Journal of Psychosornatic Research, Vol. 43, No. 1, pp.

61-71, 1997
Copyright © 1997 Elsevier Science Inc.
All rights reserved.
ELSEVIER 0022-3999/97 $17.00 + .00

S0022-3999(97)00010-X

IMPROVING STRESS ASSESSMENT: T O W A R D

A N E V I D E N C E - B A S E D A P P R O A C H TO
O R G A N I Z A T I O N A L STRESS INTERVENTIONS

R O B B. B R I N E R

Abstract--There is little evidence to suggest that the stress management interventions (SMIs) used in
organizations are necessarily effective nor is it clear why, in principle, they should be. Why then do orga-
nizations introduce SMIs? A wide variety of reasons is apparent, many focusing on reducing the pre-
sumed costs of stress and the attractiveness of what appears to be a panacea-like intervention. The cen-
tral aim of this study is to make the case for and outline an evidence-based approach to SMIs in which
data gathered from initial assessments are used to make decisions about interventions. Valid assessment
requires a relatively sophisticated and comprehensive approach to measurement; designs which permit
causal relationships between the phenomena of interest to be established; and a healthy scepticism to-
ward the claims that are made by organizations and individuals about stress. It is argued that an evi-
dence-based approach will lead to advances in theory development and intervention practices. © 1997
Elsevier Science Inc.

Keywords: Organizational stress interventions; Stress measurement; Stress audits; Stress assessment;
Evidence-based approach.

INTRODUCTION

T h e r e is n o t h i n g n o v e l o r striking in the o b s e r v a t i o n t h a t w o r k o r g a n i z a t i o n s a n d
o t h e r i n s t i t u t i o n s i n t r o d u c e p o l i c i e s a n d p r a c t i c e s in the a b s e n c e o f e v i d e n c e for
t h e i r effectiveness. I n t h e case o f policies a n d p r a c t i c e s a i m e d at d e a l i n g with stress,
h o w e v e r , this o b s e r v a t i o n s e e m s to b e p a r t i c u l a r l y a c c u r a t e . W h i l e a n u m b e r o f dif-
f e r e n t stress m a n a g e m e n t i n t e r v e n t i o n ( S M I ) t e c h n i q u e s a r e r o u t i n e l y a d o p t e d
within o r g a n i z a t i o n s a n d t h e flourishing stress m a n a g e m e n t i n d u s t r y c o n t i n u e s to
m a k e e x t r a v a g a n t claims for t h e i r benefits, t h e r e is s u r p r i s i n g l y little e v i d e n c e t h a t
s u p p o r t s t h e v a l u e o f SMIs. A f u r t h e r c o n f o u n d i n g issue is t h a t it a p p e a r s t h a t few
o r g a n i z a t i o n s u n d e r t a k e a n y k i n d o f valid a s s e s s m e n t to first e s t a b l i s h w h e t h e r o r
n o t a n S M I is a c t u a l l y r e q u i r e d o r to establish t h e p u r p o s e s o f i n t e r v e n t i o n .
T h e u l t i m a t e a i m o f this s t u d y is to d e s c r i b e s o m e of t h e f e a t u r e s o f an e v i d e n c e -
b a s e d a p p r o a c h to o r g a n i z a t i o n a l stress i n t e r v e n t i o n s in which e v i d e n c e g a t h e r e d
f r o m valid a s s e s s m e n t s is u s e d as the basis for c h o o s i n g a n d i m p l e m e n t i n g an i n t e r -
v e n t i o n . A l t h o u g h t h e n o t i o n t h a t we s h o u l d first assess b e f o r e we i n t e r v e n e in a n y

Department of Organizational Psychology, Birkbeck College, University of London, London, UK.

