Evaluating and Grading Student Achievement

Evaluating and Grading Student Achievement

1. Grade Meanings & Purposes

   • Continuous Assessment & Grading
     • Summative assessments - formal - useful for grading decisions
     • Formative Assessments - informal - making daily teaching decisions, not for grading
     • Validity - base grades on formal assessments based on standards and objectives
       • Grades are used for a variety of purposes
         • Selection decisions
         • Placement
         • Remediation
         • Administrative guidance/counseling
         • Improving teaching
         • Information for parents & students
       • Grades come from many data sources
       • Unfair to base grades on single high stakes exams
       • Excessive grading can also harm validity

2. Reporting Methods

   • Traditional Report Card
     • Letter/Number Grading
       • May combine subject matter (letter) with behavior (number)
     • Often linked to Parent/Teacher conferences
       • Infrequent in Jr/Sr. high
     • May be included in a student's permanent record
     • Convenient, but may condense too much information into a single symbol
   • Narrative Reports
     • Teacher writes detailed account of student achievement
     • Allows more detail than traditional reports
     • Particularly useful for elementary if aligned with benchmarks
     • Can be poorly written without adequate training
     • May prove confusing to some parents
     • Teachers need to be constructive without being threatening or blaming
   • Standards Referenced Grading (what my school uses)
     • Report achievements by standard instead of by subject
     • Typically uses a proficiency scale rather than traditional letter/number grades
     • Standards usually derived (simplified) from state standards
     • Usually differentiates academic achievements from behavior or attendance marks
     • Considerations
       • Focus stays on detailed learning outcomes
       • Avoids influence of non-academic factors for grading
       • Help connect grades to an overarching assessment system
       • Public can confuse traditional and SRG marks and balk at the "new grades"
         • There can be confusion and frustration about "Advanced" vs. "Proficient" marks
       • Some high schools use a compromise between SRG and traditional marks
   • Conferences
     • Traditional Parent/Teacher
     • Student-led
       • Student ownership is increased
       • Builds student communication skills and confidence
       • Students should have prior self-assessment experience in class
       • Meet with parents first (or call)
       • Students create portfolios & organize before conference
       • Can be very time consuming for teachers
       • Requires a great deal of planning
         • Some parents still may not show

3. Grading Models

   • Criterion Referencing - Absolute Standards
     • Grades are based on how well a student performs relative to standards/objectives
       • (everyone can get an A if they all meet the standard)
     • Central question: "Is it of value to know exactly what the student has learned independently of the student's own capability and the learning of others?"
   • Norm-Referencing - Relative Standards
     • Grades are based on comparative performance of students in a class.
     • Intended to use competition to bringing out student's best performance
     • Does not clearly reflect knowledge attainment, or lack thereof
       • Top/Bottom marks may be far from true mastery or failure
     • Reference group must be defined (one class, several years of classes?)
     • Mustn't be mixed with criterion referencing
   • Self-Referencing - Growth Standards
     • Compare performance with past performance or capability expectations
     • Students at various achievement levels may receive same grade based on growth
     • Intended to increase motivation and decrease competitiveness
     • Teacher judgment of capability takes a central role (for better or worse)
     • Results in unreliable assessments of capabilities and standing among peers
     • May be useful for some formative assessments
       • May help motivate some students during formative tasks
     • Not appropriate for Summative assessments (use criterion assessments instead)
     • Penalizes students who come into a class with high level of prior knowledge

4. Combining Grades to Summarize Achievement

   • Assigning Norm-Referenced Letter Grades
     • REM: both the individual & final grade need to be norm-referenced
     • Single test or assessment
       • Grade on Curve
       • Grade on a normal curve
     • Grading a composite of several scores
       • REM: Multiplying by weight can backfire due to standard deviations
         • SS Score Method
           • Keeps component weights in tact
   • Assigning Criterion-Referenced Letter Grades
     • Total Points
       • Assessment plan needs to clearly outline number of points per assessment & component in advance
       • Letter grade is assessed for all, not individual components
       • Tempts teachers to use extra credit
       • Can create difficult choices between easy assessment math vs. appropriate number of test items
       • Combining
         • Automatically graded by design
           • Just make sure your points reflect the weights you intended
     • Fixed Percentage
       • Percentage/Grade relationship is arbitrary
       • Method can unintentionally work against criterion-referenced ideals
       • Uses simple math with weights & percentages for combining
     • Quality-Level
       • Assigns a quality level to an assessment with rubrics or grading scales
       • Combinning
         • Using the Median Score
         • Standards-based: Privileging recent evidence
           • Emphasis on improvement over time
         • Minimum Attainment method
           • Bases grade on student meeting minimum standards of most important assessments
           • Allows lower performance on less important components

5. Grading Practices

   • What are the boundaries between grading levels & will exceptions be made?
     • Must have the same meaning across all assessments
     • Most teachers are willing to adjust up, but not down and when in doubt, it's better to bump up a borderline student
   • What should be graded?
     • Detailed in assessment plan
       • Assessment variables
         • Most are formative - NOT for grading
         • Reporting variables
           • Grading Variables
             • Learning focused subset of reporting variables
             • Used for final grade
             • Should describe a student's accomplishments
             • performance tasks, portfolios, projects, tests, some quizzes
             • should be the most valid and reasonable data used for determining grades
             • should not be mixed with other variables (i.e. lower grade for late work)
   • Component Grade Scales & Weights
     • Make scores compatible with the final grade configuration
     • Simple summations of scares can be invalid
     • Be precise about student differences - Use scores with accurate resolution
     • Weigh components appropriately
       • More important objectives should be weighed more heavily
       • Heaviest weight should reflect what most time was spent on
       • Higher order thinking should be weighed more heavily
       • Components w/unique objectives should weigh more than those with overlapping objectives
       • Be fair to various student groups in weighing components
       • Less reliability should = less weight
   • What does failure mean & will zeros be used?
     • Zeros can imbalance and invalidate grades
       • Alternatives:
         • use "high" failing grade (59% or 50%) instead of "lowest" failing grade of 0
         • ignore a missing grade
         • use the median grade
         • Use the same method for the whole class