Research into expressions of attitudes about drugs - What use is it?

Sections

Summary
Introduction and purpose
Methods and sample description
Results
Discussion

Details

Author: Nicholas DORN, , Anne THOMPSON
Pages: 19 to 33
Creation Date: 1976/01/01

Research into expressions of attitudes about drugs - What use is it?

Nicholas DORN,
Anne THOMPSON
Evaluation and Research Unit, Institute for the Study of Drug Dependence, London NW6 1XA, England.

Summary

The authors draw attention to an assumption often implicit in presentation and utilisation of attitude data, that attitudes are the cause of behaviour. They then present attitude data collected from English schoolchildren with various levels of familiarity with drugs - these data could be interpreted in many ways and are compatible with the hypothesis that pupils' social experiences and behaviour may precede and influence their attitudes. It is suggested that the concept of the "transitional process" between socially available roles may be useful to researchers, evaluators and programme planners.

Introduction and purpose

Although knowledge about a population's behaviour and attitudes towards illegal drugs is a necessary prerequisite for formulation of one's policy vis-a-vis that population, such knowledge in itself is not a sufficient basis for action. The policy-maker will be aware of the need for a definition not only of the current situation but also of the desired situation, and of how the former might be changed into the latter. Agreement on values and social goals may not exist among the various persons and groups to whom the policy-maker holds himself accountable - yet in spite of this lack of consensus on the preferred state of affairs, there may be a degree of agreement about disvalued aspects of an existing situation (Braybrooke and Lindblom, 1963) Collection of information about the existing situation may be a prerequisite not only for drawing up the various policy options, but also for establishing sufficient agreement about the relative values of these options compared with the revealed status quo. Thus, any useful discussion of action must start with a discussion not of values and goals in the abstract, but of the status quo. This paper attempts to do this in relation to a particular population.

However, it is not possible simply to decide to collect "knowledge" about a situation, since the products of one's discovery procedures reflect one's focus and methods of enquiry as well as one's interpretation of the results of the respondent-researcher interaction. In terms of focus of enquiry, it seems reasonable to concentrate on those areas that would assist policy-makers to choose between the options that might be available to them. The methods of enquiry should be such that statements about the generalisability and meaning of results can be made. The interpretation of results should discuss not only the meaning that the researcher prefers to attach to the results, but also any alternative meanings that the policymaker, by reason of his system of values, may prefer. The interpretation and discussion of results should also, however, aim to prevent the consumer of research from routinely interpreting the data in a way that the researcher considers improper. If the researcher questions some of the assumptions underlying the reader's interpretation and drawing of implications from the results, and if the reader; having understood the basis for such questioning either modifies of affirms support for his own original assumptions and interpretations, then the researcher has done his duty as researcher.

This preamble describes an "ideal type" of policy-orientated research rather than research as it is generally carried out. However, to have such a description in mind may be useful when reading research reports. The data described below were obtained as part of an exercise that approximated very loosely to the "ideal type" of research. The focus of enquiry, drug behaviour and attitudes toward illegal drugs by teenagers, evolved from a string of decisions and contingencies that had more to do with the emotional concerns of policy-makers and researchers than with ways of choosing between practical policy options. The method of enquiry, confidential questionnaires in schools, limited the depth and range of enquiry and ensured that results would be characterised by certain features which we discuss later. Our interpretation of results is both partial and partisan, reflecting our concern that attitudinal data has been combined with dubious background assumptions to provide support for theoretically unsound programmes, particularly in the educational area. We wish to take the opportunity, whilst presenting attitude data, of drawing attention to, if not successfully controverting, these assumptions.

Methods and sample description

The data reported here were collected as part of the second half of a one-year longitudinal and trend study in 29 state secondary schools. The data was collected from classroom groups of approximately 30 pupils at a time by fieldworkers who visited the schools. Ages of fieldworkers ranged from 24 to 27 years, and they were middle-class in dress and manner, conforming to a rule of "no straights, no freaks". They were introduced to pupils by the teacher, who remained in the room whilst questionnaires were filled in but did not approach the pupils and took no active part in the proceedings. When questionnaires were completed, they were collected by the field workers and sealed in envelopes. Guarantees of absolute confidentiality were given. The full questionnaire took approximately half an hour to complete, but only some of the answers are reported here (full report: Dorn and Thompson, 1975).

