backup: adfs::0.$.chap1

1st October 1993

 

 

Chapter 1 COMPUTERS IN EDUCATION: SOME ISSUES

 

 

From some point in the early 1980s, the microcomputer seems to have

become an increasingly conspicuous object: a potent symbol of "new

technology". In Britain, for example, 1980 was the year that Clive

Sinclair launched the first mass produced micro - the ZX80. Since

then, other manufacturers have marketed ever cheaper, smaller and more

powerful versions of this instrument. It has penetrated and occupied

the offices, shops, factories and spare bedrooms of our culture.

Moreover, within these various niches it supports a curious variety of

human activity: a situation that must be of some interest to social

scientists. Their research may help us appreciate just how this

pervasive technology is effecting our experiences of work, recreation

and social relations. Certainly, that is the very interest to be

explored in the present book. We shall consider one particular niche

occupied by computers - classrooms or, more generally, those places

where teaching and learning gets deliberately organized.

I shall concentrate on the British experience of this educational

innovation; although I am sure circumstances in many other countries

will reflect that experience quite closely (cf., Eraut, 1991; Gywnn,

1988; Plomp and Pelgrum, 1991). In the past ten years, rapid

evolution of the microcomputer has made it possible for British

schools to contemplate substantial investment in this technology.

Political pressure has encouraged them to do so. So, targeted

financial support (via the Department of Trade and Industry) allowed

schools to cope with the strain that this sudden financial commitment

entailed. Moreover, institutional structures were created to support

the curriculum development and staff training that would necessarily

follow. For example, the Microelectronics in Education Programme was

launched for schools and the Computers in Teaching Initiative for

Higher Education.

Ten years on, much of this priming activity has ended. Recent

National Curriculum documents indicate that pressure to assimilate new

technology has now been applied to most subject teaching. The

majority of teachers will have enjoyed some form of in-service

training for tackling the management and application of computers.

Most British classrooms will have reliable access to at least one

machine. So, at the time of writing, there is some sense of stability

in terms of staff development and infrastructure investment. This is

not to imply that the technology itself is not evolving: far from it,

recent developments (notably in multimedia) are impressive and hold a

special promise for educational applications. In any case, it seems a

good time to be taking stock of what has been achieved.

The first question an observer of this scene might ask is "Have

computers been any use to education?" Has this substantial commitment

allowed teachers and learners to reach their goals more efficiently,

more creatively, more agreeably - or however it is we want to express

progress?

In what follows I offer a summary of evaluative research concerned

with such questions. Evaluation defines the first of three central

"issues" arising from computer-based learning that will be reviewed in

this chapter. Discussion of the relevant research will be organized

under two headings below, but the discussion in each case will be

quite brief. This is because I am not sure that judging the outcomes

of disparate current practice is necessarily the only, or the most

urgent, kind of enterprise to be tackling. The situation is in a

great flux and we find a good deal of educational computing has an

improvised and volatile quality. Evaluation makes sense when we are

comfortable with our general aims as practitioners, and have done some

conceptual work to help understand how technology can relate to them.

Without a preliminary analysis of just what kind of educational

environment we want pupils to experience, formal evaluations of

particular computer-based ventures are unlikely to have very much

impact.

Such overarching concerns motivate a good deal of the discussion

in this book. The second of three broad issues to be considered in

the present chapter reviews prevailing models of what an educational

computer activity might "do" for a learner. I shall review these

models by reference to four metaphors that arise from them. Given my

declared interest in the social dimension of education, the

theoretical perspectives that emerge from this review will prove

somewhat discouraging: they do not express a clear place for processes

of social exchange.

This problem will be defined and explored as the final issue to be

addressed in this chapter. My remarks there set the scene for all the

discussion that then follows. Typically, computers are not regarded

as objects that contribute to the "social" quality of our lives. Yet,

I shall argue, effective educational environments are necessarily

rich in socially-organized experiences. The notion of learning as a

"collaborative" activity will be central to the present analysis. I

will be principally concerned to define how new technology can support

collaborations that can flourish within educational settings.

 

 

EVALUATION OF COMPUTER-BASED PRACTICE

Now, I would like to turn to the first issue referred to above: the

educational evaluation of computer-based practice. Because of my

concern with social psychological themes, it is natural to begin with

some observations focussing on the *people* caught up in this

innovation. In the first of the following two sections, I will

summarise difficulties associated with the reaction of practitioners

to the implementation of new technology in education. In the second

section, I shall summarize evaluative research on learning outcomes

associated with computer-based instruction where it has been

implemented.

 

 

Evaluation of implementation strategies

 

When we reflect on experience within schools and colleges over the

past ten years, it is inevitable that our judgements about progress

will be influenced by what was *expected* when investment and training

began. These expectations were often extravagant. As Maddux (1989)

observed in this context, pessimism is the familiar enemy of

innovators and, so, a degree of vigorous optimism was natural enough

in the early period of computer diffusion. Even commentary on the very

first, and most modest, examples of such educational intervention

could be fired with enthusiasm (the title of a review by Feldusen and

Szabo (1969) refers to computer-assisted instruction as the

'educational heart transplant'). A more recent judgement that is

often cited occurs in an article by Bork (1980). He comments:

We are at the outset of a major revolution in education, a revolution

unparalleled since the invention of the printing press. The computer

will be the instrument of this revolution...By the year 2000, the

major way of learning at all levels, and in almost all subject areas,

will be through the interactive use of computers (p. 53).

This prediction has more time to run but, in my view, it now looks to

have misjudged something significant in the relation between

education and new technology. It seems that the diffusion of this

technology has not been as dramatic as was expected in the early

period of microcomputer development. Certainly, within recent years,

a number of commentators (themselves sympathetic to computer-based

learning) have felt obliged to remark on the problems associated with

getting computers into active use within education (eg., Bliss,

Chandra and Cox, 1986; Collis, 1987; Cox, Rhodes and Hall, 1988;

Cuban, 1986; Hanson, 1985; Heywood and Norman, 1988; Holden, 1989;

Lepper and Gurtner, 1989; McCormick, 1992; Plomp, Pelgrum and

Steerneman, 1990).

Consequently, actual classroom usage remains limited. A recent

British government report suggests that only about 20% of teaching

time is making use of computers (DES, 1989). Similar limits on uptake

are be apparent in other countries (Dillon, 1985; Plomp and Pelgrum,

1991). Becker comments on the findings of one large scale U.S.

survey: '..in spite of the changes that computers have brought to

schools, only a small minority of teachers and students can be said to

yet be major computer users' (Becker, 1991).

What are the obstacles? It is natural to seek them within the

attitudes or strategies adopted by the teachers who manage this

technology - by looking in a focussed way at what is being done at the

classroom chalk-face. However, this would be too narrow a view.

McCormick (1992) characterises problems arising from a widespread

failure to develop a whole-school strategy. McInerney (1989), Plomp

et al (1990) and Wild (1991) identify a whole range of issues at the

institutional level that need to be confronted to make this innovation

work.

So, progress may depend, to an important extent, upon action organized

at the level of institutional practices. Research that is directed

more at the classroom level tends to dwell on teachers' lack of

self-assurance when using this technology. For example, Heywood and

Norman (1989) highlight obstacles to good practice arising from a

shortfall in (primary) teachers' confidence and their perceived

competence. In Britain at least, there was limited anticipation of

how difficult it might prove for staff unfamiliar with computers to

assimilate them into their practices. On reflection, the combination

of circumstances characterising many teachers first encounters with

this technology should have been fairly explosive. Early

configurations of classroom microtechnology were tedious and

time-consumming to prepare (often requiring the loading of programmes

off small audio cassettes). Educational software could be of very

dubious quality. All sorts of occasions were possible where the

computer would appear to fail - leaving the teacher exposed as having

lost control, (the children's more spontaneous enthusiasm having been

undermined in the process).

Politicians and educational administrators were sensitive to this

problem - if not to its scale. Certainly, some of the extra financial

support for priming this innovation was given over to staff

development. Many formal courses of in-service training (INSET) were

offered. Yet, the feeling often expressed within the profession (in

Britain, at least) is that it was not enough and, often, not of the

right character. It is now popular to challenge the faith of early

policy makers (eg. Fothergill, 1984) that in-service provision was the

quickest way to create an impact. A cascade model underpinned much of

this thinking: the hope was that those who received training on

intensive short courses would go back to their institution and pass on

their expertise. For one reason or another, if they gained any

expertise it looks as if they often kept it to themselves

(Boyd-Barrett, 1990).

The contemporary view is that a better strategy would have been to

concentrate more effort on initial training (Davis, 1992). We are

still in a situation where many new teachers remain awkwardly

unfamiliar with this technology - possibly seeing it only as

threatening (Bracey, 1988; Wellington, 1990). It remains true

that many teachers will have had only superficial pre-service exposure

to new technology; few will yet have enjoyed the experience of growing

up themselves within an established culture of computer use. At the

moment, the opportunities for teachers to gain confidence with new

technology across the period of initial training are often limited.

