01.05.2015 Views

Evaluating organizational stress-management interventions using ...

Evaluating organizational stress-management interventions using ...

Evaluating organizational stress-management interventions using ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

EUROPEAN JOURNAL OF WORK AND<br />

ORGANIZATIONAL PSYCHOLOGY<br />

2005, 14 (1), 23–41<br />

<strong>Evaluating</strong> <strong>organizational</strong> <strong>stress</strong>-<strong>management</strong><br />

<strong>interventions</strong> <strong>using</strong> adapted study designs<br />

Raymond Randall<br />

Department of Psychology, City University, London, and<br />

Institute of Work, Health and Organizations, University of Nottingham, UK<br />

Amanda Griffiths and Tom Cox<br />

Institute of Work, Health and Organizations, University of Nottingham, UK<br />

The evaluation of <strong>organizational</strong> <strong>stress</strong> <strong>management</strong> <strong>interventions</strong> has proved<br />

challenging for researchers and practitioners alike. Traditionally, researcher<br />

designed quasi-experiments have been regarded as the method for evaluating<br />

such <strong>interventions</strong>. However, relatively few such studies have been satisfactorily<br />

completed in organizations, and many of those that have did not<br />

adequately take account of intervention processes. This article presents an<br />

approach to evaluation that can help to overcome these problems. Two<br />

empirical studies are presented that demonstrate that measurement of the<br />

intervention process can be used to adapt and shape the design of the<br />

evaluation. In both studies, process evaluation incorporating the measurement<br />

of intervention exposure was used to partition participant samples (into<br />

intervention and control groups). This approach has the potential to enable<br />

and strengthen quantitative outcome evaluation in situations where controlled<br />

quasi-experimentation is not possible.<br />

Organizational-level <strong>stress</strong> <strong>management</strong> <strong>interventions</strong> are designed to deal<br />

with the sources of the problem by changing the design, <strong>management</strong>, and<br />

organization of work (Cox, Griffiths, & Rial-Gonzalez, 2000b; Semmer,<br />

2003). Organizational-level <strong>stress</strong> prevention is either implicitly or explicitly<br />

endorsed by a number of European Governments (Griffiths, 2003; Griffiths,<br />

Cox, & Barlow, 1996; Health and Safety Commission, 1999; Kompier, de<br />

Correspondence should be addressed to Dr Raymond Randall, Department of Psychology,<br />

City University, Northampton Square, London, EC1V 0HB, UK. Email: r.randall@city.ac.uk<br />

The authors would like to thank the UK Health and Safety Executive, The Royal College of<br />

Nursing, and UNISON for their support for the work presented in this article. The views and<br />

opinions expressed are those of the authors and not those of any other individual or<br />

organization.<br />

# 2005 Psychology Press Ltd<br />

http://www.tandf.co.uk/journals/pp/1359432X.html DOI: 10.1080/13594320444000209


24 RANDALL, GRIFFITHS, COX<br />

Gier, Smulders, & Draaisma, 1994) and by the European Commission<br />

(1989, Article 6:2). Because it targets the causes of work <strong>stress</strong> such a ‘‘risk<br />

assessment – risk reduction’’ strategy (Cox, Griffiths, Barlow, Randall,<br />

Thomson, & Rial-Gonzalez, 2000a) should be the most effective in the<br />

long-term (Cooper, Liukkonen, & Cartwright, 1996; Cox, 1993; Cox,<br />

Griffiths, & Randall, 2002a, 2002b; Ivancevich, Matteson, Freedman, &<br />

Phillips, 1990; Murphy, 1996; van der Hek & Plomp, 1997). However, two<br />

problems mean that this argument is, at present, difficult to justify. First, for<br />

some time there has been a dearth of adequate evaluation studies of the<br />

effectiveness of such <strong>interventions</strong> (Briner & Reynolds, 1999; Cox, 1993;<br />

Parkes & Sparkes, 1998; Reynolds, 1997; Semmer, 2003). Second, many<br />

evaluations are limited by the undermeasurement of intervention processes<br />

(Cox et al., 2000b; Griffiths, 1999; Kompier & Kristensen, 2000; Murphy,<br />

1996; Parkes & Sparkes, 1998; Semmer, 2003). The aim of this article is to<br />

describe a modified approach to evaluation that helps to address these two<br />

important interrelated problems.<br />

TRADITIONAL EVALUATION STRATEGIES<br />

Traditionally quasi-experiments have been used to evaluate <strong>interventions</strong> for<br />

work-related <strong>stress</strong> because the constraints of the <strong>organizational</strong> setting and<br />

the nature of the <strong>interventions</strong> do not support the conditions required for a<br />

‘‘true’’ experiment (Campbell, 1957; Cook & Campbell, 1979; Parkes &<br />

Sparks, 1998). However, much of the existing <strong>stress</strong> <strong>management</strong> evaluation<br />

research literature has implied that two features of the ‘‘true’’ experiment<br />

have to be retained in order to provide a robust enough evaluation (e.g.,<br />

Briner & Reynolds, 1999; Murphy, 1996; Parkes & Sparkes, 1998). The first<br />

point is that fixed (stable) study designs (based around the controlled or<br />

predictable manipulation of exposure) should be used; and the second point<br />

is that outcome evaluation should be paramount, since it has often been<br />

(erroneously) assumed that strong quasi-experimental designs make process<br />

evaluation redundant (Cook & Shadish, 1994).<br />

However, the complexity and instability of organizations tends to make it<br />

difficult (or even impracticable) to establish, and then adequately control,<br />

the delivery of <strong>interventions</strong> to achieve even the simplest of quasiexperimental<br />

study designs (Griffiths, 1999; Kompier & Kristensen, 2000;<br />

Mikkelsen, Saksvik, & Landsbergis, 2000). Moreover, the process of<br />

implementing <strong>interventions</strong> can modify indented exposure patterns (by<br />

stopping <strong>interventions</strong> reaching their intended participants and vice versa)<br />

and cannot be ignored in outcome evaluation (Griffiths, 1999).<br />

This situation presents serious problems for summative evaluation<br />

strategies (i.e., those that focus on the outcome evaluation, often at the<br />

expense of process evaluation) that rely upon fixed or predictable exposure


EVALUATION OF INTERVENTIONS 25<br />

patterns (Colarelli, 1998; Griffiths, 1999; Hartley, 2002; Heaney, Israel,<br />

Schurman, Baker, House, & Hugentobler, 1993). In unpredictable or<br />

uncontrolled settings such an approach (1) raises the risk of Type III error<br />