The schools were located in the South, the Midlands and the North of England, but are not claimed to be representative of English schools in general, whatever "representative" may be taken to mean. Given the difficulties of deciding on the criteria for representativeness and given that such criteria will change with the purpose and subject of the study, we doubt whether a claim of representativeness would be very helpful. Indeed, since drug phenomena may have little to do with informal aspects for which national norms are not available, representativeness may be an improper goal. For the purposes of delineation of the samples, however, we show the geographical areas, urban/rural locations, and formal designations of schools visited.

TABLE 1

Geographical distribution of schools visited in relation to distribution of secondary school pupils over 13 years of age (geographical areas and distribution of pupils both derived from the Registrar General)

 

Schools where fifth formers questioned in 1974

Area

Percentage of all pupils over 13 in England and Wales

Number

Percentage

London
13 5 17
S.E. (excluding London)
20 7 24
Midlands
18 6 21
S.W. and Wales
13 0 0
East Anglia
3 1 3
N.W.
14 3 10
N. and N.E.
17 7 24
  98 29 99

TABLE 2

Urban/rural locations of schools visited

 

Schools where fifth formers questioned in 1974

Urban
14 48
Sub-urban
9 31
Rural
6 21
All
29 100

TABLE 3

Types of schools visited

 

Mixed

Boys

Girls

All

 

Number

Percentage

Number

Percentage

Number

Percentage

Number

Percentage

Grammar
        2 7 2 7
Comprehensive *
16 55 1 3 2 7 19 65
Secondary modern
6 21 2 7     8 28
All
22 76 3 10 4 14 29
100

*NOTE. Previously pupils were selected at the age of 11 on the basis of an examination to go to a grammar (higher ability) or a secondary modern (lesser ability) school. These two types of selective schools are new being replaced by non-selective comprehensive schools.

Respondents were 1,796 male and female schoolchildren, mostly of ages 15 and 16 years old, but with some 17 years, in the fifth year of the secondary system (which commences at age 10 or 11 years). The fifth year is the last compulsory year of schooling, and those staying on beyond 16 years do so voluntarily. Sex, age and social class (by parental occupation) are described in tables 4, 5 and 6, and the teacher's estimates of ability of the pupils as groups are shown in table 7.

Some attitudinal and behavioural data are shown in tables 8 to 11.

DEMOGRAPHIC CHARACTERISTICS AND OTHER ATTRIBUTES OF THE FIFTH FORM SAMPLE (Sample size is 1796)

TABLE 4

Sex of pupils

Boys
43 per cent
Girls
56 per cent
N/R
1 per cent

TABLE 6

Social class of pupils (Percentages)

Executive
11
Supervisory
13
Routine white collar
14
Skilled manual
34
Unskilled manual
10
N/R
18

TABLE 8

Age at which pupils would like to leave school

Age
Percentage
15 31
16 39
17 11
18 +
14
N/R
1

TABLE 5

Age of pupils

15 52 percent
16 46 percent
17+
2 percent
N/R
1 per cent

TABLE 7

Ability level of pupils (Percentages)

High (average grammar)
3
Medium (grammar borderline, upper sec. modern)
18
Medium/low (average sec. modern)
31
Low (low sec. modern)
13
Unstreamed
29
N/R
6

TABLE 9

How well pupils feel teachers understand them (Percentages)

Very well
5
Quite well
56
Not very well
25
Very little
12
N/R
1

TABLE 10

Number of friends pupils have who smoke cigarettes (Percentages)

All
8
Most
38
About half
20
Not many
26
None
8
N/R
0

TABLE 11

Amount of cigarettes pupils smoke (Percentages)

Never tried
23
Once or twice
28
Occasionally
17
Less than 10 a week
5
More than 10 a week
5
More than 20 a week
22
N/R
1

Apparently almost a third of pupils wished to leave school at 15 years rather than at 16 years old (16 being the earliest age at which they would be permitted to leave), and 25 per cent wished to stay on beyond 16 years. Most pupils say that their teachers understand them quite well. Only a third say that not many or none of their friends smoke cigarettes. About a quarter of the pupils claim never to have tried a cigarette, just under half say that they have smoked once or twice or occasionally. A tenth smoke up to 20 cigarettes per week and just under a quarter smoke more than 20 cigarettes a week. These data give some idea of the kind of teenagers questioned.

