If two people were to both read, score and provide comments on the same student’s essay without a rubric, instructions or guidelines it is highly unlikely that they will score the essay the same way. This is because their unique personal values, expertise, and interests will be the key factors that influence how they assign a score and determine what they like and dislike about the paper. This type of scoring is considered biased and unfair because student’s progress towards a learning outcome is exclusively determined by the will and preference of whomever happens to be scoring.
Rubrics provide teachers with a uniform way of providing students with information about their progress towards specific learning goals.
Rubrics help make scoring more fair across multiple teachers and also help eliminate scoring bias. Rubrics are applied analytically or holistically depending on the type of task.
When rubrics are applied analytically, scores are assigned to each component of the rubric.
Analytical rubrics are most commonly used throughout the year to evaluate formative tasks so that students have a clear breakdown of their progress on across multiple components.
In the example below, students would receive a score from 1 to 4 for each component (Development, Focus & Organization, Language and Conventions). This information is most effective when it is paired with effective, personalized feedback so that they have clear next steps to improve.
When rubrics are applied holistically, one overall score is assigned. Rubrics with multiple components can be applied holistically. In the example above, the four components would be considered all together in order to assign one score from 1 to 4.
There are also rubrics that are designed for the purpose of being applied holistically and are most commonly used on summative tasks like state writing assessments or AP essays. Holistic rubrics are used on final exams because the scores will not be guiding students on future tasks.
Holistic rubrics help teachers be more efficient in their individual scoring, however they require more norming and calibration upfront to ensure fair scores.
Help