One survey in 1986 suggested that only 10% of students would use IT on

teaching practice (ITTE, 1987). The situation has improved recently,

although most students report they are still encountering the

technology as an isolated activity (Dunn and Ridgeway, 1991).

The urgency of this problem is hinted at by one extensive review of

the effectiveness of microcomputer work (in primary classrooms).

In a meta-analysis of recent evaluative research, Ryan (1991)

documented the effects of 40 variables on the impact of computer-based

learning experiences. Only one external variable was found to exert

any moderating effect of computer activity on pupil achievement: the

extent of teacher pre-training on the activity under study. This

draws attention to the fact that effective preparation involves more

than instilling the confidence to *motivate* implementation. The

success of computer-supported learning also depends upon teacher

contact with pedagogic ideas concerning good practice with this

technology: the enthusiastic teacher needs to be prepared in this

sense also.

Of course, any present initiative for acting at the point of initial

training is of little relevance to teachers already in post. Thus,

attention to the format of INSET experiences remains important. At

present, there is evidence that these experiences are not always

ideal. The problems are not merely a limited cascade of expertise:

the experience of course participants themselves is often one of

disappointment. The problems are illustrated in an extensive study of

in-service provision carried out by Rhodes and Cox (1990a,b). They

were able to witness the management of training programmes and to

visit the schools of staff who had attended them. So, they could

observe classroom practice as well as interview participating

teachers. A somewhat gloomy picture emerges from this comprehensive

survey. The training regimes did not appear well matched to the

experience or needs of these teachers. Consequently, they were not as

effective as the tutors had expected or hoped. Much of what was

achieved related to problems of using the technology itself, at the

expense of tackling real educational issues. Half of the sample of

teachers believed the computer resulted in an increase in their

workloads and that it made no fundamental change to the way in which

they worked - merely reinforcing existing patterns of activity.

This snapshot of practitioner experience is sobering - particularly to

those of us researchers whose (possibly selective) contact with

classroom practice may create a rosey picture of innovative

possibilities. The situation is well summarised in one survey that

revealed only 14% of primary school teachers felt competent to use a

range of IT applications without assistance (Davis, 1992). Yet how

might the general picture be made more heartening? Rhodes and Cox

identify the commitment of the Head Teacher as significant in

determining attitudes within a school more generally. But they also

urge more effective experiences for preparing and supporting teachers

in their use of this relatively unfamiliar resource. Thus, at

present, there may be few sites where a culture of computer use is

comfortably established - where the potential impact of particular

computer-based activities can be evaluated in a convincing manner.

Nevertheless, my colleague Geoff Alred and I have recently had the

opportunity to study one initiative where an effective context for

innovation was carefully cultivated: where circumstances seemed more

favourable to effective implementation of the kind that policy makers

hope for. A local education authority invited primary schools to

volunteer staff for participation in a project to evaluate Turtle

Logo. This is an activity that will be described in more detail in a

later section. Suffice to say it is a challenging exercise in

computer programming based upon controlling the movements of either a

floor robot or a screen icon ("turtle"). New equipment was supplied;

the project ran over a generous time period (at least four terms), and

it incorporated specialized in-service support (with suitable teaching

cover). In short, the conditions of the venture would seem to be very

favourable: the focal activity (Logo) is widely endorsed in early

education and the project was well supported by specialized training

opportunities. Moreover, the participants were motivated (if not

highly experienced) and could enjoy the advantage of being part of a

community of innovators. In summary, this situation seemed to us to

approximate what elsewhere has been described as the "ideal"

circumstance of an IT-related in-service provision (Owen, 1992,

p.130).

These teachers kept diaries summarising their experience and Alred and

I were able to interview them at some length towards the end of the

formal project. The emerging picture is a mixed one. Although it was

not part of our purpose to observe the classroom activity directly, it

was apparent that the children had enjoyed the Logo work: most of

the teachers had been impressed by their engagement with it. Yet, on

balance, the implementation project as a whole can not be regarded a

great success. One measure of success would be how far the activity

remained in use to become part of classroom routine for subsequent

generations of pupils to enjoy. One year after the official end of

the project, less than a quarter of the teachers were found to be

still using Logo with their new classes. As it happens, our

conversations with them had led us to expect this.

Although they recognized the innovative nature of the activity - as

well as the children's enthusiasm for it - they also were keen to

identify the practical difficulties associated with managing it on a

routine basis. Mundane problems of unreliable turtles were a major

source of disappointment. But as the activity can be supported on the

screen alone (i.e., it does not depend on a working floor robot),

this can not explain the widespread failure to consolidate the

experience beyond the life of the project. Other problems mentioned

related to the time and effort involved in preparing the computer and

ensuring its security. The teachers also remarked on the difficulty

of monitoring and supporting the activity at the same time that

classroom life was continuing as normal around it. Finally, taking

such necessary effort into account, there was some doubt as to whether

comparable academic achievements could not be reached in simpler ways.

To some extent the adequacy of the in-service provision arises again

here: certainly, this was another source of some dissatisfaction among

the participants. Many of the problems experienced *seem* as if they

should have been tractable with experienced advice and encouragement:

however, there is a danger in continually laying the blame at this

door. At some point we may have to acknowledge that the creative

deployment of this technology puts a lot more strains on the status

quo of classroom life than has been recognized. More sensitive

preparation and training might be some part of a solution to this

problem but there is clearly an invitation to reflect more carefully

on defining the optimal computer *environment* for supporting

innovation as effortlessly as we can. That is an issue to which I

shall return later in this book.

 

 

Evaluation of learning outcomes

 

Evaluation of what gets learned from such a multi-faceted innovation

is not easy. For one thing, its use has not been planned in

programmatic terms. Insofar as pressure for innovation has largely

been applied from above, practitioners may have felt little sense of

dealing in *options* for change: options that might be catalogued and

approached in a spirit of formal evaluation. The atmosphere has been

more one of seizing opportunities and improvising a way forward.

However, if we persist in seeking evaluative data, it can be sought in

respect of two broad questions. Firstly, has there been progress in

imparting to young learners some fluency in simply using new

technology? The phrase "computer literate" has evolved to capture

what this might mean. Secondly, can computers assist mastery of those

particular curriculum areas where they have been used to support

teaching: geography, maths, music, or whatever? This is a more

subject-based question.

Simply in terms of whether pupils now have the chance to encounter new

technology in schools, we can say that things have worked. Most

school children will enjoy some hands-on experience of computers.

Whether the experience leaves them adequately comfortable and

confident is another matter: it may yet be too modest in scope. This

may be a matter of some concern, given what is known about attitudes

to new technology among the current generation of adults. For that

generation, a clinical vocabulary has been felt necessary to capture

the extremes of feeling: computers are capable of inducing "phobic"

responses in some (Meier, 1985) while in others they can acquire the

status of an "addiction" (Shotton, 1989).

What can be said about the rising generation of computer users is that

such extremes of attitude are not apparent at the earlier stages of

using computers in school. Numerous reports endorse the view that

children's early experience with computers in classrooms today is a

positive one - as they themselves judge it (cf. Hughes, Brackenridge

and Macleod, 1987). It is perhaps still too soon to judge whether

such favourable early reactions predict greater confidence (and less

aversion) on completion of school. Data to be presented later in this

book suggest that some of the most academically successful of school

leavers (undergraduates) still include significant numbers who remain

(i) relative novices in terms of experience, and (ii) have very

uncertain attitudes regarding the appeal of using computers. Perhaps

ten years of gradual development is still too short a period to assess

whether educational practice is creating a natural culture of computer

use.

However, we are not just concerned with learning to use computers:

we are also interested in using computers to learn. A second

evaluation question concerns impacts associated with deploying

computers to support teaching in the traditional curriculum areas.

Can this technology help pupils learn about, say, mathematics,

geography, design, and so on? The more circumscribed nature of such

questions suggest they might allow more confident answers.

Unfortunately, the resulting picture can not be as clear as we might

like it to be.

For one thing, research in the relevant settings and under the

prevailing conditions of innovation is not easy to manage (Bork,

1991). We will usually wish to conduct studies in which different

conditions of learning are created and compared. However, explicitly

tinkering with resource allocation may be both unwelcome and

impractical. Institutions may be open to participation in research

but not if the research procedures undermine equity of opportunity for

their pupils. They will also resist being deprived of access to

leading edge technologies - where this is simply in the interest of

comparative research exercises. Thus, true *experiments* are not easy

to conduct. Moreover, "natural experiments" (in which we compare

spontaneously differing educational circumstances) bring their own

problems: the differences found in these cases may conceal other

differing population characteristics that also contribute to the

outcomes compared. These problems were identified by reviewers of

early initiatives (eg. Jamison, Suppes and Wells, 1971) and, still,

they remain an obstacle to confident evaluation.

Moreover, it is surely fanciful to suppose singular generalizations

will be found that can make sense of such diverse educational

activities. Computers support a very wide variety of learning

encounters in a very wide range of curriculum areas. We must be wary

of sweeping rulings on *the* success (or failure) of new technology.