Results

FAMILIARITY WITH DRUGS

Pupils were asked whether they knew any drug users, whether they had ever been offered a drug and whether they had ever taken a drug, and combination of the data from these questions allowed us to describe each pupil's familiarity with drugs. In the sample under consideration (data obtained from responses to several separate questions), 57.9 per cent of pupils said that they did not know any drug-taker, and 42.1 per cent did know one or more persons they recognised as a taker of illegal drugs. Of those who know a user, a quarter (10.5 per cent of sample) had taken a drug at some time, and another quarter (12.2 per cent of sample) had refused a drug at some time (we should bear in mind that these are not mutually exclusive categories, since one can refuse a drug at one time and accept it or another drug at another time). We can translate our data into mutually exclusive categories as in table 12.

More important than the percentages relating to the whole sample is the fact that the relative sizes of the four familiarity groups showed great variation from school to school. For instance, 46 per cent of pupils sampled in one (exceptional) school reported ever having used an illegal drug, but in some school groups no pupil reported such experience.

In this paper, we treat self-reported drug use as indicative of actual drug use. Several investigators have, as a result of surveying the evidence relating to validity of such questions concluded that, for example, "The evidence supports what has thus far been an assumption on the part of many researchers in the area: that there is reason to have confidence in the validity and reliability of self-reports of drug use. Studies... seem likely to have produced estimates of drug use rates that are at worst on the conservative side without too much over-reporting" (Whitehead and Smart, 1972). Whilst we ourselves have made no systematic study of this problem, experience of conducting fieldwork, interviewing students and studying protocols brings us to a similar conclusion about our work.

TABLE 12

Familiarity with illegal drugs (Number of pupils: 1796)

 

Percentage

Knowing no drug taker
57.9
Knowing one or more takers but not being offered any drug
23.5
Being offered a drug(s) on one or more occasions, but never accepting
6.3
Being offered a drug(s) on one or more occasions and accepting one or more of these offers
10.5
No response
1.8
TOTAL
100

NOTE. This table is based on pupils' reports of their level of contact with an illegal drug or drugs. Although our data allows us to say that the drug with which pupils most commonly have personal experience (i.e. acceptance of an offer) is cannabis, we have no information regarding the drugs used by other children with whom the pupil may be in contact.

Whilst the breakdown into groups with different levels of familiarity with drugs (above) and comparison of breakdowns for individual schools (Dorn and Thompson 1975) are of interest, our principal concern in introducing these categories in this paper is to use them in reporting student attitudinal data, broken down by level of familiarity with drugs.

EXPRESSED EXPECTATIONS OF PLEASURE AND OF ADDICTION AND PERCEPTIONS OF DANGEROUSNESS

Pupils were asked to evaluate the pleasurableness, addictiveness and dangerousness of cigarettes, cannabis, sedative pills, L.S.D., heroin and stimulants pills. They were asked "If you took any of these things next week-end, would you expect it to be pleasant or unpleasant when you took it?" and subsequently, "Would you expect to become addicted and have to keep on taking it again and again, or not?", and "You may think all these drugs are equally dangerous or that some are very dangerous and others less so. Please consider each drug in turn and use ticks to indicate how dangerous you think it is (when not from a doctor)." The data are shown in tables 13 to 18.

Generally speaking, expressed expectations of pleasure following use of any substance were fairly low. Even among those who have used a drug (mostly cannabis), only one in two pupils say they would expect cannabis to be "pleasant".

