|
It cannot be expected that these strategies will eliminate all the
TABLE 2.3 STRATEGIES TO IMPROVE SIZING UP ASSESSMENT __________
1. Recognize that sizing up assessment is going on; without this awareness, it is difficult to improve the process.
2. Let first impressions represent initial hypotheses to be confirmed or rejected by additional information; see behaviors at least twice before judging pupils.
3. Observe long enough to get a sense of pupils' typical behaviors.
4. Observe important pupil behaviors directly rather than inferring them from ancillary behaviors and characteristics.
5. Supplement informal observation with more formal, structured assessments such as pretests and games.
6. Pick one pupil characteristic per day and structure classroom activities to permit pupils to demonstrate that characteristic.
7. Determine whether different types and sources of information (e.g., informal observation, formal observation, tests, other teachers' comments) provide similar information about pupil characteristics.
errors that can occur in sizing up assessment; they will not guarantee perfect information. It is often said that teachers must avoid all errors in formulating perceptions of their pupils' characteristics, but this is impossible. So long as perceptions are heavily influenced by observations of behavior that must be interpreted by the observer, total elimination of error will remain a goal to be worked toward. The teacher's ethical responsibility is not to make all assessment perfect. The teacher's responsibility is to strive to keep the number of errors made as low as possible and to revise judgment of pupils when initial impressions prove wrong. The strategies suggested in this section are intended to be simple, manageable actions that teachers can take to enhance the validity and reliability of informal assessment information. They are not intended to be so formal and cumbersome that they drive the informal observation process out of existence.
1. Be Aware That the Process of Informal, Sizing Up Assessment Goes On in the Classroom and Affects Pupils. Sizing up assessment is the natural process of forming first impressions about people based on what we see and hear in interactions. The process is such a normal part of everyday behavior that teachers often are not even aware that they are doing it in their classrooms. But doing it they are—because it is a necessary prerequisite to bringing order into a social setting. If a teacher does not realize that sizing up assessment is going on in the classroom and does not recognize the dangers of forming incorrect impressions of pupils, little can be done to improve the assessment process. Teachers must be sensitive to the consequences of making poor judgments about pupils based on incomplete, biased, or erroneous observations. Awareness of sizing up assessment and its role in the classroom is the first step necessary to improve the process.
2. Treat Initial Impressions as Hypotheses to Be Confirmed or Corrected by Subsequent Observations and Information. There are many sources of information that influence the first impression a teacher forms of a pupil. Some of the information is not even observed directly by the classroom teacher but is transmitted by former teachers, the school grapevine, performance of the pupil's brothers and sisters, or knowledge of the pupil's group membership. This information is available before the teacher observes a pupil for the first time. Once pupils enter the classroom, first impressions are formed largely by their dress, posture, friends, and a few brief interactions with the teacher. Perceptions made on the basis of brief observations or interactions in the first few days of school may not represent typical pupil behavior, but establish perceptions that influence subsequent teacher-pupil interactions.
The less direct the information a teacher has about a pupil, the greater the amount of interpretation and subjective analysis the teacher must perform to arrive at a perception of the pupil's characteristics. The greater the amount of interpretation and subjective analysis required, the greater the likelihood of error in the perception. Remember, most observations are being made in the context of a functioning classroom where many activities are going on, where many pupils are being observed from one minute to the next, and where the teacher does not have a great deal of time to reflect on the meaning of an observation of one pupil before he or she has to focus attention on observation of another pupil.
Under such conditions, the teacher may misunderstand or misinterpret the initial signal from a pupil. The teacher may not fully understand the context in which the behavior took place. If a teacher turns her head just at the time Robert punches Jackie, the teacher does not have a sense of the events which led up to the punch, and therefore may be misinterpreting the observation if she automatically concludes that Robert is an aggressive, physically harmful pupil. Or the teacher may mislabel what she sees. Observing a pupil frowning during the lesson on improper fractions may be interpreted by the teacher to indicate that tbe pupil is upset and doesn't like math; in reality, the pupil may love math and is frowning because she learned improper fractions last year.
Initial impressions based on informal observations should be considered tentative hypotheses that must be confirmed or disproved
by subsequent observation and information. In areas such as achievement and ability, teachers are reasonably accurate in sizing up their pupils' performance levels. In their initial impressions of pupils' affective traits such as interest, motivation, or self-confidence, teachers are less accurate. Regardless of whether the focus is on cognitive or affective characteristics, however, no teacher's initial perceptions are correct all the time.