Address correspondence to: R. B. Briner, Ph.D., Department of Organizational Psychology, Birkbeck
College, University of London, Malet Street, London WC1E 7HX, UK. E-mail: r.briner@uk.ac.bbk.
org-psych

61
62 R.B. BRINER

system may appear to be common sense, this thinking does not yet appear to have
captured the imagination of researchers and practitioners working in the area of or-
ganizational stress. I will focus on what might constitute a valid stress assessment
and outline the limitations of currently available assessment techniques.
First, however, I will describe some of the more widely used SMIs, evaluate the
evidence for their effectiveness, and describe some of the reasons why organizations
introduce SMIs. I will also argue that some of the more cultural and social aspects
of our contemporary usage of the idea of "stress" need to be better understood if
we wish to assess and intervene effectively.

STRESS M A N A G E M E N T INTERVENTIONS

SMIs have been defined as "any activity, program, or opportunity initiated by an

organization, which focuses on reducing the presence of work-related stressors or
on assisting individuals to minimize the negative outcomes of exposure to these
stressors" ([1], p. 252). Although a somewhat simplistic classification system, such
interventions are usually described as either primary, secondary, or tertiary. Pri-
mary interventions are those which change job conditions such as the level or nature
of work demands or "stressors"; secondary, those which give people skills which
may help them to cope more effectively with job demands or other "stressors"; and,
tertiary, those which aim to treat individuals who are experiencing high levels of
negative affect or distress. Examples of each of these types of intervention will now
be discussed.
Although no data are available that describe the relative frequency with which
different types of SMIs are used by organizations, it is widely agreed that primary
SMIs are probably the least-popular choice of intervention even though they aim to
be preventative and should, in some senses, be more attractive. In an organizational
context, however, changing the nature of the job or the organization may be consid-
erably more daunting and complex than simply buying-in some of the other types
of intervention. Also, as will be discussed later, organizations introduce SMIs for a
variety of reasons--few of which may be related to the explicit goals of SMIs. Pri-
mary interventions, as indicated earlier, change some aspect of the objective work
environment and could therefore take many forms including improving communi-
cation, increasing role clarity, reducing job demands, rescheduling job demands, or
increasing the level of control employees have over their jobs.
The most common secondary level intervention is probably stress management
training (SMT). This usually takes the form of a short course which might last from
one half to perhaps several days. The content of such courses varies, but they all aim
to teach skills such as time management, relaxation, coping, and assertiveness,
which are presumed to help people to deal with stress. SMT is run both for groups
of employees who are self-selected and for whole teams, sections, departments, and
even organizations.
Counseling is the best example of a tertiary level intervention and is often pro-
vided to employees as part of an employee assistance program (EAP), which may
also offer other services such as financial and legal advice, and health checks. Initial
contact from employee to the E A P is usually by phone and, where considered desir-
able and appropriate, the employee will be referred to a counselor for time-limited
 Improving stress assessment 63
counseling. Although referrals are made in this way rather than through, say, a GP,
and the main rationale for having such services is to deal with occupational stress,
the limited evidence available suggests that sources of client difficulties are more
usually connected with life outside work [2].
One consequence of the rather broad definition of SMIs offered above is that
many different kinds of interventions can be described as SMIs. It is also noticeable
how many existing techniques and interventions such as meditation, anxiety man-
agement, time management, and counseling have been repackaged as SMIs.

ARE STRESS MANAGEMENT INTERVENTIONS EFFECTIVE?

