Download Psychological health and well-being

Vol. 20, No. 3, pp. 288-294
Printed in Great Britain
Journal of Public Health Medicine
Psychological health and well-being: why and
how should public health specialists measure
it? Part 2: stress, subjective well-being and
overall conclusions
Christopher J. Bartlett and Edward C. Coles
Abstract
Background There are cogent reasons why public health
specialists should take an active interest in and measure the
psychological health and well-being of populations. These
are discussed in Part 1 of this study, where methods of
measurement from the field of 'Psychiatric Epidemiology'
are evaluated. This paper continues the review of methods
by which psychological health and well-being might be
measured by public health specialists.
Methods The methods of the literature search and review are
described in Part 1. In Part 2, approaches from the fields of
'Stress' and 'Subjective Well-being' are examined and
evaluated.
Results Many stress questionnaires, such as those that relate
to 'life events', 'hassles' and 'perceived stress' pose conceptual problems and do not seem to have any advantages
over a simple psychiatric symptom questionnaire, such as
the General Health Questionnaire or HAD Scale. The Short
Form 36 (SF-36) is a well-being or health outcome instrument. Its mental health and vitality scales encompass both
positive and negative psychological well-being. The instrument also contains scales for aspects of physical well-being,
which make it attractive for public health use. However, it
was found that the positive aspect of these scales has not
been validated to the same extent as the negative aspect.
Conclusions Some of the methods reviewed in Parts 1 and 2
of this study can and have been used, with certain provisos,
in public health research and practice. However, a truly
valid measure of both positive and negative psychological
well-being has not yet been devised.
Keywords: psychiatric epidemiology, stress, subjective wellbeing, survey instruments
Introduction
1
In the first part of this study we describe why and how we
reviewed the literature for methods which might be adopted by
public health specialists wishing to measure the psychological
health of populations. As we wished to include instruments
from beyond the boundaries of bio-medical science, for the
initial literature search we made use of the 'Medline',
'PsychLit' and 'Current Contents: Behavioural Sciences'
databases. We searched on broad terms such as 'Psychiatric
Epidemiology', 'Well-being', 'Positive Health', 'Stress' and
'Psychological Health'. The search revealed three major fields
of academic inquiry, 'Psychiatric Epidemiology', 'Stress
Studies' and 'Subjective Well-being'. In this part we present
the results relating to studies in the fields of 'Stress' and of
'Subjective Well-being'. We summarize the results of the whole
literature review, discuss thefindingsand state our conclusions.
Results (measures of stress)
The term 'stress' is used with imprecision, both by lay people
and by professionals. It is probably best defined and conceived
of as a process. This process involves 'stressors' which make a
demand that evokes a 'stress response' or 'stress experience' in
a person.2'3 Some researchers would also claim that intrinsic
factors or 'mediators' within a person can modulate response to
a stressor.4
Life events
One line of research has focused on the measurement of the
stressors known as 'life events' (for example, losing one's job,
bereavement, housing problems, moving house) and on the
hypothesis that life events can bring about changes in a person's
mental and physical well-being.5'6 Although the 'life events
hypothesis' has been very attractive to many researchers, the
value of many studies in this field has been called into
Welsh Combined Centres for Public Health, University of Wales College of
Medicine, Abton House, Wedal Road, Cardiff CF4 3QX.
Christopher J. Bartlett, Research and Development Specialist
Department of Medical Computing and Statistics, University of Wales College
of Medicine, Heath Park, Heath, Cardiff CF4 4XN.
Edward C. Coles, Head of Department
Address correspondence to Dr C. J. Bartlett, Division of Psychiatry, University
of Bristol, 41 St Michael's Hill, Bristol BS2 8DZ.