One method to improve the adequacy of informal observations and the perceptions that result from them is not to rush too quickly to judge and label pupils on the basis of hearsay or a brief initial observation. The example of a physician's diagnosis is an appropriate one. When a doctor sees a patient, the doctor goes through a series of information-gathering stages to diagnose the patient's problem. The physician asks the patient about the problem and its symptoms. On the basis of this information the doctor forms a first, general impression of what diagnoses fit the symptoms.
But doctors do not stop with this initial information gathering. They ask further questions to narrow the range of possible diagnoses: "Does it hurt when you breathe? How long has it been like this? Does it seem to be worse? Is it worse before or after eating? Is the pain consistent or intermittent? Have you had a fever?" The synthesis of the information gathered from these two stages of questioning may lead to a diagnosis. If not, more information will be collected until the proper diagnosis is made. Doctors do not rely on first impressions or any single source of evidence to form a diagnosis, and thus provide a good model for classroom teachers when they carry out sizing up assessment. Confirm the first impression with subsequent observations and information, and do not be afraid to change an incorrect first impression.
3. Use Low-Inference Indicators to Gather Observational Information about Pupil Characteristics. Teachers must interpret their observations in order to use them in sizing up pupils. Some observations require less interpretation than others because they are closely related to the characteristic the teacher is trying to find out about. Other observations are less related to the characteristic of interest and so require a more complex inferential chain of reasoning to provide information about the pupil. Low-inference indicators are those that are closely related to the characteristic of interest, while high-inference indicators are those that are not closely related to the characteristic of interest and require a number of inferential leaps by the teacher. The closer the behavior observed is to the particular pupil trait one wishes to describe, the more confidence one can have in the observational information gathered.
Consider the following example. At the start of the school year, a teacher wanted to determine a pupil's motivation to learn. He could do this in a variety of ways. He could ask the previous year's teacher how motivated the pupil was last year. He could, as a number of teachers quoted earlier did, determine motivation by observing the pupil's dress and the group the pupil spent time with. He also could watch for behaviors such as the pupil staying after school to complete work, asking questions about topics discussed, and spending spare time reviewing lessons. There are many indicators the teacher could use.
There are differences in the quality of information provided by the different possible indicators of motivation. The pupil's dress and the group of friends the pupil spends time with are indirect, high-inference indicators of the pupil's motivation to learn. In order to make a connection between dress and motivation, the teacher has to make a number of assumptions or inferences. For example, the teacher must assume (1) that the way a person dresses tells something about their personality; (2) that for this particular pupil, clothes reflect personality rather than income level, fashion, or parental guidance; (3) that particular kinds of dress are more indicative of motivation to learn than others; (4) that the teacher knows what those styles are and what they indicate about motivation; and (5) that the style observed today is indicative of the pupil's typical style of dress. Each of these assumptions may be wrong, which would make the process of judging the student on the basis of dress flawed. The more assumptions needed to relate an observed behavior to the characteristic of interest, the more likely it is that errors of perception will occur. This is the problem with high-inference indicators.
Or suppose the teacher asks last year's teacher about the pupil's motivation to learn. He is using a lower inference indicator than the pupil's style of dress. However, by relying on the judgment of last year's teacher, he assumes (1) that last year's teacher defines motivation to learn in the same way as he does, and (2) that the pupil has not changed a great deal since last year. In the end, the best way for the teacher to determine the pupil's motivations is by direct observation of classroom activities the teacher associates with motivation to learn, such as asking questions, staying after school to complete work, and spending spare time on lesson-related topics. These are low-inference indicators because the behaviors of interest are observed directly. Note that the teacher still has to make assumptions about how typical the observed behavior is and whether the pupil is faking the apparent motivation, but direct observation of the appropriate behaviors is likely to give a better indication on which to base a perception than indirect, high-inference indicators.
The school performance of older siblings or the section of the city the pupil lives in are high-inference indicators of academic potential. They are more likely to lead to incorrect judgments than direct observation of the pupil's typical academic behaviors. If you form reading groups by asking last year's teacher what group he or she placed pupils in, you are using a higher inference indicator than if you observed pupils reading aloud and responding to questions about what they have read. In general, the more direct the observation of the trait of interest, the more straightforward interpretation of the observations will be, and the less likelihood there will be of errors in assessment.
In informal observations, teacher-pupil encounters are brief and the tendency is for the teacher to focus on superficial, high-inference characteristics such as dress, facial expression, "look in the eyes," helpfulness, general appearance, and so on. These observations are often used to generalize about broader traits like motivation, self-concept, trustworthiness, self-control, and interest. Such high-inference generalizations are apt to be wrong because, as noted, they are based on a chain of assumptions which itself is unlikely to be completely true. The moral then, is: Don't over-generalize from high-inference indicators; instead, focus your informal observations on low-inference behaviors and indicators.
