Ethical issues in interviewing

17 Accounts

Introduction

The rationale of much of this chapter is located in the interpretive, ethnographic paradigm which strives to view situations through the eyes of participants, to catch their intentionality and their interpretations of frequently complex situations, their meaning systems and the dynamics of the interaction as it unfolds. This is akin to the notion of ‘thick description’ from Geertz (1973b). This chapter proceeds in several stages: first, we set out the characteristics of the ethogenic approach; second, we set out procedures in eliciting, analysing and authenticating accounts; third, we provide an introduction to handling qualitative accounts and their related fields of network analysis and discourse analysis; fourth, we provide an introduction to handling quantitative and qualitative accounts; finally, we review the strengths and weaknesses of ethogenic approaches. We recognize that the field of language and language use is vast, and to try to do justice to it here is the ‘optimism of ignorance’ (Edwards 1976). Rather, we attempt to indicate some important ways in which researchers can use accounts in collecting data for their research.

Although each of us sees the world from our own point of view, we have a way of speaking about our experiences which we share with those around us. Explaining our behaviour towards one another can be thought of as accounting for our actions in order to make them intelligible and justifiable to our fellowmen. Thus, saying ‘I’m terribly sorry,

I didn’t mean to bump into you’, is a simple case of the explication of social meaning, for by locating the bump outside any planned sequence and neutralizing it by making it intelligible in such

a way that it is not warrantable, it ceases to be offensive in that situation (Harr´e 1978). Accounting for actions in those larger slices of life called social episodes is the central concern of a participatory psychology which focuses upon actors’ intentions, their beliefs about what sorts of behaviour will enable them to reach their goals, and their awareness of the rules that govern those behaviours. Studies carried out within this framework have been termed ‘ethogenic’, an adjective which expresses a view of the human being as a person, that is, a plan- making, self-monitoring agent, aware of goals and deliberately considering the best ways to achieve them. Ethogenic studies represent another approach to the study of social behaviour and their methods stand in bold contrast to those commonly employed in much of the educational research which we describe in Chapter 1. Before discussing the elicitation and analysis of accounts we need to outline the ethogenic approach in more detail. This we do by reference to the work of one of its foremost exponents, Rom Harr´e (1974; 1976; 1977a; 1977b; 1978).

The ethogenic approach

Harr´e (1978) identifies five main principles in the ethogenic approach. They are set out in Box 17.1.

Characteristics of accounts and episodes

The discussion of accounts and episodes that now follows develops some of the ideas contained in the principles of the ethogenic approach outlined above.

We have already noted that accounts must be seen within the context of social episodes. The

PROCEDURES IN ELICITING, ANALYSING AND AUTHENTICATING ACCOUNTS: AN EXAMPLE 385

Chapter

Box 17.1

Principles in the ethogenic approach

An explicit distinction is drawn between synchronic analysis, that is, the analysis of social practices and institutions as they exist at any one time, and diachronic analysis, the study of the stages and the processes by which social practices and

institutions are created and abandoned, change and are changed. Neither type of analysis can be expected to lead directly to the discovery of universal social psychological principles or laws.

In social interactions, it is assumed that action takes place through endowing intersubjective entities with meaning; the ethogenic approach therefore concentrates upon the meaning system, that is, the whole sequence by which a social act is achieved in an episode. Consider, for example, the action of a kiss in the particular episodes of leaving a friend’s house, the passing-out parade at St Cyr and the meeting in the garden of Gethsemane.

The ethogenic approach is concerned with speech which accompanies action. That speech is intended to make the action intelligible and justifiable in occurring at the time and the place it did in the whole sequence of unfolding and coordinated action. Such speech is accounting. In so far as accounts are socially meaningful, it is possible to derive accounts of accounts.

The ethogenic approach is founded upon the belief that human beings tend to be the kind of person their language, their traditions, their tacit and explicit knowledge tell them they are.

The skills that are employed in ethogenic studies therefore make use of commonsense understandings of the social world. As such the activities of the poet and the playwright offer the ethogenic researcher a better model than those of the physical scientist.

Source: adapted from Harr´e 1978

idea of an episode is a fairly general one. The Providing that accounts are authentic, it is concept itself may be defined as any coherent

argued, there is no reason why they should fragment of social life. Being a natural division

not be used as scientific tools in explaining of life, an episode will often have a recognizable

people’s actions.

beginning and end, and the sequence of actions that constitute it will have some meaning for the participants. Episodes may thus vary in duration

Procedures in eliciting, analysing and

and reflect innumerable aspects of life. A student

authenticating accounts: an example

entering primary school aged 7 and leaving at 11 The account-gathering method proposed by would be an extended episode. A two-minute

Brown and Sime (1977) is summarized in Box 17.2. television interview with a political celebrity