© Oxford University Press 1998
PSYCHOLOGICAL HEALTH AND WELL-BEING, PART 2
question.7 l0 For example, criticism has been made of the way
in which investigators have used survey instruments to collect
data retrospectively about life events, without giving due
attention to the serious problem of recall bias. Another
fundamental weakness seems to have been present in the
analysis of the data. Data that are not normally distributed and
that do not seem to indicate a simple linear relationship between
life event scores and illness have often been analysed in
inappropriate ways.
Sounder work has been carried out by George Brown and his
colleagues. These studies have allied a more epidemiologicaltype analysis (comparing case groups and non-case groups)
with thorough interviewing techniques originating in the social
sciences. As a result, strong evidence has been proffered that
life events (of a certain nature and in certain contexts), play a
significant role in the development of depression in women.1 '~15
However, in a public health context, life events would not be
easy to measure in large samples. To collect valid and relevant
data concerning life events, their nature and contexts, Brown
and his colleagues required specially trained interviewers and a
lengthy interview schedule, the Life Events and Difficulties
Schedule. Moreover, life events are not a measurement of
experienced positive or negative psychological health, but
could more fittingly be described as risk factors. Life events
may lead to neurotic disorder in a given individual, yet it is
possible that other factors (such as a woman having a confiding
relationship with her partner) can exercise a preventive
influence and neurotic illness does not result.
Hassles
Psychological researchers have also tried to measure stressful
influences, by concentrating on everyday 'hassles', the frequent
strains and stresses of daily living.16"18 These 'hassles' include
losing personal items, being annoyed by one's neighbours and
having trouble with one's car. Again, the investigations have
been marred by methodological weaknesses, with the work
usually being based on linear models, that are not warranted by
the nature of the data. The 'Hassles Scale', described by Kanner
et al.ls is suitable for completion by the subject, but this 'userfriendly' scale does have a major shortcoming in its design. It
appears to measure the 'stress experience' (that is the
psychiatric symptoms) as well as the 'stressors' themselves
that might possibly be causing those symptoms. In fact, Burks
and Martin19 have identified a number of items in the scale
(such as having troubled thoughts about the future) which could
be deemed to be 'outputs' rather than 'inputs' in the stress
process. This could be construed as a form of misclassification
or information bias. One might argue, however, that the Hassles
Scale provides a measure of 'perceived stress', of a person's
actual experience, even if the scale might not be suitable for use
in some aetiological investigations. Indeed, a number of such
instruments have already been developed, such as the Perceived
Stress Scale,20 the Global Assessment of Recent Stress Scale21
289
and the Index of Clinical Stress.22 Here also, it seems likely that
the instruments are measuring a mixture of external stressors,
perceptions of stressors and psychiatric symptoms. It is
doubtful, therefore, whether the Hassles Scale or any of the
perceived stress scales can offer public health specialists a more
valid or useful tool for measuring psychological health than is
offered by some of the psychiatric questionnaires examined in
Part 1. In contrast to psychiatric questionnaires, perceived stress
scales are purported to collect data about the area or 'domain' of
life from which distress originates. This might be useful to a
public health practitioner who wishes to design and direct
interventions in his or her community. However, we would
suggest that the misclassification bias, the mixing-up of cause
and effect in these scales, makes these data almost impossible to
interpret according to 'domain'. Consequently, it is difficult to
judge whether a person is feeling depressed because they are
dissatisfied with their housing, or whether they are dissatisfied
with their housing because they are depressed.
Results (subjective well-being)
The way that human beings might achieve a sense of wellbeing, particularly by adopting a specific religious or
philosophical outlook, has been a central theme in speculative
thought throughout history.23 In the twentieth century attempts
have been made to measure the construct of subjective wellbeing in populations, most notably as an aspect of social
psychology, with seminal works being produced by Gurin
et al.,24 Bradburn and Caplovitz,25 Campbell et al.26 and
Andrews and Withey.27 A good working definition of
subjective well-being can be quoted from Veenhoven:28 'The
degree to which an individual judges the overall quality of his
or her life-as-a-whole favourably.'
