Interview Preparation

This requires balancing respect for the SME's expertise with the non-negotiable requirements of the specification. Step 1: Thank the SME and acknowledge their point — "That is an interesting perspective based on your experience." Step 2: Explain the specification constraint — "The specification requires X because [reason]. I cannot change that, but how can we work within it?" Step 3: Explore if the SME's concern can be addressed without contradicting the specification. Step 4: If not, politely but firmly explain why the specification must take precedence. Step 5: Document the discussion in case it is raised later.

Step 1: Analyse the data. Are there statistically significant performance differences between groups? Check against national demographics. Step 2: Review the assessment content. Could any items introduce bias? Involve diverse reviewers. Step 3: Consult the EDI lead and compliance team. Step 4: If bias is confirmed, take corrective action — this could mean removing or adjusting items, or in serious cases, adjusting results to compensate. Step 5: Document everything and change development processes to prevent recurrence. This is not just about fairness — it is about validity. If the assessment systematically disadvantages a group, it is not measuring the construct validly.

Portfolio is appropriate when: 1) The construct involves process as well as product — you need to see development over time. 2) The skill cannot be demonstrated in a single timed session — e.g., managing projects, building relationships. 3) Authenticity is critical — evidence from real workplace practice is more valid than a simulated exam. 4) The construct involves reflection and self-evaluation. However, portfolios have limitations: they are resource-intensive to assess, susceptible to authenticity concerns (who did the work?), and require strong IQA processes. I would not use a portfolio for straightforward knowledge recall where a timed exam is more efficient and reliable.

Successful standardisation has measurable outcomes: 1) Markers' calibration scores on sample scripts fall within the tolerance range (e.g., ±2 marks of the agreed standard). 2) Inter-rater reliability statistics show agreement above 0.80. 3) During live marking, moderation sampling finds fewer than 5% of marks need adjustment. 4) Markers report confidence in applying the scheme — qualitative feedback matters too. 5) Grade boundary recommendations are consistent with statistical predictions. If any of these indicators are weak, additional standardisation or recalibration is needed.

Technical review focuses on specification alignment: Is every item mapped to the correct criterion? Do command verbs match? Are marks appropriate? Is AO coverage balanced? It is a compliance check against the specification document. SME review focuses on content: Is the information accurate? Is it current? Is it appropriate for the qualification level? Does it reflect real workplace practice? Both are essential — you can have a perfectly compliant assessment that is factually wrong, or a factually correct assessment that does not meet specification requirements.

Comparability starts with the blueprint — both papers must have identical structure, mark totals, AO weightings, and method distribution. Item writers for both papers use the same item-writing guidelines. Both papers go through the same QA stages. Post-assessment, statistical equating checks whether the papers were truly equivalent in difficulty. If Paper B was harder, grade boundaries may be adjusted to compensate. The goal: a learner of a given ability should achieve the same result regardless of which paper they took.