However, if some feeling for the state of an evaluation balance sheet

is still felt valuable, summaries of such research have been

reported. Kullick, Kullick and Bangert-Downs (1985) have conducted a

meta-analysis over a large number of published studies evaluating

computer-based teaching projects in early education. They find

reliable and recurring positive impacts - amounting, on average, to

improvements of around 0.48 standard deviations on outcome test

scores, or student movement from the 50th to the 68th percentile.

These effects seem strongest in early educational settings. More

recently, a further meta-analysis (Ryan 1991) has focused on primary

school interventions of eight weeks or longer. Similar findings are

reported. The mean effect size across the studies reviewed was 0.309.

This may be expressed by saying that the treatment groups, on average,

exceeded the performance of 62% of their matched peers in control

groups.

Kulick and Kulick (1987) report comparable effect sizes beyond early

education, although it is widely believed that outcome effects remain

strongest for interventions within the elementary sector (Niemiec and

Walberg, 1987). In a meta-analysis of such meta-reviews Niemiec and

Walberg conclude: 'The average and typical effect of CAI is to place

the average student using it at the 66th percentile of traditional

groups - a substantial though not overwhelming advantage' (1987,

p.31).

Given the scale of the underlying educational investment and upheaval,

effects of this magnitude might be judged less dramatic than we could

hope for. Of course, we must be cautious in interpreting such

findings. The statistics will conceal a great deal of variation

associated with the sheer range of possible computer-based activity.

Moreover, methodologically pure evaluation studies are more feasible

with those forms of computer-based instruction that reinforce existing

curricular concerns - and for which teacher-based control groups are

more readily defined. Certainly, studies focused on drill and skill

forms of software seem over-represented in reviews. The more radical

and innovative applications of the technology may be less easily

studied under the strict methodological requirements sought by

meta-reviews of the literature.

Disappointment in these findings may be premature for other reasons:

many practitioners will declare that, whatever evaluation research

shows, they can simply see the computer "doing good". If

post-intervention testing is failing to endorse this then, they would

argue, this may be a reflection of the short term, piecemeal nature of

the outcome measures. Our tools for assessing what is being changed

may be blunt instruments. Moreover, the modest scope and duration of

many formalized interventions could miss effects apparent in a truly

motivated and committed classroom.

If we do wish to conduct evaluations of what is learned in

computer-based contexts, we must go beyond the input-output designs

that characterize much research in the area. It may not be enough

only to expose a pupil to some software and, sometime later, do an

outcome test of understanding. The reason this is inadequate is

because any such computer experience is more or less situated in some

broader framework of teaching activity. In short, there is a risk of

casting this educational technology in terms that suggest a medical

model of how it works. Computers are unlikely to function as magic

bullets - effortlessly releasing their therapeutic effects at points

identified by teachers. The unfamiliarity and wizardry that surrounds

them may cultivate such notions, but the real impact of learning

through this technology may need to be measured with attention to how

it is assimilated into the surrounding frame of educational activity.

That will be one theme to be developed in what follows here. There is

considerable variation possible in how researchers conceptualise the

way a computer programme is being "used". This, in turn, must

influence empirical strategies favoured for the evaluating the

"effects" of some computer activity.

For this reason, the research reviewed above should not encourage too

hasty and definitive conclusions about how effective computers are in

supporting learning. Their impact may need to be judged with careful

attention to broader patterns of *use*. This demands consideration of

how computer-based experiences are integrated into the broader frame

of activities that define an organized environment for teaching and

learning. Across different settings, there may be significant

variation in how radically the same technology serves to restructure

the activity of learning. Thus, its influence will not always be

neatly contained within events at the pupil-computer interface itself.

Researchers may need to look further than this in defining the "place"

at which computers work their effects.

My discussion of evaluation has converged on this issue of

understanding the broader context of computer-based learning. The

same theme of integration will arise in the discussions below of more

theoretical perspectives available to characterise computer-based

learning. It does appear that seamless integration has been more

problematic than many committed observers had hoped. I would suggest

that there are three persistent problems that underly the dis-located

nature of many classroom computer experiences. Firstly, there is a

great deal of controversy over how this technology is best deployed:

what *sort* of role it is supposed to play in relation to existing

patterns of teaching. Secondly, there has been some teacher

resistance to computers on grounds of limited familiarity - a

circumstance that might suggest inadequate provision of pre-service

preparation and in-service staff development. Finally, there may be

deeply rooted concerns regarding the impact of technology on the

*social* quality of educational experience. This last problem is a

particular focus within what follows in this book.

These various obstacles to implementation will be explored further in

the remainder of this chapter. In dealing with the first

(conceptualising the computer as a learning resource), I shall

evaluate four prevailing metaphors: computer-as-tutor,

computer-as-pupil, computer-as-simulation and computer-as-tool. The

second problem (that of teacher preparation), I have discussed above.

The final obstacle to a seamless integration will be introduced and

defined in the last section of the chapter. There I shall consider

how an (apparently) impersonal technology relates to what is

traditionally a most socially-organized activity: classroom learning.

 

 

 

CONCEPTUALISING THE NATURE OF COMPUTER-BASED LEARNING

 

 

One imbalance within the review that follows should be acknowledged.

Much of the published research in this area has been focussed on the

concerns and practices of *early* education. To some extent this

merely reflects the special interest that psychologists and

educationalists have in the management of learning in the "formative"

years. It also may reflect the fact that computer use may be more

pervasive (Becker, 1991) and more successful (Niemiec and Walberg,

1987) in primary education than it is in the secondary sector.

However, I am confident that the substantive points arising here apply

very generally to other sectors of education.

It is not possible to review all of the achievements and arguments

that rightly belong under each of the sub-headings used in the

present section. I merely intend to establish that there are problems

to be addressed and that these are not problems that are tractable

through the application of simple evaluation experiments. Here my own

preference for a method of tackling the practical problems is

apparent. I shall argue, that what is more urgently needed is a

theoretical framework to help systematise some of the tensions and

uncertainties associated with the present state of computers applied

to education. What will actually emerge in later chapters is one

particular theoretical framework; one closely associated with certain

contemporary preoccupations within developmental psychology: namely,

the socio-cultural or socio-historical perspective.

In short, the following review of options for framing computer-based

practice will not attempt to be comprehensive; instead, it will

crystallise a number of challenges that will be recovered in later

chapters for reconsideration in terms of a socio-cultural view. This

will be particularly pertinent for discussion in Chapter 2, where the

nature of this perspective is more fully discussed.

 

 

 

(1) The tutorial metaphor: computer-as-tutor

This section concerns a form of computer software that reproduces a

traditional model of teaching and learning. I shall note that such

software is popular despite criticisms from educational theorists. I

argue that the reasons for its popularity are complex and deserve more

sympathetic reactions.

In his book *Teachers and machines*, Cuban (1986) reminds us that the

enterprise of marrying educational practice with contemporary

technologies has a long history. Yet it was not until the 1950s that

people began to entertain the concept of a "teaching machine" - in the

sense of a mechanism that directly instructs. Such a technology would

not be simply something that teachers employed to illustrate or

elaborate their teaching. To a significant degree it could take over,

wholesale, what a teacher does.

Efforts to create such machines arose from a species of applied

psychology associated with Behaviourism, a dominant theoretical

paradigm within Psychology at the time. Yet, whatever the former

authority of Behaviourism, it is fair to say that the teaching

machines of this era were neither very successful nor very popular

(see Skinner (1984) for disappointed reflection on the neglect of such

behaviourist enthusiasms). Certainly, the early realization of this

idea had virtually disappeared from education at the time that

microcomputers made their entry. However, the spirit of

machine-as-tutor has been revived: finding its modern realization in

certain controversial varieties of educational software that are in

widespread current use.

It seems that the ambition to simulate tutorial instruction does

retain a seductive quality. So, when school micros began to appear in

Britain, the education correspondent of The Times was able to headline

his copy 'A teacher on every desk' (The Times, 1984). Yet, if we are

amused by this ten years on, it is surely more for the sheer scale of

computerisation implied by such a vision. The principle of desktop

teachers probably still carries serious credibility - with many

enthusiasts supposing the obstacle is merely limited access to

adequately powerful delivery systems. Certainly, contemporary

technology is far more versatile than that available to an earlier

generation of designers: so, it will be argued, we should take

seriously the mission to embody features of teacher-pupil interaction

within the design of educational software.

In fact, the appeal of computer-as-tutor may, in part, be sustained by

one particular perspective on the nature of teacher-pupil dialogue.

Some commentators (Dillon, 1985; Levin, Kim and Riel, 1990) have drawn

a parallel between the design of much educational software and a form

of classroom interaction characterised as I-R-E sequences (Mehan,

1979; Sinclair and Coulthard, 1975). That is, verbal exchanges taking

the form of a (teacher) Initiation, a (pupil) Response and a (teacher)

Evaluation. As in:

Teacher: ... if its a pentagon, how many sides does it have?

Pupil : Five sides

Teacher: Thats right, well done.