4. Supplement Informal Observations with More Formal, Structured Activities. There is no rule that requires that only informal classroom observations be used to size up pupils at the start of the school year. In fact, as discussed earlier, the naturalistic nature of informal observations means that the teacher does not have control over the kinds of behaviors that occur, especially in the first few days of school. Many teachers recognize this limitation and introduce more structured activities to supplement their ongoing informal observations.
For example, some teachers give a review test of last year's reading or math skills to get an idea of their pupils' math and reading levels. Some teachers ask pupils to keep a journal during the first week of school or write an essay on "What I Did Last Summer." Group discussions, group projects, reading aloud, and games based on spelling words, math facts, geographical knowledge, or general information are other structured activities that can provide information about pupil knowledge, ability to work in a group, competitiveness, and so on. Games related to listening skills can provide information about pupils' abilities to follow directions and process auditory information. For example, "Write an equation that show; what I say and then solve the problem. Two times five plus four divided by two." A glance at the papers can indicate how well pupil
listen, translate what they hear into writing, and know simple math operations.
It is also useful to select one pupil characteristic per day (e.g., cooperation in groups, mastery of content prerequisites, study habits), structure classroom activities to permit pupils to exhibit the selected characteristic, and observe the characteristic for all pupils. These exercises and activities do two things. First, they provide information about pupil interests, styles, and academic performance that is not always obtainable from informal observations. Second, they require all pupils to perform the same behavior and thereby permit comparisons between pupils on desired characteristics. Such structured activities are useful supplements to informal observations in the process of sizing up pupils because they provide an additional perspective on pupil behavior and characteristics.
5. Observe Long Enough to Be Fairly Certain That You Have Perceived the Pupil's Typical Behavior. With all the factors that can distort the accuracy of informal observations, it is evident that a very important feature of good information is that it represents the typical behavior of a pupil. It is for this reason that rushing to make judgments in the first few days of school and basing perceptions solely on first impressions can produce errors in assessment. To obtain data that are dependable, the teacher must look for patterns and consistency in behaviors, not for one-time behaviors. The search is for typical pupil behaviors. If the behavior sample used to formulate perceptions of a pupil is inadequate because too few observations have been obtained or the ones that have been obtained are atypical, the assessment evidence is flawed and will lead to incorrect perceptions and decisions.
There is no standard amount of time or number of observations that can ensure that assessment data represent stable, typical pupil behavior. The amount of time a given teacher needs to be confident will vary according to the observational skills of the teacher, the number of characteristics the teacher wishes to know about, the opportunity to observe characteristics of interest, and the nature of the pupil being observed. The quieter and less demonstrative a pupil is, the longer it takes for a teacher to form an accurate perception of the pupil. The point here is not to state a time limit that always will produce stable and consistent information about pupils' typical behavior—that is not possible. It is to caution against jumping the gun and using single observations to form perceptions, without checking to be certain that the behaviors seen are the pupil's typical behaviors.
A good rule of thumb to follow is to make sure the behavior on which one wishes to base a perception is observed at least twice.
This "see it twice" principle will guard against an isolated, atypical behavior being used incorrectly to form a perception of the pupil. Don't let a single test performance, classroom incident, lunchroom behavior, or peer interaction—good or bad—define a pupil's character, ability, or personality. Give pupils the benefit of the doubt before passing judgment; give them a fair chance to show their typical behaviors and characteristics.
6. Determine Whether Different Kinds of Information Confirm Each Other. Astronomers talk about the method of "triangulation" to locate stars. Essentially, this method involves measuring a star's position relative to two other positions. The three positions (the star's and two others) form a triangle that allows the astronomer to plot the star's position in space with confidence. A form of triangulation is relevant for classroom assessment, particularly the kind that goes on in the first few days of school.
Teachers can have more confidence in their perceptions if two different types of evidence provide confirmatory information about a pupil. For example, do classroom performance and test scores indicate the same thing? Do the teacher's classroom observations of the pupil's needs correspond to those identified by last year's teacher and the pupil's parents. Do behavior patterns in the classroom persist in the lunchroom and on the playground? Are the classroom teacher's judgments of the pupil's oral reading and comprehension similar to the pupil's test performance and last year's teacher's perception?
Each of these questions represents a form of triangulation; the teacher does not rely on a single source of information about a pupil, but seeks multiple sources of information for corroboration. When multiple kinds of information indicate the same thing about a pupil, the teacher can be more confident about the pupil's true characteristics. In any situation in which the present teacher's observations or perceptions are going to be compared to those of other teachers, parents, or test scores for corroboration, it is better if the present teacher forms his or her own impressions of the pupil before obtaining the impressions of others. By doing this, the teacher's perceptions will not be influenced or biased by the perceptions of others.
