 |
|||
 |
 |
 |
 |
|
|
|||
|
|
|
|
|
|
|
||||||||||||||||
The Reading Scoring GuideA few years ago, I answered a newspaper “help wanted” ad for applicants interested in scoring student responses for the Maine reading contract. After my interview, I was asked if I had any questions about the upcoming scoring project. “Just one,” I said. “How on earth do you know what score to give a student’s response . . . ?” The answer, when I began scoring a few days later, came in the form of a stapled set of papers called a training pack, which included the test item, the scoring guide, and the reading passage. (In reading, the item consists of the question or task—in its exact wording as seen by the students—that is used to assess a particular skill or set of skills in relation to a textual passage.) Also contained in the pack were pre-scored student responses, also called anchor papers, which serve as examples of each score point. The general scoring guide/rubric furnishes the rough framework for the detailed descriptions and justifications for the score points assigned to each student response in the anchor and training sets.
Each content area has its own type of general rubric, and within the reading content area itself, there is a large variety of descriptions from item to item. Also, the degree of expectation for support must be adjusted for each grade level. Score points and their descriptions always appear in descending order, since scorers are trained from the highest score point to the lowest. Thus, a 4 score is given to thoroughly answered questions, and lower score points are assigned to essays with one or more deficiencies, such as conceptual misunderstanding, factual errors, or general-to-minimal support from the passages. Similarly, the general scoring rubric serves as a basis for benchmarking, the procedure in which student responses are chosen that best exemplify the descriptions for each score point. In benchmarking terms, a 4 score response defines what the other responses with lower score points will look like. A 4 score response thoroughly fulfills the task being tested. The response includes specific information—details—from the text provided to the student. If an item asks for a specific number of examples or details as support, the student must provide the requisite number. In addition to relevant information from the text as support, higher grade-level items often seek insightful responses from students, primarily when poetry or literary passages are used as test questions. Obviously, when a rubric description states that a 4 score response will insightfully explain what a poem means, the term is subjective. However, a good trainer will point out what makes an explanation insightful and show corresponding anchor papers. A 3 score response provides a general explanation, the student does not provide specific information or references to the passage or story. It is obvious that the student read and understood the passage, but the response lacks the specificity of details as support. A 3 score response may indeed offer specific details but they are not relevant details offering support for the student’s explanation. In addition, a score of 3 in the rubric may occasionally allow for a minor misunderstanding of the passage that a 4 score response would not. A 2 score response may also contain misunderstandings or is partial in the degree of explanation. Such a response may lack specific information or details, or if such are present, the connection to the explanation is weak. In other words, the student may understand the meaning of the passage and the question but, by providing little support, fails to fully explain or demonstrate that understanding. On the other hand, a 2 score response is limited, it is apparent that the student does not fully understand the passage or the question. A 1 score response will likewise exhibit misunderstanding of the passage and question or will briefly attempt an explanation, frequently with no supporting information, and is therefore described as being a minimal response. A 1 score response is described as having a vague understanding of the passage, meaning that it is apparent, through erroneous support, that the student barely grasps the task. This is further demonstrated when students frequently give their personal opinions on the information contained in the article or passage in lieu of accurate details from the text. If the entire student response contains nothing but personal opinion, with no relevance to the question, it is a 0 score response. Similarly, a response may attempt to answer the question, but the explanation and all supporting information, if any, is completely incorrect. Such a response earns a 0. Scorers are told that if there is neither anything absolutely correct nor relevant that could raise the score of the response to that of a 1, that response must be given a 0. Sometimes a student will place a mark in their answer space, such as an “x” or a “?” If such marks are the only indications that the student visited that part of their answer document, they receive a 0 score. Notice that the last designation in the scoring rubric is that of Blank. It would seem self-evident that if the answer space has nothing in it, the score is a blank. But what if there is a stray mark, such as a faint line across the answer space, or little dots and specks? Then the particular question is still scored a Blank, as opposed to the aforementioned “x” or “?” that earns a 0. Careful training shows scorers the difference between artifacts caused by the scanning process, such as stray marks or dots, and the conscious marking of the answer space by the student. So how do we know how to score a student’s response? We apply the wording of the scoring guide/rubric to actual student responses, collect a sufficient number of anchor papers for each score point, and compare student responses to the exemplars.
|
||||||||||||||||
| About the Author Stan Clough is an associate chief reader at Measured Progress. |
||||||||||||||||