Ethnographers of classroom discourse report such sequences are

commonplace. It is, of course, a form of talk into which pupils

become socialized and which we recognize as peculiarly characteristic

of school experience. We accept its features, perhaps without much

reflection: the ritualistic nature of the exchange, the idea of people

asking questions to which they know the answer, the expectation of

evaluative feedback, and so on. It is clearly something that happens

when teaching is in progress. It might also be something that could

happen within a dialogue arranged between a pupil and a computer: 'The

computer initiates, the student replies, the computer evaluates, the

computer initiates again, and so on' (Levin et al, 1990 p. 210).

In fact, such sparse realization of the computer-as-tutor is quite

common within the so-called 'Computer-aided instruction' (CAI)

tradition. This approach has involved preoccupation with two goals in

the creation of effective tutorial software: (i) individualization -

problems and questions tailored to the (changing) needs of particular

learners, and (ii) the delivery of constructive feedback. What this

has often entailed is refining procedures for getting the initiating

questions "right" (matched to what the current pupil knows) and

attending to the crafting of accurate and motivating feedback. At its

most modest, this may merely mean saying "no" if a pupil makes an

error, supplying correct answers where appropriate, and choosing a new

level of problem to match pupil progress.

The more advanced versions of such CAI are sometimes termed

"Intelligent Tutoring Systems" (eg., Anderson, Boyle and Franklin,

1985; Burns, Parlett and Redfield, 1991; Mandl and Lesgold, 1988;

Quere, 1986; Self, 1988; Sleeman and Brown, 1982). They illustrate

one particular enthusiasm closely associated with computer-based

education: namely, the goal of creating a strongly individualised

curriculum. Thus, the "intelligence" of an ITS system would reside

not merely in its programmed expertise for the domain of knowledge.

It would also be able intelligently to diagnose the learner's needs

and, then, intelligently to implement an individualised tutorial

dialogue.

The ITS tradition of instructional software has been more visible in

military and industrial training contexts than within formal

education. This may partly reflect the sophisticated hardware

platforms that are needed (usually beyond the financial reach of

schools). It may also reflect the fact that implementation may be

more realistic for developing skilled behaviour within fairly

circumscribed domains of action. Broader educational goals may be

more difficult to realise. Thus, at the time of writing, really

effective examples of this format are rarely encountered in

classrooms. The preoccupation of ITS enthusiasts has been with the

possibility of seriously individualising the curriculum - a teaching

technology sensitive to individual learners. Yet it has proved

surprisingly difficult to construct algorithms that diagnose

effectively a pupil's particular errors and needs. Even something as

apparently straightforward as learning subtraction turns out to be a

skill that supports a rich variety of pupil misunderstandings and

procedural "bugs" (Brown and Burton, 1978). So, even this simple

domain is not easy to "tutor" in the automated sense we are

contemplating here.

Perhaps a compromise is possible - could some of the decisions

regarding customization of problems to the learner's needs be passed

over to pupils to make for themselves? For example, pupils might

make their own choices in respect of the difficulty of problems they

are equipped to tackle. This is evidently an option; although, when

younger children confront computer-based problems, it seems we cannot

necessarily assume that they will choose levels of difficulty that are

challenging or well-matched to their state of knowledge (Crook and

Steele, 1987).

In spite of the admitted difficulty of building "intelligence" into

computer-based tutoring systems, software conceived within this

broad (tutor) model of educational computing has often been the most

readily adopted by teachers. This is particularly the case within

early education. Moreover, the most widely employed examples are

those with the least pretension to some modest "intelligence". Such

programmes - repetitively delivering discrete problems within some

domain of study - are often termed "drill and practice" or "drill and

skill" programmes. Surveys of classroom practice reveal a striking

preference for these more didactic forms of software (Becker, 1991;

Jackson, Fletcher and Messer, 1986), particularly for children under 9

years old (DES, 1991).

Conversely, these are programmes held in very poor regard among

professional commentators (eg., Papert, 1980; Self, 1985). Papert, in

particular, is a vigorous critic whose slogan that the modern pupil is

being controlled by the computer (rather than vice versa) is

frequently cited. Thus, the most widespread practical realization of

our computer-as-tutor metaphor is widely criticised by educational

theorists as constraining the learner's experience. It is judged to

offer poor approximations to what is itself a rather poor model of the

teaching process in the first place (didactic encounters guided by the

IRE pattern of dialogue). However, simply because of their widespread

use, we should reflect on these kinds of activity a little further

here. Why have programmes fashioned in this tradition been so widely

preferred by practitioners?

To answer this, we should consider an observation made by Cuban

(1986). Reviewing the history of teaching and machines, Cuban

concludes that what practitioners prefer to do with computers merely

reproduces the typical fate of any new technology applied to

education. It tends to be assimilated to prevailing traditions of

classroom practice. Cuban's documentation of this relentless pattern

is convincing but the analysis is often invoked by others in a rather

disparaging manner. We may agree that this inertia is unfortunate -

in that it reflects a failure to seize *new* opportunities - but

we should attempt to make sense of it. Simply proposing that teachers

have a limited vision of good practice is an ungenerous and, probably,

inaccurate conclusion.

Thus, we may accept Cuban's claim that the popularity of drill and

skill computer activities can serve to maintain a status quo. But I

would then suggest two particular ways to make more sense of this.

First, the strategy of assimilation to existing practice is a ready

reaction to the unwelcome demands of an imposed innovation. Second,

the strategy reflects some degree of genuine commitment to that

feature of the status quo. Opportunity for structured practice is

seen as something that is worth developing - and new technology is

identified as one way to do this.

The first of these two observations supposes that innovative adoption

of computers is more difficult to achieve than is often assumed. In

an earlier section, attention was drawn to the practical problems

associated with incorporating new technology into classrooms. Studies

of in-service training provision suggest a mismatch between what is

offered and what is needed. They draw attention to how we may

overestimate the ease with which teachers can develop confidence with

an unfamiliar technology. Research also highlights the hidden demands

entailed in the professional support of innovative work on classroom

computers. Under such pressure, teachers may well adopt the

comparatively effortless solution of focussing their commitment on

straightforward, self-contained programmes that pupils can work

through independently. These would be drill and skill software of the

sort that is found to be widely used. There is some suggestion that

this preference does represent a kind of holding response in an

uncomfortable situation (Heywood and Norman, 1989).

This reason for the popularity of more humdrum tutorial software is a

reactive one - though none the less important to uncover and

understand. However, the second reason I proposed for the popularity

of drill-based programs suggests a more actively positive attitude

towards them. It may be rash to simply exclude a place for drill and

practice within the learning process. Teachers may feel that such

experiences forms an important part of achieving certain educational

goals. Anyone who has, for example, attempted to cultivate musical

pitch, or master the differential calculus will probably appreciate

the merit in dense experience of exemplary problems from the relevant

domain. Dreyfus and Dreyfus (1984, 1986) elaborate this point in the

context of a psychological model describing the development of skilled

expertise. So, the value of furnishing opportunities for unadorned

practice may be accepted. Yet, if it is accepted, it need not presume

wholesale reduction of educational activity to the rehearsal of

discrete sub-skills.

Practitioners may, therefore, endorse the view that much of

what they teach permits some place for such focussed practice.

They may also concede that furnishing the necessary opportunities is

not the easiest or most rewarding part of their responsibility. That

is exactly where computers may offer an easy appeal. They seem to

offer some release from this problem. Computer-based versions of

practice within a domain may be particularly attractive simply because

they can offer this opportunity in an (apparently) engaging manner.

In fact, Cuban (1986) admits some sympathy with the use of computers

for drills, arguing that this is what computers are actually very good

at (see also Dreyfus and Dreyfus (1984) and Marsh, (1985)).

I am proposing that the preference for simple computer-as-tutor

implementations needs to be made sense of rather than automatically

disparaged. Part of the sense we might make of it involves

acknowledging the local ecology within which computer innovation is

being encouraged: those conditions may not always be favourable, and

skill-type programs are a ready solution to sustain some level of

computer use. But the sense we make of it might also involve an

admission that the creation of practice-oriented opportunities has

some legitimacy. Computers can raise the appeal of what might

otherwise be a less rewarding activity. In fact, when we confront

examples of the genre we may even fail to notice what their underlying

format is quite mundane. An illustration from early education may

help to make this point.

In a recent review, Scott, Cole and Engel (1992) are critical of the

computer drill tradition. Yet, an example program they furnish for

purposes of contrast - and to illustrate imaginative use of computers

- seems to have typically drill-like features. The program invites

pupils to assign values at positions on a number line; an activity

that is embedded in a game involving the harpooning of a shark. In

other words, a repetitive exercise in which the shark's position on

the computer screen is an invitation to generate a numerical

representation. This would count as a drill program in any survey of

classroom practice: it is apparently a self-contained and repetitive

exercise involving a basic mathematical skill.