It involves attention to informants, the account- would be another. The contents of an episode

gathering situation, the transformation of accounts which interest the ethogenic researcher include

and researchers’ accounts, and sets out control not only the perceived behaviour such as gesture

procedures for each of these elements. and speech, but also the thoughts, the feelings

Problems of eliciting, analysing and authen- and the intentions of those taking part. And

ticating accounts are further illustrated in the the ‘speech’ that accounts for those thoughts,

following outlines of two educational studies. The feelings and intentions must be conceived of

first is concerned with valuing among older boys in the widest connotation of the word. Thus,

and girls; the second is to do with the activities of accounts may be personal records of the events

pupils and teachers in using computers in primary we experience in our day-to-day lives, our

classrooms.

conversations with neighbours, our letters to Kitwood (1977) developed an experience- friends, our entries in diaries. Accounts serve

sampling method, that is, a qualitative technique to explain our past, present and future oriented

for gathering and analysing accounts based upon actions.

tape-recorded interviews that were themselves

386 ACCOUNTS

Box 17.2

Account gathering Research strategy

Control procedure

Informants

Definition of episode and role groups O Rationale for choice of episode and role groups representing domain of interest O Degree of involvement of potential informants

Identification of exemplars O Contact with individuals to establish motive for

Selection of individual informants participation, competence and performance

Account gathering situation

Establishing venue

Contextual effects of venue

Recording the account

Appropriateness and accuracy in documenting

Controlling relevance of account

account

Authenticating account

Accounts agenda

Establishing role of interviewer and

Negotiation and internal consistency interviewee

Degree of direction

Post account authentication

Corroboration

Transformation of accounts

Provision of working documents

Transcription reliability; coder reliability

Appropriateness of statistical and content analyses Researchers’ accounts

Data reduction techniques

Account of the account: summary, overview,

Description of research operations, explanatory interpretation

scheme and theoretical background

Source: Brown and Sime 1981: 163

prompted by the fifteen situations listed in

The total pattern of choice: the frequency of Box 17.3.

choice of various items permits some surface Because the experience-sampling method

generalizations about the participants, taken as avoids interrogation, the material which emerges

a group. The most revealing analyses may be is less organized than that obtained from a

those of the least and most popular items. tightly structured interview. Successful handling

Similarities and differences: using the same of individual accounts therefore requires the

technique as in the first method, it is possible to researcher to know the interview content

investigate similarities and differences within extremely well and to work toward the gradual

the total sample of accounts according to some emergence of tentative interpretive schemata

characteristic(s) of the participants such as age, which he then modifies, confirms or falsifies

sex, level of educational attainment, etc. as the research continues. Kitwood identifies

Grouping items together: it may be convenient eight methods for dealing with the tape-recorded

for some purposes to fuse together categories accounts. The first four methods are fairly close to

that cover similar subject matter. For example, the approach adopted in handling questionnaires,

items 1, 5 and 14 in Box 17.3 relate to conflict; and the rest are more in tune with the ethogenic

items 4, 7 and 15 relate to personal growth and principles that we identified earlier:

change.

PROCEDURES IN ELICITING, ANALYSING AND AUTHENTICATING ACCOUNTS: AN EXAMPLE 387

Chapter

Box 17.3

Experience sampling method Below are listed fifteen types of situation which most people have been in at some time. Try to think of something that

has happened in your life in the past year or so, or perhaps something that keeps on happening, which fits into each of the descriptions. Then choose the ten of them which deal with the things that seem to you to be most important, which

cover your main interests and concerns, and the different parts of your life. When we meet we will talk together about the situations you have chosen. Try beforehand to remember as clearly as you can what happened, what you and others did, and how you yourself felt and thought. Be as definite as you can. If you like, write a few notes to help you keep the situation in mind.

1 When there was a misunderstanding between you and someone else (or several others) . . . 2 When you got on really well with people . . . 3 When you had to make an important decision . . . 4 When you discovered something new about yourself . . . 5 When you felt angry, annoyed or resentful . . . 6 When you did what was expected of you . . . 7 When your life changed direction in some way . . . 8 When you felt you had done something well . . . 9 When you were right on your own, with hardly anyone taking your side . . . 10 When you ‘got away with it’, or were not found out . . . 11 When you made a serious mistake . . . 12 When you felt afterwards that you had done right . . . 13 When you were disappointed with yourself . . . 14 When you had a serious clash or disagreement with another person . . . 15 When you began to take seriously something that had not mattered much to you before . . .

Source: adapted from Kitwood 1977

Categorization of content: the content of a

The study of omissions: the researcher particular item is inspected for the total sample

may well have expectations about the kind and an attempt is then made to develop some

of issues likely to occur in the interviews. categories into which all the material will

When some of these are absent, that fact fit. The analysis is most effective when two

may be highly significant. The absence of or more researchers work in collaboration,

an anticipated topic should be explored each initially proposing a category system

to discover the correct explanation of its independently and then exchanging views to

omission.

negotiate a final category system.

