Anecdotal records McKernan (1996:67) refers to narrative-verbatim descriptions of meaningful incidents and events which have been observed in the behavioural setting. They focus on narrative, conversation and dialogue and provide short, sharp incisive summaries of points that stick in the mind after the event. Anecdotal records are treated to be useful in teacher training education because they directly observe behavioural data which enable students to see the incident and gain inside perspective. One of the key tasks for the observer is to watch for the beginning and ending of episodes of behaviour. McKernan (1996:68) sets some disadvantages of anecdotal records that are similar to diary keeping and journal as any piece of descriptive writing, such as:
- they require extensive time to observe, write and interpret;
- maintainenace of objectivity is difficult;
- observers require training in the use of anecdotes;
- they are often reported without taking accounts of setting;
- read out of context, they can be misunderstood and misinterpreted;
- some observers focus on negative or undesirable events only.
Personal action logs
Personal action logs McKernan (1996:110) defines as record sheets which document a researchers activities over a lengthy time period to get a full-blown representation of a day. Thornbury (1991:141) clarifies the purpose of log-keeping as to direct trainees attention towards areas they may have overlooked or avoided; to measure the trainees assessment against our own; to make adjustments, if necessary, to the course design and/or content. Logs may be kept in chart summary form, describing the main events with time sampling or in a more descriptive form similar to a diary. At the same time personal logs (McKernan 1996:111) are recommended to keep over a lengthy period of time and in connection with more extensive accounts, such as field notes, diaries and audio transcripts to validate findings.
The use of check-lists suggests the formulation of well-defined and clearly delineated behaviour categories, which in turn presupposes more than a superficial acquaintance with the data (Hutt and Hutt 1970:38). It is used to focus the observers attention to the presence, absence, or frequency of occurrence of each point of the prepared list as indicated by checkmarks (Hopkins and Antes 1985:467). Thus a prerequisite for obtaining reliable and valid data from check-lists is a set of clearly defined categories. For this reason a check-list would be unsuitable for recording behaviour with which the observer was not completely familiar or for recording the complete range of activities in a free-field situation. The researchers confirm that although in principle a large number of categories are feasible, in practice an observer is unable to cope reliably with more than fifteen. Different methodologists notice that as the number of categories increase, the problems involved in scanning these. That is why Hutt and Hutt (1970:69) offer from a practical view to have check-lists as compact as possible, since they are most commonly used in those situations where the observer is attempting to record unobtrusively and with the minimum of distraction to the subject.
The greatest advantage of check lists is the facility and speed with which they can be analysed, as observer just ticks off phenomenon against an appropriate category by mere observation. Measures that might be easily obtained are as follows:
- frequency with which there is a change in activity;
- number of different activities;
- number of stimuli encountered;
- duration of specific activity;
- changes in nature and duration of activities with time.
However, McKernan (1996:108) admonishes that the arrangement of the points is crucial in that sequence in task completion should be logical and sequential. An observer or designer of this instrument must ensure that:
- points to be observed are listed in their actual sequence of happening;
- all similar attributes are included in categories;
- all the relevant and specified points are listed.
Over the years numerous schemes have been developed for recording classroom interaction. Chaudron (1988:19), modifying the analysis originated by Long (1980), identifies twenty-four various schemes. In his review Chaudron (1988:17) points out that Long (1980) has included only those instruments which were designed to observe verbal interaction in a classroom, whereas the range of categories is great due to various purposes of observation. Chaudron interprets categories as
- social interactive (Allwright (1980:169) turn-taking and turn-giving, Moskowitzs (1970) jokes, praises or encourages)
- pedagogical (Jarviss (1968:336) classroom management, repetition reinforcement, or Fanselows (1977:18) solicit, respond)
- objective behaviour (Naiman, Neil, Frölich, Stern, and Todescos (1978) student hand-raising, student callout, or Moscowitzs (1970) student response -choral)
- semantic or cognitive content of behaviours (Fanselows (1977:31) characterize)
- type and grouping of participants (Mitchell et al. (1981:19) whole class, individuals doing the same task)
For teacher training purpose Chaudron (1988:18) recommends to apply eleven schemes among which Capelle, Jarvilla, and Revelle (n.d.), Moskowitzs (1970), Politzer (1980), Seliger (1977) are conducted in real time coding and categories of schemes refer to low degree of inference.
Advantages of interaction schemes as the basis of reflection in experiential knowledge are described by Wallace (1991:121) and he claims that these systems
- objectify the teaching process;
- provide a reliable record (by a trained observer);
- promote self-awareness in the teacher;
- provide a meta-language, which enables teachers to talk about their profession;
- make teacher training more effective by improving the quality of teaching.
At the same time systematic observation schemes have some critics. Delamont and Hamiltons (1976:3) main critique is levelled at the use of pre-specified categories to code or classify the behaviour of teachers and pupils, which can not capture and reflect the whole complexity of classroom life.
Delamont and Hamilton (1976:8) identify seven criticisms of systematic observational systems:
- Systematic observation provides data only about average or typical classrooms, teachers and pupils.
- All the interactional analysis systems ignore the temporal and spatial context in which the data are collected as most systems use data gathered during very short periods of observation the observer is not expected to record information about the physical setting.
- Interaction analysis systems are usually concerned only with overt, observable behaviour. In the case if intentions lay behind the direct behaviour an observer must himself impute the intention.
- Interaction analysis systems are concerned with what can be categorized or measured (Simon and Boyer 1986:1). They may obscure, distort or ignore the qualitative features which they claim to investigate, by having ill-defined boundaries between the categories.
- Interaction analysis systems focus on small bits of action or behaviour rather than global concepts (Simon and Boyer 1986:1). Delamont and Hamilton clarifies that there is a tendency to generate a superabundance of data which must be linked either to the complex set of descriptive concepts or to a small number of global concepts.
- The systems utilize pre-specified categories.
- Placing arbitrary boundaries on continuous phenomena obscures the flux of social interaction.
Walker and Adelman (1976: 136) emphasize the problems of recording child-child talk and objectivity of incorporating this kind of talk into the normal flow of teacher-centred classroom. They illustrate that there is no research instrument to code the spontaneous talk or social function of jokes and humour. Talk is seen to be a highly complex, problematic activity, rich in contradictory and bizarre meanings and frequently with difficulties and confusions (Walker and Adelman 1976: 137). This organisation is taken for granted in observation schemes.
McKernan (1996:118) reviews various styles of rating scales category, numerical, graphic and pictorial. They all share the common feature of having a rater place an object, person or idea along a sequential scale in terms of estimated value to the rater. Rating scales are treated as helpful instrument to measure non-cognitive areas where an observer is interested in cooperativeness, industriousness, tolerance, enthusiasm, group skills. At the same time McKernan (1996:119) notes that all rating sheets need to
- include observable behavior;
- rate significant outcomes as opposed to m