There is a positive relationship between expectations of pleasure from use of any substance on the one hand, and level of familiarity with illegal drugs on the other. The more contact a pupil has had with illegal drugs, the more likely he or she is to express an expectation that use of any substance would be pleasant rather than unpleasant. The strength of the relationship between expectations of pleasure from cannabis use and familiarity with an illegal drug is particularly striking: (4 per cent of the "no-contact" group would expect cannabis use to be pleasant and 71 per cent would expect it to be unpleasant, while 52 per cent of the group who report past use of an illegal drug would expect cannabis use to be pleasant and 18 per cent would expect it to be unpleasant). The comparative strength of this relationship may be partly explained by the fact that the drug most commonly reported as being used is cannabis.

Rather than the ever-users having a unique expectation of degree of pleasure to be obtained from future use of substances, we find that there is a progressive increase in expectations of pleasure as we move from the no-contact group, through the pupils who know a user, through those who have been offered and refused, to the ever-users. This suggests that the tendency to report expectations of pleasure is related to degree of familiarity with drug-takers and drugs rather than to personal experience of drug effects: pupils may be forming these expectations on the basis of past social behaviour, rather than personal experience of drug effects. One might have expected students who have rejected offers to be more likely to hold expectations of an unpleasant experience than students who have not been offered a drug. About a quarter of all pupils would certainly not expect to become addicted if they tried a drug, irrespective of the type of drug and experience of the pupil. Approximately one-quarter to one-third feel that they probably or certainly would become addicted if they tried any substance.

The greater the pupil's familiarity with an illegal drug or drugs, the more likely he or she is to say that he/she would certainly expect to become addicted to cigarettes. Although almost all of the group of ever-users of illegal drugs (93.1 per cent) have smoked at least once, a higher proportion of this group than of the other familiarity groups report an expectation of certain addiction (21 per cent of ever-users as opposed to 7 per cent of the no-contact group). Interviews we have carried out in the course of a separate project suggest that pupils may report dependence on cigarettes even though they may have smoked only a few times and this may function as a rationalisation or "excuse" for current smoking behaviour. (Although there is a positive relationship between familiarity with illegal drugs and expectations of certain addiction to cigarettes, it is not the case that the majority of ever-users expect to become addicted to cigarettes, merely that a higher proportion of this group than of the other familiarity groups have this expectation.)

It will be noted that those with no contact are more likely than others to say that they would certainly not become addicted to any drugs. This may or may not reflect a feeling that being offered or taking any drug (and thus having the chance of becoming addicted) is inconceivable.

The more familiarity pupils report having had with an illegal drug or drugs (the drug most often reported as having been used being cannabis), the less dangerous pupils feel cannabis to be: the belief that cannabis is not very dangerous or fairly harmless increases progressively with familiarity with illegal drugs. Of ever-users of an illegal drug 60 per cent say that cannabis is "fairly harmless" or "not very dangerous" as opposed to only 5 per cent of the no-contact group. Only 14 per cent of pupils who have ever used an illegal drug think cannabis can kill, compared with half of those who report no-contact with illegal drugs or drug takers. However, the relationship observed between pupils' familiarity with an illegal drug and their expressed beliefs about dangerousness of cannabis does not extend to beliefs about substances other than cannabis; the four familiarity groups do not differ consistently in their evaluations of dangerousness of drugs other than cannabis.

TABLE 13

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction, perceptions of dangerousness in relation to cigarettes

 

Report use

Have been offered

Know takers

No contact

Pleasant
63 53 39 23
Partly pleasant, partly unpleasant
13 18 25 24
Unpleasant
12 20 19 32
Could be either pleasant or unpleasant
12 9 16 22
  100 100 99 101
Certainly become addicted
21 18 11 7
Probably become addicted
19 20 19 14
Might become addicted
18 29 26 29
Probably not become addicted
19 17 25 24
Certainly not become addicted
23 16 20 27
  100 100 101 101
Very dangerous, can kill
14 9 9 9
Very dangerous
4 3 4 5
Dangerous
29 20 34 39
Not very dangerous
36 45 40 35
Fairly harmless
17 23 14 12
  100 100 101 100

TABLE 14

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction,' perceptions of dangerousness in relation to cannabis

 