Reconstruction of a social life-world: this

Tracing a theme: this type of analysis transcends method can be applied to the accounts of the rather artificial boundaries which the items

a number of people who have part of their themselves imply. It aims to collect as much

lives in common, for example, a group of data as possible relevant to a particular topic

friends who go around together. The aim is regardless of where it occurs in the interview

to attempt some kind of reconstruction of material. The method is exacting because it

the world which the participants share in requires very detailed knowledge of content

analysing the fragmentary material obtained and may entail going through taped interviews

in an interview. The researcher seeks to several times. Data so collected may be further

understand the dominant modes of orienting analysed along the lines suggested in the fourth

to reality, the conceptions of purpose and the method above.

limits to what is perceived.

388 ACCOUNTS

Generating and testing hypotheses: new hypothe- tests arose when interpretations of interview data ses may occur to the researcher during the

were taken back to participating teachers for their analysis of the tape-recordings. It is possible to

comments. Similarly, pupils’ scores on certain self- do more than simply advance these as a result

concept scales were discussed individually with of tentative impressions; one can loosely apply

respondents in order to ascertain why children the hypothetico-deductive method to the data.

awarded themselves high or low marks in respect This involves putting the hypothesis forward as

of a range of skills in using computer programs. clearly as possible, working out what the verifi- able inferences from it would logically be, and testing these against the account data. Where

Network analyses of qualitative data

these data are too fragmentary, the researcher Another technique that has been successfully may then consider what kind of evidence and

employed in the analysis of qualitative data method of obtaining it would be necessary for

is described by its originators as ‘systematic more thorough hypothesis testing. Subsequent

network analysis’ (Bliss et al. 1983). Drawing upon sets of interviews forming part of the same

developments in artificial intelligence, Bliss and piece of research might then be used to obtain

her colleagues employed the concept of ‘relational relevant data.

network’ to represent the content and structuring of a person’s knowledge of a particular domain.

In the light of the weaknesses in account Essentially, network analysis involves the gathering and analysis (discussed later), Kitwood’s

development of an elaborate system of categories (1977) suggestions of safeguards are worth

by way of classifying qualitative data and mentioning. First, he calls for cross-checking

preserving the essential complexity and subtlety between researchers as a precaution against

of the materials under investigation. A notational consistent but unrecognized bias in the interviews

technique is employed to generate network-like themselves. Second, he recommends member

structures that show the interdependencies of the tests, that is, taking hypotheses and unresolved

categories as they are developed. Network mapping problems back to the participants themselves or

is akin to cognitive mapping, 1 an example of which to people in similar situations to them for their

can be seen in the work of Bliss et al. (1983). comments. Only in this way can researchers be sure that they understand the participants’ own grounds for action. Since there is always the possibility that

What makes a good network?

an obliging participant will readily confirm the Bliss et al. (1983) point out that there cannot be researcher’s own speculations, every effort should

one overall account of criteria for judging the

be made to convey to the participant that one merits of a particular network. They do, however, wants to know the truth as he or she sees it, and

attempt to identify a number of factors that ought that one is as glad to be proved wrong as right.

to feature in any discussion of the standards

A study by Blease and Cohen (1990) used cross- by which a network might fairly be judged as checking as a way of validating the classroom

adequate.

observation records of co-researchers, and member First, any system of description needs to be valid tests to authenticate both quantitative and

and reliable: valid in the sense that it is appropriate qualitative data derived from teacher and pupil

in kind and, within that kind, sufficiently complete informants. Thus, in the case of cross-checking,

and faithful, reliable in the sense that there exists the classroom observation schedules of research

an acceptable level of agreement between people assistants and researchers were compared and

as to how to use the network system to describe discussed, to arrive at definitive accounts of the

data.

range and duration of specific computer activities Second, there are properties that a network occurring within observation sessions. Member

description should possess such as clarity,

DISCOURSE ANALYSIS 389

Chapter

completeness and self-consistency. These relate book. Habermas argues that utterances are never to a further criterion of ‘network utility’, the

simply sentences (Habermas 1970: 368) that sufficiency of detail contained in a particular

are disembodied from context, but, rather, their network. A third property that a network should

meaning derives from the intersubjective contexts possess is termed ‘learnability’. Communicating

in which they are set. A speech situation has

the terms of the analysis to others, say the authors,

a double structure, the propositional content is of central importance. It follows therefore that

(the locutionary aspect – what is being said) and much hinges on whether networks are relatively

the performatory content (the illocutionary and easy or hard to teach to others. A fourth aspect

perlocutionary aspect – what is being done or of network acceptability has to do with its

achieved through the utterance). For Habermas ‘testability’. Bliss et al. (1983) identify two forms

