Assembly Assessment Model
The Assembly Assessment Model is structured to handle all the key concepts of assessment, making the associated data as easy as possible to move, manage and analyse. The model consists of 8 main parts:
The assessment is the grouping that knits together a range of concepts. The name of the assessment also refers to the source of the result (for example, a Key Stage national assessment, or a standardised assessment provider).
2. Assessment Family
The Assessment Family groups together multiple assessments. In cases where Assembly gathers assessment data using multiple gradesets for the same test (e.g. a raw score, a standardised score and a percentile), the Assessment Family is the grouping that ties those multiple asessments together.
For example, the "KS1 Reading Official" Assessment Family comprises the "KS1 Reading Scaled Score" and "KS1 Reading Raw Score" assessments.
All results belong to one of the below four facets, with each facet reflecting a different type of grade. Facets can be used to compare 2 grades of the same assessment.
|Achieved||This is used to represent a result for a completed assessment, and is the most commonly occurring Aspect.|
|Target||A target relates to a possible outcome for an assessment that has not yet been sat by the student, and may be somewhat aspirational.|
|Prediction||A prediction relates to the likely outcome for a student (in the eyes of the student's teacher) in a future assessment at a given point in time.|
|Baseline||The baseline represents a student's proficiency level in the relevant subject at the beginning of a given assessment point. It may be compared to an "Achieved" grade in order to work out progress.|
4. Assessment Point
An assessment point represents a point in the school key stage, year, term or half-term that results can be attached to. All assessment points on the Assembly Platform are static and fall in to one of these categories.
|Assessment Point Type||Rank||Examples|
|Key Stage||1-6||"Key Stage 1", "Key Stage 2", "Key Stage 4"|
|Year||7-22|| "Year 1", "Year 7", "Year 12"
|Term||23-70||"Year 1 Spring", "Year 3 Autumn", "Year 7 Summer"|
|Half-term||71-166|| "Year 2 Autumn 1", "Year 5 Autumn 2", "Year 9 Spring 1", "Year 11 Summer 2"
Gradesets offer an acceptable range of values for any given assessment. The grades in a gradeset have the following:
- ID: an integer that represents a given grade. All grades have a unique ID in the Platform, and this should be used to send back results.
- Name: the name of the grade as given to a student.
- Value: this represents the value of the grade in a numerical format. For national exams, it links to the values that the DfE associates with a grade (e.g. the below GCSE values are the equivalent Attainment 8 points for 2017/2018). When a scale does not relate to a national examination, where possible the value is linked to a standardised scale (e.g. Percentile values are the percentile rank in numerical form).
|Grade ID||Grade Name||Grade Value|
The subject links the result to a standard list of subjects (where relevant, tied to the National Curriculum subject). All subjects have an associated three letter code.
|Subject Code||Subject Name|
|GPS||Grammar, Punctuation and Spelling|
A result combines all of the above principles and attaches them to a student. Full guidance on how to write results back to the Platform will follow soon.
Components are subdivisions of results either by strand, sub-part, or individual question. This functionality is in private beta, so we do not yet document how we handle components. For more information or to discuss how components could be used with your app, please contact us at firstname.lastname@example.org.