Inherent in the concept of subjective well-being is the notion
that it is possible to measure degrees of positive well-being.
Much of the research, however, has focused on how satisfied
people are with their place in society, for example, with their
marriage, housing, neighbourhood, workplace, government or
personal finances. In the review, therefore, we concentrated on
instruments that measure degrees of positive and negative
psychological well-being being experienced in a person's
emotional life, without reference to any particular 'domains' of
life.
The Bradburn Affect Balance Scale (BABS)
The Bradburn Affect Balance Scale (BABS)29 is a short
inventory of ten items that measures positive and negative
affect on respective sub-scales, each containing five items. The
items relate to pleasant and unpleasant experiences of the
preceding few weeks, such as receiving a compliment, enjoying
something, boredom, and restlessness, although the context of
the emotional experience is deliberately unspecified. Bradburn
290
JOURNAL OF PUBLIC HEALTH MEDICINE
recommended that an overall score, the actual 'affect balance'
of the two sub-scales, should be calculated.
Validation of the scale took place by means of interview in
five samples of adults from communities in the United States.
Over 2000 people were studied for a period of one year. As few
statistical assumptions could be made about the data, the
analyses which took place employed non-parametric methods,
some of which are not in common use in contemporary public
health research and practice. The gamma statistic of association
was used to measure association between the BABS and
choices of interviewees on Likert-type scales attached to
statements about happiness. Statistically significant associations between happiness statements and positive affect on the
BABS, and between happiness statements and negative affect
on the BABS were reported. Bradburn also made use of 'ridit
analysis', explaining that the ridit is an indication that, once the
element of chance has been removed, scores in a specific group
are greater than those from a reference population. Average
ridits for the cross-tabulations of happiness levels and the affect
scores confirmed these associations and their directions.
Gamma coefficients of association that ranged from 0.29 to
0.52 were found between negative affect on the BABS and
various measures of worry, anxiety and physical symptoms.
Associations of positive affect on the BABS with these same
measures ranged from 0.05 to —0.13. All items in the scale had
a high external reliability over three days as denoted by Qs of
association in the range 0.86-0.96. This stability test took place
in a much smaller sample of 174 people. The BABS seems to
have reasonable psychometric properties, although the validation exercises relied greatly on measures of the strength of
association, rather than on measures of the nature of
association. The instrument has not been extensively used
since its creation in the 1960s. Although Bradburn was working
in the psychiatric field, the scale has rarely been used by
psychiatric epidemiologists. It has been referred to more often,
although not used more frequently, in work by investigators in
the field of social science.
The Oxford Happiness Inventory,30'31 devised by the social
psychologist Michael Argyle, can be viewed as a belated
successor of the BABS. The inventory measures degrees of
positive happiness. It is a 29-item questionnaire based mainly
on reversing the negative affect items of the Beck Depression
Inventory.32 It gives a choice of four responses to each item,
three of which register different degrees of positive happiness.
Argyle reported an association between Oxford Happiness
Inventory scores and the judgement of friends in the form of a
correlation of r = 0.43 and also an inverse association between
the new Inventory and the Beck Depression Inventory of
r= —0.52. The instrument seems to be relatively stable with an
external reliability reported of r = 0.78 over seven weeks and,
in a separate exercise, r = 0.67 over five months. Although the
researchers would have strengthened their case by augmenting
these measures of association with some indications of the
nature of the relationships described, this work suggests that the
Oxford Happiness Inventory might prove to be a useful measure
of a form of positive well-being. As the Inventory deals with
more than a few types of emotional episode and allows for a
number of degrees of response, it has, at least in theory, a
superiority to the BABS, although we are not aware that this has
ever been empirically tested.
