As long as trained evaluators are in place, differentiation across six levels may be quite adequate for admission into institution or placement into courses. For classroom instructional purposes, holistic scores provide very little information. In most classroom setting where a teacher whises to adapt a curriculum to the needs of a pasticular group of students, much more differentiated information across subskills is desirable than is provided by holistic scoring.
Primary trait scoring
A second menthod of scoring, primary trait, focuses on “how well students can write within a narrowy defined range of discourse” (weight,2002,p.110). This type of scoring emphasizes the task at hand and assign a score based on the effectiveness of the text’s achieving that one goal. For example, if the purpose or function of an essay is to persuade the reader to do something, the score for the writing would rise or fall on the accomplishment of that function. If a learned is asked to exploit the imaginative function of language by expressing personal feelings, then the response would be evaluated on that feature alone.
For rating the primary trait of the text, Lloyd-Jones (1997) suggested a fourpoint scale ranging from zero (no response or fragmented response) to 4 (the purpose is unequivocally accomplished in a convincing fashion).it almost goes without saying that organization, surpotting details, fluency, syntactic variety, and other features will implicitly be evaluated in the process of offering a primary trait score. But the advantage of this menthod is that it allows both writer and evaluator to focus on function. In summary, a primary trait score would assess
• The accuracy of the account of the original (summary)
• The clarity of the steps of the procedure and the final result (lab report),
• The description of the main features of the graph (graph description), and
• The expression of the writer’s opinion (response to an article).
For classroom instruction, holistic scoring provides little washbackinto the writers further stages of learning. Primary trait scoring focuses on the principal function of the text and therefore offers some feedback potential, but no washback for any of the aspects of the written production that enhance the ultimate accomplishment of the purpose. Classroom evaluation of oearning is best served trought analytic scoring, in wich as many as six major element of writing are scored, thus enabling learners to home in on weaknesses and to capitalize on strengths.
Analytic scoring mayh be appropiatelly called analytic assessment in order to capture its closer association with classroom language instruction that with formal testing. Brown and bailey (1984) designed an analytical scoring scale that specified five major categories and a description of five different levels in each category, ranging from unacceptable to excellent (see table 9.2)
at first glance, borwn and bailey’s scale may look similar to the TWE holistical scale discussed earlier; for each scoring category there is a description that encompasses several subsets. A closer inspection, however, reverals much more detail in the analutic method. Instead of just six descriptions, there are 25, each subdivided into a number of contributing factors.
The order in wich the five categories (organization, logical development of ideas, grammar, punctuatuion/spelling/mechanics, and style and quality of expression) are listed may bias the evaluator toward the greater importance of organization and logical development as opposed to punctuation and style. But the mathematical assignment of the 100 point scale gives equal weigh (a maximum og 20 point) to each of the five major categories. Not all writing and assessment specialists agree. Tou might, for example, consider the analytic scoring profile suggested by Jacobs et.al. (1981), in which five slightly different categories were given the point values show on page 246.
As your curricular goals and students need vary, your own analytical scoring of essay may be appropriately tailored. Level of proficienly can make a significant difference in emphasis; at the intermediate level, for example, you might give more emohasis to syntax and mechanics, while advanced levels of writing may call for a stong push toward organization and development. Genre can also dictate variations in scoring. Would a summary of an article require the same relative emphasis as a narrative essay? Most likely not. Certain of tyes writing, such as lab reports or interpretations of statistical data, may even need additional-ar at least redefined-categories in order to capture the essential components of good writing those genres.
Analiyic scoring of composition offers writers a little more washback than a single holistic or primary trait score. Scores on five or six major element will help t call the writers attention to areas of needed improvement. Practicality is lowered in that more time is required for teacher to attend to details within each of the categories in order to render a dinal score alone, however are still not sufficient for enabling student to become proficient writers, as we shall see in the next section.
Beyond scoring: responding to extensive writing
Formal testing caries with the burden od designeing a practical and instrument that assessment its intended criterion accurately. To accomplish that mission, designers of writing tests are charged with the task of profiding as objective a scoring procedure as possible, and one that in many cases ca be easily interpreted by agents beyond the learner. Holistic, primary trait, and analityc scoring all satisfy those ends. Yet beyone mathematically calculate scores lies a rich domain od assessment in wich of developing writers is coached from stage to stage in a process of building a storehouse of writing skills. Here in the classroom, in the tutored relationship of teacher and student, and in the community of peer learners, most of the hard work of assessing writing is carried out. Such assessment is formal, formative, and replete with washback.
Most writing specialist agree that the best way to each writing is a hands-on approach that stimulates student output and then generates a series of self-assessment, peer editing and rvision, and teacher response and conferencing (Raimes, 1991, 1998; Reid, 1993; Seow, 2002).