Report use

Have been offered

Know takers

No contact

Pleasant
52 20 11 4
Partly pleasant, partly unpleasant
19 18 12 6
Unpleasant
18 46 49 71
Could be either pleasant or unpleasant
12 16 27 19
  101 100 99 101
Certainly become addicted
11 16 18 17
Probably become addicted
12 16 16 19
Might become addicted
23 26 27 17
Probably not become addicted
23 16 14 9
Certainly not become addicted
31 27 26 37
  100 101 101 99
Very dangerous, can kill
14 27 32 48
Very dangerous
9 25 27 31
Dangerous
17 21 26 16
Not very dangerous
30 17 11 3
Fairly harmless
30 10 5 2
  100 100 101
100

TABLE 15

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction, perceptions of dangerousness in relation to sedatives

 

Report use

Have been offered

Know takers

No contact

Pleasant
17 15 11 3
Partly pleasant, partly unpleasant
36 22 25 16
Unpleasant
26 44 39 57
Could be either pleasant or unpleasant
21 19 26 23
  100 100 101 99
Certainly become addicted
7 8 7 8
Probably become addicted
18 15 16 17
Might become addicted
28 28 30 23
Probably not become addicted
22 29 23 20
Certainly not become addicted
25 20 24 33
  100 100 100 101
Very dangerous, can kill
15 13 15 19
Very dangerous
12 15 17 22
Dangerous
35 38 43 38
Not very dangerous
27 27 20 18
Fairly harmless
11 7 5 4
  100 100 100 101

TABLE 16

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction, perceptions of dangerousness in relation to LSD

 

Report use

Have been offered

Know takers

No contact

Pleasant
21 8 5 3
Partly pleasant, partly unpleasant
18 13 11 7
Unpleasant
35 58 59 75
Could be either pleasant or unpleasant
26 20 26 16
  100 99 101 101
Certainly become addicted
13 20 16 151
Probably become addicted
22 22 23 21
Might become addicted
25 29 24 19
Probably not become addicted
18 9 12 9
Certainly not become addicted
22 19 26 36
  100 99 101 100
Very dangerous, can kill
30 39 39 45
Very dangerous
28 31 35 35
Dangerous
30 26 24 18
Not very dangerous
6 2 1 1
Fairly harmless
7 2 2 1
  101 100 101
100

TABLE 17

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction, perceptions of dangerousness in relation to heroin

 

Report use

Have been offered

Know takers

No contact

Pleasant
9 8 4 3
Partly pleasant, partly unpleasant
18 8 9 5
Unpleasant
46 66 64 76
Could be either pleasant or unpleasant
26 18 23 17
  99 100 100 101
Certainly become addicted
24 31 22 21
Probably become addicted
30 18 21 18
Might become addicted
19 22 20 16
Probably not become addicted
8 9 11 8
Certainly not become addicted
20 20 26 37
  101 100 100 100
Very dangerous, can kill
62 65 63 62
Very dangerous
21 21 24 26
Dangerous
13 12 12 10
Not very dangerous
1 1 1 1
Fairly harmless
3 2 1 1
  100 101 101 100

TABLE 18

Percentages of pupils with particular levels of familiarity with an illegal drug or drugs who report expectations of pleasure, expectations of addiction perceptions of dangerousness in relation to stimulants

 

Report use

Have been offered

Know takers

No contact

Pleasant
31 14 10 5
Partly pleasant, partly unpleasant
24 26 17 10
Unpleasant
26 41 48 64
Could be either pleasant or unpleasant
19 19 25 20
  100 100 100 99
Certainly become addicted
9 10 10 13
Probably become addicted
19 20 21 20
Might become addicted
28 36 30 20
Probably not become addicted
22 14 16 12
Certainly not become addicted
21 21 23 36
  99 101 100 101
Very dangerous, can kill
16 18 23 30
Very dangerous
16 25 31 35
Dangerous
35 33 34 27
Not very dangerous
24 21 8 7
Fairly harmless
9 3 3 2
  100 100 99
100

Although many pupils unfamiliar with illegal drugs believe that cannabis can kill, and a fair proportion (between 30 per cent and 45 per cent of all familiarity groups) believe LSD can kill, few pupils believe that cigarettes (9-14 per cent of each group) and sedatives (13-19 per cent of each group) can kill. These beliefs may be at odds with the facts as understood by many workers in the field, and are certainly different from the authors' beliefs: LSD cannot kill (although pupils may have been referring to deaths caused by accident) whereas cigarettes can kill (slowly) and so can sedatives (quickly, by overdose).