Before we can consider whether or not SMIs are effective, we need to be able to
specify the aims of SMIs. First, the general aims of SMIs will be considered and then
the aims of particular types of SMIs. Next, I will consider their effectiveness both
from an empirical and rational perspective.
What then are the general aims of SMIs? Returning to the definition of SMIs pro-
vided earlier it appears that all SMIs have one (or perhaps both) of two basic aims,
which are: (i) to reduce the presence of work-related stressors; and (ii) to assist indi-
viduals to minimize the negative outcomes of exposure to these stressors. To what
extent, then, do each of the three types of SMIs discussed above meet these aims?
This question cannot be answered without first considering the extent to which the
three SMIs discussed here actually do have these aims.
Only primary SMIs, which aim to change the work environment or job conditions,
aim to reduce the presence of stressors. SMT and counseling, although clearly more
focused on assisting individuals to cope with negative outcomes, may also, to some
extent, aim to help individuals to minimize the presence of stressors by encouraging
them to change objective work conditions and hence reduce the presence of stres-
sors. The teaching of time management skills, for example, may result in an em-
ployee refusing to take on additional work to reduce the presence of stressors. As
indicated earlier, the distinction between primary, secondary, and tertiary is not
necessarily a useful one.
From an organization's point of view, in addition to these two basic aims of SMIs,
a more fundamental organizational aim is to improve the levels of those phenomena
which are assumed to be caused by stress including absence, turnover, and lowered
motivation and performance. Hence, in evaluating the effectiveness of these inter-
ventions we will also consider their effectiveness in terms of the presumed aims of
each SMI and also in terms of this more fundamental organizational aim.
So, do SMIs work? It is difficult to begin a discussion of the empirical evidence
for SMIs without first stating an extremely important fact about the research litera-
ture in this area, which is that we have little sound evidence for the effectiveness or
otherwise of SMIs. For example, Ivancevitch et al. ([1], p. 259) state that "present
knowledge about SMIs is largely based on anecdotes, testimonials, and method-
ologically weak research." In a review of the organizational stress field, Kahn and
Byosiere ([3], p. 632) write that "the research e v i d e n c e . . , is unfortunately meagre
and the methodological problems are serious." In a report about stress interven-
tions, Cox ([4], p. 73) comments that "evaluation data are relatively rare," and that
the "jury is still out on stress management training." Unfortunately, there is simply
64 R . B . BRINER

not enough evidence to allow us to subject it to a conclusive review or detailed dis-

cussion. However, the limited evidence that is available will be briefly considered.
Although there is little evidence about the effectiveness of primary interventions,
there are some indications from research on job redesign about what we might ex-
pect to find when we change objective job conditions. A good example of a typical
job redesign intervention is one in which employees are provided with greater au-
tonomy or control over the way in which their work is arranged and scheduled. Two
studies, which evaluated such an intervention in a manufacturing context, show
that, whereas some employee feelings, attitudes, and behaviors of interest such as
intrinsic job satisfaction or job commitment did improve as a result of the interven-
tion, others, such as motivation or mental health did not change, and yet others,
such as absence and turnover, became worse [5, 6]. In a recent review of such inter-
ventions, Briner and Reynolds [7] argued that this pattern of results is found in all
evaluations of job redesign and that it is likely that primary SMIs are therefore
likely to have costs and limitations as welt as perhaps some benefits.
The limited evidence for the effectiveness of SMT suggests that it has no long-
term effects and that the short-term effects on psychological well-being may be
more connected with a nonspecific "feel-good" factor following the training than
the specific content of the training (e.g., [8]). One problem may, however, be that
SMT is evaluated as though it is a treatment when it should be considered as a pre-
ventative intervention. In other words, researchers tend to look for changes in mea-
sures of psychological well-being following SMT when there is no reason to assume
that a preventative intervention will, in itself, produce any immediate benefits.
One exception to the lack of evidence for the effectiveness of SMIs is evidence
for general benefits counseling. It is reasonably well-established that people who
are experiencing high levels of negative affect do benefit from counseling or psycho-
therapy (see, e.g., [8]). As mentioned earlier, it appears to be the case that counsel-
ing provided through work may, in most cases, be dealing with problems and diffi-
culties in the client's life which have nothing to do with work.
Although a conclusive review of the empirical evidence about the effectiveness
of SMIs is not currently possible, it is helpful to consider in rational or theoretical
terms whether or not SMIs could, in principle, be effective.
Primary SMIs could, in principle, be effective if those aspects of the objective job
environment that were changed in such an intervention were actually the cause of
lowered psychological well-being among most employees. A key issue is the extent
to which employees do react similarly to objective job conditions. If there are large
individual or group differences, changing any part of the work environment will
have mixed effects, some of which may be detrimental to some employees' well-
being. Of course, these interventions could, in principle, also only be effective if the
initial assessment techniques used were capable of clearly establishing causal links
between job conditions and employee well-being. These issues will be discussed
later in more detail.
In principle, SMT could be effective as it aims to teach skills that are supposed
to minimize the negative effects of stressors. However, this reasoning is based on
a number of assumptions that are difficult to support. First, is the assumption that
skills such as coping are more important than, say, personality, in minimizing the
effects of stressors. It is also not clear, for example, that having a wide repertoire
 Improving stress assessment 65
or particular styles of coping responses is necessarily beneficial. The second assump-
tion is that such skills can actually be taught effectively in an SMT format (or, in-
deed, that they can be taught at all). Again, this assumption is difficult to support
as assertiveness and time management, for example, are complex skills which, even
where individuals are intensively and thoroughly trained, may not transfer into ac-
tual behaviors in the workplace.
It is possible to view counseling as something other than an SMI as it will not nec-
essarily have any effect on the presence of stressors. Nor will it usually help an indi-
vidual to minimize negative effects of exposure to stressors unless: (i) we choose to
label any client's problem brought to counseling as one caused largely by "stres-
sors" as opposed to some other cause; (ii) those stressors are work-related; and (iii)
the counseling has long-term benefits. Counseling may therefore be regarded as an
SMI only in these very particular circumstances.
In general, then, SMIs do not appear to be highly effective. In part, this is because
they are unlikely, in principle, to be effective. In other cases, it may be that SMIs
can only work under certain conditions. Given such limitations their popularity is
somewhat puzzling. To better understand this popularity, I will now consider some
of the reasons why organizations use SMIs. It will be argued later that such an un-
derstanding is a prerequisite for making effective assessments in organizations.