(1979; 1984) each utterance has to abide by the of testability, the first having to do with testing a

criteria of legitimacy, truth, rightness, sincerity network as a ‘theory’ against data, the second with

and comprehensibility. His concept of the ‘ideal testing data against a ‘theory’ or expectation via

speech situation’ argues that speech – and, for

a network. our purposes here – discourse, should seek to be Finally, the terms ‘expressiveness’ and ‘persua-

empowering and not subject to repression or siveness’ refer to qualities of language used in

ideological distortion. His ideal speech situation developing the network structure. And here, the

is governed by several principles, not the least authors proffer the following advice: ‘Helpful as

of which are mutual understanding between the choice of an expressive coding mood or neat

participants, freedom to enter a discourse, an use of indentation or brackets may be, the code

equal opportunity to use speech acts, discussion to actually says no more than the network distinguishes’

be free from domination, the movement towards (Bliss et al. 1983, our italics).

consensus resulting from the discussion alone and To conclude, network analysis would seem to

the force of the argument alone (rather than have a useful role to play in educational research

the position power of speakers). For Habermas, by providing a technique for dealing with the

then, discourse analysis would seek to uncover, bulk and the complexity of the accounts that are

through ideology critique (see Chapter 1), the typically generated in qualitative studies.

repressive forces which ‘systematically distort’ communication. For our purposes, we can take from Habermas the need to expose and interrogate

Discourse analysis

the dominatory influences that thread not only Discourse researchers explore the organization

through the discourses which researchers are of ordinary talk and everyday explanations

studying, but also the discourses that the research and the social actions performed in them.

itself produces.

Collecting, transcribing and analysing discourse Various developments in discourse analysis data constitutes a kind of psychological ‘natural

have made important contributions to our history’ of the phenomena in which discourse

understanding of children’s thinking, challenging analysts are interested (Edwards and Potter 1993).

views (still common in educational circles) of Discourses can be regarded as sets of linguistic

‘the child as a lone organism, constructing a material that are coherent in organization and

succession of general models of the world as content and enable people to construct meaning in

each new stage is mastered’ (Edwards 1991). social contexts. The emphasis on the construction

Rather than treating children’s language as of meaning indicates the action perspective of

representative of an inner cognitive world to discourse analysis (Coyle 1995: 245).

be explored experimentally by controlling for a Further, the focus on discourse and speech

host of intruding variables, discourse analysts treat acts links this style of research to Habermas’s

that language as action, as ‘situated discursive critical theory set out at the start of this

practice’.

390 ACCOUNTS

By way of example, Edwards (1993) explores need for the researcher to be highly sensitive discourse data emanating from a visit to a

to the nuances of language (Coyle 1995: 247). greenhouse by 5-year-old pupils and their teacher,

In discourse analysis, as in qualitative data to see plants being propagated and grown. His

analysis generally (Miles and Huberman 1984), analysis shows how children take understandings

the researcher can use coding at an early stage of of adults’ meanings from the words they hear and

analysis, assigning codes to the textual material the situations in which those words are used. And

being studied (Parker 1992; Potter and Wetherall in turn, adults (in this case, the teacher) take

1987). This enables the researcher to discover from pupils’ talk, not only what they might mean

patterns and broad areas in the discourse. With but also what they could and should mean. What

this achieved the researcher can then re-examine Edwards describes as ‘the discursive appropriation

the text to discover intentions, functions and of ideas’ (Edwards 1991) is illustrated in Box 17.4.

consequences of the discourse (examining the Discourse analysis requires a careful reading

speech act functions of the discourse, e.g. to impart and interpretation of textual material, with

information, to persuade, to accuse, to censure, to interpretation being supported by the linguistic

encourage etc.). By seeking alternative explanations evidence. The inferential and interactional aspects

and the degree of variability in the discourse, it is of discourse and discourse analysis suggest the

possible to rule out rival interpretations and arrive

Box 17.4

Concepts in children’s talk 81 Sally

Cuttings can grow to plants.

82 Teacher

[writing] ‘Cuttings can grow–,’ instead of saying ‘to

to plants.’

84 Sally

= You wrote Chris

tina.

85 Teacher

Oops. Thank you. I’ll do this again. ‘Cuttings can 86 grow into plants’. That’s also good. What is a cutting, 87 Christina?

88 Christina

A cutting is, umm, I don’t know.

89 Teacher

Who knows what a cutting is besides Sally? Sam.

90 Sam

It’s when you cut off a–, it’s when you cut off a piece 91 of a plant.

92 Teacher Exactly, and when you cut off a piece of a plant, what do 93 you then do with it to make it grow? If you leave

95 X Put it in soil.

96 Teacher

Well, sometimes you can put it in soil.

97 Y 98 Teacher

wait, what else could you put it in? 99 Sam

But what –,

Put it in a pot?

100 Teacher

Pot, with soil, or. . . ? There’s another way.

101 Sally

I know another way. =

102 Teacher

= Wait. Sam, do you know? No? =

103 Sam

= Dirt.