CHAPTER SUMMARY
This chapter discussed the role of classroom assessment at the start of the school year. The purpose of such assessment is to help teachers "size up" pupils and learn about their characteristics so that
communication between the teacher and pupils can commence and the classroom can be organized into a social and learning setting. Teachers obtain the information they use to form initial impressions of pupils from many sources: the school grapevine, comments of prior teachers, test scores and information, pupil dress and manners, and their own observations. The assessment information that is gathered in the first week or so of school is synthesized by the classroom teacher into a general perception and character sketch of each pupil. These perceptions are formed early in the year and remain quite stable over time.
Initial assessments are very important in forming teachers' perceptions about pupils and can influence teacher-pupil interactions throughout the school year. Initial assessments also rely heavily on teacher observations that are idiosyncratic, subjective, and occur in a naturalistic, unplanned manner. For these reasons, it is important to be certain that the informal assessments made at the start of the school year are valid and reliable. Two sources of assessment invalidity were described: observer bias and logical error. Observer bias occurs when initial observations or information predetermine how the teacher will view the pupil in all subsequent observations. Logical error occurs when the behavior a teacher observes is not a direct indicator of the characteristic the teacher wishes to know about. In this case, the teacher must create a logical sequence of inferences that relates the observed behavior to an unobserved characteristic. The longer the inferential sequence, the more likely the teacher's perception of the pupil is to be wrong.
Reliability is also a concern in informal assessment, particularly since the information is gathered and the perceptions formed during the first few days of school. The reliability issue involves the extent to which a few observations made in the first few days of school represent the typical behavior of the pupil. If the information gathered is not representative of the pupil's typical behavior, perceptions of the pupil will be in error.
Six recommendations for enhancing the validity and reliability of informal assessments were made. These recommendations will not eliminate all errors in informal assessment, but their practice should eliminate many of the commonest errors. The recommendations were: (1) be aware that sizing up assessment is going on in the classroom; (2) confirm initial perceptions with subsequent observations and information; (3) try to observe pupil characteristics directly, rather than inferring them from behaviors that may not be related to the characteristic of interest; (4) supplement informal information-gathering methods with more formal, structured methods; (5) observe long enough to be certain that the pupil has had an opportunity to show you his or her typical behavior; and (6) determine whether different kinds of information about a pupil gave
the same results. Understanding the process of informal assessment at the start of school and the consequences it can have for pupils and teachers should make one more sensitive to the need to carry out valid and reliable sizing up assessments.
REFERENCES
Airasian, P. W. (1989). Classroom assessment and educational improvement. In L. W. Anderson (ed.), The effective teacher (pp. 333-342). New York: Random House.
Airasian, P. W. (1991). Perspectives on measurement education. Educational Measurement: Issues and Practice, 10, 1.
Airasian, P. W., Kellaghan, T., and Madaus, G. F. (1977). The stability of teachers' perceptions of pupil characteristics. Irish Journal of Education, 11,(1 & 2), 78-84.
Airasian, P. W., Kellaghan, T., Madaus, G. F., and Pedulla, J. (1977). Proportion and direction of teacher rating changes of pupil progress attributable to standardized test information. Journal of Educational Psychology, 69(6), 702-709.
Anderson, L. M., and Evertson, C. M. (1978). Classroom organization at the beginning of school: Two case studies. Paper presented at the annual meeting of the American Association of Colleges for Teacher Education, Chicago, February.
Asch, S. (1946). Forming impressions of personality. Journal of Abnormal and Social Psychology, 41, 258-290.
Brophy, J. E., and Good, T. L. (1974). Teacher-student relationships. New York: Holt, Rinehart and Winston.
Brophy, J. E., and Rohrkemper, M. M. (1981). The influence of problem ownership on teachers' perceptions of and strategies for coping with problem students. Journal of Educational Psychology, 73, 295—311.
Calderhead, J. (1983). A psychological approach to research on teachers' classroom decision making. British Educational Research Journal, 7(1), 51— 57.
Cooper, H. M., and Berger, J. M. (1980). How teachers explain students' academic performance: A categorization of free response academic attributions. American Educational Research Journal, 67, 274—277.
Downey, M. (1977). Interpersonal judgments in education. London: Harper & Row.
Evertson, C, Brophy, J., and Good, T. (1972). Communication of teacher expectations: First grade (91). Research and Development Center for Teacher Education, University of Texas at Austin.
Дата добавления: 2015-11-04; просмотров: 30 | Нарушение авторских прав
<== предыдущая лекция | | | следующая лекция ==> |