W H Y D O O R G A N I Z A T I O N S I N T R O D U C E STRESS
MANAGEMENT INTERVENTIONS?

No data are currently available that would allow us to answer this question pre-
cisely. However, through observation of the marketing tactics of the stress manage-
ment industry, an analysis of stress researchers' rhetoric, and from listening to prac-
titioners and managers discuss stress issues, it is possible to identify seven of most
widely cited reasons for the introduction of SMIs.
The first reason for introducing SMIs is to improve effectiveness. Organizations
may use SMIs because they believe employee behaviors that are detrimental to ef-
fectiveness and performance, such as absence and lowered motivation, are caused
by stress. It is therefore assumed that, by removing or reducing stress, there will be
a corresponding reduction in the levels of these behaviors. The predominant way
of marketing stress management products is through claims that stress is costly
to organizations.
A second important reason is that SMIs may be introduced in response to a dra-
matic event or series of events for which stress is blamed such as a manager's prema-
ture death, a number employees suffering heart-attacks, a sudden increase in the
numbers of people taking long-term sickness absences, or a serious accident. In this
case, the organization is jolted into believing that it has to do something about
stress.
Third, it may that the organization is a caring and responsible employer. In some
organizations, paternalistic and philanthropic motives may be sufficient reason for
the organization to help employees with stress or indeed any problems they may be
experiencing in relation to work or even outside work. For such organizations, SMIs
may simply be seen as part of general welfare provision.
A fourth reason is that the organization wants to be seen to be a caring and re-
66 R.B. BRINER