(erroneously concluding an intervention is ineffective when it is actually its<br />

implementation that is faulty; Dobson & Cook, 1980) and (2) limits<br />

explanatory yield (e.g., inconsistent intervention effects remain difficult to<br />

explain; see Cox et al., 2000b; Parkes & Sparkes, 1998). Moreover,<br />

controlled or predictable intervention exposure patterns occur so infrequently<br />

that there is a need for alternative ways of managing quantitative<br />

evaluation 1 that are viable in the face of unpredictable and uncontrollable<br />

exposure patterns (Colarelli, 1998; Kompier, Aust, van den Berg, & Seigrist,<br />

2000a). In summary, the identification of causal relationships may be<br />

hindered unless study designs are adapted to reflect true, but uncontrollable<br />

and unpredictable, patterns of intervention exposure.<br />

ADAPTED STUDY DESIGNS AS AN EVALUATION<br />

STRATEGY<br />

Applied social scientists (such as those evaluating public health promotion<br />

or large-scale community education programmes) have achieved good<br />

results by being flexible in their application of the principles of study design<br />

(Fitzgerald & Rasheed, 1998; Harachi, Abbot, Catalano, Haggerty, &<br />

Fleming, 1999; Lipsey & Corday, 2000). When, for example, working in<br />

community settings they have adapted study designs, through the use of<br />

process evaluation, to reflect actual intervention exposure patterns (Lipsey,<br />

1996; Lipsey & Corday, 2000). On-going or post hoc measures of<br />

intervention exposure (i.e., process evaluation: see Kompier & Kristensen,<br />

2000; Yin, 1994, 1995; Yin & Kaftarian, 1997) have been used to identify or<br />

adapt the evaluation design so that the evaluation can ‘‘work backward from<br />

the target clientele and what they actually receive/experience, not forward<br />

from the intervention activities and what the intervention agents purportedly<br />

deliver’’ (Lipsey, 1996, p. 301).<br />

Given the sometimes insurmountable difficulties associated with intentionally<br />

introducing and controlling intervention exposure, this flexible<br />

approach to evaluation offers a practical means of evaluating <strong>stress</strong><br />

<strong>management</strong> <strong>interventions</strong>. Data on exposure to <strong>interventions</strong> can be<br />

obtained through an intervention process evaluation (i.e., questioning<br />

participants about their experiences and triangulating those data with<br />

documentary information and by interviewing those involved in planning<br />

and implementing <strong>interventions</strong>; Griffiths, 1999; Nytro, Saksvik, Mikkelsen,<br />

1 The authors recognize that qualitative methods also offer viable alternative approaches to<br />

evaluation in chaotic <strong>organizational</strong> settings (see Kompier et al., 2000a).


26 RANDALL, GRIFFITHS, COX<br />

Bohle, & Quinlan, 2000; Saksvik, Nytro, Dahl-Jorgensen, & Mikkelsen,<br />

2002). This process evaluation can then be used to adapt outcome<br />

evaluation by <strong>using</strong> it to determine whether each participant is more<br />

appropriately placed in a intervention/exposed or control/not exposed group.<br />

Measured exposure patterns can thus be exploited as an evaluation design<br />

variable. This approach is different to that used in ‘‘natural experiments’’<br />

where exposure patterns are predictable and controlled (see Jackson, 1983).<br />

Rather, it is a constructive use of the manipulation check: The study design<br />

is adapted to reflect actual exposure patterns.<br />

Treating uncontrolled and unpredictable exposure patterns as a natural<br />

manipulation may help to expand the size of the ‘‘pool’’ of <strong>organizational</strong><br />

intervention evaluation research and make informative evaluation possible<br />

in chaotic <strong>organizational</strong> settings.<br />

THE PRESENT STUDY<br />

This article presents two empirical studies drawn from the authors’ research<br />

on risk <strong>management</strong> and work-related <strong>stress</strong> (Cox et al., 2000a; Cox,<br />

Randall, & Griffiths, 2002b). Together they illustrate that through a simple<br />

exploration of the intervention processes, exposure patterns can be identified<br />

and used to adapt the final study design and analysis. This achieves two<br />

things: It measures actual exposure to the intervention to allow a valid<br />

evaluation of its effectiveness (thus controlling Type III error), and it<br />

permits informative evaluation where exposure patterns cannot be planned<br />

or tightly controlled. Both studies examine, ad hoc, the actual ‘‘<strong>organizational</strong><br />

penetration’’ (Cox et al., 2000a) of an intervention into a participant<br />

group. This strategy enables evaluation without the need to (1) make<br />

possibly untenable assumptions about stable intervention exposure patterns,<br />

(2) to control exposure to an intervention, or (3) to rely on events or<br />

<strong>organizational</strong> structures to coincidentally support predictable and stable<br />

exposure patterns. Both studies have one central hypothesis:<br />

The intervention effect (improved well-being) will only be apparent when measures<br />

of actual intervention exposure are used to partition the participant group prior to<br />

analysis (i.e., actual intervention exposure is a moderator of change).<br />

METHOD<br />

Design<br />

In both studies, the <strong>interventions</strong> were designed by stakeholders from within<br />

the participating organizations, through the feedback of risk assessment<br />

data and discussions facilitated by the research team. These discussions were


EVALUATION OF INTERVENTIONS 27<br />

based on an initial problem analysis carried out <strong>using</strong> a risk assessment<br />

questionnaire survey of working conditions. Both studies began with a<br />

simple and traditional pre (Time 1) – post (Time 2) longitudinal intervention<br />

design. Because of the constraints operating within the organizations<br />

involved it was not possible to allocate participants to ‘‘intervention’’ and<br />

‘‘control’’ groups: each intervention was intended to reach all study<br />

participants. In both studies, participants were asked to report on their<br />

awareness of (Study 1) or their involvement in (Study 2) the intervention. In<br />

both studies the dependent variable was self-reported well-being (measured<br />

in terms of levels of exhaustion). Time 2 measures of well-being and<br />

intervention exposure were taken 18 months after the Time 1 measures of<br />

well-being. The evaluation design was adapted by partitioning the<br />

participants groups according to their reported exposure to the intervention<br />