The Short Form 36 (SF-36)
The physical and mental health status measure, the Short Form36 or SF-36, 33 has many features that make it appropriate for
review in the present enquiry and it is well documented in the
literature. It was originally developed as a way of measuring the
outcome of different types of health service delivery in the
United States. 34 " 36 In recent years the SF-36 has been
administered in clinical, non-clinical and general population
samples by public health specialists. 37 ~ 40 The items allow for
Likert-type responses (such as 'All of the time', 'Most of the
time') and are suitable for self-administration. One of its scales
relates to mental health, and a second ('vitality') also has
psychological implications. The authors of the instrument
conceived of the construct of mental health, as measured by the
mental health scale, as being a dimension of psychological
well-being with both a negative and a positive pole.
However, a number of criticisms have been made of the
SF-36. Much of the research published by the authors involved
analysis of data for the 36 items, produced when these items
were administered as part of a much larger 'parent' battery of
items; this approach weakens the validity of these analyses
considerably. 41 ' 42 Despite the use of the instrument in outcome
studies, there has been relatively little emphasis in validation
studies on measuring change in groups or on establishing the
instrument's external reliability. 43 ' 44
Evidence for the validity of the SF-36 mental health scale
has been offered by McHorney et a/.,35 who demonstrated that
the mental health scale can distinguish between patient groups
in terms of mean score. A group with minor medical conditions
had a mean score of 83, a serious medical condition group a
mean of 78, a serious medical condition with psychiatric comorbidity group a mean score of 57, and a psychiatric group
had a mean score of 53 (n = 916). The scale is also capable of
distinguishing, in terms of mean scores, between groups of
patients with symptomatic depression and with more severe
clinical depression. These are reasonable steps to take in
demonstrating the validity of such scale, but people who were
not receiving medical attention of any sort (the 'well' or
comparatively 'well') were not included.
Work with the SF-36 has also been undertaken in the United
Kingdom. 39 ' 40 ' 45 " 49 Using a UK version of the SF-36, Garratt
et al.45 have reported a downward gradient of mental health
scores that accords with severity ratings for patients given by
family doctors. These patients did not have mental disorders (at
least not conspicuous disorders that had been clinically
recognized), so this might be interpreted as moderate evidence
PSYCHOLOGICAL HEALTH AND WELL-BEING, PART 2
that the SF-36 mental health scale is sensitive to gradations of
mental health in people without overt disorder. On the other
hand, the 'well' or comparatively 'well' were not included.
Lyons et al.39 have reported finding a 'distinctive profile' of
SF-36 scores for a number of conditions in a community
sample. In a group of subjects diagnosed as suffering from
anxiety and in a group of subjects diagnosed as suffering from
depression, the mean mental health scale scores were 28 points
below the mean score for the rest of the sample. Asthma
patients and stroke patients, for example, had means on the
mental health scale only eight and 11 points below the mean for
the rest of the sample. These figures had been adjusted for age
and sex. In a sample of 16400 people from the general
population, Brazier et al.46 measured the external reliability of
the different scales over two weeks. The mental health scale
showed a mean difference of less than one point. As the range
of scores extends up to 100, this level of change does not seem
to have any clinical significance and is a favourable result. The
vitality scale also showed a similar level of stability.
We would argue that the mental health scale rests upon a
basic conceptual weakness, however. It is implied by the
authors of the SF-36 that the mental health items (as well as the
vitality items) measure a form of positive mental health. A high
score means that the respondent: 'feels peaceful, happy and
calm all of the time, past four weeks'.34 Although, as we have
shown, there is evidence that the scale does measure a certain
amount of variation in mental well-being, it is not clear (1) how
the authors of the SF-36 established that very high scorers
were enjoying this extremely positive state, or (2) how they
calibrated the scale so that the different scores near the top of
the scoring range represent different degrees of peacefulness,
happiness and calmness. Initial validation exercises in the
United States were based upon comparisons of different patient
groups, some with psychiatric conditions.35 These studies were
a comparison of 'more ill' versus 'less ill'. Of course, this does
not mean that these items completely fail to measure a form of
positive mental health, but that their measurement of degrees of
positive mental health was not adequately researched.