sponsible employer. In addition to and perhaps at the same time as actually caring
about its employees, organizations may implement SMIs to create a positive image
of themselves, not only with the internal audience of trades unions and employees
but also with the external audience of customers, clients, and shareholders.
The fifth reason for introducing SMIs arises from the concern that the organiza-
tion's employees may take legal action. As organizations have a duty of care to pro-
vide a safe working environment, if they do not do so the employer may be liable
for damages (e.g., [9]). Some organizations therefore introduce SMIs in the hope
that these will safeguard them from litigation and the payment of damages. Al-
though in the UK very few such cases have come to court, those that have and in
which employees won substantial damages are used widely within the SMI market-
ing literature in the attempt to convince organizations to implement some form
of SMI.
A sixth and important reason (which also underlies many organizational actions)
is that everyone else is doing something about stress. The importance of fashion and
fad in determining what an organization does cannot be overestimated. This seems
particularly important in the case of SMIs where it is not uncommon for one organi-
zation to introduce SMIs or a particular type of SMI simply because another organi-
zation is doing so. Managers from different organizations discuss what they are do-
ing about stress, and those selling SMI products use the names of other
organizations with whom they have done business to convince potential customers
of the soundness of their products.
The seventh and final reason to be identified here is that stress and SMIs are easy
issues to sell within an organization. A recently developed approach to thinking
about why organizations introduce policies and practices suggests that they may be-
come adopted as a consequence of certain individuals' success in convincing key de-
cisionmakers that the issue they are "selling" is one that requires attention and ac-
tion [10]. An important reason for trying to sell issues to others is for purposes of
individual advancement within the organization. For this reason, particularly diffi-
cult or complex issues are unlikely to be taken on by issue-sellers. Because of its
popularity, and its flexibility in meaning, stress may be a particularly appealing and
easy issue to sell within an organization [11].
A wide range of rather diverse reasons for the introduction of SMIs can be identi-
fied. A somewhat odd mixture of ethical, legal, social and pragmatic influences can
be identified in the decision to use an SMI.
It is also important to consider the process by which these reasons operate to
bring about the decision to use an SMI. They are commonly sold to organizations
using the appeal of cost reduction. As discussed previously, it is assumed that such
costs could be ongoing (e.g., absence, lowered productivity) or potential (i.e., legal
costs). The argument presented to organizations is therefore stunningly simple: (i)
stress is expensive; (ii) SMIs will reduce stress; and (iii) SMIs will save you money.
This argument is sometimes also put in more positive terms: (i) levels of absence
and productivity could be improved; (ii) stress causes absence and low productivity;
and (iii) reduce stress and levels of absence and productivity will improve. SMIs can
therefore take on the appearance of a panacea in that they offer the hope of dealing
with a wide range of organizational problems.
SMIs have become the "solution" to a wide range of nondefined or ill-defined
 Improving stress assessment 67

"problems." Almost every organization wants to improve productivity, reduce ab-

sence, avoid costly legal proceedings, and do something about stress. In fact, none
of these may be a problem as such for any particular organization, but the allure of
an SMI, which promises to solve all of these problems, apparently simultaneously,
may be sufficient for an organization to decide to use one.
Perhaps the central problem in the area of SMIs is the failure to assess the nature
and extent of whatever problems may exist as a result of making largely false as-
sumptions about what stress is, what it does, and how it can be reduced. Although
the stress management industry, business consultants, and organizational psycholo-
gists (both academics and practitioners) are partly to blame, it is also the more gen-
eral popularity of the notion of stress and its broader cultural significance in provid-
ing an "explanation" for physical and psychological dysfunction. This issue will be
discussed in more detail in what follows.
It is as a consequence of the virtual absence of valid assessment within the area
of SMIs that I am suggesting that SMIs, like any other intervention, should be evi-
dence-based. By evidence-based I mean that the nature of the causal relationships
between the phenomena of interest are first established. The central premise of
work on organizational stress is that there are causal links between work conditions
and employee well-being. An evidence-based approach would in each case examine
the nature and truth of such a premise by first conducting a thorough assessment.
But what constitutes a valid assessment? How can good evidence that would inform
the design and nature of an intervention strategy be obtained?

WHAT CONSTITUTES VALID STRESS ASSESSMENT?

