OfS Condition B2: Why a Satisfaction Score Isn't Support Evidence

Bottom line: England's Office for Students makes the resources, support and student engagement that students "need to succeed" a hard regulatory requirement under Condition B2 (in force 1 May 2022). B2 is not satisfied by a high satisfaction mean — it asks whether provision was sufficient for a high-quality academic experience and for students to succeed in and beyond higher education. Course evaluation is the only routine instrument that asks students that question directly — but only if you read its support items as B2 evidence rather than as a customer-experience score.

What B2 actually requires

The OfS's revised quality conditions came into force on 1 May 2022 and split "quality" across several B-conditions. B1 covers the academic experience; B3 covers student outcomes (continuation, completion, progression). Condition B2 — Resources, support and student engagement — sits between them. Its substantive limb (B2.2) requires a provider to take all reasonable steps to ensure that each cohort on each higher education course receives:

resources sufficient to ensure a high-quality academic experience and to help students succeed in and beyond higher education;
support to the same standard; and
effective engagement with each cohort, again to the same standard.

Two features of this wording matter for evaluation. First, the test is sufficiency for a purpose, judged by the regulator — not student contentment. A cohort could rate a module 4.4 out of 5 and still, on the OfS's assessment, have lacked sufficient academic or wellbeing support for the students who most needed it. Second, the OfS explicitly foregrounds students "from disadvantaged backgrounds" — the students least likely to be heard in a low-response voluntary survey.

Why a satisfaction mean is the wrong evidence

The reflex when a regulator asks "did students get enough support?" is to point at the satisfaction score. But B2 is an absolute sufficiency standard, and the average Likert response is the wrong instrument for at least three reasons.

It measures the wrong construct. "How satisfied were you with the module?" blends teaching, workload, grades and mood; it does not isolate whether the library provision, the personal tutor, the disability support or the feedback turnaround were sufficient. As we have argued elsewhere, averaging Likert scores discards exactly the distributional information — the tail of students who struggled — that a sufficiency test cares about.

It is silent on the students B2 names. Non-response in course evaluation is not missing at random: the students who disengage from the survey are disproportionately the students who disengaged from the course, and often the disadvantaged students B2 puts first. A mean computed over the respondents flatters the picture exactly where the regulator looks hardest, a problem compounded by differential non-response.

It cannot show engagement. B2.2's third limb is about effective engagement with each cohort. A one-way satisfaction form is, definitionally, not engagement — it collects a rating and closes the loop nowhere. Evidence for B2 has to show that the institution heard, interpreted and acted.

Where Koji fits

Koji turns evaluation from a satisfaction thermometer into a B2 evidence base. Its AI-moderated conversational interviews ask directly about the three limbs — were the resources adequate, was the academic and pastoral support there when you needed it, did you feel engaged — and probe beyond the number when a student says support was "fine" or "hard to get." The six structured question types let you pair a sufficiency scale ("Was disability support sufficient for you to succeed?") with the open narrative that explains a "no," and automatic thematic analysis surfaces where support gaps concentrate — which cohorts, which services, which points in the year.

Because the moderation is standardised, every student is asked consistently rather than depending on which tutor ran the conversation, and formative mid-cycle collection means a support gap can be fixed during the module, not diagnosed after the cohort has left — the difference between managing B2 and reporting a breach. Koji's action-tracking and programme- and institution-level reporting give exactly the audit trail the OfS looks for: heard, interpreted, acted, closed the loop. Because the analysis reads distributions and open text rather than a single mean, it can foreground the disadvantaged students B2 names instead of averaging them away. Teams doing wider service and support research beyond the classroom use the same interview engine on the main koji.so platform.

But isn't this just gold-plating a compliance exercise?

A fair objection: B2 is an English regulatory condition, and turning course evaluation into "B2 evidence" risks compliance theatre — collecting for the regulator, not the student. Two responses. First, the sufficiency-of-support question is worth asking whether or not a regulator requires it; B2 simply makes explicit what any serious quality process should already know, and institutions outside England face the same substantive question under ESG and ENQA or a national framework in different words. Second, the honest limit: course evaluation evidences students' experience of support, not the objective adequacy of resourcing. Whether the library budget or the counselling waitlist is genuinely sufficient is a management judgement that student voice informs but does not settle. Koji surfaces and mitigates the gap between the support you think you provide and the support students say they received; it does not, by itself, discharge the condition. Used that way — as the student-experience half of a B2 case sitting alongside B1 and B3 evidence — it is the difference between a number and an argument.

Reading a module for sufficiency, not contentment

The practical shift B2 demands is small to describe and large in consequence: stop asking "were students happy?" and start asking "did each group get what it needed?" Concretely, that means disaggregating — commuter students, disabled students, first-in-family students, part-time students — and looking for the cohorts where a support item drops, rather than reporting a single figure that averages them into invisibility. It means treating a "support was hard to access" comment as a compliance-relevant finding, not a grumble. And it means keeping the trail: what the cohort said, how the programme interpreted it, what changed, and how students were told. A satisfaction score offers none of this; a sufficiency-oriented evaluation is designed around it, and it is exactly the material an OfS assessment of B2 would expect a provider to be able to produce on request.

Frequently asked questions

When did OfS Condition B2 come into force?

The revised B-conditions, including B2 ("Resources, support and student engagement"), came into force on 1 May 2022 as part of the OfS's regulatory framework for quality and standards in England.

Does a high satisfaction score demonstrate compliance with B2?

No. B2 is a sufficiency standard judged by the regulator — whether resources, support and engagement were sufficient for a high-quality experience and for students to succeed. A satisfaction mean measures contentment, not sufficiency, and typically excludes the disengaged and disadvantaged students the condition prioritises.

How is B2 different from B1 and B3?

B1 concerns the academic experience (course content, delivery, rigour); B3 concerns student outcomes (continuation, completion, progression). B2 sits between them: the resources, support and engagement that make the experience and the outcomes possible. Evaluation evidence for each is different, which is why a single overall score cannot serve all three.

What evidence does the OfS expect for B2?

The OfS looks for evidence that provision is sufficient and that the provider engages effectively with students — which implies a heard-interpreted-acted-closed-the-loop trail, disaggregated by cohort, rather than an institution-wide average. Action tracking and programme-level reporting matter as much as the raw feedback.

Does this apply outside England?

B2 is an English condition, but the underlying question — was support sufficient, especially for the students most at risk — applies under ESG/ENQA and every national quality framework. The instrument and the argument transfer even where the regulatory citation does not.

Want your course evaluation to answer the question B2 actually asks? See how Koji for Education builds a support-sufficiency evidence base, not just a score.