JCQ Version 2.0
(DRAFT, not for circulation, until Board approval December 1, 2004)
III. Minutes of the JCQ Workshop, March 19, 2003, Toronto
"A Preparatory Discussion for Revision of the JCQ,"
Minutes, as recorded by: Karasek, Kerr, Landsbergis, and Houtman/Tsutsumi
1. Workshop Process: Goals, Sections, and Progress
2. Revised Discussion Topic List and Priority Votes, March 19th, 2003
B. Full Texts of Workshop Minutes
1. Minutes recorded by Robert Karasek (Plenary, All Topics)
2. Minutes recorded by Micky Kerr (Plenary, Five High Priority Topics)
3. Minutes recorded by Paul Landsbergis (Decision Latitude Plenary, and Subsection)
4. Minutes recorded by Irene Houtman , w/ Akizumi Tsutsumi. (Psyc. Dem. Plenary, and Subsection)
Introduction
The Toronto meeting marked the second significant Job Content Questionnaire International Workshop in a two year period devoted to making progress toward a revised instrument for psychosocial work hazard assessment. It was the third in three years including the presentations of the scientific summaries in Brisbane ICBM conference in November 2000. Thus, there is substantial momentum in this international dialogue process, and a substantial group of both sophisticated and energetic new researchers has accumulated. Several participants remarked on the extraordinary range of relevant topics that were pulled into an integrated plan during the Toronto workshop. Expectations that a cutting edge, internationally ratified, " JCQ 2.0" will eventually emerge are substantial.
Discussion participation was very disciplined in the long main meeting, with participants consistently limiting the length of their own comments and facilitating other's participation. As a result, it became possible to review,- in plenary session - a very large number of issues at three different logical levels: (a) the level of initial scientific summary review of scale validities, (b) the level of priority ranking between topics, and (c) the level of specific questions within the high priority scale areas. The discussion kept these topics - on the one hand sufficiently distinct - and on the other hand formatted to allow integration - and consequently derived a priority plan for broad discussions about revising the JCQ in the future, complete with specific information about alternative directions. Given the multiple social and medical scientific contributions and the multiple wellbeing foci; given the broad range of methodological issues; given the broad range of political implications and social goals of participants; and further given the multiple connections between each of these topics, the meeting was a unique effort in creative, collective, and self-disciplined contribution (one intense, complex six hour meeting with 32 international participants and a few breaks, followed by scale subgroup meetings) - achieving at the end its very ambitious goal.
These minutes record the commentary in those important meetings, to provide a "map" recalling how topics were integrated and prioritized.
Thanks are due to many here - all of the Workshop participants, and especially Paul Landsbergis, Irene Houtman, Mickey Kerr, and Akizumi Tsustumi. For helping make possible the JCQ Web site, soon to host this discussion, special thanks are due to Tania Araujo, Bong Kyoo Choi, Hanne Sønderborg, and Christine Lewis.
1. The Workshop Process: March 19, 2003
The Workshop Goal was to develop discussion themes for a future "decision process" on alternative design ideas for a new JCQ 2.0. This program was intended to shift the focus from preceding reviews of scale reliability and validity data, toward "instrument design." The challenge would be to evolve a unifying "map" for organized discussion of the diverse JCQ revision issues, so that the next step - JCQ User decisions about trade-offs and alternative designs - could be made in the following workshop.
The Workshop was based on a four-part agenda, and took place at the Institute for Work and Health in Toronto (preceding the APA/NIOSH "Stress at Work" Conference) at 9:00 am. The program structure included: (a) three brief Research Summary Discussions of scientific evidence reviewed in Dusseldorf, 2002, one for each scale area (15 minutes each ); then (b) a brief presentation of 16 topics relating to JCQ revision for priority ranking - based on a previously formatted list sent from the JCQ Center (March 7, 2003). This was to allow selection, by vote, of a limited number of high priority topics for future in further discussion in the workshop. (c) The high priority topics were discussed , and (d) a set of scale subgroup meetings folowed, mainly about papers.
The freely structured Priority Discussion opened up a number of new vistas. More discussion was clearly needed in the area of "Goals of the JCQ:" relating to decision of types of research and users involved (company practitioners, social policy experts, scale psychologists, etc), and to the power base of the JCQ, and to the types of data to be collected in the future, and the possibilities of separate modules for detailed research, modules for separate industry groups.
By consensus the workshop was extended to 2:30, creating a single 6-hour meeting (with a brief buffet lunch) - an extension essential to the achievement of the Workshop expansive goals. Although the list of 15 topics prioritized the main topics, the plenary devoted some time at the end to cover all listed topics. Unfortunately, there was little discussion in the area of "trade-offs," cutting existing areas to allow new areas. In all topic areas subgroups were appointed for the next step in development. IWH Facilitator Rhoda Reardon was helpfully able to jointly chair the meeting overall, with Board expert input periodically.
