

Standards that endure

A shared standard should be stable enough to trust—but flexible enough to remain useful.

That is the principle behind Steady State.

Once a ruler has been created and approved, new work can be placed against it without automatically changing the standard. This preserves a consistent reference point for assessors, teachers, examiners and moderators. RM Compare describes this as using established rulers to support repeatable assessment activity while retaining control over when a standard is reviewed or refreshed.

Over time, organisations may need to revisit the evidence behind a standard. Tasks can change. Specifications can evolve. New forms of work may need to be represented. Some examples may need to be added, restricted or retired.

The answer is not to rebuild everything from scratch, or to allow a standard to drift every time it is used.

It is to maintain the standard deliberately: review the evidence, control access, keep track of ownership and approval, and refresh the ruler through governed activity when the organisation decides it is needed.

This is what makes a standards repository different from a simple collection of files. It provides an organisational memory of how quality has been understood, established and maintained.

How RM Compare fits together

Shared Standards connects RM Compare’s three modules into one workflow:

Module	Role in Shared Standards
Studio	Creates comparative ranks through Adaptive Comparative Judgement
Hub	Holds and governs the evidence behind shared standards, including mark-derived and comparative ranked reference sets
Live	Enables authorised users to apply approved rulers to new work on demand

A typical journey might begin with a team selecting authentic work from the repository. That evidence may include existing marked and annotated examples, or work selected for comparative judgement. The team creates or curates a rank, reviews the evidence and approves a ruler for a defined purpose.

Hub then makes that standard available to the right people. In Live, new work can be placed against the approved ruler, helping users apply a shared interpretation of quality without recreating a complete assessment event.

Explore Shared Standards

Shared Standards is a new direction for RM Compare. It is built on a straightforward principle: the evidence and expertise created through assessment should not disappear when a session ends.

Whether a standard begins with rubric-based marking, senior-examiner approval or comparative judgement, RM Compare can help organisations retain the evidence behind it, govern its use and make it available to support consistent professional judgement.

The following articles explore the model in more depth:

1. From Item Repositories to Standards Repositories: Making Shared Quality Reusable
   Why a searchable collection of work is only the starting point—and how organisations can make evidence of quality into a governed, reusable asset.
2. From Marked Evidence to Reusable Rulers: A Practical Path for Awarding Organisations
   How rubric-marked exemplars can become governed reference sets, and when selected evidence can be strengthened through comparative calibration.
3. Steady State, Not Static: Maintaining Assessment Standards When Questions Change
   How a stable core of reference items can be maintained and tested as assessment tasks, cohorts and contexts change.

For practical guidance, explore the Help Centre:

• Shared Standards - an introduction to RM Compare’s shared-standards model.
• The Standards Repository - how Hub provides the governed evidence base behind a standard.
• Ranks and Rulers - how comparative ranks become reusable rulers.
• Sharing Standards with Live - how authorised users apply shared rulers to new work.
• Maintaining Standards in Steady State - how to maintain, review and refresh shared standards over time.
  
• Shared Standards helps organisations bring together the evidence they already trust - whether created through marking or comparative judgement—then govern, strengthen and reuse it wherever consistent professional judgement is needed.