Although items appear in the mental health scale which are
evident symptoms of positive psychological well-being ('Have
you been a happy person?', for example) and produce a positive
score, it is a notable feature of this scale that the scores of items
describing negative symptoms are actually 'reversed' to
produce positive mental health scores. Thus, a score of two
out of six ('a little of the time') on the question 'Have you felt
so down in the dumps that nothing could cheer you up?' is
transformed into a positive score of four out of six. The
assumption here is that not often feeling 'down in the dumps'
means that a person is enjoying a relatively positive state of
mental health, in respect of affect, most of the time. In contrast,
Bradburn29 concluded that, in terms of affect and moods at
least, positive and negative well-being could be regarded as
separate dimensions, that they were not opposite ends of the
same construct. A person who is feeling down in the dumps
291
only a little of the time is not necessarily feeling happy the rest
of the time. The creators of the SF-36, however, have
uncritically interpreted absence of 'ill-being' as implying the
presence of a predictable quantum of positive 'well-being'.
This undermines the validity of their measurement, if they truly
wish to measure a form of positive well-being.
Some of the criticisms made of the mental health scale of the
SF-36 and of validation studies involving it can also be made of
the vitality scale. Absence of 'tiredness' or 'feeling worn out' is
converted into a positive score for some items.33 No attempt has
been made to calibrate the vitality scale by using subjects with a
high degree of positive wellness. However, as vitality might be
expected to vary in people with different degrees of physical
illness, the SF-36 studies cited above do offer direct evidence
for the validity of the vitality scale. Garratt et a/.45 reported how
vitality mean scores are lower than, and differ in varying
degrees from, a general population mean in different groups of
patients. The groups examined were suffering from, respectively, low back pain, menorrhagia, suspected peptic ulcer, and
varicose veins. In short, the mental health and vitality scales
have produced some reasonable results in recent public health
research in the United Kingdom in which their validity and
external reliability have been examined, but there must be some
doubt as to whether these two scales are an accurate measure of
positive psychological well-being. More recent research has
shown how a mental health summary score can be produced
from a more convenient 12-item version of this questionnaire.50
However, the issue of the concept of mental health being
measured, was not addressed and the analysis only involved
patients with physical disorders.
The characteristics of main instruments reviewed in Parts 1
and 2 of this study are given in Table 1.
Discussion
It became evident in the course of the review that some
instruments which have been used to measure the psychological
health of populations have serious weaknesses with regard to
their validity. In some cases, for example that of the Diagnostic
Interview Schedule, the intention of the authors may have been
too ambitious. Some instruments, such as the Hassles Scale and
the Perceived Stress Scale, seem to measure a mixture of
constructs and so suffer from an information bias that would
make their application in public health research problematical.
In a comparison of instruments from a number of fields the
General Health Questionnaire fares remarkably well. As stated
in Part 1, the General Health Questionnaire is already well
established as a screening instrument for neurotic disorder, and
it has the benefit of providing a distribution of scores for a
population. The revised version of the GHQ, the Chronic
General Health Questionnaire or C-GHQ, is probably more
suitable for public health practice, because it can also produce
scores for people who have been experiencing symptoms for
some time. It is not as well researched as the GHQ, however.
Table 1 Summary of characteristics of main instruments reviewed in Parts 1 and 2
Positive and
negative
measurements?
Instrument
Constructs measured
Validity and external
reliability
Diagnostic Interview Schedule*
Symptoms of neurotic and psychotic
disorder
Validity varies by diagnostic
category; reliability not known
Negative only
Designed to detect those with mental
disorder; needs trained interviewer
Revised Clinical Interview Schedule*
Symptoms of neurotic disorder
More research needed
Negative only
Distribution of scores can be produced;
needs trained interviewer
General Health Questionnaire
(GHQ)* (also C-GHQ)
Symptoms of neurotic disorder
Very good and well researched
Negative only
Distribution of scores can be produced;
prevalence can be estimated; completion
by subject
HAD Scale*
Symptoms of neurotic disorder
Good but not as well researched
as GHQ
Negative only
Distribution of scores can be produced,
prevalence can be estimated; completion
by subject
Life Events and Difficulties
Schedule!