REASONS GIVEN FOR AND AGAINST "PEOPLE" TAKING DRUGS

Pupils were asked to choose the three most likely reasons for "people" taking and being put off drugs from two lists of 13 alternatives, based on responses to an open-ended question used in previous research (Swift, Dorn, Thompson, 1974). A problem with interpretation of answers to these questions is that the referent "drugs" may be taken to refer to different substances or classes of substances by different pupils (tables 19-20).

We can see that whilst enjoyment is given as one of the three reasons for taking drugs by half the ever-users, it is given by a quarter or less of other pupils. "To see what it is like" is given almost as often by ever-users and by all other pupils. "Can't cope with life" is given by only a quarter of the ever-users, but more often by other pupils. It would appear that ever-users and other pupils would agree about the role of curiosity, but the ever-users would cite enjoyment where the other pupils would cite inability to cope.

TABLE 19

Percentages of pupils who select, from a list of alternatives, particular reasons why some people take drugs

 

Ever-users

Offered

Know takers

No contact

Number of pupils
188
114
422
1040
Enjoy it
48 25 19 18
To see what it's like
45 41 45 42
Because friends do it
25 36 34 39
Can't get on with people
10 11 15 13
Can't cope with life
26 40 45 47
Makes them more of a person
4 5 3 4
Because it's offered
10 3 4 6
To help them mix with people
9 4 8 5
Fed up
8 10 11 10
To get away from their worries
31 41 36 41
To rebel
11 7 7 5
Bored
18 10 14 14
Showing off
16 25 26
23

TABLE 20

Percentages of pupils who select, from a list of alternatives, particular reasons why some people do not take drugs

 

Ever-users

Offered

Know takers

No contact

Number of pupils
188
114
422
1 040
Effects might be unpleasant
47 34 39 34
Don't want to become addicted
45 52 50 44
Makes them less of a person
6 10 9 8
Because it's illegal
13 11 7 10
Bad for their health
18 18 18 30
Damages their minds
22 31 29 28
Might have an accident
9 5 5 5
Don't want to hurt family and friends
19 11 16 13
Don't need them
21 25 22 19
Have other things to
7 7 5 5
Don't want to ruin life
18 31 35 34
Might die
15 18 24 29
It's expensive
31 18 14 9

This disagreement about the role of pleasure is also apparent in the reasons offered against drug use. Almost half the ever-users, but rather less of the other pupils, cite unpleasant effects as a reason for not taking drugs. About half of each familiarity group cites addiction. Three out of ten of ever-users, as opposed to between one and two out of ten of other pupils, cite expense as a reason against. Rather than expense, non-users cite "don't want to ruin life".

The reasons given more often by the ever-users ("pleasure" as a reason for taking; "unpleasant effects", "expense" as a reason for not taking) have to do with the immediate situation and experience, whilst reasons given by those with less familiarity with drugs have to do with potential longer-term hazards. "Bad for health" is given as a reason for not using drugs by three out of ten of the no contact group, but fewer of the other pupils. Illegality is rarely offered as a reason for not using by any of the pupils.

Discussion

These data show that when asked to formulate and report expectations of drug effects and rationales for and against use, those pupils with personal experience of drugs tend to focus on immediate pleasurable/unpleasurable effects, whilst those who have less or no familiarity with drugs and drug-takers, tend to focus on inability to cope and on longer-term consequences, such as "ruins life". The greatest disagreement is over the drug that is most likely to be taken if anything is taken-cannabis. The greater the degree of familiarity with drugs, the more pleasant and less likely to "kill" cannabis appears to the pupil. On the other hand, the pupils who have taken an illegal drug in the past are also more likely than the pupils who are less familiar with drugs to report expectations of certain addiction to cigarettes, and slightly more likely to report that cigarettes can kill. (This is not to say that the majority of the ever-users expect certain addiction, nor believe that cigarettes can kill, rather, that more of this group than of the other familiarity groups do so.) These findings are interesting in the light of the facts that ever-users of illegal drugs are also more likely than other pupils to smoke cigarettes (indeed to smoke a lot of cigarettes) and also more likely to expect cigarette smoking to be pleasant.