(<strong>using</strong> a Measured exposure 6 Time interaction term) in a repeated<br />

measures analysis of covariance (ANCOVA).<br />

Participants and <strong>interventions</strong><br />

Study 1: Railway staff. Thirty-seven station managers from a railway<br />

transport company provided data at both Time 1 and Time 2. This<br />

represented a response rate of approximately 50% (based on the number of<br />

supervisors providing both Time 1 and Time 2 data). Response rates for<br />

Times 1 and 2, taken separately were higher (68% and 64%). All were male<br />

with an average age of 40 years (SD = 8.0) and average tenure of 14 years<br />

(SD = 7.1) at Time 1. The supervisors worked at different sized stations:<br />

Some stations were larger and busier than others. Inspection of company<br />

records indicated that the supervisors returning questionnaires came from a<br />

representative sample of stations and that they did not differ from the whole<br />

station manager population in terms of average age or length of service.<br />

Some months before Time 1, a central part of the station managers’ role<br />

had been changed. Because of budgetary constraints, their responsibility for<br />

managing the repair of faulty station equipment (including reporting faults<br />

and authorizing and managing repairs 2 ) had been removed. At Time 1<br />

senior managers, not supervisors, managed equipment repair. However,<br />

when the risk assessment data collected at Time 1 was reported, the<br />

organization interpreted these as indicating that the removal of roles and<br />

responsibilities had not been well received by staff. In an attempt to improve<br />

staff satisfaction and well-being, responsibility for managing the repair of<br />

faulty station equipment was returned to station supervisors (i.e., super-<br />

2 This does not refer to track and trains, but rather to equipment found in the station<br />

building. Inspection and repair of track equipment was carried out by other specialist members<br />

of staff.


28 RANDALL, GRIFFITHS, COX<br />

visors were instructed to resume reporting faults and instigating their<br />

resolution). The intention was that this change be communicated to all<br />

station supervisors through a number of levels of <strong>management</strong> through two<br />

media: via (1) written memos from senior <strong>management</strong> delivered through<br />

established communication routes and (2) verbal communications in a<br />

variety of forums (e.g., individual and team meetings).<br />

Study 2: Hospital staff. Participants were 31 senior paediatric nursing<br />

staff with significant managerial and administrative responsibilities in<br />

addition to a specialist clinical workload. They worked in a large urban<br />

hospital. A response rate of 52% was achieved (70% at Time 1 and 66% at<br />

Time 2 respectively). All participants were female, with the majority (56%)<br />

being aged between 36 and 45 years, with most (also 56%) having worked in<br />

the hospital for more than 11 years. 3 The nurses worked in 15 different<br />

wards, each with its own specialty including oncology, orthopaedic, and<br />

outpatients. Comparisons between the demographic data obtained from the<br />

risk assessment questionnaire (age, length of service, and size of ward<br />

worked in) and the hospital’s records indicated that the nurses completing<br />

questionnaires were representative of the whole sample.<br />

The rationale driving the intervention in this study was relatively<br />

straightforward. The risk assessment identified that there were few<br />

computing facilities on the wards at Time 1. There were a handful of<br />

computers shared between the 15 wards. As a consequence, access to these<br />

facilities was erratic. Staff needed to use computers for many aspects of their<br />

administrative and managerial work, and were often unable to progress<br />

tasks because of a lack of access to them. Further, it was well recognized<br />

that communication within such a large and diverse department (comprising<br />

15 different wards) was difficult: It was felt that providing staff with access to<br />

intranet and email facilities would significantly improve the flow of<br />

information within the department. The agreed intervention plan was to<br />

introduce fully functional computing facilities (an internet-ready computer<br />

with email, word-processing, and spreadsheet capabilities) in each ward over<br />

the 6 months after the Time 1 measures. Each participant’s involvement in<br />

(exposure to) the intervention was determined purely by the progress that<br />

had been made on the installation of new computer technology.<br />

Measures<br />

In both studies data on demographic variables age, gender, length of service,<br />

and work location were gathered by self-report. A correlate of the emotional<br />

3 Age and length of service measured <strong>using</strong> categorical items to protect participant<br />

anonymity.


EVALUATION OF INTERVENTIONS 29<br />

experience of work <strong>stress</strong> (work-related well-being) was measured <strong>using</strong> the<br />

exhaustion scale of the General Well-Being Questionnaire (GWBQ; Cox &<br />

Griffiths, 1995; Cox, Thirlaway, Gotts, & Cox, 1983). This measure was<br />

used as the dependent variable to examine the likely impact of the<br />

intervention on work-related well-being. The exhaustion scale is a 12-item<br />

self-report measure of nonspecific symptoms of general malaise relating to<br />

fatigue, cognitive confusion, and emotional irritability. It has been shown to<br />

be sensitive to the fluctuations in well-being associated with the emotional<br />

experience of <strong>stress</strong> at work (Cox & Gotts, 1987; Cox et al., 1983).<br />

Participants recorded their experience of these symptoms <strong>using</strong> a 5-point<br />

frequency scale of 0 (never) to 4 (always) with a time window of<br />

measurement set as the preceding 6 months: The higher the participant’s<br />

score on the questionnaire, the ‘‘poorer’’ their well-being. The scale was<br />

found to be reliable in both studies (Cronbach’s alphas: preintervention<br />

= .82 (Study 1) and .85 (Study 2); postintervention = .89 (Study 1) and<br />

.83 (Study 2). Examination of the standard errors for skewness and kurtosis<br />

showed that scores were normally distributed at both measurement points.<br />

Existing data shows that for employees in managerial posts the normative<br />

score on this measure is approximately 17 (Cox & Gotts, 1987; Cox et al.,<br />

2000). Preintervention, both participant groups appeared more exhausted<br />

than the normative group (mean Study 1 = 20.3, SD = 8.7; mean Study<br />

2 = 18.9, SD = 7.5). These high levels were, in part, justification for both<br />

risk <strong>management</strong> projects.<br />

Measures of intervention exposure were taken at Time 2 and were<br />

designed to tap into the active ingredient of the each intervention, i.e., the<br />

aspect of exposure hypothesized to be the driver of change (Kompier,<br />

Cooper, & Geurts, 2000b) in exhaustion scores. In Study 1, the active<br />

ingredient was awareness of the new guidelines and procedures for fault<br />

reporting. It was predicted that being aware of the new guidelines would<br />

make a difference. Participants were asked to indicate their awareness of<br />

(exposure to) the intervention through a single dichotomous item: ‘‘Indicate<br />

whether or not you are aware of: the return of fault reporting to your<br />

control’’ (0 = ‘‘No’’, 1 = ‘‘Yes’’). To test their reliability and validity these<br />