Comment on use in population studies
Acceptable validity
Negative events
and context only
Risk factor rather than health status being
measured; needs trained interviewer
Everyday hassles
Suffers from information bias
Negative only
Interpretation difficult; completion by
subject
Perceived everyday stress
Suffers from information bias
Negative only
Interpretation difficult; completion by
subject
Life events in context
Hassles Scalet
Perceived Stress Scalet
Positive and negative affect
Bradburn Affect Balance Scalet
Good, though methods used
limited
Small number of
positive and
negative degrees
Health implications not clear; completion
by subject
c
50
Z
>
r
O
•n
"0
C
»
r
n
x
m
>
r
H
X
S
tn
O
n
z
Oxford Happiness Inventory!
Positive affect
Good, though methods used
limited
Positive only
Would need negative scale to be useful;
completion by subject
SF-36 Mental Health and Vitality
Scalest
Mental health; vitality
Reasonable results, but rests
on a conceptual weakness
Positive and
negative mixed up
Gives distribution of scores for
population; problem of accuracy;
completion by subject
•Reviewed in Part 1.
tReviewed in Part 2.
O
rn
PSYCHOLOGICAL HEALTH AND WELL-BEING, PART 2
On the other hand, the GHQ in its various forms does not
measure any form of positive psychological well-being,
although Lewis51 has suggested that there is a latent potential
within the instrument to do this. The measurement of a form of
positive well-being does seem to have been the intention of the
authors of the SF-36. The mental health scale of the SF-36
appears to measure a dimension of psychological well-being
that extends from a negative to a positive pole. Doubts have
been expressed about the way the SF-36 was developed, but in
subsequent research the scale has performed fairly well. We
would suggest that the concepts used in the SF-36 scales are so
basic to the experience of people that the instrument has been
able to display an unexpected degree of quality. As part of a
convenient omnibus instrument, the use of the mental health
and vitality scales would seem to be reasonable in public health
research, if general monitoring of both the physical and mental
health of a population is being undertaken. However, as
individual instruments for the precise measurement of negative
and positive psychological well-being and for research into
those constructs, these scales cannot be recommended. The
work of Bradburn indicated that positive and negative
psychological well-being were probably two separate dimensions. The authors of the SF-36 did not take this into account
and assumed that the absence of 'ill-being' was equivalent to
the presence of 'well-being'. In a public health context, the
results of detailed research using the SF-36 could be
misleading. To offer a hypothetical example, if there were a
sub-group whose members were not neurotically disordered,
but who smoked cigarettes because they lacked positive wellbeing, their mean SF-36 mental health scale score would
probably be erroneously high and favourable. It is likely that the
SF-36 would miss many subtleties in the pattern of the health of
populations, particularly where psychological health is part of
that pattern.
Conclusion
At present, in terms of public mental health, we can say how ill
a population is and how illness is distributed within it, but it is
difficult to say how well it is. The complete picture is still
eluding us. We conclude that there is a need for a new
instrument which would have the strengths of a simple
instrument such as the GHQ or C-GHQ, but which would
also have an additional scale for the measurement of positive
psychological well-being in populations.
References
1 Bartlett CJ, Coles EC. Psychological health and well-being:
why and how should public health specialists measure it?
Part 1: rationale and methods of the investigation, and
review of psychiatric epidemiology. J Publ filth Med 1998;
20: 281-287.
293
2 Cox T. Stress. 1978. Basingstoke: Macmillan.
3 Cox T, McKay C. A psychological model of occupational
stress. A paper presented to the MRC meeting 'Mental
Health in Industry', London, November 1985.