104 Teacher No, it doesn’t have to do with s –, it’s not a solid, it’s 105 106

Source: Edwards 1993

ACCOUNT GATHERING IN EDUCATIONAL RESEARCH: AN EXAMPLE 391

Chapter

at a fair reading of what was actually taking place is usually undertaken by researchers. It is they, in the discourse in its social context.

generally, who read ‘between the lines’ and The application of discourse analysis to our

‘within the gaps’ of classroom talk by way of understanding of classroom learning processes is

interpreting the intentionality of the participating well exemplified in a study by Edwards and Mercer

discussants (O’Neill and McMahon 1990).

(1987). Rather than taking the classroom talk as evidence of children’s thought processes, the researchers explore it as

Analysing social episodes

A major problem in the investigation of discourse itself is the educational reality and the issue

contextualized dialogue with the teacher. The

that natural unit of social behaviour, the ‘social episode’, has been the ambiguity that

becomes that of examining how teacher and children surrounds the concept itself and the lack of an construct a shared account, a common interpretative

framework for curriculum knowledge and for what acceptable taxonomy by which to classify an

interaction sequence on the basis of empirically happens in the classroom. quantifiable characteristics. Several quantitative

studies have been undertaken in this field. For Overriding asymmetries between teachers and

(Edwards 1991)

example, McQuitty (1957), Magnusson (1971) pupils, Edwards (1991) concludes, both cognitive

and Ekehammer and Magnusson (1973) use (in terms of knowledge) and interactive (in terms

factor analysis and linkage analysis respectively, of power), impose different discursive patterns and

while Peevers and Secord (1973), Secord and functions. Indeed Edwards (1980) suggests that

Peevers (1974) and Forgas (1976; 1978) use teachers control classroom talk very effectively,

multidimensional scaling and cluster analysis. reproducing asymmetries of power in the classroom by telling the students when to talk, what to talk

Account gathering in educational

about, and how well they have talked. Discourse analysis has been criticized for its

research: an example

lack of systematicity (Coyle 1995: 256), for its The ‘free commentary’ method that Secord and emphasis on the linguistic construction of a social

Peevers (1974) recommend as a way of probing for reality, and the impact of the analysis in shifting

explanations of people’s behaviour lies at the very attention away from what is being analysed and

heart of the ethnographer’s skills. In the example towards the analysis itself, i.e. the risk of losing the

of ethnographic research that now follows, one independence of phenomena. Discourse analysis

can detect the attempt of the researcher to get risks reifying discourse. One must not lose sight

below the surface data and to search for the of the fact that the discourse analysis itself is a

deeper, hidden patterns that are only revealed text, a discourse that in turn can be analysed for

when attention is directed to the ways that group its meaning and inferences, rendering the need for

members interpret the flow of events in their reflexivity to be high (Ashmore 1989). 2 lives.

Edwards and Westgate (1987) show what substantial strides have been made in recent years in the development of approaches to

Heath: ‘Questioning at home and at school’

the investigation of classroom dialogue. Some Heath’s (1982) study of misunderstandings methods encourage participants to talk; others

existing between black children and their white wait for talk to emerge and sophisticated

teachers in classrooms in the south of the United audio/video techniques record the result by

States brought to light teachers’ assumptions whatever method it is achieved. Thus captured,

that pupils would respond to language routines dialogue is reviewed, discussed and reflected upon;

and the uses of language in building knowledge moreover, that reviewing, discussing and reflecting

and skills just as other children (including their

392 ACCOUNTS

own) did. Specifically, Heath (1982) sought to In particular, she focused upon the ways in which understand why these particular children did

‘the children learned to use language to satisfy not respond just as others did. Her research

their needs, ask questions, transmit information, involved eliciting explanations from both the

and convince those around them that they were children’s parents and teachers. ‘We don’t talk

competent communicators’ (Heath 1982). This to our children like you folks do’, the parents

involved her in a much wider and more intensive observed when questioned about their children’s

study of the total fabric of life in Trackton, the behaviour. Those children, it seemed to Heath,

southern community in which the research was were not regarded as information givers or as

located. She comments that she was able to appropriate conversational partners for adults.

collect data from a wide range of contexts and That is not to say that the children were

situations, tracking children longitudinally and in excluded from language participation. They did,

several contexts, taking care to record language in fact, participate in a language that Heath

used and the social contexts of the language, and describes as rich in styles, speakers and topics.

the communicative competence of participants. 3 Rather, it seemed to the researcher that the teachers’ characteristic mode of questioning was ‘to pull attributes of things out of context,

Problems in gathering and analysing

particularly out of the context of books and name

accounts

them – queens, elves, police, red apples’ (Heath The importance of the meaning of events and 1982). The parents did not ask these kinds of

actions to those who are involved in them is questions of their children, and the children

now generally recognized in social research. The themselves had their own ways of deflecting

implications of the ethogenic stance in terms such questions, as the example in Box 17.5 well

of actual research techniques, however, remain illustrates.