data were triangulated. Senior managers (n = 3), and a sample of the<br />

managers responsible for implementing the intervention (n = 6) were asked<br />

to comment on its implementation. Records of written communications<br />

were also examined for evidence of implementation.<br />

In Study 2, the ‘‘active ingredient’’ was involvement in the programme of<br />

updating computer equipment in the wards (i.e., a measure of whether each<br />

participant had used new computer equipment recently installed in their<br />

ward). Participants were asked to indicate their involvement in the<br />

intervention through a single dichotomous item: ‘‘Indicate whether or not<br />

you have been involved in: the use of new computer equipment in your


30 RANDALL, GRIFFITHS, COX<br />

ward’’ (0 = ‘‘No’’, 1 = ‘‘Yes’’). These data were triangulated. Senior<br />

managers (n = 4) were asked to identify wards which had received new<br />

computer equipment. Documents detailing the progress of the installation<br />

programme were also examined.<br />

Analysis<br />

A two-stage analytical procedure was used in both studies. First, the initial<br />

pre – post design (assuming 100% exposure) was used. Each intervention<br />

was evaluated through inspection of descriptive data and the use of pair<br />

sampled t-tests and repeated measures ANOVAs to examine changes in<br />

well-being over time. Covariates were not considered to allow for more<br />

liberal testing of intervention effects thus reducing the chance of Type II<br />

error. Traditionally this design would be used to test for the impact of the<br />

intervention assuming 100% exposure.<br />

In the second stage of analysis, exposure to the intervention was included<br />

as a design variable. A repeated measures ANCOVA was used with one<br />

between-subject variable (‘‘exposed’’ to the intervention or ‘‘not exposed’’ to<br />

the intervention) and one within-subject variable (time) with two measurement<br />

levels (Time 1 and Time 2). Covariates (age and length of service) were<br />

included because of their potential influence over growth in the dependent<br />

variable and to control for nonequivalence in the between-group portion of<br />

the design and reduce the chance of Type I error (Cook & Campbell, 1979;<br />

Tabachnick & Fiddell, 2001). The impact of exposure to the intervention<br />

was assessed by examining the significance of the two-way interaction term<br />

Intervention exposure 6 Time. Interaction effects were then explored in two<br />

ways. First, paired sample t-tests were used to test for changes in exhaustion<br />

scores within each group. Second, comparisons between the exposed and not<br />

exposed groups were also made at both Time 1 and Time 2 <strong>using</strong> ANOVAs.<br />

RESULTS<br />

Study 1<br />

Exposure to the intervention. The measure of awareness of the<br />

intervention showed that most (25 supervisors) reported that they had been<br />

made aware that they had regained control of fault reporting and repair<br />

authorization. However, a significant proportion (32%) was not aware of<br />

the change. There were no significant differences between the ‘‘aware/<br />

exposed to the intervention’’ and the ‘‘not aware/not exposed to the<br />

intervention’’ group in terms of demographic details (age, length of service,<br />

the size of station they were based at), nor their exhaustion scores at Time 1,


EVALUATION OF INTERVENTIONS 31<br />

F(1, 36) = 1.27, p 4 .05. When triangulated with other methods of process<br />

evaluation this finding appeared robust. Stakeholder analysis indicated that<br />

some senior managers had resisted informing staff of the intervention<br />

because of the budgetary constraints placed on particular groups of stations.<br />

This was supported by two other pieces of data: (1) There was no evidence<br />

of the intervention in communication records for some stations, and (2)<br />

those who were unaware of the intervention shared the same communication<br />

routes.<br />

Evaluation of the intervention. Using the traditional nonadaptive pre –<br />

post study design, there was no significant change in exhaustion scores,<br />

t = 0.64, p 4 .05; F(1, 36) = 0.24, p 4 .05, over the intervention period (see<br />

TABLE 1<br />

Changes in exhaustion score (Study 1, whole sample)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Mean<br />

change t F<br />

Exhaustion score 20.3 (SD = 8.7) 18.9 (SD = 9.4) – 1.4 0.64* 0.24*<br />

N = 37.<br />

*p 4 .05.<br />

TABLE 2<br />

Repeated measures analysis of covariance (Study 1)<br />

Type III sum of squares 1<br />

F<br />

Within-subjects effects<br />

Time (pre – post intervention) 67.26 2.22<br />

Time 6 Length of service 49.03 1.62<br />

Time 6 Age 97.99 3.24<br />

Time 6 Exposure to intervention 206.79 6.83**<br />

Between-subjects effects<br />

Age 566.21 6.13*<br />

Length of service 1.79 0.02<br />

Awareness of the intervention 352.47 3.82<br />

N = 37.<br />

*p 5 .05; **p 4 .01.<br />

1 Used here to take into account the discrepancy in sample size between the group exposed to<br />

the intervention and the group not exposed to the intervention.<br />

Box’s M statistic was nonsignificant, F(3, 10402) = 0.68, p 4 .05, and Levene’s test of the<br />

equality of error variance was nonsignificant: for preintervention exhaustion scores,<br />

F(1, 35) = 0.19, p 4 .05; for postintervention exhaustion scores, F(1, 35) = 0.05, p 4 .05.


32 RANDALL, GRIFFITHS, COX<br />

Table 1). This would lead naturally to the conclusion that the intervention<br />

was ineffective in improving well-being.<br />

Preanalysis checks confirmed the suitability of the data for repeated<br />

measures ANCOVA analysis (see Table 2): The usual significance level of<br />

p 5 .05 could be applied to the testing of effects. The results of the repeated<br />

measures analysis of covariance are presented in Table 2. After controlling<br />

for variance in the dependent variable accounted for by age, F(1, 33) = 6.13,<br />

p 5 .05, the test of within-subjects effects revealed one significant interaction:<br />

Exposure to the Intervention 6 Time, F(1, 32) = 6.83, p = .01; etasquared<br />

.17. None of the other within-subject effects (the main effect of time<br />

and the other interaction effects), were significant, F 4 3.24, p 4 .08. This<br />

interaction (with adjusted means) is shown in Figure 1. None of the other<br />

between-subjects effects were significant.<br />

Table 3 shows the changes in exhaustion scores for the exposed and not<br />

exposed groups separately. Exploration of the interaction term indicated<br />

that from similar preintervention worn-out scores, F(1, 36) = 1.27, p 4 .05,<br />

the scores of the two groups diverged to result in a significant difference in<br />

postintervention exhaustion scores, F(1, 36) = 10.3, p 5 .01. This divergence<br />

Figure 1.<br />

Interaction effect (Study 1: Railway staff).


