Are you interested in collecting and submitting samples of student work from AAC Performance Tasks?
Learn More

Q & A: Assessment Tools

Click on the questions to reveal answers.

Q & A: Rubrics

Q1. How can students understand the language of the rubric?

It is essential that students understand the language of the rubric in order to have a clear vision of the learning target. It is also important not to resort to oversimplification of terms in an effort to meet young students at their vocabulary level.

The use of a variety of faces (happy, sad) might be appropriate for very young children but as soon as students have a wider vocabulary, these symbols should be replaced with more precise vocabulary.

The same principle applies to the overused terms such as excellent, very good and so forth. While teachers may have an intuitive sense of what is excellent, students need to know specifically what makes a piece of work excellent. Exemplars of student work at varying levels of quality will help clarify this for students.

CLICK HERE to access the Professional Learning Module: Understanding the Language of the Rubric

Q2. Is there a preferred number of levels in a rubric?

Some educators are concerned about the problems of central tendency – the gravitation of scores to the middle that can exist when using rubrics with odd number scales. This really is only a problem in contexts where quotas are placed on various levels of student scores, a practice which is only appropriate in within large populations, and not appropriate for general classroom or school use.

What is most important is to determine what is needed in order to make judgements regarding student performance. For skills where there is a wide range of observable levels of quality, a 4 or even a 5 or 6 point scale might be most appropriate. For other skills, a 3 point scale might be appropriate and for some, a checklist (2 point) might be the best choice.

What is most important is that teachers achieve clarity on how student performance recorded on the rubric is communicated to students and reported to parents/guardians whether through achievement grades, comments, and/or in the personal growth section of the report card.

Q3. Quantitative descriptors seem less subjective. Why don't we use quantitative descriptors?

Unless a numerical value is specifically mentioned in the learner outcome it is inappropriate to assign a quantitative value to the rubric descriptor on which the grade will be based.

For example, if a learner outcome indicates that students must access multiple sources of information during the research process, it is inappropriate to determine that students who consult 4 sources will receive a score of 4 on a 4 point scale (or A), 3 sources will yield a score of 3 (or B) and so forth. This system also has the unintended consequence of allowing students to opt out of their learning by deciding to take a lower score in order to do less work.

Rubric descriptors need to reflect the quality of how students demonstrate the skill. In this example, a teacher may determine that between 4 and 6 sources are needed in order to gather evidence of student ability to demonstrate this skill. Students who submit work without the required number of sources may require additional support in locating additional sources before completing their assignment; they would not, however, be penalized by lower grades. The rubric descriptor is then created based on what quality looks like at various levels.

The challenge in attempting to describe quality is in coming to a shared understanding between teacher and students about what various levels of quality look like. Exemplars of student work representing various levels of quality are useful in helping students understand the learning target.

Q4. How can rubric scores be converted to report card grades?

Three main methods are available for converting rubric scores to report card grades. While there may be limitations with any one method selected, rubrics are only one method of recording student performance. In a balanced classroom assessment program a range of assessment strategies will be used; thus rubric scores will exist in a grade book alongside test scores and other evaluation evidence.

For further information on this topic, consult the AAC publication Smerging Data: Grading . . . More Than Just Number Crunching (2001).

1. Frequency of Scores

Pre-determined logical rules are created and shared with students. Two examples of logical rules:

  • To get an A, half of the scores must be a 4 (on a 4-point scale) with no scores below a 3
  • Mostly 3s and 4s translates to an A; mostly 2s and 3s translate to a B, and so forth

2. Total Points

  • points earned divided by total possible points yields a percentage score
    e.g. Five criteria statements on a 4-point rubric would enable the highest possible to be 20
  • certain categories can be weighed more heavily than others if desired

3. Mid-point Scoring

  • If A = 80 – 100, level 4 on the rubric will be assigned the midpoint score in that category, i.e. 90%

For more information on this method, see “Using Rubrics to Promote Thinking and Learning,” by Heidi Goodrich Andrade, Educational Leadership 57, 5 (2000), pp. 16–17. Alexandria, VA: Association for Supervision and Curriculum Development.

 

Q & A: Checklist and Rating Scale

Q1. What is the effect of extreme scores?

Assigning numerical scores for checklist items can lead to grade distortion. Items in the ‘yes’ column would be scored as 100%, whereas items in the ‘not yet’ column would be scored as 0%. The impact of a zero grade is exaggerated by typical report card scales. For example, most report cards have a 10 – 15 point range from one performance level to the next with the exception of the failing grade which has a 50 point range. Attempting to translate checklist scores into a numerical value can distort the student weaknesses and keep the score artificially low.

Checklists are best used as formative tools to help students identify areas for improvement prior to submitting work to the teacher for grading. The positive phrasing of the ‘not yet’ category highlights the ongoing nature of the work in progress; while the work might not yet meet the acceptable standard, with more work and time, the potential exists for the work to be brought up to standard.

For a detailed discussion on the impact of zeros on grading, see the following:

  • A Repair Kit for Grading (O’Connor, 2007)
  • The Case Against the Zero (Reeves, Phi Delta Kappan, Dec 2004)
  • Alternatives (GuskeyPrincipal Leadership; Oct 2004)

Q2. Should students be 'grading' the work of their peers?

Many references to teachers’ professional responsibilities for assessment can be found in the Teacher Quality Standard. While peer feedback processes have been found to be useful in supporting student learning, these do not replace teacher professional responsibility for grading. Peer feedback should be provided in a formative context only.

Descriptive feedback, checklists and rating scales can be useful tools to help students provide peer feedback. Rubrics, with the numerical values removed can also be used. When used purposefully, the results of peer feedback can help students make decisions about how to improve their work prior to submitting an assignment to the teacher for summative grading.

Q3. How can scores from a rating scale be transferred to a traditional grade book?

Rating scales are most useful as formative tools to help students identify areas for improvement of work in progress. As such, a 3-point rating scale can be used, even though a report card may use 4 categories. While this may lead to most ratings falling in the middle range of the rating scale, teacher/student comments provide the detail needed for students to make adjustments to their work (see checklist samples and checklist templates).

However, it may be desirable to use the information obtained from a rating scale to transfer to a grade book for reporting purposes. Simply adjust the categories in the rating scale to align with the grading categories on the report card. It is essential, however, to ensure that the criteria contained on the rating scale are linked to the learner outcomes.

Rating scales are often used for behavioural characteristics that are not appropriate for grading but may be more applicable in the section of the report card dealing with work habits and personal growth. Again, the categories in the rating scale can be adjusted to align with the report card categories.

 

Q & A: Descriptive Feedback

What does the research say about the value of descriptive feedback?

A research study conducted by Ruth Butler* (1988) examined the results of three different means of providing feedback to students about their work. The table below summarizes the feedback method and the results:

 

*as summarized by Black & Wiliam* Butler, R. "Enhancing and Undermining Intrinsic Motivation." 
British Journal of Educational Psychology
 58 (1988): 1–14.

[print-me target="#block-84"]