4 Cohen S, Edwards J. Personality characteristics as moderators of the relationship between stress and disorder. In:
Neufeld R, ed. Advances in the investigations of psychological stress. Chichester: John Wiley and Sons, 1989: 235272.
5 Holmes TH, Rahe RH. The Social Readjustment Rating
Scale. J Psychosom Res 1967; 11: 213-218.
6 Holmes TH. Masuda M. Life change and illness susceptibility. In: Dohrenwend BS, Dohrenwend BP, eds. Stressful
life events: their nature and effects. Chichester: John Wiley
and Sons, 1974: 45-72.
7 Rabkin JG, Struening EL. Life events, stress and illness.
Science 1976; 194: 1013-1020.
8 Mechanic D. Some problems in the measurement of stress
and social readjustment. J Human Stress 1975; 1: 43-48.
9 Brown G. Meaning, measurement and stress of life events.
In: Dohrenwend BS, Dohrenwend BP, eds. Stressful life
events: their nature and effects. Chichester: John Wiley and
Sons, 1974: 217-244.
10 Brown G. Life events, psychiatric disorder and physical
illness. J Psychosom Res 1981; 25: 461-473.
11 Brown G, Harris T, Copeland J. Depression and loss. Br J
Psychiat 1977; 130: 1-18.
12 Brown G, Harris T. Social origins of depression. A study of
psychiatric disorder in women. London: Routledge, 1978.
13 Brown G, Sklair F, Harris T, Birley J. Life events and
psychiatric disorders. Part 1: some methodological issues.
Psychol Med 1973; 33: 74-87.
14 Brown G, Harris T, Peto J. Life events and psychiatric
disorders. Part 2: Nature of causal link. Psychol Med 1973;
3: 159-176.
15 Brown G, Andrews B, Harris T, et al. Social support, selfesteem and depression. Psychol Med 1986; 16: 813-831.
16 Delongis A, Coyne J, Dakof G, et al. Relationship of daily
hassles, uplifts and major life events to health status. Health
Psychol 1982; 1: 119-136.
17 Lazarus R. Puzzles in the study of daily hassles. J Behav
Med 1984; 7: 375-389.
18 Kanner AD, Coyne JC, Schaefer C, Lazarus RS. Comparison of two modes of stress measurement: daily hassles
and uplifts versus major life events. J Behav Med 1981; 4:
1-39.
19 Burks N, Martin B. Everyday problems and life change
events. Ongoing versus acute sources of stress. / Human
Stress 1985; 11: 27-35.
294
JOURNAL OF PUBLIC HEALTH MEDICINE
20 Cohen S. Contrasting the hassles scale and the perceived
stress scale. Am Psychol 1986; 41: 716-718.
21 Linn MW. A global assessment of recent stress scale. Int J
Psychiat Med 1985-1986; 15: 47-59.
22 Abell N. The Index of Clinical Stress: a brief measure of
subjective stress for practice and research. Social Work Res
Abstr 1991; 27: 12-15.
23 Diener E. Subjective well-being. Psychol Bull 1984; 95:
542-575.
24 Gurin G, Veroff J, Feld S. Americans view their mental
health. New York: Basic Books, 1960.
25 Bradburn NM, Caplovitz D. Reports on happiness. Chicago:
Aldine, 1965.
26 Campbell A, Converse P, Rodgers W. The quality of
American life. New York: Russell Sage Foundation, 1976.
27 Andrews FM, Withey SB. Social indicators of well-being.
New York: Plenum, 1976.
28 Veenhoven R. Questions on happiness; classical topics,
modern answers, blindspots. In: Strack F, Argyle M,
Schwarz N, eds. Subjective well-being: an interdisciplinary
perspective. Oxford: Pergamon Press, 1991: 7-26.
29 Bradburn NM. The structure of psychological well-being.
Chicago: Aldine, 1969.