Before discussing in detail some of the ways in which stress assessment can be
made more valid and an evidence-based approach to SMIs can be developed, the
main limitations of self-report measures, which is currently the most widely used
method, will be considered.
A wide range of self-report measures of stressors, job characteristics, physical and
psychological well-being, coping strategies, and stress-related personality character-
istics is available. Most of these have been developed for research purposes, but
some, often called stress audits or stress profiles, have been developed for commer-
cial use. It is claimed that one purpose of stress audits is to provide information
which can then be used as the basis for selecting a SMI. At the same time, these
audits can also be used to provide individual profiles that are intended to help the
individual better understand and deal with their particular stress problems.
Although self-report measures are extremely popular as commercial products,
and as research tools, they could only be effective as stress assessment techniques
within an evidence-based approach to SMIs if certain conditions are held true.
These measures would be useful if, for example; (i) people were able to accurately
report those work conditions which cause negative effects and which were actually
"stressors" (as opposed to those work conditions they merely do not like or that be-
come a symbolic focus of attention); (ii) self-report measures were not strongly in-
fluenced by the respondent's concurrent affective state; (iii) self-report measures
were not strongly influenced by individual differences; and (iv) self-report measures
68 R.B. BRINER
provide an accurate assessment of the respondent's physical and psychological
well-being.
Clearly, there are numerous rational and empirical reasons why these conditions
cannot be met: reports of job conditions are influenced by many other factors which
have nothing to do with objective job conditions ]12]; current affective state can in-
fluence reporting of physical and psychological symptoms (e.g., [13]); individual dif-
ferences influence many self-report measures (e.g., [14]); self-report measures of
physical and psychological symptoms may have little to do with well-being in an or-
ganic or clinical sense (e.g., [15, 16]).
One starting place for improving stress assessment is to first consider what the
components of "stress" and its presumed effects might be. It seems that there are
at least three broad components, each of which has objective and subjective ele-
ments. First is subjective and objective well-being or "strain," or the levels of psy-
chological well-being or other aspects of health, among employees. By objective
well-being I mean other criteria that could be used to assess psychological well-
being which are not self-report (e.g., behaviors, ratings by coworkers or family, indi-
cators of functioning).
The second component contains subjective and objective job conditions. One of
the major limitations of self-report measures of job conditions is that they seem, in
many cases, to have little to do with objective job conditions. Spector [12], for exam-
ple, suggests that only 20% of the variance in subjective ratings can be explained
by objective conditions. This has profound implications for intervention; that is,
even where we find a causal relationship between subjective reports of a job condi-
tion and well-being, intervening to change the objective nature of the job will have
little impact on subjective reports and hence little impact on well-being. For exam-
ple, if a causal relationship is found between perceptions of high workload and anxi-
ety, reducing objective levels of workload will have little or no impact on anxiety
as there is no strong relationship between perceptions of workload and objective
levels of workload. In this case, where there is little relationship between subjective
and objective workload, the intervention would have to be aimed at changing sub-
jective perceptions rather than objective levels of workload. This is one reason why
any valid stress assessment must take into account both elements of job conditions.
The third component includes measurement of subjective and objective levels of
those phenomena thought to be caused by stress (e.g., absence, poor performance).
It is extremely common that organizations believe they have a problem with, say,
absence, but have no reliable data on which to base such a judgment.
Any assessment should ideally gather reliable data on all these components in
such a way that causal relations between them can be established and individual,
group, or location comparisons can be made. Although the only way causal relation-
ships can be examined is through longitudinal designs, there are a number of differ-
ent types of design that could be deployed. One is through naturally occurring ex-
periments in which assessments could be made over a period during which there
were natural fluctuations in objective job characteristics such as during the end of
the financial year or in seasonal work. A second type of longitudinal design would
be one in which, following an initial assessment period, an intervention based on
evidence from the assessment would be made and its effects observed. A third de-
sign, suggested by Briner and Reynolds [17], would utilize quantitative daily diaries
 Improvingstress assessment 69
in which self-reports of well-being and job conditions in employees would be taken
along with more objective assessments of job conditions. It would then be possible
to examine within-person correlations that indicate the extent to which employees'
well-being fluctuates, over time, with changes in objective job conditions. In this
way, individual and group profiles of the reactivity and responsiveness to changes
in work conditions could be developed.
A number of general principles of an evidence-based approach to SMIs have been
outlined in this work. There are also a number of practical steps that need to be
taken to develop this approach further. These involve the development of new and
better use of existing assessment techniques and changing our approach to dealing
with stress claims within organizations.
Many of the commercial and research-oriented self-report measures used in orga-
nizational stress are simply not based on best measurement practice. For example, it
is now widely agreed that affective states have at least two dimensions, often labeled
negative affect and positive affect, yet most measures use unidimensional measures
of mood or feelings. Similarly, as discussed earlier, scores on symptom measures are
open to a wide variety of interpretations and, in general, are unlikely to be good
indicators of either well-being or particular medical or clinical conditions. Although
self-report measures are both necessary and desirable, they should be based on
what is currently known about the ways in which people respond to self-reports,
rather than taking them simply at face value (cf. [17]).
In terms of developing a more sophisticated understanding of the components of
stress and, in particular, making sensible interpretations of data from other sources,
interviews and other kinds of qualitative data would be invaluable. Exit interviews
and absence interviews could be used to examine the extent to which stress is
thought to play a role in these behaviors and, more generally, the way in which em-
ployees construe their well-being and its relation to their work environment.
Although organizations already have data that could potentially be used in an as-
sessment, such as absence or performance figures, it is often the case that these data
have not been gathered or recorded systematically. An evidence-based approach
would also involve collecting such data and encouraging organizations to take a
more systematic approach.
An evidence-based approach would also include treating the stress claims made
by organizations with caution. Stress is often treated as though it is a purely scien-
tific phenomenon, somehow immune from cultural and social influence. Stress, like
any other explanation or label society offers to "explain" illness or negative affect,
is highly influenced by cultural conditions (e.g., [19]). The reason employees make
complaints about stress may often be highly symbolic and, for example, in response
to feeling undervalued or politically powerless [20]. Hence, an effective assessment
technique, particularly in interpreting the meanings of results obtained through as-
sessment, must take such factors into account. There is also a broader social context,
in that employees may come to believe, through exposure to popular media ac-
counts of stress, that the organization should do something about stress. Here too,
it may be that the complaint is more symbolic, and reflects a desire to be cared for
by the organization, or a desire to label the organization as uncaring.
Given the popularity of stress, and the "solution looking for a problem" phenom-
enon of SMIs, any effective assessment of stress also needs to take account of the
70 R.B. BRINER