problematic. Menzel (1978) discusses a number of Heath (1982) elicited both parents’ and

ambiguities and shortcomings in the ethogenic teachers’ accounts of the children’s behaviour

approach, arising out of the multiplicity of and their apparent communication ‘problems’ (see

meanings that may be held for the same behaviour. Box 17.6). Her account of accounts arose out

Most behaviour, Menzel (1978) observes, can be of periods of participation and observation in

assigned meanings and more than one of these may classrooms and in some of the teachers’ homes.

very well be valid simultaneously. It is fallacious

Box 17.5

‘Ain’t nobody can talk about things being about theirselves’ This comment by a 9-year-old boy was directed to his teacher when she persisted in interrogating him about the story he

had just completed in his reading group. Teacher:

What is the story about? Children:

(silence) Teacher:

Uh . . . Let’s . . . Who is it the story talks about? Children:

(silence) Teacher:

Who is the main character? . . . What kind of story is it? Child:

Ain’t nobody can talk about things being about theirselves. The boy was saying ‘There’s no way anybody can talk (and ask) about things being about themselves’.

Source: adapted from Heath 1982

STRENGTHS OF THE ETHOGENIC APPROACH 393

Chapter

Box 17.6

Parents and teachers: divergent viewpoints on children’s communicative competence

Parents The teachers won’t listen. My kid, he too scared to talk, ’cause nobody play by the rules he know. At home, I can’t shut ’im up.

Miss Davis, she complain ’bout Ned not answerin’ back. He say she asks dumb questions she already know ’bout. Teachers

They don’t seem to be able to answer even the simplest questions. I would almost think some of them have a hearing problem; it is as though they don’t hear me ask a question. I get blank

stares to my questions. Yet when I am making statements or telling stories which interest them, they always seem to hear me.

The simplest questions are the ones they can’t answer in the classroom; yet on the playground, they can explain a rule for a ballgame or describe a particular kind of bait with no problem. Therefore, I know they can’t be as dumb as they seem in my class.

I sometimes feel that when I look at them and ask a question I’m staring at a wall I can’t break through. There’s something there; yet in spite of all the questions I ask, I’m never sure I’ve gotten through to what’s inside that wall.

Source: adapted from Heath 1982

therefore, he argues, to insist upon determining Menzel’s (1978) plea is for the usefulness of ‘the’ meaning of an act. Nor can it be said that the

an outside observer’s account of a social episode task of interpreting an act is done when one has

alongside the explanations that participants identified one meaning of it, or the one meaning

themselves may give of that event. A similar that the researcher is pleased to designate as the

argument is implicit in McIntyre and Macleod’s true one.

(1978) justification of objective, systematic

A second problem that Menzel (1978) raises observation in classroom settings. Their case is is to do with actors’ meanings as sources of bias.

set out in Box 17.7.

How central a place, he asks, ought to be given to actors’ meanings in formulating explanations of events? Should the researcher exclusively and

Strengths of the ethogenic approach

invariably be guided by these considerations? To The advantages of the ethogenic approach to do so would be to ignore a whole range of potential

the educational researcher lie in the distinctive explanations which few researchers would wish to

insights that are made available to her through the see excluded from consideration.

analysis of accounts of social episodes. The benefits These are far-reaching, difficult issues though

to be derived from the exploration of accounts are by no means intractable. What solutions does

best seen by contrasting the ethogenic approach Menzel (1978) propose? First, we must specify ‘to

with a more traditional educational technique whom’ when asking what acts and situations mean.

such as the survey that we discussed in Chapter 9. Second, researchers must make choices and take

There is a good deal of truth in the assertion responsibility in the assignment of meanings to

of the ethogenically oriented researcher that acts; moreover, problem formulations must respect

approaches that employ survey techniques such the meaning of the act to us, the researchers. Third,

as the questionnaire take for granted the very explanations should respect the meanings of acts

things that should be treated as problematic in to the actors themselves but need not invariably

an educational study. Too often, the phenomena

be centred around these meanings. that ought to be the focus of attention are taken

394 ACCOUNTS

Box 17.7

Justification of objective systematic observation in classroom settings

When Smith looks at Jones and says, ‘Jones, why does the blue substance spread through the liquid?’ (probably with a particular kind of voice inflection), and then silently looks at Jones (probably with a particular kind of facial expression), the observer can unambiguously categorize the event as ‘Smith asks Jones a question seeking an explanation of diffusion in a liquid.’ Now Smith might describe the event as ‘giving Jones a chance to show he knows something’, and Jones might describe the event as ‘Smith trying to get at me’; but if either of them denied the validity of the observer’s description, they would

be simply wrong, because the observer would be describing at least part of what the behaviour which occurred means in English in Britain. No assumptions are made here about the effectiveness of classroom communication; but the assumption is made that . . . communication is dependent on the system of conventional meanings available within the wider culture. More fundamentally, this interpretation implies that the systematic observer is concerned with an objective reality (or, if one prefers, a shared intersubjective reality) of classroom events. This is not to suggest that the subjective meanings of events to participants are not important, but only that these are not accessible to the observer and that there is an objective reality to classroom activity which does not depend on these meanings [our emphasis].

