EDST5443 Chap.6 Assessment Data for Teaching Decisions
Assessment Data for Teaching Decisions
Assessment data become useful through a disciplined chain from observation to interpretation, action and follow-up. This chapter shows how to inspect patterns without turning a score into a verdict. It distinguishes what evidence directly shows from a plausible explanation, matches the scale of a response to the resolution of the data and treats each teaching decision as a hypothesis to be checked.
It also considers validity, privacy, opportunity to learn and the organisational conditions that shape responsible data use.
What this chapter covers
- 01
Starting with the teaching or learning decision that evidence may inform
- 02
Separating an observed pattern from an interpretation and a causal claim
- 03
Using several relevant sources to examine reasoning, performance and participation
- 04
Matching broad indicators to broad inquiry and close evidence to instructional action
- 05
Considering validity, access, opportunity to learn and assessment conditions
- 06
Selecting a proportionate response tied to the pattern rather than generic remediation
- 07
Protecting privacy and avoiding fixed labels when discussing learner data
- 08
Collecting follow-up evidence to test the response and refine the original explanation
Worked example · free
Interpret a response pattern without overclaiming
- +1Describe the two error patterns without assigning fixed ability labels to learners.
- +1State limits: the task is narrow, and correct labels may not reveal the quality of reasoning.
- +1Use a whole-class borderline example, with an optional sorting scaffold and required justification.
- +1Check the interpretation with a fresh rubric excerpt and inspect explanations as well as labels.
Key terms
- Data pattern
- A recurring feature in results, work, talk or participation that may warrant closer investigation.
- Inference
- An interpretation drawn from evidence rather than a fact directly contained in the evidence.
- Triangulation
- Examination of an interpretation through relevant evidence from more than one source.
- Opportunity to learn
- The extent to which learners had access to the content, support and conditions assumed by an assessment.
- Proportionate response
- An action whose specificity and intensity fit the strength and resolution of the evidence.
- Follow-up evidence
- New information collected after a response to test whether learning or the interpretation changed.
Assessment Data for Teaching Decisions FAQ
Does a lower score identify the cause?
No. A score describes performance under particular conditions. Possible explanations may include conceptual difficulty, task interpretation, access, motivation or assessment design. Closer evidence is needed before choosing a targeted response.
Is more data always better?
No. Collect evidence relevant to a defined decision and consider its quality. Excess data can obscure the question, increase workload and create privacy risks without improving the inference.
Can external results be used formatively?
Yes, when they generate questions, guide a proportionate inquiry and lead to actions whose effects are checked. Their broad grain often makes them better for identifying an area to investigate than prescribing a classroom technique.
Why examine distributions rather than only averages?
An average can conceal distinct patterns, including a misconception held by a smaller group. Distribution, item-level work and learner reasoning can reveal instructional information that a single summary number removes.
Assessment move
Practise a four-column routine with headings for observation, possible explanation, response and follow-up evidence. Begin with cases from the course materials and invent no learner identities or official results. In the observation column, record only what the work or response pattern shows. In the explanation column, list at least two plausible accounts and identify what additional evidence would distinguish them.
Select an action that addresses the visible pattern while remaining useful to the whole group, then design a fresh check that can reveal whether the action helped. For critical reflection, resist causal language unless the evidence warrants it.
Phrases such as ‘the pattern suggests’ and ‘a plausible explanation is’ can mark an inference honestly, but they should be followed by concrete reasoning rather than used as protective decoration. When evaluating a dashboard or external assessment, ask what was sampled, who participated, what access learners had and whether the measure is fine-grained enough for the proposed decision.
Include organisational conditions such as collaborative time, trust, roles and technology when they materially affect action. End every data discussion with a decision boundary: state what you will do now, what you will not conclude and what evidence would make you revise the plan. Add an uncertainty column when practising with aggregate results.
Record a rival explanation, a source of measurement error and an access condition that could change the pattern. Decide which uncertainty must be investigated before acting and which can be monitored through a low-risk response. Compare a class average with item-level reasoning to see how a stable summary can conceal a useful misconception.
Finally, write a short privacy note identifying who genuinely needs the information and how learner identities will be protected. These habits keep analysis proportional while still allowing evidence to inform a timely teaching decision.
Working through Assessment Data for Teaching Decisions in EDST5443? Sia is AskSia’s AI Education tutor — ask any EDST5443 Assessment Data for Teaching Decisions question and get a clear, step-by-step explanation grounded in how EDST5443 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