Yet, Scott et al may be right to cite such a program with some

approval. It sounds engaging enough and, surely, a teacher with the

task of helping children cultivate strategies of estimation will find

it attractive. Moreover, there are plenty of other examples of such

circumscribed activities that could be judged equally appealing. Some

of them have evolved from careful work within cognitive psychological

theory (eg., Resnick and Johnson, 1988). In the face of temptation

from interesting drill-based programmes such as that described above,

a tolerant attitude to the genre seems possible - especially if

Dreyfus and Dreyfus' caution is respected: '...the only danger in the

use of the computer for drill and practice and for diagnosis arises

from the temptation to overemphasize the sort of training in which the

computer works, precisely because it works so well' (1986, p.133).

Coming to believe that domains of knowledge could be reduced *only* to

packaged exercises is one possibility we should remain alert to. I

shall mention one more before closing this section on

computer-as-tutor: one that is more central to my concern here with

social context.

This flirting with computer activities - incorporating them at the

periphery as vehicles-for-practice - is a strategy with its own

problems. In particular, marginalising certain activities in the way

that can happen with computers, may serve to undermine their impact.

Something of value may be lost where such activities are not knitted

into a mainstream of class learning. The limit to computers

functioning as tutors arises not just because tutorial dialogue is

hard to simulate at the moment-to-moment level of conversation.

Computers are also limited in this role because 'tutoring' talk is

something that is organized at levels superordinate to that of the

current moment. In other words, effective tutorial dialogues are

embedded in more extensive contexts of shared classroom experience.

Such dialogues are normally made possible by the history of this

experience. Their effectiveness may depend on creating a natural

continuity of reference with it - a richness of context that will be

very hard to reproduce mechanically. This is a concern I shall

elaborate further in Chapter 4 where it will be approached in terms of

our current theme of collaboration.

A failure to involve activities of the kind we have been discussing

within a fuller collaborative pattern is one basis for concern with

the popular computer-as-tutor strategy. Another has already been

alluded to: namely, that this preferred use of computers causes us to

miss an opportunity for innovation. The misfortune here is more a

failure to recognize that a particular technology has potential beyond

that which represents its more accessible and effortless applications.

More generously, and as argued above, we may grant that the potential

may be recognized by teachers, but the reality of implementation may

be more demanding than educational commentators appreciate.

There are grounds to feel disappointment that circumstances are

apparently jeopardizing opportunities to do genuinely new things

through the mediation of computers. The technology is capable of

supporting distinctively challenging and innovative activities, as

well as re-formatting more familiar tasks. This is an observation

commonly deployed to justify and encourage implementations conceived

according to our second metaphor, the one to be considered next.

(2) The construction metaphor: compter-as-pupil

 

We encountered above Papert's (1980) widely-cited concern that to

regard the computer as a form of teacher was to risk a situation in

which children were "controlled" by the technology. Yet, the survey

statistics do reveal that software in the simple tutorial tradition

commands considerable appeal. Perhaps the rhetoric of "control" is

too provocative here to be fully persuasive. Teachers recognize that

these computer-based activities often parallel traditional classroom

activity; they conform to the purposes pursued elsewhere in the class;

Moreover, they may be more engaging in format. If this is to be

conceptualized as "control", then it runs deeply within educational

culture and, thereby, the critique is somewhat defused.

On the other hand, there may be teachers who are uneasy about that

culture and impatient with the style of practice they often find

themselves party to. They may then recognize in new technology a

potential for subverting the status quo rather than upholding it. In

particular, they may sense a possibility of shifting from more

teacher-centred to more pupil-centred practices. For many, Papert's

vision of the computer in education offered something of this

potential. In his book *Mindstorms* (1980), Papert not only denied

the tradition of computers in control of pupils, he proposed an

alternative in which that relation would be reversed: pupils would

control the machines. But, most importantly, the ways in which they

might come to do this would entail especially potent learning

experiences. The idea is captured well in his notion of a

microworld. This is a setting in which learners can apply principled

knowledge to effect genuinely creative activity. They *construct* new

understandings through their exploratory activity.

Papert's proposal is driven by a compelling image. If you wish to

learn to speak French, he argues, you go to France. This surely makes

good sense to us. But if France is where you go to command French,

where do you "go" to command, say, Mathematics? What must be

discovered in that case is a sort of "Mathsland". In such a place,

the learner might encounter situations where various opportunities are

offered for hypothesis-testing explorations; where existing knowledge

must be mobilized to solve urgent and motivated problems. It would be

a versatile platform from which one's initially modest base of

knowledge could be meaningfully exercised.

This describes an environment of discovery-based learning. So, it

suggests an important theoretical influence: namely, the ideas of

Piaget on cognitive development. First, Piaget's notion of the

learner as necessarily active is strongly endorsed. A microworld is a

place where things are getting *done*: in Papert's realizations, such

action amounts to teaching computers to do something interesting - the

computer is thus the "pupil" in these encounters. Another important

principle defining the microworld environment is that it should be

somewhere that maximises the experience of discovery. So, it might be

a place where teachers can more effectively respect Piaget's warning:

'Each time we prematurely teach a child something he would have

discovered for himself, the child is kept from inventing it and

consequently from understanding it completely' (Piaget, 1970, p. 715).

Papert furnished a particularly striking example of a microworld

environment with his Turtle Logo. This is an activity that can be

described from two perspectives. Firstly, it is a vehicle for

introducing certain concepts that are central to writing computer

programmes. Logo is a computer language that has a number of

important features. In particular, it incorporates procedures

(sub-units of code performing discrete functions), it is extensible

(the user can define these units to be used as primitives), and it

embodies recursion (such procedures can be self-referring). For

Papert, these exemplify 'powerful ideas': ideas that can be mobilized

very generally for problem solving. He supposes that the opportunity

to exercise them in some concrete (computer-based) activity allows

them to surface sufficiently clearly that the learner can directly

contemplate them. Such reflection will help these skills become more

readily available in other problem solving domains. Treating the

computer as a "pupil" in Papert's sense (programming it) is, thus,

taking an opportunity to cultivate general problem solving skills.

So, Logo is about acquiring such skills via programming. It is also

about controlling the particular settings which act as vehicles for

the programming. The vehicle that Papert promoted most successfully

was turtle geometry. Within this microworld, the learner can issue

instructions (a programme) that cause movement of a floor robot

(turtle). The same instructions can, more simply, be used to create

patterns on a computer screen. Here, then, we find something

corresponding to the elusive "Mathsland": arithmetic, algebra and

geometry all encountered in the purposeful task of generating movement

or creating patterns. The computer offers a meaningful and motivating

environment in which principled knowledge can be applied towards

creative goals. The learner is using her understanding to "teach" -

the computer is acting as pupil.

Turtle maths is not the only outlet for Logo. For example, in another

instantiation of the basic programming environment, some interesting

possibilities can be created to support work in the language

curriculum. However, it was turtle Logo that truly captured the

imagination of educators. It could be said that it remains the most

striking realization of what Papert so appealing defined as a

"microworld". Certainly, it must be unique in the sheer energy and

enthusiasm that it drew from the community of users. Magazines and

special interest groups have flourished in support of the Logo

teaching community.

Ten years on, some of that energy arguably has dissipated. One reason

is an important one to mark here. Logo attracted more focussed

evaluation research than any other computer-based enterprise. The

strength of the claims made for it and its obvious appeal among pupils

naturally invited careful assessment. On the whole, the evaluations

have not been as positive as enthusiasts might have hoped for. A

number of critical reviews have now appeared (eg., Dudley-Marling and

Owston, 1988; Pea and Kurland, 1987; Pea, Kurland and Hawkins, 1987;

Simon, 1987). Moreover, apart from the example of Logo, studies of

pupils learning other computer languages have not furnished very

persuasive evidence for the generalizability of programming skills

(Dalbey, Tournaire and Linn, 1986; Palumbo, 1990).

Yet the emerging picture is not one of unqualified disappointment. In

fact, in terms of the arguments developed elsewhere in this book, the

findings of Logo evaluations are very instructive. Pea and his

colleagues, in particular, have made a useful characterization of the

circumstances under which this kind of learning environment may and

may not be effective. It seems that situations where outcomes from

Logo experience have been less successful are those where participants

have too enthusiastically adopted Piaget's dictum (quoted above) about

avoiding "premature" teaching. These less successful ventures may

illustrate too great a faith in the principle characterised by Perkins

(1985) as 'the opportunity does the teaching by itself' - the

idea that simply using it is enough. Except that, in this case, such

faith is likely to be amplified by witnessing conspicuous engagement

among the learners.

Enthusiasm for such pupil-led opportunities is complemented by natural

distrust of the opposing alternative: more teacher-controlled

situations. Perkins comments (sceptically) 'Often it is even urged

that direct teaching may do mischief by forcing the issue in an

unmotivated and acontextual manner' (1985, p. 13). In all this there

may be too careless a polarization of the options. There *is* an

issue associated with inauthentic learning and we will wish to remain

vigilant in respect of it. But 'direct teaching' is not the exclusive

alternative to a discovery-based microworld encounter. In the ideal

situation, involvement of teachers may need to be very much "indirect"

in manner, but nevertheless their involvement is crucial. Around

Logo-learning pupils, there are important things to be done (and said)

by others who are themselves more confident with the relevant

concepts. The challenge is to discover more of how this supportive

function is to be defined: that, in turn, may require us to consider

the "collaborations" that computers-as-pupils afford.

