Annie is a sixth grade teacher who successfully changed the targets and tasks of the Containers assessment to include a graphing exercise. Annie wanted to see if students could create a graph with a descriptive title, appropriately labeled axes (temperature for the y-axis and time for the x-axis), appropriate scales for each axis, the correct points plotted, and straight lines drawn between points. She also expected students' discussion of insulation to reflect results from their graphs. Thus, in addition to answering the questions from the original task, she asked her students to graph the change in temperature of the water over time for each container. Question 2 from the original Containers task stated, "Look at the table. Which container keeps a hot drink warm for the longest time?" Annie modified it so instead it stated, "Use the data in the table to create a graph of your data. (Use three different colors to graph the results for each container.) Based on your results from your data table and your graph, which container keeps a hot drink warm for the longest time?"

To reflect these changes in the targets and task, she modified Item 2 of the rubric. She changed the original Item 2 from:

Identify container that keeps hot drink warm longest.

to the revised version:

Identify container that keeps hot drink warm longest.

Although Annie successfully changed the targets and task of the Containers assessment, the minor word change she made to the rubric does not help her adequately measure the targets she wanted to assess. She planned to evaluate specific aspects of students' graphing skills and her modification makes it difficult for her to measure any of them.

Her targets suggest that there are many aspects of graphing that she planned to assess such as labeling axes and plotting correct points. Most likely, students will make mistakes in some parts of the graphing problem. Annie's targets suggest that she wants to know what these mistakes are so that she really knows about her students' graphing skills. Unfortunately, her change to the rubric is not detailed enough to evaluate all of these individual aspects of graphing.

Another problem with Annie's modification to the rubric is that it does not allow her to measure separately the consistency between: (1) the table and the container students identified; and (2) the graph and the container identified. For example, it is possible that the container a student selects as the best insulator matches results from the table but not from the graph because she has trouble with graphing. Annie's rubric is only appropriate for students who have both a table and a graph that is consistent with their choice of container or for students who have both a table and a graph that is inconsistent with their choice of container. Annie's revised rubric cannot measure student work in which the table matches the container identified, but the graph does not and vice versa.

Finally, Annie's modification to the rubric does not reflect the importance she placed on graphing in her targets and in the task. Clearly, Annie believes graphing is a very valuable skill. Additionally, graphing is time consuming and quite challenging for students who have little experience or practice with graphing. Despite the apparent value she placed on graphing skills in the targets and task, the points on the scale of the rubric do not appropriately reward students who are good at graphing. More points should be awarded for a high quality performance on the graphing activity.

An appropriate modification of the rubric based on Annie's changes to the targets and the task would include:

1. specifying the desired aspects of student graphs

2. distinguishing between consistency between table and choice of best insulating container and consistency between graph and choice of best insulating container

3. awarding the appropriate amount of points to the task based on level of importance and amount of work required by students to complete the graphing task

Generally, good rubrics measure all of the assessment targets, provide separate scores for each important behavior, and provide points for performance that reflect the relative importance of each behavior assessed. When rubrics are designed to be continuous, parallel, and coherent and use highly descriptive language, they are more likely to be valid and reliable measures of student performance.

Peter is a fourth grade science teacher who adapted the Containers task to assess a district standard in addition to the National Science Education Standards. The district suggests that students write about the scientific process when conducting experiments. To assess students' writing about the scientific process, Peter asked his students to keep a laboratory notebook. In this notebook they needed to write a few sentences about the purpose of the experiment and any hypotheses they might have, list the materials used in the experiment and copy the necessary procedures. The targets to be assessed were: clearly describes the purpose of the experiment, develops appropriate hypotheses, and accurately copies materials and procedures. To incorporate this material into the original Containers task, Peter provided students with an extra 30 minutes to write in their laboratory notebooks prior to conducting the experiment.

To assess his students' ability to write about the scientific process, Peter added the following item to the rubric:

Item 1 - Writing about the scientific process. Student is scored on clearly describing the purpose of the experiment, developing appropriate hypotheses, and accurately copying materials and procedures.

Clearly describes the purpose of the experiment.

Writes central purpose of experiment in his/her own words.

Peter's modification to the rubric is appropriate because it clearly communicates performance goals with highly descriptive language. He specifically defines the behaviors he expects from his students. Additionally, points awarded for correct responses reflect the amount of work needed to achieve each goal. Peter's modification allows the rubric to remain a valid and reliable measure of students' understanding of the relationship between properties of materials and heat transfer, and makes the rubric a valid and reliable measure of students' ability to write about the scientific process.