Recording and Summarizing Evidence
Recording and Summarizing Evidence
Recording evidence
- A situation must allow a record of the behaviour to be kept, not just evoke it.
- Methods: a written record, an observer’s description, motion pictures, sound recording, or a checklist.
Summarizing the record
- Decide the terms or units that summarize the behaviour.
- The summary should parallel the nature of the behaviour itself.
Judging an instrument
- Objectivity of scoring, reliability of the sample, and validity of the method.
Finding a situation that evokes a behaviour is only half the work. If the behaviour vanishes the moment it happens, there is nothing to appraise. So the procedure must also record the behaviour and then summarize it in a form that means something.
Recording evidence
Once situations are chosen to get evidence of the desired behaviours, it is necessary to devise a means of recording the learner’s behaviour in that situation. Some situations record themselves. In a written examination, learners make their own record in their writing, so keeping a record is not a serious problem. Other situations do not. A grade-one activity that gives children a chance to play and work together is a good situation for evidence of personal-social adjustment, but the behaviour is gone as soon as it occurs unless someone keeps a record of the children’s reactions, so it can be appraised afterward.
Record keeping can be done in several ways. It may involve a detailed description of the reaction by an observer, or it may call for motion pictures, sound recording, or a checklist as a means of getting a satisfactory record. This step must be considered for each test situation, to be sure the situation not only evokes the desired behaviour but allows a record to be obtained that can be appraised later.
A situation must allow a usable record, not just evoke the behavior
A written exam records itself, but live behavior like cooperation vanishes unless captured by an observer’s description, film, sound recording, or checklist for later appraisal.
Summarizing the record
A pile of recorded behaviour still has to be made sense of. The next step in developing an instrument is to decide the terms or units used to summarize and appraise the record of behaviours obtained. The method of appraising the behaviour should parallel the implications of the behaviour itself; the units must fit what the behaviour actually is.
Reading interest shows how. If the objective defines reading interest as the development of increasingly broad and mature interests, then the record of a learner’s reading must be summarized in units that show breadth and maturity:
- Breadth can be indicated by a number showing the different categories of reading material a learner takes in, such as detective and adventure stories alongside religious and historical reading.
- Maturity can be indicated by classifying the reading under different levels of maturity, such as an average level.
The units chosen, breadth and maturity, mirror the objective itself. That is the rule: summarize behaviour in terms that match the behaviour you set out to develop.
The summary units should parallel the behavior the objective defines
If reading interest means broad and mature interest, the record is summarized by breadth (categories read) and maturity (levels of reading). The units must mirror what the objective set out to develop.
Judging an instrument: objectivity, reliability, validity
Finally, an evaluation instrument is judged by three qualities:
- Objectivity: the scoring does not depend on who does it.
- Reliability: the sample of behaviour is dependable and consistent.
- Validity: the method genuinely provides evidence of the desired behaviour, and not something else.
Validity is the deepest of the three. An instrument can be perfectly objective and reliable and still measure the wrong thing. The question that matters most is whether the method actually captures the behaviour the objective named.
Objectivity, reliability, and validity
Objectivity: scoring does not depend on the scorer. Reliability: the sample is consistent. Validity: the method actually measures the desired behavior. Validity is deepest, since an instrument can be reliable yet measure the wrong thing.
How was this article?