page 1 - 2 - 3 - 4

An introduction to systematic observation

Observation in classrooms
Observation of what goes on in classrooms has uncovered differences in time allocations which suggest that high progress children get more opportunities to learn than low progress children. Studies have produced evidence of how the successful children tend to get better and better, drawing away from the average or below average children whose progress proceeds at an apparently slower rate. However, this slow rate of learning can occur because children do not get the kinds of help they need to learn at faster rates.

Observing individual progress
If we attend to individual children as they work, and if we focus on the progressions in learning that occur over time, such observations can provide feedback to our instruction. Observations which lead us to modify our instruction are particularly appropriate in the formative stages of new learning, as in beginning reading, beginning writing and beginning mathematics.

I have tried to observe individual children at work, reading and writing, and to capture evidence of the
progress that they make. Science is based on systematic observation of phenomena under known conditions. Physicists or chemists in laboratories, botanists and zoologists in the field, and behavioural scientists in psychology, sociology, linguistics and cultural anthropology all use observation to get research data. Despite some lingering mistrust of observation in educational research, it is becoming more acceptable direct observation as a method for data collection, particularly in the years of early childhood education (Genishi, 1982).

With good reason, educators have relied on systematic testing rather than systematic observation of learning. The measurement theories that are used to guide test construction and research analyses lead to better interpretation of test and examination results. But as we have improved our testing strategies we have placed undue emphasis on testing to the point where we can deprive teachers and administrators of valuable information about learners and their learning. There is a seductive efficiency about final assessment scores. Yet a funny thing happens on the way to those final assessments: day-to-day learning takes place. In education, we need to pay more attention to the evaluation of learners who are on the way to those final assessments. One of the things that a class teacher needs to know is what occurs over time for the individual learner in a particular classroom programme.

We do need research endeavour which looks for explanations of what causes what, or what conditions bring about differences, and such questions call for the use of a variety of research paradigms, but for effective instruction we also need answers to two basic questions:
‘What typically occurs for children like those I teach as learning takes place over the school year?’ and ‘How is this individual child changing over time in relation to what typically occurs?’ Teachers who have answers to these questions will be more responsive to the daily learning of their pupils and will deliver more appropriate feedback.

In developmental psychology young children were always studied by direct observation. Studies of how children learn to speak have been exciting, and so have the more recent studies of young children learning to read and write. But teachers must go beyond reactions like ‘Ooh! Ah!’, or ‘I am surprised!’ and ‘Isn’t that cute’ and try to understand what is happening and why it is happening.

Measuring outcomes
Historically, most assessment has been directed to the outcomes of instruction. We wait until the end of the instruction sequences before we assess. We

When we measure the outcomes of teaching with important tests the instruction of the learners is already over. The test score is, in legal terms, after the fact. It is almost too late to change the fate of the students because of what we find out. The opportunity has gone.
We do not have to test all children to assess educational standards at the national, state or district level; we can use sampling methods to get information on standards in the same way that we take public opinion polls. We do not need to test every child to know whether the school system is producing satisfactory outcomes. And in many countries there is some agreement that this measurement of the education system can be done best from the age of nine years and upwards.
Measures of outcome tell us where the achievement levels of the education system lie. They do not tell us what brought about those achievement levels. We do not know from the assessments how the high achievement levels were obtained, or why the low achievement levels occurred. If we try to use those results to improve instruction,