EVALUATION OF INTERVENTIONS 33<br />

TABLE 3<br />

Descriptive statistics for exhaustion scores exploring the interaction term (Study 1)<br />

Exposed/aware group<br />

(n = 25)<br />

Not exposed/not aware group<br />

(n = 12)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Mean exhaustion 19.3 16.3 22.8 25.3<br />

scores (range 0 – 48) (SD = 8.2) (SD = 9.1) (SD = 8.1) (SD = 7.6)<br />

was attributable to a significant drop (3.0 scale points) in exhaustion scores<br />

in the group exposed to the intervention, t = 2.13, p 5 .05; F(1, 26) = 5.23,<br />

p 5 .05, alongside a nonsignificant average rise of rise of 2.5 scale points in<br />

the group not exposed to the intervention, t = – 1.53, p 4 .05; F(1,<br />

11) = .61, p 4 .05. These results indicated that the group exposed to the<br />

intervention experienced an improvement in well-being, while well-being in<br />

the group not exposed to the intervention remained relatively stable. This is<br />

a very different conclusion from that reached when the measure of<br />

intervention exposure was not considered in the traditional nonadaptive<br />

analysis.<br />

Study 2<br />

Exposure to the intervention. There was an approximate 50:50 split<br />

among the group of paediatric nurses in relation to the intervention: Fifteen<br />

reported having used new <strong>using</strong> computer facilities on their ward, while<br />

sixteen reported not having access to new computer facilities on their own<br />

ward. As in Study 1, there were no significant differences between the<br />

‘‘involved in/exposed to the intervention’’ and the ‘‘not involved in/not<br />

exposed to the intervention’’ group in terms of demographic details or<br />

preintervention exhaustion scores. This finding also appeared robust when<br />

triangulated. Senior managers reported that progress on installing new<br />

computers had been slow because of a lack of specialist computer staff<br />

within the hospital. Only around 60% of the computers had been fully<br />

installed and were operational.<br />

Evaluation of the intervention. Table 4 shows that <strong>using</strong> the traditional<br />

nonadaptive, pre – post study design there was no significant change in<br />

exhaustion scores over the intervention period, t = 0.32, p = .75; F(1,<br />

30) = 0.03, p 4 .85. As in Study 1 this would lead to the conclusion that the<br />

intervention was ineffective.


34 RANDALL, GRIFFITHS, COX<br />

TABLE 4<br />

Changes in exhaustion score (Study 2, whole sample)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Mean<br />

change t F<br />

Exhaustion score 18.9 19.5 0.6 0.32* 0.03*<br />

(SD = 7.5) (SD = 6.3)<br />

N = 31.<br />

*p 4 .05.<br />

TABLE 5<br />

Repeated measures analysis of covariance (Study 2)<br />

Type III sum of squares<br />

F<br />

Within-subjects effects<br />

Time (pre – post intervention) 0.20 0.01<br />

Time 6 Length of service 0.79 0.04<br />

Time 6 Age 0.03 0.03<br />

Time 6 Involvement in intervention 100.78 4.83*<br />

Between-subjects effects<br />

Length of service 0.68 0.01<br />

Age 2.95 0.04<br />

Involvement in the intervention 4.73 0.06<br />

N = 31.<br />

*p 5 .05.<br />

Box’s M statistic, F(3, 177953) = 0.48, p 4 .05, and Levene’s test of the equality of error<br />

variance were nonsignificant: for preintervention exhaustion scores, F(1, 29) = 0.39, p 4 .05; for<br />

postintervention exhaustion scores, F(1, 29) = 0.01, p 4 .05.<br />

Preanalysis checks showed that the data was suitable for repeated<br />

measures ANCOVA analysis (see Table 5). The results of the repeated<br />

measures analysis of covariance are presented in Table 5.<br />

No significant adjustments needed to be made to the dependent variable<br />

before testing the Exposure 6 Time interaction term. This interaction was<br />

significant,<br />

F(1, 27) = 4.83, p = .04; eta squared = .16 (see Table 5). None of the<br />

other within-subject effects (the main effect of time and the other interaction<br />

effects) were significant, F 4 0.06, p 4 .81. This interaction showed a<br />

crossover effect and is shown (with adjusted means) in Figure 2.<br />

Table 6 shows the changes in exhaustion scores for the exposed and<br />

not exposed groups separately. Exploration of the interaction term<br />

indicated that there was no significant difference between the two groups


EVALUATION OF INTERVENTIONS 35<br />

Figure 2.<br />

Interaction effect (Study 2: Nurses).<br />

TABLE 6<br />

Descriptive statistics for exhaustion scores exploring the interaction term (Study 2)<br />

Exposed/involved group<br />

(n = 15)<br />

Not exposed/not involved group<br />

(n = 16)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Preintervention<br />

(Time 1)<br />

Postintervention<br />

(Time 2)<br />

Mean exhaustion 20.5 17.4 18.7 20.8<br />

scores (range 0 – 48) (SD = 7.8) (SD = 7.2) (SD = 6.4) (SD = 5.8)<br />

exhaustion scores preintervention, F(1, 30) = 0.43, p = .52. After the<br />

intervention, the scores for the group not involved in the in intervention<br />

had risen, though not significantly, t = – 1.5, p = .15; F(1, 15) = 1.5,<br />

p = .24, while the mean exhaustion score for the group involved in the<br />

intervention had dropped, though again not significantly, t = 1.74,<br />

p = .10; F(1, 14) = 3.10, p = .10. After the intervention the group


36 RANDALL, GRIFFITHS, COX<br />

involved in the intervention reported lower exhaustion scores than the<br />

group not involved in the intervention, but this difference only<br />

approached significance, F(1, 30) = 2.24, p = .14.<br />

Taken together, however, the effect of the changes within the two groups<br />

was significant: Interaction terms are more sensitive than separate withingroups<br />

analysis (Tabachnick & Fidell, 2001). As in Study 1, these results<br />

suggested that exposure to the intervention impacted on participants’ wellbeing:<br />

This was only apparent when measures of exposure were used to<br />

adapted the design of the study during analysis.<br />

DISCUSSION<br />

The significant Measured exposure 6 Time interaction effects in both<br />

studies indicated that exposure to the intervention predicted changes in<br />

exhaustion scores over time. The central hypothesis of this article was<br />

confirmed: Nonadaptive designs underestimated the impact of the intervention<br />

on well-being with significant change only becoming apparent when<br />

adapted study designs were used. Both studies indicated that Type III error<br />