What conclusions can be drawn from these results? We suggest that there is no clear conclusion to be drawn. We cannot simply say that ever-users of illegal drugs have an over-all tendency to "whitewash" their actions - they are softer on pot but harder on nicotine than other pupils (even though they are likely to use and to say that they anticipate pleasure from the use of both substances). There are several possible explanations for this: ever-users may be less likely to say cannabis can kill and more likely to say that cigarettes can kill because they are better informed than non-users; or ever-users may say that nicotine is dangerous because, by pointing out the relative dangers of a legal drug, they place cannabis in a relatively more positive light; or both these explanations, or some other explanations may apply.

It is sometimes assumed that drug-users use drugs because they hold certain attitudes, and that non-users are prevented by their attitudes from using drugs. It is this assumption that our attitudes are responsible for our behaviour, and that a more anti-drug attitude (sometimes dressed up as a "sensible" or "healthy" attitude) makes drug use less likely, that underpins much current drug education. However, the idea that attitudes lead to behaviour is an idea derived from "common sense" rather than from social science and the evidence for it is extremely weak. What evidence there is suggests that the relationship between characteristics of the individual and behaviour is more complex than this, and that a particular behavioural choice cannot be attributed purely to the characteristics of an individual and abstracted from the social context in which it occurs (for a review see Warner and DeFleur, 1969).

There is also evidence that the relationship may be the other way around: behaviour change may precede and shape attitude change. These considerations caution us from assuming, just because a person who has behaved in a particular way expresses a particular attitude, that the attitude came first or "caused" the behaviour.

So we can see that although "common sense" would lead us to accept that drug use is caused by something "inside" the person (either a disease state or an "unhealthy" attitude), this is not necessarily the case, and serious enquiry into why and how people came to take a drug is pre-empted by acceptance of such assumptions. We must adopt a more critical, enquiring stance, noting that if factors A and B tend to occur together, then A may cause B, B may cause A, or A and B may both be caused by a third factor or set of factors, and that the system of intertwining factors within which these relations occur may be in a stable state or in flux, rendering the relationships relatively permanent or transitory.

The sentiments expressed in the questionnaire answers are at least partially the result of the pupils' perception of the expectations of the fieldworkers and of the norms of the institution (school) in which the answers are given. As far as we could establish by observation and talking with pupils, we were viewed as separate from, but holding values in harmony with the schools, and the values of the school as seen by pupils were more anti-drug than pro-drug. Thus, we expect all the reasons, expectations, etc. given above, to have been influenced towards the antidrug and to reflect rather more culturally-approved beliefs about drugs than would have been the case had those reasons and expectations been given out of school and to a non-adult group. The question of whether these expressions of sentiment are the cause of, the result of, or in a reciprocal relationship with the levels of familiarity with drugs, is irresolvable within this project. But we definitely cannot make the automatic assumption that expectations and attitudes and reasons offered for and against drug use have in the past or will in the future determine behaviour.

We would hypothesise, first, that sentiments expressed in the school environment are poor predictors of behaviour in the out-of-school environment; second, that behaviour in any situation, whether it be verbal response in the school research situation or drug use in the peer situation, is largely situationally determined according to the normative constraints that then seem to apply. These are hypotheses which cannot be tested with the data reported here, but they are hypotheses that require urgent attention. For if, as we believe, self-reported drug attitudes of the kind reported here neither directed students' past behaviour nor will be a reliable guide to their future behaviour, then many drug education and intervention programmes are predicated upon a false assumption. In our opinion, researchers who present attitudinal data in a manner that allows themselves or consumers of their work to interpret this data as predictive of behaviour, without either establishing the validity of such interpretation nor using cautions against it, perform a disservice for programme planners and policy-makers. Whilst we take the unusual position of accepting at face value our self-reported drug use data whilst looking askance at our attitude data, we recognise the importance of behavioural prediction on a sound theoretical and empirical basis (e.g. Ajzen and Fishbein, 1969). Successful prediction of future behaviour, we suggest, may rest more on prediction of the subject's perception of the actual choice-situation, when it occurs, than on measurement of his or her willingness to express normative beliefs about drugs and drugtakers.

