Measuring Complex Achievement

1
Measuring Complex Achievement

• Performance Based Assessments

2
Performance Based Assessment

• The Emphasis is on DOING, not merely Knowing
  on PROCESS as well as PRODUCT.
• Many highly valued learning outcomes emphasize
  the actual performance of task in realistic
  settings.

3
Types of Performance Based Assessment

• A variety of task may be used to assess the
  skills students have at making and recording
  observations, summarizing the observations and
  reaching conclusions.
• Essays
• Oral Test
• Realistic Problem Solving
• Experiments
• Artistic Presentations
• Performance Tasks
• Product Creation

4
Authenticity

• Authenticity is an important goal of performance
  assessment.
• Providing realistic context can make problems
  more engaging for students and help the teacher
  evaluate whether a student who can solve a
  problem in one context can solve it in another.
  Hence, it is desirable to increase the
  authenticity of task to whatever extent possible.

5
Objective vs. Performance Based Assessment

• Objective test items are generally more efficient
  and more reliable for measuring factual knowledge
  and the ability to solve well-structured
  problems.

• Problem formulation, the organization of ideas,
  the integration of multiple types of evidence,
  and originality are all important aspects of
  performance that may not be adequately assessed
  by objective testing.

6
Group Activity

• Restricted-Response Performance Task

7
Group Activity

• Extended Performance Task

8
Advantages and Limitations

• Advantages
• Can clearly communicate instructional goals that
  involve complex performances.
• Can measure complex learning outcomes that cannot
  be measured by other means.
• Provides a means of assessing process or
  procedure as well as the product that results
  from performing a task.
• Modern Living Theory.
• Views students as active participants in the
  construction of meaning.

9
Limitations

• Unreliability of ratings of performances across
  teachers or across time for the same teacher.
• Time consuming.

10
Suggestions for Constructing Performance Task

• Should be used to assess learning outcomes that
  are not adequately measured by less time
  consuming approaches.
• Limit the resources that students can use in
  performing a task.
• Avoid focusing skills that are irrelevant to what
  is being assessed.

11
Continued. . .

• Provide the necessary scaffolding for student to
  be able to understand the task and what is
  expected.
• Construct task directions so that the students
  task is clearly indicated.
• Clearly communicate performance expectations in
  terms of the scoring rubrics by which the
  performances will be judged.

12
Scoring Rubrics

• Analytic Scoring Requires identification of
  different dimensions of performance that are
  rated separately.
• Holistic Scoring Provides descriptions of
  different levels or overall performance.

13
Rating Scales

• Rating scales may take many forms.
• Numerical Rating Scale Rater checks or circles a
  number to indicate the degree to which a
  characteristic is present.
• Constant-Alternative Rating Scale Using words
  that range from never to always, you indicated
  the students performance.

14
Continued. . .

• Descriptive Graphic Rating Scale Uses
  descriptive phrases to identify points on a
  graph. The descriptions are how students should
  perform at each level. More specific than using
  just words or numbers.
• When using rating scales, have a definition
  associated with each level so that students can
  see what each level expects.

15
Uses of Rating Scales

• As a matter of convenience, these uses may be
  classified into two assessment areas (a) process
  or procedure and (b) product.

16
Common Errors in Rating

• Personal Bias Occur when there is a general
  tendency to rate all individuals at approximately
  the same position on the scale.
• Generosity Error Occurs when raters tend to use
  only the top part of the scale.
• Severity Error Occur when raters tend to use
  only the bottom part of the scale.
• Central Tendency Rater rates everyone as
  average.

17
Continued. . .

• Halo Effect Occurs when a raters general
  impression of a person influences their rating.
• Logical Error Results when two characteristics
  are rated as more alike or less alike then they
  actually are because of the raters beliefs of
  their relationship.

18
Effective Rating

• Rating scales should be in harmony with the
  school and classrooms objectives.
• Separate rating corresponding to each learning
  can enhance the value of the feedback to
  students.
• Characteristics should be directly observable.
• Characteristics on the scale should be clearly
  defined.

19
Continued. . .

• Select the type of rating scale that is
  appropriate.
• Between three and seven rating positions should
  be provided.
• Ratings from several observers should be combined.

20
Checklist

• A checklist is similar in appearance and use to
  the rating scale. The basic difference between
  them is in the type of judgment needed. On a
  rating scale, one can indicate the degree to
  which a characteristic is present or the
  frequency the behavior occurs. The checklist
  calls for a simple yes or no judgment. It is
  basically a method of recording whether a
  characteristic is present or not.

21
Student Participation in Rating

• Most devices used for recording the teacher's
  observations also can be used by students to
  judge their own progress.
• Benefits
• Better understanding of learning objectives.
• Recognize the process towards objectives.
• Can see strengths and weaknesses.
• Develop increased skill in self-assessment.

22
Conclusion

• Both the PROCESS and the PRODUCT resulting form
  the performance can be assessed. The emphasis on
  performance assessment should be on measuring
  complex achievement that cannot be measured well
  by objective test.