(Dobson & Cook, 1980; Harachi et al., 1999) may be minimized by <strong>using</strong> the<br />

results of a robust process evaluation (in this study one that centred on the<br />

triangulation of exposure data) to adapt outcome evaluation. This<br />

protection against Type III error is particularly important given that the<br />

psychological components of work design may exert a relatively modest<br />

influence over general well-being in the short term (Zapf, Dormann, &<br />

Frese, 1996). Moreover, in both studies measures of intervention exposure<br />

identified hidden and ‘‘unintended’’ between-groups designs that facilitated<br />

much-needed and significant improvements in methodological adequacy (see<br />

Beehr & O’Hara, 1987; Murphy, 1996). In Study 1, the between-group<br />

differences found at Time 2 reflected a worsening of the situation for the<br />

‘‘not aware’’ group cooccurring with stability in the ‘‘aware’’ group,<br />

suggesting that the return of responsibilities to them protected supervisors<br />

from the effects of problems associated with not being able to report faults.<br />

This ‘‘protective effect’’ has been observed in other intervention studies (e.g.,<br />

Terra, 1995). The pattern of change in Study 2 indicated a significant<br />

intervention effect when the small changes in the intervention and control<br />

groups cooccurred.<br />

Clearly, and for good reasons, the methodological adequacy of both<br />

studies reflected the constraints placed on them by the research setting. Like<br />

almost all <strong>stress</strong> <strong>management</strong> intervention evaluation studies they do not<br />

play exactly by the methodological rules (Kompier et al., 2000a). However,<br />

measuring and capitalizing on uncontrolled and unpredictable exposure<br />

patterns did extend the established principles of the ‘‘natural experiment’’.<br />

Indeed, in the majority of situations where complete control over


EVALUATION OF INTERVENTIONS 37<br />

intervention exposure is not possible, this approach offers a practical and<br />

informative method of evaluation. Of course, it should only be used when<br />

rigorous process evaluation yields a strong case for adapting the study<br />

design: The methodology should not be used as a justification for ‘‘fishing’’<br />

in data for significant results (Yin, 1994). The small sample sizes provided by<br />

the majority of organizations (as was the case in both of the studies<br />

presented here) make theory-driven analysis and preanalysis checks on the<br />

data particularly important.<br />

Naturally, this approach is not without its problems. Unpredictable<br />

exposure patterns create an extreme form of nonequivalent study design and<br />

may be affected by selection biases (Campbell & Stanley, 1963; Cook &<br />

Campbell, 1979). In both of the studies presented here, intervention and<br />

control groups were comparable across a range of demographic variables,<br />

and appeared well-matched at Time 1 on the measure of the dependent<br />

variable. However, when this is not the case possible bias can be controlled<br />

for by the analysis of covariates, or by <strong>using</strong> ‘‘blocking’’ designs that match<br />

cases on an ad hoc basis (Cook & Campbell, 1979; Cook & Shadish, 1994).<br />

Further, the list of active ingredients (Kompier & Kristensen, 2000) that are<br />

indicative of intervention exposure across a range of different types of<br />

intervention is not yet established. Measuring involvement may be<br />

appropriate for <strong>interventions</strong> that require the participant to actively engage<br />

in the intervention for it to work (e.g., a training intervention, teambuilding<br />

intervention, staff consultation process, etc.). Measuring awareness may be<br />

more appropriate for <strong>interventions</strong> that have a more passive mechanism (e.g.,<br />

the receipt of information, changes in guidelines, or a redefinition of roles<br />

and responsibility). Further work is needed to establish the modus operandi<br />

of a variety of <strong>organizational</strong> <strong>interventions</strong> in order to ensure that<br />

appropriate measures of intervention exposure are used to adapt study<br />

designs.<br />

Triangulating the self-report data <strong>using</strong> process evaluation helped to<br />

guard against threats to the validity of the self-report intervention exposure<br />

measure, and greatly enhanced understanding of the intervention effects. In<br />

Study 1, it appeared that staff who were managed by particular senior<br />

managers had not received the intervention. In Study 2, exposure to the<br />

intervention was determined by the speed at which the hospital’s computer<br />

technicians could install and make fully operational the computer<br />

equipment. These findings indicated that reported exposure to the<br />

intervention was driven by implementation factors rather than by any<br />

individual differences (e.g., negative affect: Brief, Burke, George, Robinson,<br />

& Webster, 1988; Watson & Clark, 1984) or methodological artifacts (e.g.,<br />

common method variance; Spector, 1994). Even in controlled exposure<br />

studies such data about the change process should be gathered to enhance<br />

the explanatory yield of outcome evaluation (Cook & Shadish, 1994).


38 RANDALL, GRIFFITHS, COX<br />

As has been recommended, semistructured interviews were used alongside<br />

the questionnaire surveys to identify the mechanisms of change<br />

(Griffiths, 1999; Kompier et al., 2000b; Yin, 1995). Analysis of the data<br />

from the 12 interviews carried out with participants in Study 1 indicated that<br />

the intervention increased participation in decision making, improved<br />

control over the <strong>management</strong> of the station environment, and enhanced<br />

control over the allocation of work in the station they managed. In the 14<br />

interviews carried out with paediatric nurses a wide variety of mechanisms<br />

were identified. These included: an increase in uninterrupted time to deal<br />

with administrative work and with tasks requiring concentrated thought;<br />

being better able to meet project deadlines; having more control over the<br />

<strong>management</strong> of one’s own time; and a reduction in the amount of work<br />

completed at home. Almost all of the changes mentioned have been shown<br />

elsewhere to have an impact on work-related well-being (Bond & Bunce,<br />

2001; Heaney & Goetzel, 1997; Jackson, 1983; Kompier et al., 2000a;<br />

Kompier & Cooper, 1999; Landsbergis & Vivona-Vaughan, 1995; Mikkelsen<br />

et al., 2000; Parker & Wall, 1998; Parkes & Sparkes, 1998; Schaubroeck,<br />

Ganster, Sime, & Ditman, 1993). The variety of mechanisms mentioned in<br />

participants’ discourses also indicates that the same intervention may<br />

operate through multiple or different mechanisms, and is a finding that<br />

merits further investigation (Meijmen, Mulder, & Cremer, 1992; Randall,<br />

2002).<br />

In conclusion, adapted designs built around process evaluation appeared<br />

to offer a means of strengthening the evaluation of <strong>stress</strong> <strong>management</strong><br />

intervention in complex and unpredictable environments. Combining<br />

process and outcome evaluation in this way offers a rigorous quantitative<br />

evaluation of outcomes in situations hostile to study designs built around<br />

controlled exposure patterns. In time, this has the potential to enable a<br />

larger and more informative evaluation research literature to emerge.<br />

REFERENCES<br />

Beehr, T. A., & O’Hara, K. (1987). Methodological designs for the evaluation of occupational<br />