30 Argyle M. Happiness as a function of personality and social
encounters. In: Forgas JP, Innes JM, eds. Recent advances in
social psychology: an international perspective. Amsterdam:
North Holland, Elsevier, 1989: 189-203.
31 Argyle M. The social psychology of everyday life. London:
Routledge, 1992: 284-285.
32 Beck A, Mendelson M, Mock J, et al. Inventory for
measuring depression. Arch Gen Psychiat 1961; 4: 561 —
571.
33 Ware JE. SF-36 Health Survey. Manual and interpretation
guide. Boston, MA: Health Institute, New England Medical
Center, 1993.
34 Ware JE, Sherbourne CD. The MOS 36-item Short Form
Health Survey (SF-36): I. Conceptual framework and item
selection. Med Care 1992; 30: 473-481.
35 McHorney CA, Ware JE, Raczek A. The MOS 36-item
Short Form Health Survey (SF-36): II Psychometric and
clinical tests of validity in measuring physical and mental
health constructs. Med Care 1993; 31: 247-263.
36 McHorney CA, Ware JE, Rachel Lu JF, et al. The MOS 36item Short Form Health Survey (SF-36): III Tests of data
quality, scaling assumptions and reliability across diverse
patient groups. Med Care 1994; 32: 40-66.
37
Usherwood T, Jones N, Hanover Project Team. Selfperceived health status of hostel residents: use of the SF36D health survey questionnaire. J Publ Hlth Med 1993; 15:
311-314.
38 West Glamorgan Health Authorities. Annual report of the
Director of Public Health Medicine. Swansea: West
Glamorgan Health Authorities, 1993.
39 Lyons R, Lo S, Littlepage B. Comparative health status of
patients with 11 common illnesses in Wales. J Epidemiol
Commun Hlth 1994; 48: 388-390.
40 Lyons R, Fielder H, Littlepage B. Measuring health status
with the SF-36: the need for regional norms. J Publ Hlth
Med 1995; 17: 46-50.
41 Hunt S, McKenna S. Letter. SF-36 misses the mark. BrMed
J 1993; 307: 125.
42 Hill S, Harries U, Popay J. Letter: Unanswered questions
remain. Br Med J 1993; 307: 449.
43 Ruta D, Garratt A, Abdalla M. Letter: the SF-36 health
survey questionnaire: a valid measure of health status. Br
MedJ 1993; 307: 448.
44 Sheldon T. Letter: Reliability of SF-36 remains uncertain.
BrMedJ 1993; 307: 125-126.
45 Garratt A, Ruta D, Abdalla M, et al. The SF-36 health
survey questionnaire: an outcome measure suitable for
routine use in the NHS? BrMedJ 1993; 306: 1440-1444.
46 Brazier JE, Harper R, Jones NMB, et al. Validating the
SF-36 health survey questionnaire: a new outcome measure
for primary care. Br Med J 1992; 305: 160-164.
47 Jenkinson C, Coulter A, Wright L. Short Form 36 (SF-36)
health survey questionnaire: normative data for adults of
working age. Br MedJ 1993; 306: 1437-1440.
48 Lyons R, Lo S, Littlepage B. Perception of health amongst
ever-smokers and never-smokers: a comparison using the
SF-36 Health Survey Questionnaire. Tobacco Control 1994;
3: 213-215.
49 Lyons R, Perry H, Littlepage B. Evidence for the validity of
the Short-form 36 Questionnaire (SF-36) in an elderly
population. Age Ageing 1994; 23: 102-104.
50 Jenkinson C, Layte R, Jenkinson D, et al. A shorter form
health survey: can the SF-12 replicate results from the SF-36
in longitudinal studies? J Publ Hlth Med 1997; 19: 179-186.
51 Lewis G. Dimensions of neurosis. Psychol Med 1992; 22:
1011-1018.
Accepted on 3 March 1998