This section on computers-as-pupils has reached one conclusion very

similar to that reached in the discussion computers-as-tutors above.

In each case, the implementation of the computer activity may too

easily encourage a distancing of teacher involvement; or, more

generally, a dislocation from the normally rich context of class-based

activity and discussion. As was commented at the end of the previous

section, this is a threat to the collaborative quality of learning

experiences: one to which we shall give more attention in later

chapters.

Conceptions of computers as tutors and as pupils have been important

in determining the most common patterns of use. However, there are at

least two further metaphors with wide appeal. I shall comment more

briefly on each of these next.

(3) Simulations

 

It is characteristic of Papert's microworlds that they should have an

open-ended quality. A set of powerful ideas are available, the

application of which will support a rich variety of creative activity.

A relatively clean working surface is supplied and pupils direct

action towards it. In this version of computer-based learning, the

power of the computer to manage symbolic activity is harnessed to

allow the pupil real control over a given domain of knowledge.

However, this form of control can be realized in more circumscribed

ways. Learners can interact with more closed systems and, in doing

so, instructive experiences are made to occur within them.

In such cases, the symbol-manipulating power of the technology is

exploited to offer simulations of real-world systems. Computers

might, for example, simulate a stock exchange for students of

economics; or the behaviour of cell membranes for students of

biology. The promise of such scaled-down experiences lies in their

capability for offering the learner control over the operating

parameters of some system. In this way, a system's characteristics

may be explored through experimentation. It is likely that simulation

software will become more commonplace as multimedia technology

develops.

An early indication of what is possible with such resources - and a

theoretical context for evaluating it - is available in the work of

the Cognition and Technology Group at Vanderbilt (1990). This group

have employed videodisc technology to create "macrocontexts" for

learning. They encourage the idea of "anchored instruction": a form

of educational practice that is strongly oriented to exercising

knowledge in rich and meaningful settings of authentic practice. The

learner is invited to control some domain of interest in a manner that

more resembles the experience we associate with apprenticeship mode of

learning.

Exposure to simulations is seen by some as a solution to slow progress

being made within the computer-as-tutor tradition. A limitation of

many intelligent tutoring systems may be their approach to the

representation and the communication of expert knowledge in the domain

being taught. Such representation is typically guided by ideas from

artificial intelligence. AI stresses the *rule*-based nature of a

knowledge domain. This perspective encourages learning resources that

tend to support the integration of rules describing families of "if x

then y" relationships (thought to describe some domain of knowledge).

Riesbeck and Schank (1991) have argued that this form of

representation may be impractical for domains of any complexity. More

important, they argue (1989) that rule-based reasoning may not be a

good model of everyday intelligent thinking.

Instead, they propose that human reasoning can be described as

"case-based" (Schank, 1982). Expertise arises from dense experience

with representative and discrete problems from particular domains of

practice. Such human reasoning demands a particular kind of resource

to support new learning: access to a library of concrete situations

(cases) in which problem solving can be exercised. Creative teaching

would then partly be located in the effective 'indexing' (Reisbeck and

Schank, 1991) of these experiences - such that they were readily

retrievable for reference in dealing with novel circumstances. The

principle of case-based reasoning suggests that traditions of

computer-as-tutor and computer-as-simulation could move closer

together. At present the relevant theoretical debates underlying this

possibility remain in progress.

One worry aspect of learning from simulations concerns the possibility

of over-simplifying complex systems. This is often inevitable if a

simulation must meet the finite possibilities of delivery on

microcomputers. The necessary simplifications may convey a

misleadingly straightforward impression of how a multi-variate,

open-ended system works in the real world. This is particularly

problematic in situations where attempts are made to simulate systems

whose real behaviour is governed by significant human agency. The

microcomputer may force too rigid a representation of how typical

human intervention is organized. So, it may be seriously misleading

if simulations imply that systems incorporating substantive social

management are, for that reason, governed by rule-like and planful

processes: often it is clear that they are not (cf. Suchman, 1987).

Finally, it is apparent that these forms of computer-based learning

environments must share some of the characteristics (and attractions)

noted above for microworld environments. They offer the novice a

strong discovery-oriented experience and they are engagingly

interactive. They have not yet attracted a great tradition of

evaluative research. But, as with microworlds, we are likely to be

anxious as to the necessary role of experts (teachers) in sustaining

and consolidating this kind of learning and contextualising it within

a broader classroom experience. Sheingold, Kane and Endreweit (1983)

show that such integration does not typically happen. Moreover,

Laurillard (1992) has highlighted the poor showing of simulation

software in situations where it is not carefully integrated into a

broader context of socially-organized teaching.

 

 

(4) The toolbox metaphor: computer-as-tool

 

It is common to characterize the computer as a "general purpose"

machine. We say this because it can be fashioned (programmed) to

serve a wide range of human purposes. In other words, there is an

important sense in which it can be said to provide a "toolbox".

Indeed, when we now think of computers most of us are likely to think

first of word processors, spreadsheets, databases, applications for

graphics or design and so on. This naturally supplies a

straightforward reason why we might encourage such use of computers

within education. The tools they create are in widespread use within

everyday contexts. Thus, quite simply, children must be helped to

control and understand them within the preparatory settings of school.

Surveys reveal that teachers are increasingly seeing the educational

potential of the technology in terms of these tool-like

characteristics (Becker, 1991).

However, strategies for fostering experience with computer-based tools

will be motivated by more than just narrow vocational concerns. The

falling cost of new technology has had the effect of creating a

greater continuity between school and work - in that powerful tools

that would previously only have been encountered in specialist

settings are now accessible in classrooms. It would be disappointing

if these tools were only used in ways that mirrored too literally the

particular demands of the world of work. Fortunately, the *content*

of problems tackled with these powerful new tools can reflect

interests that are actually nearer the world of childhood. Thereby,

it is possible to respect the widely-accepted educational principle

that problems posed for children should be authentic - drawn from

their own experience and reflecting their own concerns. Children may

well be more engaged by information-managing activities if what they

discover (through summarising, systematising, communicating etc)

describes something immediate to their own experience. The

possibility of provoking such vivid discoveries means that computers

offer teachers a valuable opportunity: to foster within even young

children powerful skills relating to the organization and

communication of data.

However, the educational deployment of computers in this guise is more

controversial than that characterisation might imply. Here, I shall

identify two troublesome perspectives associated with the

computer-as-tool conception. The first is a worry that pupils enjoy

less direct encounters with the world they are learning about. The

second is a particular theoretical perspective on learning that the

tool metaphor encourages - one that I find problematic.

The arguments arising from the first of these concerns are located

around the concept of "mediated experience". Mediation is one way to

capture the multi-faceted character of this technology. The computer

mediates our action - it exists between us and the world and

transforms our activity upon the world. For one thing, it encourages

us to act upon that more elusive quantity: information. This mediated

quality of computer-based activities has attracted some critical

commentary, particularly as it is realized in the contexts of early

education. Thus, in a collection of sceptical essays edited by Sloan

(1984), Cuffaro (1984) voices a recurrent concern that computer-using

pupils are being deprived of the opportunity for "direct" manipulative

activity on their world. Interactions with computers are supposed to

render their experience increasingly "indirect". Most worrying

perhaps, the more powerful these tools, the more they can be

accessible to younger and younger children (Crook, 1992b).

Cuffaro, and other critics in this collection, make frequent reference

to the Piagetian interpretation of cognitive development. Piaget's

commitment to the importance of direct manipulative activity on the

world is cited with approval as the foundation of discovery learning.

It is this foundation that is seen as threatened by the mediational

status of computer tools. Evidently, it is right to be cautious here.

As noted in remarks above regarding computers furnishing full-blown

simulations: access to a simulation should not undermine efforts to

give pupils more direct encounters with the system modelled. Not

enjoying direct interactions may create a misleading impression as to

the simplicity and self-contained nature of that system.

However, the force of this critique turns on cases where computers are

made to function as *alternative* tools to those traditionally used in

children's exploratory activity. Certainly, if computer graphics

packages became a commonplace *substitute* for work with, say, paints

or charcoal, then we might feel valuable creative opportunities were

being lost to pupils. However, we should allow the possibility that

these computer tools can offer a different and distinctive kind of

experience in graphic media (to take the present example): an

experience that will complement others. For instance, my own

research (1991b) on young children using screen painting programmes

suggest these tools may cultivate a more editorial attitude towards

graphic creations. Thus, the opportunity to delete or "undo" painting

strokes seems to reinforce active review and revision of a developing

composition - much as a word processing tool does for the editing of

text. This may be just one distinctive feature of experience with

graphic work in the computer medium.