<strong>stress</strong> <strong>interventions</strong>. In S. V. Kasl & C. L. Cooper (Eds.), Stress and health: Issues in research<br />

methodology (pp. 79 – 112). Chichester, UK: Wiley.<br />

Bond, F., & Bunce, D. (2001). Job control mediates change in a work reorganization<br />

intervention for <strong>stress</strong> reduction. Journal of Occupational Health Psychology, 6, 290 – 302.<br />

Brief, A. P., Burke, M. J., George, J. M., Robinson, B. S., & Webster, J. (1988). Should negative<br />

affectivity remain an unmeasured variable in the study of job <strong>stress</strong>? Journal of Applied<br />

Psychology, 73, 193 – 198.<br />

Briner, R., & Reynolds, S. (1999). The costs, benefits and limitations of <strong>organizational</strong> level<br />

<strong>stress</strong> <strong>interventions</strong>. Journal of Organizational Behavior, 20, 647 – 664.<br />

Campbell, D. T. (1957). Factors relevant to the validity of experiments in social settings.<br />

Psychological Bulletin, 54, 297 – 312.


EVALUATION OF INTERVENTIONS 39<br />

Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for<br />

research. Chicago: Rand-McNally.<br />

Colarelli, S. M. (1998). Psychological <strong>interventions</strong> in organizations: An evolutionary<br />

perspective. American Psychologist, 53, 1044 – 1056.<br />

Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis issues for<br />

field settings. Chicago: Rand McNally.<br />

Cook, T. R., & Shadish, W. R. (1994). Social experiments: Some developments over the last<br />

fifteen years. Annual Review of Psychology, 45, 545 – 580.<br />

Cooper, C. L., Liukkonen, P., & Cartwright, S. (1996). Stress prevention in the workplace:<br />

Assessing the costs and benefits to organisations. Dublin, Ireland: European Foundation for<br />

the Improvement of Living and Working Conditions.<br />

Cox, T. (1993). Stress research and <strong>stress</strong> <strong>management</strong>: Putting theory to work. Sudbury, UK:<br />

HSE Books.<br />

Cox, T., & Gotts, G. (1987). The General Well-Being Questionnaire manual. Nottingham, UK:<br />

Department of Psychology, University of Nottingham.<br />

Cox, T., & Griffiths, A. J. (1995). The nature and measurement of work <strong>stress</strong>: Theory and<br />

practice. In J. Wilson & N. Corlett (Eds.), The evaluation of human work: A practical<br />

ergonomics methodology. London: Taylor & Francis.<br />

Cox, T., Griffiths, A. J., Barlow, C. A., Randall, R. J., Thomson, L. E., & Rial-Gonzalez, E.<br />

(2000a). Organisational <strong>interventions</strong> for work <strong>stress</strong>. Sudbury, UK: HSE Books.<br />

Cox, T., & Griffiths, A. J., & Randall, R. (2002a). The assessment of psychosocial hazards at<br />

work. In M. J. Schabracq, J. A. M. Winnubst, & C. L. Cooper (Eds.), Handbook of work and<br />

health psychology (pp. 191 – 207). Chichester, UK: Wiley & Sons.<br />

Cox, T., Griffiths, A. J., & Rial-Gonzalez, E. (2000b). Research on work-related <strong>stress</strong>.<br />

Luxembourg: Office for Official Publications of the European Communities.<br />

Cox, T., Randall, R., & Griffiths, A. (2002b). Interventions to control <strong>stress</strong> at work in hospital<br />

staff. Sudbury, UK: HSE Books.<br />

Cox, T., Thirlaway, M., Gotts, G., & Cox, S. (1983). The nature and assessment of general wellbeing.<br />

Journal of Psychosomatic Research, 27, 353 – 359.<br />

Dobson, L. D., & Cook, T. J. (1980). Avoiding Type III error in program evaluation: Results<br />

from a field experiment. Evaluation and Program Planning, 3, 269 – 376.<br />

European Commission. (1989). Council framework directive on the introduction of measures to<br />

encourage improvements in the safety and health of workers at work 89/391/EEC. Official<br />

Journal of the European Communities, 32( L183), 1 – 8.<br />

Fitzgerald, J., & Rasheed, J. M. (1998). Salvaging an evaluation from the swampy lowland.<br />

Evaluation and Program Planning, 21, 199 – 209.<br />

Griffiths, A. (2003). Actions at the workplace to prevent work <strong>stress</strong>. Science in Parliament, 60,<br />

12 – 13.<br />

Griffiths, A. J. (1999). Organizational <strong>interventions</strong>: Facing the limits of the natural science<br />

paradigm. Scandinavian Journal of Work, Environment and Health, 25, 589 – 596.<br />

Griffiths, A. J., Cox, T., & Barlow, C. A. (1996). Employers’ responsibilities for the assessment<br />

and control of work-related <strong>stress</strong>: A European perspective. Health and Hygiene, 17, 62 – 70.<br />

Harachi, T. W., Abbot, R. D., Catalano, R. F., Haggerty, K. P., & Fleming, C. B. (1999).<br />

Opening the black box: Using process evaluation measures to assess implementation and<br />

theory building. American Journal of Community Psychology, 27, 711 – 731.<br />

Hartley, J. (2002). Organizational change and development. In P. Warr (Ed.), Psychology at<br />

work (pp. 399 – 425). London: Penguin.<br />

Health and Safety Commission. (1999). Management of health and safety regulations. London:<br />

Her Majesty’s Stationery Office.