Source: McIntyre and Macleod 1978

as given, that is, they are treated as the starting point of the research rather than becoming the centre of the researcher’s interest and effort to discover how the phenomena arose or came to be important in the first place. Numerous educational studies, for example, have identified the incidence and the duration of disciplinary infractions in school; only relatively recently, however, has the meaning of classroom disorder, as opposed to its frequency and type, been subjected to intensive investigation. Unlike the survey, which is a cross-sectional technique that takes its data at a single point in time, the ethogenic study employs an ongoing observational approach that focuses upon processes rather than products. Thus it is the process of becoming deviant in school which would capture the attention of the ethogenic researcher rather than the frequency and type of misbehaviour among k types of ability in children located in n kinds of school.

A note on stories

A comparatively neglected area in educational research is the field of stories and storytelling. Bauman (1986: 3) suggests that stories are oral literature whose meanings, forms and functions are situationally rooted in cultural contexts,

scenes and events which give meaning to action. This recalls Bruner (1986) who, echoing the interpretive mode of educational research, regards much action as ‘storied text’, with actors making meaning of their situations through narrative. Stories have a legitimate place as an inquiry method in educational research (Parsons and Lyons 1979) and, indeed, Jones (1990), Crow (1992), Dunning (1993) and Thody (1997) place them on a par with interviews as sources of evidence for research. Thody (1997: 331) suggests that, as an extension to interviews, stories – like biographies – are rich in authentic, live data; they are, she avers, an ‘unparalleled method of reaching practitioners’ mindsets’. Thody (1997: 333–4) provides a fascinating report on stories as data sources for educational management research as well as for gathering data from young children.

Thody (1997: 331) indicates how stories can

be analysed, using, for example, conventional techniques such as categorizing and coding of content; thematization; concept building. In this respect stories have their place alongside other sources of primary and secondary documentary evidence (e.g. case studies, biographies). They can be used in ex post facto research, historical research, as accounts or in action research; in short they are part of the everyday battery of research

A NOTE ON STORIES 395

Chapter

instruments that are available to the researcher. is difficult (unless other people were present to The rise in the use of oral history as a legitimate

verify events reported), stories, being rich in the research technique in social research can be seen

subjective involvement of the storyteller, offer an here to apply to educational research. Although

opportunity for the researcher to gather authentic, they might be problematic in that verification

rich and ‘respectable’ data (Bauman 1986).

18 Observation

Introduction

The distinctive feature of observation as a re- search process is that it offers an investigator the opportunity to gather ‘live’ data from naturally occurring social situations. In this way, the re- searcher can look directly at what is taking place in situ rather than relying on second-hand ac- counts. The use of immediate awareness, or direct cognition, as a principal mode of research thus has the potential to yield more valid or authentic data than would otherwise be the case with mediated or inferential methods. And this is observation’s unique strength. There are other attractions in its favour: as Robson (2002: 310) says, what people do may differ from what they say they do, and observation provides a reality check; observation also enables a researcher to look afresh at every- day behaviour that otherwise might be taken for granted, expected or go unnoticed (Cooper and Schindler 2001: 374); and the approach with its carefully prepared recording schedules avoids prob- lems caused when there is a time gap between the act of observation and the recording of the event – selective or faulty memory, for example. Finally, on a procedural point, some participants may prefer the presence of an observer to an intru- sive, time-consuming interview or questionnaire.

Observation can be of facts, such as the number of books in a classroom, the number of students in

a class, the number of students who visit the school library in a given period. It can also focus on events as they happen in a classroom, for example, the amount of teacher and student talk, the amount of off-task conversation and the amount of group collaborative work. Further, it can focus on behaviours or qualities, such as the friendliness of the teacher, the degree of aggressive

behaviour or the extent of unsociable behaviour among students.

One can detect here a putative continuum from the observation of uncontestable facts to the researcher’s interpretation and judgement of situations, which are then recorded as observations. What counts as evidence becomes cloudy immediately in observation, because what we observe depends on when, where and for how long we look, how many observers there are, and how we look. It also depends on what is taken to be evidence of, or a proxy for, an underlying, latent construct. What counts as acceptable evidence of unsociable behaviour in the example above requires an operational definition that is valid and reliable. Observers need to decide ‘of what is the observation evidence’, for example: is the degree of wear and tear on a book in the school library an indication of its popularity, or carelessness by its readers, or of destructive behaviour by students? One cannot infer cause from effect, intention from observation, stimulus from response.