Both the data presented here and by Hindmarch et al. (1975) are consistent with the hypothesis that respondents tend to attribute to themselves (and to others) attitudes that "fit" or make sense of their social behaviour and experience (Kelley, 1967). An hypothesis that attitudes or belief systems internal to the individual may act as an effective "barrier" to certain forms of behaviour is less than convincing if those individuals who do nevertheless come to exhibit the proscribed behaviour have to be explained away as being inconsistent or lacking in cognitive integrity (Hindmarch et al., 1975, pp. 32). Such a hypothesis may have wide appeal due to its attribution of consistency and integrity (rather than lack of adequate opportunity or greater exposure to normative constraints) to non-users. However, we should not consider a hypothesis tested and proven merely because it makes us feel good.

Attaining any particular position vis-a-vis drugs (including legal drugs) is the result of a number of "transitional processes" (Dorn and Thompson, 1975), from not knowing a user, to coming to know a user, to being offered the substance, to accepting an offer. These steps may span many years, may all occur within sixty seconds, or may never occur. Further (alternative) transitional process include becoming a regular user, irregular user or ex-user, becoming a treatment or social work client, being "busted", becoming a dealer, etc. The transitional processes are the linking mechanisms between the various roles available to the person, and require the co-operation of a wide range of other persons who, deliberately or not, sustain such roles. Understanding of the transitions cannot be based upon the reasons people give for their past experience and behaviour nor upon their current attitudes, since these reasons and attitudes may largely be shaped by the need to make sense of, and/or present as desirable the past as viewed from the present. Transitional processes cannot be understood by reference only to the individual who was a participant and is now a product of them: we have to look at nature of each transition, as viewed by the individuals currently at those points.

These considerations open up the possibility of a major reorientation of perspective by social researchers, evaluators, and programme planners. The following become questions of interest: "What transitional processes are relevant for the population in question?". "What is the structure of the network of roles within which the transitions occur?". "What is the nature of the constraints at each choice-point in this network and how might the balance of constraints at a particular choice-point be modified by a careful transitional intervention?". The value of asking such questions is at present unknown.

Bibliography

Ajzen I., and M. Fishbein. The Prediction of Behavioural Intentions in a Choice Situation, in Thomas K., Attitudes and Behaviour, Penguin, England, 1971. (See also Fishbein M. (Ed.), Readings in Attitude Theory and Measurement, John Willey & Sons, New York/London/Sydney, 1967.)

Braybrooke D., and C. Lindblom. A Strategy of Decision: Policy Evaluation as a Social Process, Free Press, New York/Collier-Macmillan, London, 1963.

Dorn N., and A. Thompson. A Comparaison of 1973 and 1974 Levels of Experimentation with Illegal Drugs in Some Schools in England, pp. 88, ISDD, London, 1975.

Hindmarch I., I. Hughes, and R. Einstein. Attitudes to Drug Users and to the Use of Alcohol, Tobacco and Cannabis on the Campus of a Provincial University, Bulletin on Narcotics, vol. XXVII: 1,pp. 27-36, 1975.

Kelley H. H. Attribution Theory in Social Psychology, in Levine D. (Ed.) Nebraska Symposium on Motivation. Lincoln U. of Nebraska Press, 1967.

Swift B., N. Dorn and A. Thompson. Evaluation of Drug Education, pp. 296, ISDD, London, 1974.

Warner G. L., and M. L. DeFleur. Attitude as an Interactional Concept: Social Constraint and Social Distance as Intervening Variables Between Attitudes and Action, pp. 153-169, Amer. Sociol. Rev., vol. 34, No. 2, April 1969.

Whithehead P. C., and R. G. Smart. Validity and Reliability of Self-Reported Drug Use, in S. Einstein and S. Allen (Eds.), Student Drug Surveys, Farmingdale, N. Y. Baywood, 1972.