40 RANDALL, GRIFFITHS, COX<br />

Heaney, C. A., & Goetzel, R. Z. (1997). A review of health-related outcomes of multicomponent<br />

worksite health promotion programs. American Journal of Health Promotion,<br />

11, 290 – 308.<br />

Heaney, C. A., Israel, B. A., Schurman, S. J., Baker, E. A., House, J. S., & Hugentobler, M.<br />

(1993). Industrial relations, worksite <strong>stress</strong> reduction, and employee well-being: A<br />

participatory action research investigation. Journal of Organizational Behavior, 14, 495 – 510.<br />

Ivancevich, J. M., Matteson, M. T., Freedman, S. M., & Phillips, J. S. (1990). Worksite <strong>stress</strong><br />

<strong>management</strong> <strong>interventions</strong>. American Psychologist, 45, 252 – 261.<br />

Jackson, S. (1983). Participation in decision-making as a strategy for reducing job-related<br />

strain. Journal of Applied Psychology, 68, 3 – 19.<br />

Kompier, M., de Gier, E., Smulders, P., & Draaisma, D. (1994). Regulations, policies and<br />

practices concerning work <strong>stress</strong> in five European countries. Work and Stress, 8, 296 – 318.<br />

Kompier, M. A. J., Aust, B., van den Berg, A., & Siegrist, J. (2000a). Stress prevention in bus<br />

drivers: Evaluation of 13 natural experiments. Journal of Occupational Health Psychology, 5,<br />

11 – 31.<br />

Kompier, M. A. J., & Cooper, C. L. (Eds.). (1999). Preventing <strong>stress</strong>, improving productivity:<br />

European case studies in the workplace. London: Routledge.<br />

Kompier, M. A. J., Cooper, C. L., & Geurts, S. A. E. (2000b). A multiple case study approach<br />

to work <strong>stress</strong> prevention in Europe. European Journal of Work and Organizational<br />

Psychology, 9, 371 – 400.<br />

Kompier, M. A. J., & Kristensen, T. (2000). Organisational work <strong>stress</strong> <strong>interventions</strong> in a<br />

theoretical, methodological and practical context. In J. Dunham (Ed.), Stress in the<br />

workplace: Past, present and future. London: Whurr Publishers.<br />

Landsbergis, P. A., & Vivona-Vaughan, E. (1995). Evaluation of an occupational <strong>stress</strong><br />

intervention in a public agency. Journal of Organizational Behavior, 16, 29 – 48.<br />

Lipsey, M. W. (1996). Key issues in intervention research: A programme evaluation perspective.<br />

American Journal of Industrial Medicine, 29, 298 – 302.<br />

Lipsey, M. W., & Corday, D. S. (2000). Evaluation methods for social intervention. Annual<br />

Review of Psychology, 51, 345 – 375.<br />

Meijman, T., Mulder, G., & Cremer, R. (1992). Workload of driving examiners: A psychosocial<br />

field study. In H. Kragt (Ed.), Enhancing industrial performance: Experiences of integrating<br />

the human factor. London: Taylor & Francis.<br />

Mikkelsen, A., Saksvik, P. O., & Landsbergis, P. (2000). The impact of a participatory<br />

<strong>organizational</strong> intervention on job <strong>stress</strong> in community health care institutions. Work and<br />

Stress, 14, 156 – 170.<br />

Murphy, L. R. (1996). Stress <strong>management</strong> in work settings: A critical review of health effects.<br />

American Journal of Health Promotion, 11, 112 – 135.<br />

Nytro, K., Saksvik, P. O., Mikkelsen, A., Bohle, P., & Quinlan, M. (2000). An appraisal of key<br />

factors in the implementation of occupational <strong>stress</strong> <strong>interventions</strong>. Work and Stress, 14,<br />

213 – 225.<br />

Parker, S., & Wall, T. (1998). Job and work design: Organizing work to promote well-being and<br />

effectiveness. Thousand Oaks, CA: Sage.<br />

Parkes, K. R., & Sparkes, T. J. (1998). Organizational <strong>interventions</strong> to reduce work <strong>stress</strong>: Are<br />

they effective? A review of the literature. Sudbury, UK: HSE Books.<br />

Randall, R. J. (2002). Organisational <strong>interventions</strong> to manage work-related <strong>stress</strong>: Using<br />

organisational reality to permit and enhance evaluation. Unpublished PhD thesis, University<br />

of Nottingham, UK.<br />

Reynolds, S. (1997). Psychological well-being at work: Is prevention better than cure? Journal of<br />

Psychosomatic Research, 43, 93 – 102.


EVALUATION OF INTERVENTIONS 41<br />

Saksvik, P. O., Nytro, K., Dahl-Jorgensen, C., & Mikkelsen, A. (2002). A process evaluation of<br />

individual and <strong>organizational</strong> occupational <strong>stress</strong> and health <strong>interventions</strong>. Work and Stress,<br />

16, 37 – 57.<br />

Schaubroeck, J., Ganster, D. C., Sime, W. E., & Ditman, D. (1993). A field experiment testing<br />

supervisory role clarification. Personnel Psychology, 46, 1 – 25.<br />

Semmer, N. (2003). Job <strong>stress</strong> <strong>interventions</strong> and the organization of work. In J. Quick & L.<br />

Tetrick (Eds.), A handbook of occupational health psychology (pp. 325 – 353). Washington,<br />

DC: American Psychological Association.<br />

Spector, P. E. (1994). Using self-report questionnaires in OB research: A comment on the use of<br />

a controversial method. Journal of Organizational Behavior, 15, 385 – 392.<br />

Tabachnick, B., & Fidell, L. (2001). Using multivariate statistics (3rd ed.). New York:<br />

HarperCollins.<br />

Terra, N. (1995). The prevention of job <strong>stress</strong> by redesigning jobs and implementing selfregulating<br />

teams. In L. R. Murphy, J. J. Hurrell, S. L. Sauter, & G. P. Keita (Eds.), Job<br />

<strong>stress</strong> <strong>interventions</strong> (pp. 265 – 281). Washington, DC: American Psychological Association.<br />

Van der Hek, H., & Plomp, H. N. (1997). Occupational <strong>stress</strong> <strong>management</strong> programmes: A<br />

practical overview of published effect studies. Occupational Medicine, 47, 133 – 141.<br />

Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience aversive<br />

emotional states. Psychological Bulletin, 96, 465 – 490.<br />

Yin, R. K. (1994). Case study research: Design and methods (2nd ed.). Thousand Oaks, CA:<br />

Sage.<br />

Yin, R. K. (1995). New methods for evaluating programs in NSF’s Division of Research,<br />

Evaluation and Dissemination. In J. A. Frechtling (Ed.), Footprints: Strategies for nontraditional<br />

program evaluation (pp. 25 – 36). Arlington, VA: National Science Foundation.<br />

Yin, R. K., & Kaftarian, S. J. (1997). Introduction: Challenges of community-based program<br />

outcome evaluations. Evaluation and Program Planning, 20, 293 – 297.<br />

Zapf, D., Dormann, C., & Frese, M. (1996). Longitudinal studies in <strong>organizational</strong> <strong>stress</strong><br />

research: A review of the literature with reference to methodological issues. Journal of<br />

Occupational Health Psychology, 1, 145 – 169.<br />

Manuscript received December 2003<br />

Revised manuscript received July 2004

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!