social and political processes which underlie the desire by particular members of an
organization, or outside agents, to sell stress as an issue. In other words, as there
are considerable pressures to identify stress as an issue, irrespective of available evi-
dence, the claims made by organizations and employees about stress should not be
taken simply at face value.
An evidence-based approach to SMIs must also start with the premise that, fol-
lowing assessment, it may be that stress is found not to be a problem, and that other
problems and other solutions may emerge. Although this might be seen (particu-
larly by those keen to sell the stress issue and SMIs) as something of a disappoint-
ment, such an approach will ultimately be of greater value to the organization. For
example, it may be that absence is a problem but has nothing to do with "stress"
in any sense. In this case, interventions should be aimed at managing absence (see,
e.g., [21]) rather than managing stress.
Valid stress assessment requires the following: a relatively sophisticated and com-
prehensive approach to measurement; designs that permit causal relationships be-
tween the phenomena of interest to be established; a healthy scepticism toward the
claims that are made by organizations about stress; and a willingness to allow other
types of problems and solutions to emerge. This approach to assessment, which is
the foundation of the evidence-based approach to stress management intervention
in organizations outlined here, is certainly more elaborate and costly than current
practices. However, without such an approach, SMIs simply will not work.

CONCLUSIONS
Taking an evidence-based approach to SMIs is likely to raise some new issues.
First, what happens if our assessments show there are very limited or indeed no
links between work conditions and employee well-being? If the absence of links is
accepted I would hope that we would start to think again about the nature and ex-
tent of the influence of work conditions on employee well-being. Clearly, such links
of some kind do exist but perhaps traditional organizational stress models are sim-
ply inadequate: work may affect well-being but perhaps not in the term of "stres-
sors" causing "strain" (see [22]). Another possible benefit of not finding the links
we hope to find is that organizations focus on more specific problems with more spe-
cific solutions.
A second issue, already mentioned briefly, which is likely to emerge, is the extent
to which causal associations between job conditions and well-being have to be
shared or common among employees and the implications of this for intervention.
For example, if it was found that 50% of the workforce show a strong negative reac-
tion to reductions in control or autonomy, and the other 50% do not, or even show
a positive reaction to a reduction in control, increasing levels of autonomy by chang-
ing job design may be of benefit for only half the workforce and perhaps detrimen-
tal to the rest. If there are large individual differences in responses to work condi-
tions, which seems likely, then interventions that change job conditions for
everyone in the same way will not generally be effective. As a result of more sophis-
ticated assessment, more sophisticated types of intervention will need to be de-
veloped.
The available empirical evidence and the rational analysis presented earlier sug-
 Improving stress assessment 71

gest that SMIs as they are currently used are not particularly effective nor could
they be in principle until an evidence-based approach, such as the one proposed
here, is adopted.