Deploying computers in this way - to *extend* the experience of

drawing, writing, classifying or calculating - seems an exciting

enterprise. The mediating status of this technology is something we

may come to terms with: we may become sensitised to its effective

management. The second controversy arising from the tool metaphor

concerns certain more overarching theoretical attitudes that it

encourages. It encourages a line of theorising that I find

problematic: in particular, the view claiming that such experience

with computer tools have very general effects on the thinking of those

who use them. We can better see this possibility with respect to a

working example.

In writing the present text, I am using a computer in its capacity as

word processor. The tool-like character of the device allows me to

engage in various useful manipulative activities. Thus, I can

certainly have my spelling checked to some advantage. But I can also

"manage" the overall text in a more flexible manner. So, I might

organize material into headed sections: then, by using a pointer on an

index of headings at the start of a document, I can easily move around

my text. Or, I can shuffle sections into a more optimal organization.

I can refer to files containing notes or references; both of which are

visible in windows "behind" my main text. I can import material

between these areas. Is imparting to students fluency with this kind

of tool "merely" a preparation for employment in settings where it

will be expected? Or does it create more far-reaching cognitive

impacts?

One widely-acknowledged consequence of access to computer tools is the

freeing up of "space" for parallel cognitive activity. Thus, in using

a word processor, I could be said to relieve some of the (humdrum)

burden of text management and release cognitive resources for other

creative work. The idea being that I have a finite reserve of such

resources. The quality of my thinking might benefit if some of it can

be rescued from more routine cognitive commitments and, thereby,

become focussed more effectively. So far, this does not entail a

particularly radical claim about the impact of experience in using

such tools. It could be said merely to clarify just why experience

with such technology will be prized in the world of work: the tool is

effective and useful. But there is a more radical claim that can be

made about the experience of coming to control it.

Suppose that expertise with certain tools leaves us equipped with new

tools of *thought*. This point is made energetically in one recent

book reviewing the effective application of computers to early

education (Underwood and Underwood, 1990). These authors propose

that an important focus of computer-based activities is 'to equip

children with a toolkit of basic mental skills' (p. 29). This

perspective is not without its problems. However, it is a compelling

idea; one that has been argued with particular effect in a recent

paper by Salomon, Perkins and Globerson (1991). They make a useful

distinction relating to the "effects" associated with computers as

students use them. The various products of student activity can be

seen in terms of effects achieved *with* computers or in terms of the

effects *of* computers. In the first case, we note that some creative

product has depended upon a working partnership with a machine. That

machine has taken over lower level activities associated with the task

and allowed the student to do the whole thing more economically, more

efficiently, more imaginatively, or whatever (cf. the example of text

processing rehearsed above). However, the whole enterprise may

generate what Salomon et al (1991) refer to as a "cognitive residue".

In which case, students walk away from the experience with new (or

more finely honed) tools in their cognitive toolkits (Salomon, 1988a,

1993). So, in other settings - including those that do not

incorporate the computer as a prop - more powerful intellectual work

can get done.

Returning to the example of using word processing technology: we might

suppose that a cognitive residue is imparted by coming to control

the device. This supports the subsequent manipulation of text -

including work done *off* the computer - and will be to the benefit of

cognitive skills associated with the general management of ideas in

written form, including their effective communication. We might even

consider that the experience of acting upon text in this way heightens

our sensitivity to the written word as a manipulable quantity and, for

example, cultivates a richer sense of audience as we compose.

In summary, the metaphor of computer-as-tool is a powerful one and

provocative in at least two ways. Firstly, it draws our attention to

the mediating role of technology. Some commentators have expressed

concern about this: worrying that it deflects the learner's experience

from concrete exploratory activity. However, viewing this

technology as a mediational means will be a useful idea to

which I shall want to return later. Secondly, others have suggested

that the consequences of experience with computer-based tools may

include cognitive residues - new tools of an intellectual kind for

interpreting the world. I shall also return to this idea; although

with a less positive response towards it.

Both of these themes will be taken up in later Chapters, along with

others that have been highlighted in the present general review of

implementation metaphors. I shall conclude the present chapter by

highlighting a recurring concern that integrates several problematic

issues mentioned in so far - namely, the relation of new technology to

the social quality of educational experience.

 

 

THE SOCIAL FRAMEWORK OF COMPUTER EXPERIENCE

 

Under this heading we confront some of the core concerns that motivate

the present book. A good number of the implementation controversies

that have been reviewed above can be usefully considered in terms of

tensions between new technology and the *social* quality of

educational settings. Numerous commentators have cautioned against

technological determinism in applying computers to education. For

example, Bowers (1988) and Noble (1991) each explore within

substantial monographs the non-neutrality of information technology as

it has been developed within education. They argue that this

technology does not "simply" serve human interests in some benign

fashion: it actively transforms human relations.

It is important to acknowledge and act upon this state of affairs. My

own view is that its consequences are not inevitably to be regretted.

Indeed, a significant challenge is to recognize the transformational

effects of new technology and, thereby, mobilize them towards

realizing goals that we regard as precious. If we miss this challenge

there is some danger that the medium will be seized and used to

support forms of educational practice that many may find

unwelcome. In this section, I shall review some straightforward ways

in which there can be said to exist a "social" dimension to the

development of computers for educational purposes.

 

(i) The reproduction of inequalities

An early but recurring strand of critical commentary relating to

computers in schools dwells upon the irony of equipping so many

children with skills for jobs that will be scarcely available in the

world beyond school (cf. Noble, 1991; Robins and Webster, 1987).

Yet this is a concern that might be voiced very widely in relation to

educational practice: the *special* status of information technology

in this respect is no longer so obvious. For most pupils, the

learning they do that involves computers is more through them than

about them. That is, the technology impinges very broadly on the

curriculum, its use involving more than the teaching of specialized

knowledge about technology itself. However, its generic quality gives

rise to another strand of criticism.

If computers really are a wide-ranging resource for learners, then we

must be wary that their deployment does not serve to *amplify*

existing patterns of disadvantage (Olson, 1988). Evidently, there is

a straightforward way in which the technology is likely to be divisive

in this sense: it demands significant financial investment and the

opportunities for funding will be distributed unevenly within

education systems. Surveys of the present distribution of computers

in schools reveal a 10:1 ratio of variation between the best and least

well equipped schools (POST, 1991).

Moreover, this is a technology that can be differentially available as

a resource in children's own homes. Educational advantage arising

from such domestic access is evidently a possibility. However, we

should note two observations that suggest it may not be a significant

source of inequality. Firstly, ownership of home computers appears to

have reached a peak - at least, for the present generation of

technology. The UK General Household Survey indicates that around 20%

of households own a computer but that this figure has barely changed

between 1985 and 1991 (OPCS, 1991). Because prices have come down

across this period, these statistics are telling. They suggest

consumers have discovered that general-purpose computers do not much

enrich family life. One form of such enrichment might have involved

the support of educational agendas within children's home experience.

Yet there is some suggestion that this is not a role the technology is

currently playing. Giacquinta and Lane (1989) studied how computers

were used in 51 (US) families with school-aged children. Of the 113

children involved, most had no access at all to educational software

at home: those that did made very rare use of it. This is a

provocative observation that partially allays our present concern

about reproducing inequalities - although the research could usefully

be replicated with wider samples and in other cultures.

In the end, however, these are all observations that are very

generally true for educational resourcing at home and school; it

remains to be seen whether computers emerge as a particularly telling

differential. There are more subtle senses in which the technology may

reproduce inequalities.

One is in respect of gender. A number of surveys reveal that girls do

not perceive computers as being so much "for them" as do boys

(Durndell, 1991; Fife-Shaw, Breakwell, Lee and Spencer, 1986; Hoyles,

1988; Hughes et al, 1987). Moreover, their attitude to using the

technology may become increasingly negative as they proceed through

school. Various surveys show the percentage of women pursuing

computer science as an undergraduate subject is actually falling

(Newton and Beck, 1993).

Scott, Cole and Engel (1992) provide a review of work on this topic

and articulate a widespread concern regarding its implications.

Chivers (1987) identifies the problem as present in a number of

contemporary cultures and also discusses some of the possible measures

that we might adopt to tackle it. There is some doubt just how early

this differential perception and interest sets in. Lipinski, Nida,

Shade and Watson (1986) report that boys in a preschool setting spent

more time on a computer activity than girls. On the other hand, Essa

(1987) finds no such distinction among preschoolers and Crook and

Steele (1987) report no gender differences in time spent using a

cafeteria-style computer activity maintained in the Reception class of

a primary school. The problem invites more research. However, we may

say that gender-based attitude differences are not *convincingly*

present at the start of schooling: they must somehow be cultivated

within the early school years.

Scott et al (1992) consider a further sense in which the application

of this technology may be socially divisive (see also LCHC (1989)).

They note that there is some evidence to suggest that the type of

software favoured in different educational settings can reflect the

educational advantage or disadvantage of the pupils (cf. Becker and

Sterling, 1987). They caution, in particular, against a trend whereby

the less innovative software (eg., drill and skill programmes) are

over-represented in the experience of disadvantaged communities. Both

this discrimination and that associated with gender should certainly

be viewed carefully and attract more comprehensive documentation.