Observational data are sensitive to contexts and demonstrate strong ecological validity (Moyles 2002). This enables researchers to understand the context of programmes, to be open-ended and inductive, to see things that might otherwise

be unconsciously missed, to discover things that participants might not freely talk about in interview situations, to move beyond perception- based data (e.g. opinions in interviews) and to access personal knowledge. Because observed incidents are less predictable there is a certain freshness to this form of data collection that is often denied in other forms, e.g. a questionnaire or a test.

Observations (Morrison 1993: 80) enable the researcher to gather data on:

INTRODUCTION 397

Chapter

the physical setting (e.g. the physical environ- ment and its organization)

the human setting (e.g. the organization of people, the characteristics and make up of the groups or individuals being observed, for instance, gender, class)

the interactional setting (e.g. the interactions that are taking place, formal, informal, planned, unplanned, verbal, non-verbal etc.)

the programme setting (e.g. the resources and their organization, pedagogic styles, curricula and their organization).

Additionally, observational data may be useful for recording non-verbal behaviour, behaviour in natural or contrived settings, and longitudinal analysis (Bailey 1994: 244). On the other hand, the lack of control in observing in natural settings may render observation less useful, coupled with difficulties in measurement, problems of small samples, difficulties of gaining access and negotiating entry, and difficulties in maintaining anonymity (Bailey 1994: 245–6). Observation can

be a powerful research tool, but it is not without its difficulties, and this chapter exposes and addresses these.

Patton (1990: 202) suggests that observational data should enable the researcher to enter and understand the situation that is being described. The kind of observations available to the researcher lie on a continuum from unstructured to structured, responsive to pre-ordinate. A highly structured observation will know in advance what it is looking for (i.e. pre-ordinate observation) and will have its observation categories worked out in advance. A semi-structured observation will have an agenda of issues but will gather data to illuminate these issues in a far less predetermined or systematic manner. An unstructured observation will be far less clear on what it is looking for and will therefore have to go into a situation and observe what is taking place before deciding on its significance for the research. In a nutshell,

a structured observation will already have its hypotheses decided and will use the observational data to conform or refute these hypotheses. On the other hand, a semi-structured and, more

particularly, an unstructured observation, will

be hypothesis-generating rather than hypothesis- testing. The semi-structured and unstructured observations will review observational data before suggesting an explanation for the phenomena being observed.

Though it is possible to argue that all research is some form of participant observation since we cannot study the world without being part of it (Adler and Adler 1994), nevertheless Gold (1958) offers a well-known classification of researcher roles in observation, that lie on a continuum. At one end is the complete participant, moving to the participant-as-observer, thence to the observer-as-participant, and finally to the complete observer (see http://www.routledge.com/textbooks/ 9780415368780 – Chapter 18, file 18.1.ppt). The move is from complete participation to complete detachment. The mid-points of this continuum strive to balance involvement with detachment, closeness with distance, familiarity with strangeness. The role of the complete observer is typified in the one-way mirror, the video-cassette, the audio-cassette and the photograph, while complete participation involves researchers taking on membership roles (overt or covert) (see http://www.routledge.com/textbooks/ 9780415368780 – Chapter 18, file 18.2.ppt).

Traditionally observation has been character- ized as non-interventionist (Adler and Adler 1994: 378), where researchers do not seek to manipulate the situation or subjects, they do not pose questions for the subjects, nor do they deliberately create ‘new provocations’ (Adler and Adler 1994: 378). Quantitative research tends to have a small field of focus, fragmenting the observed into minute chunks that can subsequently be aggregated into a variable. Qualitative research, on the other hand, draws the researcher into the phenomenological complexity of participants’ worlds; here situations unfold, and connections, causes and correlations can be observed as they occur over time. The qualitative researcher aims to catch the dynamic nature of events, to see intentionality, to seek trends and patterns over time.

If we know in advance what we wish to observe, i.e. if the observation is concerned to

398 OBSERVATION

chart the incidence, presence and frequency of elements and maybe wishes to compare one situation with another, then it may be more efficient in terms of time to go into a situation with a prepared observation schedule. If, on the other hand, we want to go into a situation and let the elements of the situation speak for themselves, perhaps with no concern with how one situation compares with another, then it may

be more appropriate to opt for a less structured observation. The former, structured observation, takes much time to prepare but the data analysis is fairly rapid, the categories having already been established, while the latter, less structured approach, is quicker to prepare but the data take much longer to analyse. The former approach operates within the agenda of the researcher and hence might neglect aspects of the four settings above if they do not appear on the observation schedule, i.e. it looks selectively at situations. On the other hand, the latter operates within the agenda of the participants, i.e. it is responsive to what it finds and therefore, by definition, is honest to the situation as it unfolds. Here selectivity derives from the situation rather than from the researcher in the sense that key issues emerge from the observation rather than the researcher knowing in advance what those key issues will be. Structured observation is useful for testing hypotheses, while unstructured observation provides a rich description of a situation which, in turn, can lead to the subsequent generation of hypotheses.

Flick (1998: 137) suggests that observation has to be considered along five dimensions: