KNOWLEDGE ABOUT WHERE TO FIND A PRELIMINARY ANALYSIS
HELP:
ALEC G. PEMBERTON Trinity College, Oxford, U.K.
and F. A. WHITLOCK University of Queensland. Brisbane. Australia
and I’. R. WILSON University of Queensland. Brisbane, Australia Abstract-Medical sociology has neglected the theoretical and practical importance of community member’s knowledge about where to find help for their problems. This is a report of a preliminary analysis of data from an Australian study of knowledge about helping facilities. Replies of survey respondents to questions probing their level of information about where to find help for psychiatric, marital and financial problems were analysed. This revealed a surprisingly high and quite specific level, of information. A “knowledge of resources” index was constructed and further cross-tabulation revealed a relationshio between knowledge and a number of social variables. Soecificallv. socio-economic status was related to knowledge: higher status persons were more likely io be high scorers on the knowledge index than low social status respondents.
Do members
of the community know where help for psychiatric and personal problems?
to find
In this paper we intend to look at an important but sorely neglected aspect of the field of social problemsknowledge about where to seek help in time of distress. One of the most rapidly growing areas in the sociology of medicine and welfare has been in the domain that David Mechanic called “illness behaviour and the help-seeking process”. Something of the essence of this work ma) be gauged from his statement that
Lll: On the most simple and obvious level, it is plain that symptoms are differentially perceived, evaluated and acted upon (or not acted upon) by different kinds of people and
in different kinds of situations. In social psychiatry. some of the principal theoretical and empirical concerns have been to study knowledge about illness and the recognition of symptoms. attitudes towards medical care and health agencies. patient-physician relations and the sick role [Z]. For social workers. recent interest has focussed on what might be called a “theory of clients” concerned with the following problems: who are they and what are their characteristics’? Does social status (or age, religion or sex) influence their decision to seek help and their selection of a particular agency?[3] However. while such research has emphasised the problematical nature of psychiatric treatment or social work intervention-seeking help is not seen as an “automatic”. inevitable response to illness or stress-little or no work has been done on whether. in the first place. persons know where to find any assistance they might require. or further. whether information about the
availability of treatment is differentially distributed in the population-for example in certain socio-economic, ethnic or age groups. A review of the literature in the sociology of medicine reveals a dearth of empirical studies on the problem. Indeed, Mechanic devotes only a few lines to the topic [4]: it is probable that once having defined the need for treatment, patients in the various class groups have different information and knowledge concerning how to arrange psychiatric intervention.
Yet it is clearly no trivial matter: knowledge about where to find assistance for mental and personal problems is an initial and indispensable step in the process of receiving help. Would members of the community know where to seek the services of a psychiatrist if the need arose? Do people know where to turn in case of a marital or financial crisis‘? And, equally important, what are the social correlates of knowledge: that is, what is the influence of socioeconomic status or age or sex? In our opinion it is essential that social researchers tackle these issues. for they have crucial implications at both the immediate, everyday level for the laymen in trouble and for administrators and policy-makers concerned with the planning and provision of services to the community at large. The data reported here are intended as a provisional first step towards providing some evidence on these questions. the figures coming from a preliminary analysis of some information gathered in a wider project on social pathology (crime. suicide, mental illness and road accidents) in two suburbs in Brisbane. an Australian city. 433
43-I
A.
THE C.ARA\ AbGREEN SCOPE
G.
PEMBERTOV.
F.
A.
WHITLOCK
and
P.
R. WILSOV
END STLDY-
AND METHODS
Our data come from an investigation conducted during 1970 and 1971 of two Brisbane suburbs -Green End (population 9500 1971 Census) a predominantly middle class area and Caravan (population 8 500 197 I Census) a working class. largely State Housing Commission suburb. Caravan and Green End were singled out for study because their markedly different socio-economic composition made them ideally suited for an examination of the relationship between socio-cultural factor+especially social class, our principal concern-and various forms of aberrant behaviour. Green End is a middle class suburb. attractive and modern. occupying a distinctive physical space (between ranges of hills) on the outskirts of Brisbane. the administrative capital of the state of Queensland. It is an area with a unique suburban identity, fostered by the lively local progress association and by real estate firms. Green End is a pleasant, rapidly growing area and a large proportion of its population are young adults with young children. On the other hand, Caravan is a working class suburb with a fairly stable population, of which the biggest proportion is in the age range of 35-55 yr and with children in an older age group than those ai Green End. Set in the suburban sprawl of Brisbane, it is a nondescript kind of place. flat and treeless. taking its character from the caravan park and from the Housing Commission estate that forms a large part of it. Datu collrction
Two principal forms of data collection were employed, a psychiatric census and a social survey. The Psychiatric Morbidity Census was conducted to estimate the prevalence of reported mental impairment in the two suburbs. This involved a careful counting of all new cases of mental disturbance within the period of one year to be found at the local suburban general practitioners, private psychiatrists in practice in Brisbane city, out-patient departments of general public hospitals. public psychiatric outpatient facilities and public psychiatric hospitals in the Brisbane metropolitan area. A research social worker was employed for the specific purpose of ensuring an accurate and comprehensive enumeration of the prevalence of treated mental illness in the two places. Initial analysis of these data reveal striking differences in numbers from the two locations with more than twice as many cases coming from Caravan than were found at Green End. A social survey based on a questionnaire interview with samples from the two suburbs was the other main form of data gathering. Sumpliy: Following extensive pilot testing of the schedule and training of the interviewers to deal with “sensitive” topics like crime and insanity. two large quota samples were drawn from the residents-500 persons from Caravan (a response rate of 62”,:), 596 from Green End (73’1,; response rate) [S]. Each interviewer was assigned to a street-block within a suburb, with instructions to begin interviewing at a specified place on that block and to work clockwise around it. Only persons between 21 and 70yr of age were
to be interviewed [6]. In order to maximise the size of samples, interviewers were instructed to obtain as many interviews as possible on their assigned block, but to make only one interview per household. They were encouraged to work during the evenings and weekends when it would be most likely to find the whole family-especially working males, at home, and they were told to make as many “call-backs” as possible when people were not found at home on the first
Knowledge about where to find help
435
3
appeared to be a slight trend for Green End residents to be more knowledgeable than their counterparts KNcmLEDGE ABOUT LVHERF TOFIND HELP PclR PSYCHIx~LTRIC m&RITPL from Caravan about where to find help for marital AND F1NliNCD.L PROBLEPIS: -VW lWlD and financial problems. GREEN MDSAIIPLES Unfortunately there are no comparative figures ,perCS%tC+e*, +ither Australian or overseas-with which to P.YChl*t=LC Pr‘ObLzmn llaritll ProblemSFlnMS*alProblUU match these data. This of course raises the whole question of what is an acceptable or desirable level of community knowledge about such matters. Given the extent of mental impairment in most Western industrial societies, between one person in ten and one person in twelve will be hospitalised in their life time, it looks as though community members will be quite well placed in terms of knowing where to turn for help-or so the Caravan-Green End results would suggest. But it is not an easy question to answer. In the case of mental illness (though the point applies attempt. Most important. interviewers were told to equally to marital and financial problems) part of the obtain a balance across age and sex categories and answer lies in the experiences and understanding of they were given a card on which. at the conclusions mental health professionals; whether, in fact. they feel of the interview. they recorded the age and sex of there is an adequate level of knowledge about where the respondent. The social composition of the sample to find psychiatric treatment among the general pubis to be found in Tables 1 and 2 [7]. lic. Part of the answer, also, lies in epidemiological T/w questionnaire. Items on the questionnaire research to assess the number of persons in the comcovered a range of issues including the recognition munity with untreated illnesses because of their own and evaluation of simulated cases of psychiatric dislack of knowledge of where to seek aid or because order. experiences with the mentally ill and other their family or friends are uninformed. social problems such as crime, marital discord and Following the query “would you know where to road accidents, as well as the questions probing find a psychiatrist?’ the obvious question to ask next sample member’s knowledge about where to find help is; “where would you go?” [9] Consequently, sample for psychiatric. marital and financial troubles. The members who had answered “yes” were asked. standard personal and demographic questions completed the questionnaire. Our analysis opens with a What would you do? How would you go about finding consideration of the respondents’ level of information a psychiatrist‘? about where to obtain help. TABLE
Their replies are given in Table 4 and it will be most fruitful to consider these figures in the light of the
RESULTS
A good way to begin is with a simple tabulation of replies to the question Now. suppose you did have a problem and wanted to talk to a psychiatrist. Would you know where to find one? and to our queries about knowledge of where to find help for marital and financial problems [S]. The data are presented in Table 3 and the important thing to note is the high level of knowledge claimed by respondents. Also it is interesting to see that while there were no differences between Caravan (the blue collar area) and Green End (the mostlv middle class suburb) on information about psychiatric services. there
sou*ce
Asklocal or familyd0Ct.X t0 r+cononend one cc t0 the neare*tP”bXC hospital IOC.1government p*ychiatricclinic COn*“ltDipSyChiatrl*r directly Other,don’r*“OX+ NOLlnSYe*
caTa”.S” (7
GreenEnd 58
19
?
II
12
1,
13
07:1
,,::,
patterns of actual help-seeking observed from the Psychiatric Morbidity Census of the two suburbs. 19% of the Caravan sample compared to only 7% from Green End stated that they would go to the nearest public hospital and these suburban differences were reflected in the figures for the actual use of psychiatric services--Caravan residents were much more likely to be found at general public hospitals than persons from Green End. Similarly, as Table 4 shows, sample members from Caravan (14%) and Green End (1276) were about equally likely to have said they would go to the nearest government psychiatric clinic, a situation matched by their actual patterns of use of that service in Brisbane. On the other hand, although 14:; from Caravan and 13:,2 from Green End said they would consult a private psychiatrist directly, the Morbidity Census prevalence figures showed that, in fact, residents from Green End were much more likely to be found in treatment with a urivate usvchiatrist. There were suburban differences observed i’n the frequency with which respondents stated they would use their local or family doctor to refer them to a psychiatrist. This is by far the most popular resource (for sample members from both places) and it is very likely that the area differences (47% at Caravan and 58y0 from Green End) express the greater likelihood of residents of the mostly middle class area. Green End. being members of one of the Australian subsidised. contributory health schemes which requires
436
h. G. PEMBERTON. F. A. WHITLOCKand P. R. WILSOS
that consultation with a specialist such as a psychiatrist be referred through a general practitioner. Finally. it is pleasing to find that only two respondents (one from each suburb) said they “didn’t know” where they would go to find a psychiatrist (they have been placed in a single category with the “other” responses and “no answers” in Table 4). It is likely then that the figures in Table 3 represent an accurate picture of the respondents’ knowledge about where to find a psychiatrist. When pressed for details of how they would, in fact, do so, the overwhelming majority of those who had claimed knowledge were able to describe the specific steps they would take. These data give good reason to believe that in a psychiatric emergency (and we would argue. in times of marital and financial distress) sample members would be likely to take steps to obtain help which are in line with their verbal behaviour [lo]. To sum up, it seems that for the Australian situation at least, Kadushiti was being unduly pessimistic when he speculated that: individual psychotherapists and psychiatric clinics are quite unknown to the general public and merely finding a good therapist (or any therapist) is a serious problem for psychiatric patients [I 11. The data from the Caravan-Green End study indicate a relatively high level of information about where to locate psychiatric treatment and help for marital and financial problems. Some 70:; of the total sample claimed they would know where to find a psychiatrist if they needed one and further questioning revealed a wide and realistic range of options favoured by respondents, a pattern which generally matched the actual help-seeking among patients located by the Psychiatric Census. THE KNOWLEDGE
INDEX
So much then for the overall pattern of responses to each of the knowledge items. The immediate question is: what does this mean for an individual’s general level of knowledge about where to seek help? Does it mean that a person who knows where to find a psychiatrist. for example. will also know where to seek help in a marital crisis or, if the need arises. find financial assistance? In short, it is the problem of whether a single dimension or common .element runs through the three items-knowledge about various kinds of facilities. If there is consistency in. the response pattern we would expect that those who had indicated a high level of knowledge on one item to be more likely to indicate a high level of knowledge on any other item than those indicating low knowledge on the first item [12]. Consequently we tested for item homogeneity by calculating item total correlations for each item with the total score over all items [13]. This provides us with a measure of the extent to which individuals who responded “positively” to any one item responded similarly to the other items. Respondents were scored 1 for each time they said “yes” to one of the three knowledge items, all other. responses were scored 0. Thus they could score from zero to three on the index; a 0 represents a respondent who was not able to say he could find a psychia-
TP.BLE
5
trist. marital help or financial assistance. while a 3 represents the “knowledgeables” on all three items. Then a set of item-total correlations were calculated to determine the extent to which there was a correlation between item score and total score. Table 5 presents the item-total correlations and they are all significantly different from zero at better than the 001 level of confidence-fairly good evidence for concluding that a single dimension (knowledge about facilities) runs through the items. The distribution of respondents over the Index is presented in Table 6. As can be seen, only 63 respondents (69; of the total) scored 0 and a further 203 (99);) scored 1. For the purposes of analysis the index will be trichotomised: those who scored 0 and 1 will be combined into a single category. Obviously the distribution is skewed toward the “high” knowledge end of the index. A significant proportion of the Caravan-Green End samples (459,) would know where to find help for all three kinds of problems covered in the survey, while another third of the respondents would know where to go for help for at least two of the problems. Lastly, it is worth mentioning that there were area differences on the index: 30”,, of the Caravan sample were low scorers while 2?, from Green End were in that category or. put the other way around, 41”~ from Caravan were high scorers compared with 494, from Green End. SOCIO-ECONOMIC
STATIJS 4ND
SCORE ON THE KNOWLEDGE INDEX
Many studies have demonstrated the salience of socio-economic status for a range of health-related &haviour [14], so consequently we tested to ascertain whether it was operating in our sample to influence knowledge of facilities. First. scores on the knowledge index were cross-tabulated with the respondent’s educational level (the details are in Table 7). The influence of education is clearly visible. Higher status respondents-the ones who had attended or TABLEh DISTRIBUTION OFSCORES ONTHEKNOKEDCE INDEX iN = 1,096)
Knowledge about TABLE
7
completed university-were most likely to be high scorers. followed by those with secondary schooling and then, the group least likely to be high on the knowledge index, those with only primary school education. But the effects of suburban residence showed through as well; across each level of education, Green End respondents were most likely to be high scorers than their Caravan counterparts. Predictably, similar results were obtained when the cross-tabulation was run again for the influence of the occupational level of the head of the respondent’s household (Table 8). Sample members from professional and administrative home backgrounds were most likely to be high scorers, next were respondents from other white collar households (such as clerical workers) and, least likely to be high scorers were those from blue collar backgrounds. Once again though. the influence of suburban residence was apparent. The data on socio-economic status and knowledge about facilities may be summed-up thus: the higher the respondent’s socio-economic position, the greater the likelihood of him knowing where to seek help in times of a psychiatric, marital or financial emergency. Thus the data lend useful empirical weight to Kaplan’s commonsense but unsupported assertion that. not only are highly educated people more likely to define their problems in mental health terms and more likely to be favourably ing help
disposed
towards
seek-
the decision to take such action is facilitated because more highly educated people are also likely to have information on the availability of resources [ 161.
where
437
to find help
For many members of the working class obtaining help for psychiatric. marital or financial problems will be made difficult simply because of their lack of knowledge of the whereabouts of the appropriate facilities. The implications of this should not be missed: if an adequate response is to be made to social problems such as mental illness, marital discord and poverty. then mental health and welfare institutions must become more visible to those in need of their services. Because of their lack of knowledge of such things, blue collar members of the community may be unlikely to make good use of the existing resources -or so it appears from the Caravanereen End data. Finally it is worth mentioning some other social correlates of knowledge about helping resources [I 73. Cross-tabulation revealed that sex. national origin and amount of contact with the mentally ill were found to be associated with knowledge about where to seek help. Women, the Australian born and those who had relatives or friends who had been treated for a mental or nervous disorder were the most likely to be high scorers. On the other hand, the respondent’s age, whether or not they grew up in a rural or urban environment and whether they had ever needed professional help for a problem were all unrelated to scores on the knowledge index.
SUMMARY
AND CONCLUSIONS
In summing up, we would want to make the point that the issues dealt with here are important in both a theoretical and practical sense. Despite the methodological limitations of the Caravan-Green End study, it is abundantly clear from the data that the issue of knowledge about where to find a psychiatrist and other kinds of help is a crucial one. In their writing and research, sociologists have been concerned to make the point that obtaining treatment depends very much on the perception of the seriousness of symptoms and the recognition of illness. Yet in the case of both “self and “other” defined illness (to borrow Mechanic’s terms) whether or not people actually know where to obtain aid is clearly a vital factor in the help-seeking process. But it is a problem that has largely been ignored by both academics and practitioners alike. There is no mention of knowledge of resources in the two recent landmark studies in medical sociology-the Wadsworth Butterfield and Blaney study 01 the perception of physical illness and the use of ser\ IL’C‘S in London. or the investigation of the quality and ,l(.livery of psychiatric services in .New York by Kolb .III,I his associates [lS]. Overall, the level ol hnowlcdge about where to find a psychiatrist and help for marital and financial problems was quite high among the Caravan-Green End sample members. Indeed, a rather pleasing number (some 707; of the total sample) said they would know where to consult a psychiatrist (and seek aid in times of marital or financial crisis). Furthermore. when respondents were asked specifically where they would obtain help (for the three sorts of problems) they were able to provide the interviewer with accurate and realistic details of the steps they would take-such as asking their local or family doctor to recommend one.
43x
A. G. PFNBERTO~. F. A. WHITLOCI< and P. R.
The three questions probing levels of information were used to form an “index of knowledge of facilities”. Cyoss-correlation of the index with divers social variables revealed that knowledge about the various helping facilities was related to social status; respondents high on the socio-economic ladder were more likely to be knowledgeable than persons of lower social standing in the community. Because they may be relatively uninformed about where to find the appropriate services. blue collar members of the community may have considerable difficulty in obtaining help for their problems-or so the Caravan-Green End data suggests. The methodological limits of the present studyrestricted as it is to two suburbs in a single Australian city-make it imperative that further research be done on broader based. more representative samples. Data is required urgently from samples drawn from class and ethnic groups in a range of Western industrial countries to supplement these Australian figures. In addition, there is a need to investigate whether community members share the same meanings of terms like “problems”. “marital difficulties” and the like. It may well be that these general labels conceal real and important shades of meaning (with significant consequences for action) among the general public. Finally. (keeping in mind the tentative nature of the findings) there is, as Hempel reminds us, a “why” question to be asked; it is not enough in social science to demonstrate a relationship between phenomena. they require an explanation as well. Why are higher status members of the community more likely than people of lower social rank to be knowledgeable about where to find help’! One possibility worth exploring is that white collar families, especially those from the professional and managerial tiers of society. tend to have the highest exposure to mass media sources of mental hygiene information that advocate psychiatric consultation for troubles of a socio-emotional kind. In this context, information is likely to be available about the location of psychiatric and other services. Thus white collar people are more likely to be “psychologically-minded”‘(as the mental health jargon goes) and to know where to find help--but this is very much an hypothesis for further testing rather than an empirically grounded statement of fact [ 191. We might also try to account for the socio-economic differences in knowledge in terms of the attitudes and opinions of blue collar classes. For instance, it is possible that the residents of Caravan with its older and less mobile population. may have a “lay-referral” or informal support system which reduces their need for knowledge about and demand for formal services [ZO]. Hollingshead and Redlich in their original study in the 1950s on social class and mental disorder advanced a more provocative thesis. Writing about the relevance for psychiatric treatment of lower class attitudes towards community facilities they said; “A deep-seated distrust of authority pervades class V persons from childhood to old age. Suspicion is directed toward police. clergymen. teachers. doctors. public otlicials, public health nurses, and social workers. Institutions for care of the disabled and the ill are believed to be run for money and one has to have
Wrrso\
“pull” to get into them” [-11]. These remarks Marc written about the .American situation nearI> 30 !r ago. qet they may provide some insights into the relatively low level of knobvledge which blue collar people have about helping st‘r\ US: their negati\c i‘upt‘rlences and attitudes arc unlikeI) to lead them actliely to seek out information about these facilities. But this too is clearly a matter to be pursued in furthcl research. REFERE\CES
1 1. Mechanic.
D. whirl.: and Freidson. E. T/NJ Pro/c,.\\~or~ Chap. 13. Dodd Mead. New York. 147.: Mayer. J. E. and Timms. N. Tilt, Client SprtrLs: Ilo~+ r,xl Cluss /,,I~OJSSIO,I.SO/ C~/.X,>I~OI.I\Chap I. Routledge & Kegan Paul. 1970: and Timms. N. (Editor) Tilt, Rrcuicir~q Ed. Routledge & Kegan Paul. 1973. Mechanic. D. op. clr.. p. 150. The somewhat lower than usual response rate wab most probably due to two factors beyond the control of the researchers. At the time the survey was conducted there was conslderable unfa.vourablc publicIt> in the mass media about so-called “Intrusions” from social surveys (particularly those carried out under commercial auspices). Perhaps even more important was the bad publicity in the press given to the unlversity because of disruptions caused b! student radicals. The lower limit of 71 qr of age was decided upon to facilitate data analysis with regard to one of the pnnclpal concerns of the project. crimes against households. It was felt that ?I yr of age was the most hkely lower age limit for heads of households and home owners. though it is recognised here that other forms of social deviance such as mental illness and marital breakdown may be found in younger age ranges. Because most of the interviewing was carried out b) student interviewers keen to do as many interviews as possible. much of the work was done in the daytlme. consequently. the final samples were somewhat underrepresentative of males in the suburbs. slightly underrepresentative of the lower socio-economic ranks and over-weighted for persons higher up the social scale at Caravan and Green End. and. tinall>. adequate on the age and religious distributions. The precise wording of the yuestlon was: la) ‘.Whut about a husband and wife having marital difficulties” If they were friends of yours. would you be able to tell them where they could receive help if they asked you”‘? (b) “Now. imagine you were in financial dificulties. Would you know where to go for help”‘? Similar follow-up questions were asked of those respondents who claimed knowledge about where to find help for financial and marital problems. The data analysis revealed (as was the case for knowledge 01 psychiatric treatment) that respondents were able to nominate specific and realistic services for both marital and financial ditiicultxs. The data are available from the authors on request. As Deutscher has recently pointed out however. surve! rescarchers must always be on guard against gaps between respondent’s attltudcs or opinions and thruactual behavlours: \CL‘ D~utschcr. I. (Fditor) Cl’lrclr Proplt~ %I.. CZ/lrrl Pcoplc Do. Scott & Foresman. Glenvllle. Illinois. 1973. Kadushin. C. U’/I \‘ PCC~I/C,(;o 77, P~~c~/~rc~rrr,sr\ p. 3 Ii Athcrton. New York. IYhY. Clearly. if we cannot dcmonstratc thts consistency In response WC cannot assert with much confidence that we arc examining a unitary scale of knowledge about where to tind help. oJ’.Lfctiicirw
3
3 5
6
,
‘.
9.
10,
,, I’.
Knowledge about where to find help 13. Guildford, J. P. Fur&~mvntal Statistics in Psychology ur~d Edwcrriorl p. 502. McGraw-Hill. New York. 1965. 14. For a review. see Fried, M. Social differences in mental health. In Pouerry and Health: A Sociological A~~alysis (Edited by Kosa, J. ~‘t al.) pp. 1I3- 167. Harvard University Press, Cambridge, 1969. 15. Our occupational ranking is derived from Broom, L. et al. An occupational ranking of the Australian work force. Ausr. N.Z. J. Social. (I) supplement, 1965. 16. Kaplan. H. B. Mental Illness as a Social Problem. In Haudhook on thr Study of Social Problems (Edited by Smigel. E. 0.) p. 342. Rand McNally, Chicago. 1971. I?. These data are available from the authors on request. At a later date it is intended to pursue the analysis in greater depth. For example in the preliminary analysis presented in this paper, tests of significance have not been used because (as the “Columbia School” researchers point out) it is important in the exptoratory phase of research not to limit productive lines of enquiry by insisting that the findings prove to be significant; in other words, when the task at hand is
439
to explore trends and to generate hypotheses. 11 is not fruitful to disregard or omit results because they fail to meet stringent criteria of significance. Set Coleman. .J. S. Statistical Problems. In Union Dernocruc~ (Edited by Lipset. S. M. et nl.) pp. 427-432. Fret Press. New York. 1956. 18. Wadsworth. M. at rrl. Heu/rh arld Sickr~rss: 2% Chorw of Trwrwm. Tavistock. London. 1961: and Kolb. L. ~‘r(11.Urban Challrnyes to Psychiatry. Little. Brown & Co.. Boston, 1969. 19. Data on socio-economic differences in exposure to mass media messages including mental health information are available in; Rossi. P. H. and Blum. Z. D. Class. Status and Poverty. In 011 Undcvsfundiug Pouerry (Edited by Moyniham, D. P.) p. 53. Basic Books. New York, 1969. 20. See Meyer and Timms. op. cir., for a discussion of “lay referral” in the British context. We are indebted to Dr. Peter McEwan for bringing this to our attention. 21. Hollingshead. A. B. and Redlich. F. C.. So&l Class and Mrntu/ fllwss. p. 130. Wiley, New York. 1967: and see also pp. 176 and 181.
HELP:
ALEC G. PEMBERTON Trinity College, Oxford, U.K.
and F. A. WHITLOCK University of Queensland. Brisbane. Australia
and I’. R. WILSON University of Queensland. Brisbane, Australia Abstract-Medical sociology has neglected the theoretical and practical importance of community member’s knowledge about where to find help for their problems. This is a report of a preliminary analysis of data from an Australian study of knowledge about helping facilities. Replies of survey respondents to questions probing their level of information about where to find help for psychiatric, marital and financial problems were analysed. This revealed a surprisingly high and quite specific level, of information. A “knowledge of resources” index was constructed and further cross-tabulation revealed a relationshio between knowledge and a number of social variables. Soecificallv. socio-economic status was related to knowledge: higher status persons were more likely io be high scorers on the knowledge index than low social status respondents.
Do members
of the community know where help for psychiatric and personal problems?
to find
In this paper we intend to look at an important but sorely neglected aspect of the field of social problemsknowledge about where to seek help in time of distress. One of the most rapidly growing areas in the sociology of medicine and welfare has been in the domain that David Mechanic called “illness behaviour and the help-seeking process”. Something of the essence of this work ma) be gauged from his statement that
Lll: On the most simple and obvious level, it is plain that symptoms are differentially perceived, evaluated and acted upon (or not acted upon) by different kinds of people and
in different kinds of situations. In social psychiatry. some of the principal theoretical and empirical concerns have been to study knowledge about illness and the recognition of symptoms. attitudes towards medical care and health agencies. patient-physician relations and the sick role [Z]. For social workers. recent interest has focussed on what might be called a “theory of clients” concerned with the following problems: who are they and what are their characteristics’? Does social status (or age, religion or sex) influence their decision to seek help and their selection of a particular agency?[3] However. while such research has emphasised the problematical nature of psychiatric treatment or social work intervention-seeking help is not seen as an “automatic”. inevitable response to illness or stress-little or no work has been done on whether. in the first place. persons know where to find any assistance they might require. or further. whether information about the
availability of treatment is differentially distributed in the population-for example in certain socio-economic, ethnic or age groups. A review of the literature in the sociology of medicine reveals a dearth of empirical studies on the problem. Indeed, Mechanic devotes only a few lines to the topic [4]: it is probable that once having defined the need for treatment, patients in the various class groups have different information and knowledge concerning how to arrange psychiatric intervention.
Yet it is clearly no trivial matter: knowledge about where to find assistance for mental and personal problems is an initial and indispensable step in the process of receiving help. Would members of the community know where to seek the services of a psychiatrist if the need arose? Do people know where to turn in case of a marital or financial crisis‘? And, equally important, what are the social correlates of knowledge: that is, what is the influence of socioeconomic status or age or sex? In our opinion it is essential that social researchers tackle these issues. for they have crucial implications at both the immediate, everyday level for the laymen in trouble and for administrators and policy-makers concerned with the planning and provision of services to the community at large. The data reported here are intended as a provisional first step towards providing some evidence on these questions. the figures coming from a preliminary analysis of some information gathered in a wider project on social pathology (crime. suicide, mental illness and road accidents) in two suburbs in Brisbane. an Australian city. 433
43-I
A.
THE C.ARA\ AbGREEN SCOPE
G.
PEMBERTOV.
F.
A.
WHITLOCK
and
P.
R. WILSOV
END STLDY-
AND METHODS
Our data come from an investigation conducted during 1970 and 1971 of two Brisbane suburbs -Green End (population 9500 1971 Census) a predominantly middle class area and Caravan (population 8 500 197 I Census) a working class. largely State Housing Commission suburb. Caravan and Green End were singled out for study because their markedly different socio-economic composition made them ideally suited for an examination of the relationship between socio-cultural factor+especially social class, our principal concern-and various forms of aberrant behaviour. Green End is a middle class suburb. attractive and modern. occupying a distinctive physical space (between ranges of hills) on the outskirts of Brisbane. the administrative capital of the state of Queensland. It is an area with a unique suburban identity, fostered by the lively local progress association and by real estate firms. Green End is a pleasant, rapidly growing area and a large proportion of its population are young adults with young children. On the other hand, Caravan is a working class suburb with a fairly stable population, of which the biggest proportion is in the age range of 35-55 yr and with children in an older age group than those ai Green End. Set in the suburban sprawl of Brisbane, it is a nondescript kind of place. flat and treeless. taking its character from the caravan park and from the Housing Commission estate that forms a large part of it. Datu collrction
Two principal forms of data collection were employed, a psychiatric census and a social survey. The Psychiatric Morbidity Census was conducted to estimate the prevalence of reported mental impairment in the two suburbs. This involved a careful counting of all new cases of mental disturbance within the period of one year to be found at the local suburban general practitioners, private psychiatrists in practice in Brisbane city, out-patient departments of general public hospitals. public psychiatric outpatient facilities and public psychiatric hospitals in the Brisbane metropolitan area. A research social worker was employed for the specific purpose of ensuring an accurate and comprehensive enumeration of the prevalence of treated mental illness in the two places. Initial analysis of these data reveal striking differences in numbers from the two locations with more than twice as many cases coming from Caravan than were found at Green End. A social survey based on a questionnaire interview with samples from the two suburbs was the other main form of data gathering. Sumpliy: Following extensive pilot testing of the schedule and training of the interviewers to deal with “sensitive” topics like crime and insanity. two large quota samples were drawn from the residents-500 persons from Caravan (a response rate of 62”,:), 596 from Green End (73’1,; response rate) [S]. Each interviewer was assigned to a street-block within a suburb, with instructions to begin interviewing at a specified place on that block and to work clockwise around it. Only persons between 21 and 70yr of age were
to be interviewed [6]. In order to maximise the size of samples, interviewers were instructed to obtain as many interviews as possible on their assigned block, but to make only one interview per household. They were encouraged to work during the evenings and weekends when it would be most likely to find the whole family-especially working males, at home, and they were told to make as many “call-backs” as possible when people were not found at home on the first
Knowledge about where to find help
435
3
appeared to be a slight trend for Green End residents to be more knowledgeable than their counterparts KNcmLEDGE ABOUT LVHERF TOFIND HELP PclR PSYCHIx~LTRIC m&RITPL from Caravan about where to find help for marital AND F1NliNCD.L PROBLEPIS: -VW lWlD and financial problems. GREEN MDSAIIPLES Unfortunately there are no comparative figures ,perCS%tC+e*, +ither Australian or overseas-with which to P.YChl*t=LC Pr‘ObLzmn llaritll ProblemSFlnMS*alProblUU match these data. This of course raises the whole question of what is an acceptable or desirable level of community knowledge about such matters. Given the extent of mental impairment in most Western industrial societies, between one person in ten and one person in twelve will be hospitalised in their life time, it looks as though community members will be quite well placed in terms of knowing where to turn for help-or so the Caravan-Green End results would suggest. But it is not an easy question to answer. In the case of mental illness (though the point applies attempt. Most important. interviewers were told to equally to marital and financial problems) part of the obtain a balance across age and sex categories and answer lies in the experiences and understanding of they were given a card on which. at the conclusions mental health professionals; whether, in fact. they feel of the interview. they recorded the age and sex of there is an adequate level of knowledge about where the respondent. The social composition of the sample to find psychiatric treatment among the general pubis to be found in Tables 1 and 2 [7]. lic. Part of the answer, also, lies in epidemiological T/w questionnaire. Items on the questionnaire research to assess the number of persons in the comcovered a range of issues including the recognition munity with untreated illnesses because of their own and evaluation of simulated cases of psychiatric dislack of knowledge of where to seek aid or because order. experiences with the mentally ill and other their family or friends are uninformed. social problems such as crime, marital discord and Following the query “would you know where to road accidents, as well as the questions probing find a psychiatrist?’ the obvious question to ask next sample member’s knowledge about where to find help is; “where would you go?” [9] Consequently, sample for psychiatric. marital and financial troubles. The members who had answered “yes” were asked. standard personal and demographic questions completed the questionnaire. Our analysis opens with a What would you do? How would you go about finding consideration of the respondents’ level of information a psychiatrist‘? about where to obtain help. TABLE
Their replies are given in Table 4 and it will be most fruitful to consider these figures in the light of the
RESULTS
A good way to begin is with a simple tabulation of replies to the question Now. suppose you did have a problem and wanted to talk to a psychiatrist. Would you know where to find one? and to our queries about knowledge of where to find help for marital and financial problems [S]. The data are presented in Table 3 and the important thing to note is the high level of knowledge claimed by respondents. Also it is interesting to see that while there were no differences between Caravan (the blue collar area) and Green End (the mostlv middle class suburb) on information about psychiatric services. there
sou*ce
Asklocal or familyd0Ct.X t0 r+cononend one cc t0 the neare*tP”bXC hospital IOC.1government p*ychiatricclinic COn*“ltDipSyChiatrl*r directly Other,don’r*“OX+ NOLlnSYe*
caTa”.S” (7
GreenEnd 58
19
?
II
12
1,
13
07:1
,,::,
patterns of actual help-seeking observed from the Psychiatric Morbidity Census of the two suburbs. 19% of the Caravan sample compared to only 7% from Green End stated that they would go to the nearest public hospital and these suburban differences were reflected in the figures for the actual use of psychiatric services--Caravan residents were much more likely to be found at general public hospitals than persons from Green End. Similarly, as Table 4 shows, sample members from Caravan (14%) and Green End (1276) were about equally likely to have said they would go to the nearest government psychiatric clinic, a situation matched by their actual patterns of use of that service in Brisbane. On the other hand, although 14:; from Caravan and 13:,2 from Green End said they would consult a private psychiatrist directly, the Morbidity Census prevalence figures showed that, in fact, residents from Green End were much more likely to be found in treatment with a urivate usvchiatrist. There were suburban differences observed i’n the frequency with which respondents stated they would use their local or family doctor to refer them to a psychiatrist. This is by far the most popular resource (for sample members from both places) and it is very likely that the area differences (47% at Caravan and 58y0 from Green End) express the greater likelihood of residents of the mostly middle class area. Green End. being members of one of the Australian subsidised. contributory health schemes which requires
436
h. G. PEMBERTON. F. A. WHITLOCKand P. R. WILSOS
that consultation with a specialist such as a psychiatrist be referred through a general practitioner. Finally. it is pleasing to find that only two respondents (one from each suburb) said they “didn’t know” where they would go to find a psychiatrist (they have been placed in a single category with the “other” responses and “no answers” in Table 4). It is likely then that the figures in Table 3 represent an accurate picture of the respondents’ knowledge about where to find a psychiatrist. When pressed for details of how they would, in fact, do so, the overwhelming majority of those who had claimed knowledge were able to describe the specific steps they would take. These data give good reason to believe that in a psychiatric emergency (and we would argue. in times of marital and financial distress) sample members would be likely to take steps to obtain help which are in line with their verbal behaviour [lo]. To sum up, it seems that for the Australian situation at least, Kadushiti was being unduly pessimistic when he speculated that: individual psychotherapists and psychiatric clinics are quite unknown to the general public and merely finding a good therapist (or any therapist) is a serious problem for psychiatric patients [I 11. The data from the Caravan-Green End study indicate a relatively high level of information about where to locate psychiatric treatment and help for marital and financial problems. Some 70:; of the total sample claimed they would know where to find a psychiatrist if they needed one and further questioning revealed a wide and realistic range of options favoured by respondents, a pattern which generally matched the actual help-seeking among patients located by the Psychiatric Census. THE KNOWLEDGE
INDEX
So much then for the overall pattern of responses to each of the knowledge items. The immediate question is: what does this mean for an individual’s general level of knowledge about where to seek help? Does it mean that a person who knows where to find a psychiatrist. for example. will also know where to seek help in a marital crisis or, if the need arises. find financial assistance? In short, it is the problem of whether a single dimension or common .element runs through the three items-knowledge about various kinds of facilities. If there is consistency in. the response pattern we would expect that those who had indicated a high level of knowledge on one item to be more likely to indicate a high level of knowledge on any other item than those indicating low knowledge on the first item [12]. Consequently we tested for item homogeneity by calculating item total correlations for each item with the total score over all items [13]. This provides us with a measure of the extent to which individuals who responded “positively” to any one item responded similarly to the other items. Respondents were scored 1 for each time they said “yes” to one of the three knowledge items, all other. responses were scored 0. Thus they could score from zero to three on the index; a 0 represents a respondent who was not able to say he could find a psychia-
TP.BLE
5
trist. marital help or financial assistance. while a 3 represents the “knowledgeables” on all three items. Then a set of item-total correlations were calculated to determine the extent to which there was a correlation between item score and total score. Table 5 presents the item-total correlations and they are all significantly different from zero at better than the 001 level of confidence-fairly good evidence for concluding that a single dimension (knowledge about facilities) runs through the items. The distribution of respondents over the Index is presented in Table 6. As can be seen, only 63 respondents (69; of the total) scored 0 and a further 203 (99);) scored 1. For the purposes of analysis the index will be trichotomised: those who scored 0 and 1 will be combined into a single category. Obviously the distribution is skewed toward the “high” knowledge end of the index. A significant proportion of the Caravan-Green End samples (459,) would know where to find help for all three kinds of problems covered in the survey, while another third of the respondents would know where to go for help for at least two of the problems. Lastly, it is worth mentioning that there were area differences on the index: 30”,, of the Caravan sample were low scorers while 2?, from Green End were in that category or. put the other way around, 41”~ from Caravan were high scorers compared with 494, from Green End. SOCIO-ECONOMIC
STATIJS 4ND
SCORE ON THE KNOWLEDGE INDEX
Many studies have demonstrated the salience of socio-economic status for a range of health-related &haviour [14], so consequently we tested to ascertain whether it was operating in our sample to influence knowledge of facilities. First. scores on the knowledge index were cross-tabulated with the respondent’s educational level (the details are in Table 7). The influence of education is clearly visible. Higher status respondents-the ones who had attended or TABLEh DISTRIBUTION OFSCORES ONTHEKNOKEDCE INDEX iN = 1,096)
Knowledge about TABLE
7
completed university-were most likely to be high scorers. followed by those with secondary schooling and then, the group least likely to be high on the knowledge index, those with only primary school education. But the effects of suburban residence showed through as well; across each level of education, Green End respondents were most likely to be high scorers than their Caravan counterparts. Predictably, similar results were obtained when the cross-tabulation was run again for the influence of the occupational level of the head of the respondent’s household (Table 8). Sample members from professional and administrative home backgrounds were most likely to be high scorers, next were respondents from other white collar households (such as clerical workers) and, least likely to be high scorers were those from blue collar backgrounds. Once again though. the influence of suburban residence was apparent. The data on socio-economic status and knowledge about facilities may be summed-up thus: the higher the respondent’s socio-economic position, the greater the likelihood of him knowing where to seek help in times of a psychiatric, marital or financial emergency. Thus the data lend useful empirical weight to Kaplan’s commonsense but unsupported assertion that. not only are highly educated people more likely to define their problems in mental health terms and more likely to be favourably ing help
disposed
towards
seek-
the decision to take such action is facilitated because more highly educated people are also likely to have information on the availability of resources [ 161.
where
437
to find help
For many members of the working class obtaining help for psychiatric. marital or financial problems will be made difficult simply because of their lack of knowledge of the whereabouts of the appropriate facilities. The implications of this should not be missed: if an adequate response is to be made to social problems such as mental illness, marital discord and poverty. then mental health and welfare institutions must become more visible to those in need of their services. Because of their lack of knowledge of such things, blue collar members of the community may be unlikely to make good use of the existing resources -or so it appears from the Caravanereen End data. Finally it is worth mentioning some other social correlates of knowledge about helping resources [I 73. Cross-tabulation revealed that sex. national origin and amount of contact with the mentally ill were found to be associated with knowledge about where to seek help. Women, the Australian born and those who had relatives or friends who had been treated for a mental or nervous disorder were the most likely to be high scorers. On the other hand, the respondent’s age, whether or not they grew up in a rural or urban environment and whether they had ever needed professional help for a problem were all unrelated to scores on the knowledge index.
SUMMARY
AND CONCLUSIONS
In summing up, we would want to make the point that the issues dealt with here are important in both a theoretical and practical sense. Despite the methodological limitations of the Caravan-Green End study, it is abundantly clear from the data that the issue of knowledge about where to find a psychiatrist and other kinds of help is a crucial one. In their writing and research, sociologists have been concerned to make the point that obtaining treatment depends very much on the perception of the seriousness of symptoms and the recognition of illness. Yet in the case of both “self and “other” defined illness (to borrow Mechanic’s terms) whether or not people actually know where to obtain aid is clearly a vital factor in the help-seeking process. But it is a problem that has largely been ignored by both academics and practitioners alike. There is no mention of knowledge of resources in the two recent landmark studies in medical sociology-the Wadsworth Butterfield and Blaney study 01 the perception of physical illness and the use of ser\ IL’C‘S in London. or the investigation of the quality and ,l(.livery of psychiatric services in .New York by Kolb .III,I his associates [lS]. Overall, the level ol hnowlcdge about where to find a psychiatrist and help for marital and financial problems was quite high among the Caravan-Green End sample members. Indeed, a rather pleasing number (some 707; of the total sample) said they would know where to consult a psychiatrist (and seek aid in times of marital or financial crisis). Furthermore. when respondents were asked specifically where they would obtain help (for the three sorts of problems) they were able to provide the interviewer with accurate and realistic details of the steps they would take-such as asking their local or family doctor to recommend one.
43x
A. G. PFNBERTO~. F. A. WHITLOCI< and P. R.
The three questions probing levels of information were used to form an “index of knowledge of facilities”. Cyoss-correlation of the index with divers social variables revealed that knowledge about the various helping facilities was related to social status; respondents high on the socio-economic ladder were more likely to be knowledgeable than persons of lower social standing in the community. Because they may be relatively uninformed about where to find the appropriate services. blue collar members of the community may have considerable difficulty in obtaining help for their problems-or so the Caravan-Green End data suggests. The methodological limits of the present studyrestricted as it is to two suburbs in a single Australian city-make it imperative that further research be done on broader based. more representative samples. Data is required urgently from samples drawn from class and ethnic groups in a range of Western industrial countries to supplement these Australian figures. In addition, there is a need to investigate whether community members share the same meanings of terms like “problems”. “marital difficulties” and the like. It may well be that these general labels conceal real and important shades of meaning (with significant consequences for action) among the general public. Finally. (keeping in mind the tentative nature of the findings) there is, as Hempel reminds us, a “why” question to be asked; it is not enough in social science to demonstrate a relationship between phenomena. they require an explanation as well. Why are higher status members of the community more likely than people of lower social rank to be knowledgeable about where to find help’! One possibility worth exploring is that white collar families, especially those from the professional and managerial tiers of society. tend to have the highest exposure to mass media sources of mental hygiene information that advocate psychiatric consultation for troubles of a socio-emotional kind. In this context, information is likely to be available about the location of psychiatric and other services. Thus white collar people are more likely to be “psychologically-minded”‘(as the mental health jargon goes) and to know where to find help--but this is very much an hypothesis for further testing rather than an empirically grounded statement of fact [ 191. We might also try to account for the socio-economic differences in knowledge in terms of the attitudes and opinions of blue collar classes. For instance, it is possible that the residents of Caravan with its older and less mobile population. may have a “lay-referral” or informal support system which reduces their need for knowledge about and demand for formal services [ZO]. Hollingshead and Redlich in their original study in the 1950s on social class and mental disorder advanced a more provocative thesis. Writing about the relevance for psychiatric treatment of lower class attitudes towards community facilities they said; “A deep-seated distrust of authority pervades class V persons from childhood to old age. Suspicion is directed toward police. clergymen. teachers. doctors. public otlicials, public health nurses, and social workers. Institutions for care of the disabled and the ill are believed to be run for money and one has to have
Wrrso\
“pull” to get into them” [-11]. These remarks Marc written about the .American situation nearI> 30 !r ago. qet they may provide some insights into the relatively low level of knobvledge which blue collar people have about helping st‘r\ US: their negati\c i‘upt‘rlences and attitudes arc unlikeI) to lead them actliely to seek out information about these facilities. But this too is clearly a matter to be pursued in furthcl research. REFERE\CES
1 1. Mechanic.
D. whirl.: and Freidson. E. T/NJ Pro/c,.\\~or~ Chap. 13. Dodd Mead. New York. 147.: Mayer. J. E. and Timms. N. Tilt, Client SprtrLs: Ilo~+ r,xl Cluss /,,I~OJSSIO,I.SO/ C~/.X,>I~OI.I\Chap I. Routledge & Kegan Paul. 1970: and Timms. N. (Editor) Tilt, Rrcuicir~q Ed. Routledge & Kegan Paul. 1973. Mechanic. D. op. clr.. p. 150. The somewhat lower than usual response rate wab most probably due to two factors beyond the control of the researchers. At the time the survey was conducted there was conslderable unfa.vourablc publicIt> in the mass media about so-called “Intrusions” from social surveys (particularly those carried out under commercial auspices). Perhaps even more important was the bad publicity in the press given to the unlversity because of disruptions caused b! student radicals. The lower limit of 71 qr of age was decided upon to facilitate data analysis with regard to one of the pnnclpal concerns of the project. crimes against households. It was felt that ?I yr of age was the most hkely lower age limit for heads of households and home owners. though it is recognised here that other forms of social deviance such as mental illness and marital breakdown may be found in younger age ranges. Because most of the interviewing was carried out b) student interviewers keen to do as many interviews as possible. much of the work was done in the daytlme. consequently. the final samples were somewhat underrepresentative of males in the suburbs. slightly underrepresentative of the lower socio-economic ranks and over-weighted for persons higher up the social scale at Caravan and Green End. and. tinall>. adequate on the age and religious distributions. The precise wording of the yuestlon was: la) ‘.Whut about a husband and wife having marital difficulties” If they were friends of yours. would you be able to tell them where they could receive help if they asked you”‘? (b) “Now. imagine you were in financial dificulties. Would you know where to go for help”‘? Similar follow-up questions were asked of those respondents who claimed knowledge about where to find help for financial and marital problems. The data analysis revealed (as was the case for knowledge 01 psychiatric treatment) that respondents were able to nominate specific and realistic services for both marital and financial ditiicultxs. The data are available from the authors on request. As Deutscher has recently pointed out however. surve! rescarchers must always be on guard against gaps between respondent’s attltudcs or opinions and thruactual behavlours: \CL‘ D~utschcr. I. (Fditor) Cl’lrclr Proplt~ %I.. CZ/lrrl Pcoplc Do. Scott & Foresman. Glenvllle. Illinois. 1973. Kadushin. C. U’/I \‘ PCC~I/C,(;o 77, P~~c~/~rc~rrr,sr\ p. 3 Ii Athcrton. New York. IYhY. Clearly. if we cannot dcmonstratc thts consistency In response WC cannot assert with much confidence that we arc examining a unitary scale of knowledge about where to tind help. oJ’.Lfctiicirw
3
3 5
6
,
‘.
9.
10,
,, I’.
Knowledge about where to find help 13. Guildford, J. P. Fur&~mvntal Statistics in Psychology ur~d Edwcrriorl p. 502. McGraw-Hill. New York. 1965. 14. For a review. see Fried, M. Social differences in mental health. In Pouerry and Health: A Sociological A~~alysis (Edited by Kosa, J. ~‘t al.) pp. 1I3- 167. Harvard University Press, Cambridge, 1969. 15. Our occupational ranking is derived from Broom, L. et al. An occupational ranking of the Australian work force. Ausr. N.Z. J. Social. (I) supplement, 1965. 16. Kaplan. H. B. Mental Illness as a Social Problem. In Haudhook on thr Study of Social Problems (Edited by Smigel. E. 0.) p. 342. Rand McNally, Chicago. 1971. I?. These data are available from the authors on request. At a later date it is intended to pursue the analysis in greater depth. For example in the preliminary analysis presented in this paper, tests of significance have not been used because (as the “Columbia School” researchers point out) it is important in the exptoratory phase of research not to limit productive lines of enquiry by insisting that the findings prove to be significant; in other words, when the task at hand is
439
to explore trends and to generate hypotheses. 11 is not fruitful to disregard or omit results because they fail to meet stringent criteria of significance. Set Coleman. .J. S. Statistical Problems. In Union Dernocruc~ (Edited by Lipset. S. M. et nl.) pp. 427-432. Fret Press. New York. 1956. 18. Wadsworth. M. at rrl. Heu/rh arld Sickr~rss: 2% Chorw of Trwrwm. Tavistock. London. 1961: and Kolb. L. ~‘r(11.Urban Challrnyes to Psychiatry. Little. Brown & Co.. Boston, 1969. 19. Data on socio-economic differences in exposure to mass media messages including mental health information are available in; Rossi. P. H. and Blum. Z. D. Class. Status and Poverty. In 011 Undcvsfundiug Pouerry (Edited by Moyniham, D. P.) p. 53. Basic Books. New York, 1969. 20. See Meyer and Timms. op. cir., for a discussion of “lay referral” in the British context. We are indebted to Dr. Peter McEwan for bringing this to our attention. 21. Hollingshead. A. B. and Redlich. F. C.. So&l Class and Mrntu/ fllwss. p. 130. Wiley, New York. 1967: and see also pp. 176 and 181.
