For classroom observation as a learning tool Richards (1998:143) proposes three perspectives on a lesson for pre-service training to develop a deeper understanding of how and why teachers teach the way they do and the different ways teachers approach their lessons. They are:

1) Teacher-centered focus: the teacher is primary focus; factors include the teacher’s role, classroom management skills, questioning skills, presence, voice quality, manner, and quality instructions.

2) Curriculum-centered focus: the lesson as an instructional unit is the primary focus; factors include lesson goals, opening, structuring, task types, flow, and development and pacing.

3) Learner-centered focus: the learners are the primary focus; factors include the extent to which the lesson engaged them, participation patterns, and extent of language use.

Wallace (1998:68) substitutes the focus on the curriculum with the focus on the context in which the teacher teaches: the classroom layout, the teaching aids available and how they are used.

Low-inference and high-inference categoreis

The presentation of items involves constructing sets of categories into which occurrences must be coded unambiguously. In this respect Long (1980:3) introduces low-inference and high-inference measures. Low-inference categories include things that can be counted or coded without the observer having to infer their meaning from observable behaviour. Such categories according to Allwright and Bailey (2000:73) involve the number of times the student raises her/his hands, or the frequency with which the teacher uses the student’s name. High-frequency items demand that the observer make a judgement that goes beyond what is immediately observed. The samples of this type of categories cover factors like learner’s attention, or the social climate. I can conclude that observation data should cover categories of observable behaviour that does not require much interpretation.

Typology of observation

Typology of classroom observation instruments is worked out by Wallace (1991:66) and he presents the following oppositions:

1. system-based, ethnographic or ad-hoc

2. global or specific

3. evaluative, formative or research-related

4. teacher-focused, learner-focused or neutral in focus

5. quantitative or qualitative

He admits that some of the oppositions are not clear-cut and overlap. For example, observation techniques which are primarily evaluative may be employed for formative purposes, ethnographic approach is treated as global and qualitative. System based approach can focus on teacher’s activity and learners’ activities. System-based (systematic), ethnographic and ad-hoc approaches encompass other characteristics of the classification provided. Thus, I outline the features of the first opposition.

System-based approach

By system-based observation Wallace (1991:67) means the observation that is based on a system of fixed and pre-specified categories. They are global in nature, i.e. ‘they are intended to give general coverage of the most salient aspects of the classroom process’ (Wallace 1991:110). Any system contains a finite array of categories. The endeavour of all system-based observation instruments is the analysis of teacher-class interaction. The two most influential systems are devised by Bellack (1966:267) and by Flanders (1970:314). Wallace (1991:112) has identified the characteristic features of the first system as:

1) the data are measured from a transcript, i.e. the data have to be first recorded and then transcribed;

2) the central place of labelled units of discourse are structure, solicit, response, reaction.

In the ‘Flanders tradition’ there is a form of documented recall where tallies are made every three minutes under one range of categories. In chapter 2.6 the analysis of a range of interaction schemes, their advantages and disadvantages are presented with more details. They are widely used by researchers as they are ready-made, well known and ‘it does not to be trialled and validated’ (Wallace 1991:111).

Ethnographic approach

The observation techniques share many of qualities of ethnographic practices. Ethnography is a detailed sociological observation of people which immerses the researcher in an intense period of observation ‘which guides and informs all subsequent data gathering’. (Radnor 2002:49)

Ethnographical approach is originally developed from the methodologies of field anthropologists and sociologists concerned with studying human behaviour within the context in which that behaviour would naturally occur. Methodologically, ‘anthropological’ classroom studies are based on participant observation, during which the observer immerse him/herself in the ‘new culture’. Initial data gathered by the ethnographer are open-ended and relatively unstructured that ‘allows and encourages the development of new categories’ (Delamont and Hamilton 1976:13). An ethnographer uses a holistic framework. S/he makes no attempt to manipulate, control or eliminate variables. At the same time s/he reduces the breadth of research problems systematically to give more concentrated attention to the emerging salient issues.

The great strength of the ethnographic research is that it gets away from the simplistic behavioural emphasis of the pre-specified codes. (Delamont and Hamilton 1976:37).

The main purpose of the ethnographic approach is the search for meaning and is based on the description of the studied phenomenon. However, Lutz (1986:112) warns that not everyone who can write a paragraph describing an encounter between a teacher and a student is an ethnographer, and he points out that an observer should be trained in ethnographic methods, particularly participant-observer field methods.

Ad-hoc approach

The term ‘ad-hoc’ is used to describe something that has been devised for a particular purpose, ‘with no claims to generality’ (Wallace 1991:113). The ad-hoc approach relates to structured approaches but the categories derive from a particular problem or research topic. That is why this system is more popular with practising teachers. What is more this approach is flexible and eclectic, and involves both quantitative and qualitative data where each seems appropriate. Wallace (1991:113) assumes that each different area of concern will yield a different system of analysis. Ad-hoc approach is considered to be the most appropriate in teacher-training education as it is basically guided discovery approach that drive student-teachers to focus and reflect on an important area of language teaching, and provide a meta-language with which to discuss. The instrument of ad-hoc approach is known as observation tasks (Wajnryb 1992) and is described in Chapter 2.6.2.