Despite its significance, this sense of "social context" is not

central to the more interpersonal concerns of the present book.

Issues identified in the following two subsections however are .

 

(ii) Computers and social development

Whether encountered within education or elsewhere, at least two

features of computers may exert unwelcome influence on children's

early social development - that is, on the development of their

capacities for entering into a world of social relationships. The

first feature is the apparently compulsive attraction this technology

can exert over many users (Levy, 1984; Kidder, 1981; Shotton, 1989;

Turkle, 1984; Weizenbaum, 1976). The second arises from the quality

of "intelligence" that we tend to identify within computer

interaction. Contact with this, it is feared, might encourage a

mechanistic interpretation of *human* activity or, more generally,

blur important distinctions between ourselves and our machines. At

least two consequences arise in relation to social development:

children's' social cognitions (their *thinking* about the social

domain) may be overly influenced by the mechanistic or computation

metaphor (Boden, 1981; Brod, 1984; Papert, 1980). In addition, they

may be drawn towards too dedicated or absorbing an engagement with

this highly interactive and responsive technology. They thereby risk

entering a socially reclusive world (Boden, 1977; Bontinck, 1986;

Simons, 1985).

I have reviewed these possibilities elsewhere (Crook, 1992b) and

concluded that such fears have been overstated. Firstly, many of them

depend upon a particular conception of computer use: one

that indeed can evoke self-contained and compulsive involvement -

namely, computer programming. It is true that absorption in the

writing of programming code may have preoccupied an earlier generation

of 'hollow-eyed youths' (Weizenbaum, 1976), but the present generation

of users will encounter computers in a form more akin to Norman's

(1986) conception of 'convivial tools'. Secondly, there is some

question as to whether children really are so readily seduced by

psychological metaphors in their thinking about computers (cf. Hughes

et al, 1987). Finally, Shotton (1989) has documented an intimate

study of dedicated (adult) computer hobbyists and paints a picture of

their involvement that is merely suggestive of very many other

innocent recreational enthusiasms.

These remarks relate very generally to young people's experience of

computers. However, there is a species of this general concern that

is rather more focussed on computers encountered in educational

contexts. Murphy and Pardeck (1985), for example, dwell upon the

dangers of a mechanistic model of mentality being inadvertently

fostered within classroom experience. Sloan (1984) cautions the

danger of 'relegating feelings to the realm of the peripheral in

education' (p. 543) and regrets the mechanistic imagery that

characterises computer culture. My own view is that vulnerability to

these dangers has been greater among professional psychologists than

among school children. Whatever may have been the curriculum emphasis

in earlier educational applications, contemporary encounters with

classroom technology are now much richer in their variety. They less

easily invite pupils into a focussed preoccupation with metaphysical

issues.

Nevertheless, there remains a real strand of concern regarding

children's social experience that does arise in the particular context

of computers within education. It is more directly concerned with the

process of learning itself and I will consider it in the final section

of the present overview of issues.

(iii) The social quality of learning

Kreuger, Karger and Barwick (1989) identify the solitary quality

of much microcomputer-based learning when they warn against the

cultivation of 'thought in isolation' (p. 113). They comment: 'What is

learned, then, is passivity and alienation from oneself and others,

and that the most fruitful relationships with people will be as

passive and impersonal as the solitary interaction with the computer'

(114). Cuban (1986) expresses similar concern:

In the fervent quest for precise rationality and technical

efficiency, introducing to each classroom enough computers to tutor

and drill children can dry up that emotional life, resulting in

withered and uncertain relationships.

This worry over the isolation of the computer-based *learner* is

commonplace in critical commentaries of how educational technology is

currently deployed (eg. Baker, 1985; Moore, 1993).

In respect of one problematic issue discussed above - the slow uptake

of innovative practices based on new technology - Cuban further

suggests a tension that exists for teachers confronting pressure to

develop computer-based work. That tension arises from their

perception of how the technology is to be used: its deployment seems

at odds with a strong professional commitment to the *interpersonal*

quality of education.

Reflecting on the models of implementation reviewed above, it is

apparent that such concerns are well founded. The computer-as-tutor

metaphor seems quite explicit in its implication that the teacher's

role might be vulnerable to substitution. In its popular

representation, this is the technology of the journalist's "desktop"

teacher. It is the image that will haunt any teacher who encountered

the stultifying example of the behaviorist's teaching machine.

However, the perceptions of journalists are stimulated by the

commentary of educationalists themselves. The often-cited view

expressed by Suppes still has some appeal:

One can predict that in a few more years millions of school children

will have access to what Philip of Macedon's son Alexander enjoyed as

a royal prerogative: the personal services of a tutor as well-informed

and responsive as Aristotle. (Suppes, 1966, p. 207).

Surely another misjudged prediction, but the popular interest in

computers as substitutes for the services of a tutor remains real

enough.

The computer-as-pupil metaphor also leaves underspecified the place of

social interaction between teachers and learners. This style of

computer use is firmly located within the discovery learning tradition

of educational theory - as has been commented above in relation to the

ubiquitous example of Logo. The interactive opportunities of much

open-ended software of this kind may be particularly effective in

sustaining pupils' task engagement. This, in turn, may suggest less

tutorial involvement on the part of a teacher. Thus, Papert's

presentation of Logo in *Mindstorms* (1980) includes sparse

consideration of how teachers do participate productively within Logo

learning. The stronger impression that is felt from his account is

one of the affordances within the software for spontaneous discoveries

on the part of the independent learner. We shall see that this

dislocation of the activity from the interpersonal dimension of

learning is not what was intended by Papert. But perhaps his vigorous

challenge to the "controlling" image of traditional educational

software was bound to shift concerns towards stressing the *autonomy*

of the learner.

We shall see that the image of the solitary learner sketched in the

quotations earlier has been resisted by teachers in practice. This is

certainly the case in early education, where the usual impression is

one of computers being absorbed into the familiar bustle of primary

class life. The main strategy to achieve this has been a preference

to organize computer activities as group work (Jackson et al, 1986,

1988). However, the "social" character of effective educational

experience is a more delicate quality than that conveyed by bustle

alone. The view to be developed later in this book is that computers

require us to consider more carefully the nature and scope of

"collaboration" as it may be organized within education. In

particular, to judge how far the social energy that might be visible

in classrooms incorporates a kind of interaction that has great

significance for learning and cognitive development. To pursue this,

we must outline a theoretical perspective that puts the social

dimension at the centre point of educational experience. This will be

taken up in the next chapter.

 

 

CONCLUDING COMMENTS

 

In the very short space of the past 10 years, considerable effort has

been invested in establishing the microcomputer as a significant

resource within education. This effort has been evident within

primary, secondary and tertiary sectors. There must be few

educational media that have been promoted with such energy (and

funding). The reasons for this seem to be a mixture of two

commitments on the part of educational politicians and practitioners.

First, it is considered important that children should now encounter

powerful new information technologies at school - where they may

become comfortable and confident with the medium before entering a

working world where it so pervasive. The second reason for all this

effort rests on a belief that the technology can transform learning

and teaching across a variety of existing curriculum areas: it is a

very general educational resource.

There is a current of opinion amongst close observers of this scene

that the impact so far has been modest. One measure of a shortfall is

the poor outcome of training efforts aimed at encouraging teachers to

incorporate new technology into their work. Many teachers are slow to

gain confidence in making active use of computers. Perhaps there is a

(fatal) irony in asking teachers - a profession where self-assurance

about what one knows and does is paramount - to incorporate a

technology of such patent complexity. Mastering it only at the level

of running isolated applications is bound to feel unsatisfactory,

especially in the (all too likely) event of unscripted technical

problems.

In my own experience, it has always seemed that the educational

application of new technology can too often have a "bolted on" feel of

just that kind. Some of the literature reviewed in this chapter lends

support to this as a general view of present practice. The

educational thinking that lies behind the main categories of

application do little to challenge this setting apart of computer

activities. If the computer is conceptualized as another kind of

tutor, it is highly likely to be put to work independently of any

human tutors in the environment - such an economy is the implicit

advantage in this conception. However, the other models of computer

implementation that have been discussed here invite similar

marginalising. In those cases, the problem may arise from the

powerful interactivity the medium offers: its simply too easy for

pupils to sustain independent activity.

I am drawing attention to the way in which computer-based learning

might readily become decoupled from the mainstream of classroom life.

In fact, we may have mixed reactions to this. So, for example, to

point out that a setting for learning is effective in sustaining

independent activity might be taken to define a quite desirable state

of affairs. Yet, I believe that many practitioners will also become

uneasy about learning that readily excludes the involvement of others

- teachers or peers. Surveys of teachers reacting to the introduction

of classroom computers tend to support this (eg. Bliss et al, 1986;

Lichtman, 1979; Woodrow, 1987). Cuban (1986), in particular, traces

much of the suspicion among teachers to this concern for the isolating

property of computer-based learning. I believe our thinking about

this tension between independent and socially-organized learning needs

to be guided by an overarching theoretical perspective. I shall offer

such a framework in the following chapter.