REFERENCES

1. Ivancevitch JM, Matteson MT, Freedman SM, Phillips JS. Worksite stress management interven-
tions. Am Psychol 1990;45:252-261.
2. Highley JC, Cooper CL. An assessment of employee assistance and counselling programmes in Brit-
ish organizations. Paper presented at the 1995 annual British Psychological Society Occupational
Psychology Conference.
3. Kahn RL., Byosiere P. Stress in organizations. In: Dunnette M, ed. Handbook of industrial and orga-
nizational psychology. Chicago: Rand-McNally 1991:571-650.
4. Cox T. Stress research and stress management: putting theory to work. Health and Safety Executive
Contract Research Report No. 61/1993. Suffolk, UK: HSE Books 1993.
5. Wall TD, Kemp NL Jackson PR, Clegg CW. Outcomes of autonomous workgroups: a long-term
field experiment. Acad Mgt J 1986;29:280-304.
6. Cordery JL, Mueller WS, Smith LM. Attitudinal and behavioural effects of autonomous group work-
ing: a longitudinal field study. Acad Mgt J 1991;34:464-476.
7. Briner RB, Reynolds S. On the costs, benefits, and limitations of organizational level stress interven-
tions. Paper submitted for publication.
8. Reynolds S, Briner RB. Stress management at work: with whom, for whom, and to what ends? Br
J Guid Counsel 1994;22:75-89.
9. Earnshaw J, Cooper CL. Workers' compensation in stress-related claims: some thoughts for employ-
ers in the UK. Work Stress 1991;5:253-257.
10. Dutton JE, Ashford SJ. Selling issues to top management. Acad Mgt Rev 1993;18:397-428.
11. Leigh C, Briner RB. Identifying and selling stress as an organizational issue. Paper presented at the
1995 annual British Academy of Management Conference.
12. Spector PE. A consideration of the validity and meaning of self-report measures of job conditions.
In: Cooper CL, Robertson IT, eds. International review of industrial and organizational psychology.
Chichester: Wiley 1992:123-151.
13. Salovey P, O'Leary A, Stretton MS, Fishkin SA, Drake CA. Influence of mood on judgements about
health and illness. In: Forgas JP, ed. Emotion and social judgement. Oxford: Pergamon
1991:241-262.
14. Burke M J, Brief AP, George JM. The role of negative affectivity in understanding relations between
self-reports of stressors and strains: a comment on the applied psychology literature. J Appl Psy-
chol 1993;78:402-412.
15. Pennebaker JW. The psychology of physical symptoms. New York: Springer New York 1982.
16. Coyne JC. Self-reported distress: analog or ersatz depression. Psychol Bull 1994;116:29-45.
17. Briner RB, Reynolds S. Occupational stress: effectiveness of organizational interventions. Research
proposal submitted to the Health and Safety Executive: research project reference RSU3203/R54.61.
University of London, Birkbeck College 1994.
18. Jex SM, Beehr TA. Emerging theoretical and methodological issues in the study of work-related
stress. Res Person Human Res Mgt 1991;9:311-365.
19. Pollock K. On the nature of social stress: production of a modern mythology. Soc Sci Med
1988;26:381-392.
20. Barley SR, Knight DB. Toward a cultural theory of stress complaints. Res Org Behav 1992;14:1-48.
21. Briner RB. Absence from work. BMJ 1996;313/7061:874-877.
22. Briner RB. Feeling for the facts. People Mgt 1997;3:34-37.