2. JCQ Workshop Revised Discussion Topic List and Priority Votes, March 19th, 2003
Votes Rank Topic#
15 3 1. Macro-level decision latitude; at individual or collective level
19 2 2. Psychological demands – generic vs. specific; coverage: eg. cognitve, emotion.
17 4 3.+ 9. Social capital - integral vs. derived variables
- Social support at collective level
12 5 4. Fairness and respect; short ERI – intrinsic component?
5 5. Separate skill discretion & decision authority
1 6. Drop repetitive work
0 7. Drop conflicting demands
0 8. Differentiate supervisor & coworker support
6 10. Response set - quantitative vs. agreement formats
0 11. Avoid self-reflexive – intrinsic ERI + some JCQ items
24 1 12.+18. Delete/expand on psychometric properties only, or use policy goals also
- Goals of the JCQ – core set only or modules, intervention aides, etc.
10 13. How to validate in future – objective measures?
7 14. Gather data only from individuals or also at organizational level
7 15.+ 19. Dealing with job insecurity
- Key phenomena, like downsizing issues, adequately captured
3 16. New outcomes – beyond health eg active learning
2 17. Occupational or workplace specific items - any scale
B. Full Texts of Workshop Minutes
KARASEK's Minutes from the JCQ Workshop, Toronto, March 19, 2003
•Minutes of Plenary Discussion, All topics
Prepared by Robert Karasek (3-19-03, integrated 10-04)
I. Introductory Remarks and Goal of the Meeting from JCQ Center ("Preparation for Design" -see above)
a. Karasek noted the social policy goals for assessing the "soft costs" of work stressors on health and wellbeing in the global economy represented a level of sophistication not yet been attained, but an important and reasonable goal for JCQ 2.0.
b. Discussion of the potential need for a basic "core" and "module" concept: a short basic instrument with more detailed modules for specialized research purposes added (Copenhagen instrument-like)
c. Predictive validity is actually the main criteria, given theoretical coherence.
d. Questions arise about how to tailor the questionnaire to local work sites.
e. A list of topics for workshop discussion was quickly solicited (Theorell noted the absolute need for brevity here).
A. Summary: Decision latitude: The work by a group lead by Paul Landsbergis had become very sophisticated on a three year period, with many sophisticated contributors (see summary document from Landsbergis et al). However, some concerns were expressed that the review was conservative (i.e., it was not true that decision latitude has no strong predictive associations). A question was raised whether the committee's review was "just science, not policy: -too internal." (Jeff Johnson). However, the future importance of macro-decision latitude and neo-liberalism-job insecurity issues was obvious (from Landsbergis, et al's very interesting review March 2002, Dusseldorf) - and would come up below.
B.Summary: Psychological Demands. Tsutsumi and Houtman: here noted clear problems, but not so clear a path forward. The scales were criticized as performing less reliably (lower alphas - indeed borderline), inconsistent inter-scale correlations, and some divergent questions ("conflicting demands"). Problems with "work fast" and work hard" were noted: are they only mental demands? Physical demands also appear to be captured. Thus the label of the current dimensions needs discussion. Is it sympathetic arousal? It is certainly closest to "quantitative demands." Cognitive demands are not being included and could possibly be difficult (predictive validity problems). If fact, the JCQ Center recommended dropping four "new" cognitively focused questions recommended in 1895 for this reason, reverting to a five question scale. Inclusion of emotional demands had long been called for. Predictive validity was often discussed as inferior to decision latitude, but not necessarily so (Kupper's reanalysis of Whitehall, MSD and burnout literatures). Is the scale "tight" enough to be used predictively?
C. Summary: Social Support. Here a brief summary was presented by Per-Olof Ostergren (no summarized version was available since there had been no open Dusseldorf discussion). Problems with internal scale reliability had rarely been noticed, and face validity and theoretical relevance and function (Jeff Johnson's work) was good - but objective and concurrent validity were hard to discuss. Predictive validity was much less clear than for the other two broad dimension areas. Since the detailed discussion had not proceeded as far as with the other scales, a broad questions of what to do with related questions emerged.
II. Priority Topic Discussions
After the Summary Reports the original list of discussion topics was voted on for discussion priority. The goal was to select the top four topics to devote most time for in a plenary discussion. Five topics were ultimately selected, and several of the remaining topics were grouped and discussed very briefly (see: Discussion Topic Priority Vote List). The issue was raised: why should we go through this extended process: shouldn't the JCQ Center just produce a recommended instrument, and then solict discussion?
A. Detailed Discussions: JCQ Instrument Goals
1. The instrument should be useful for: a. Health studies; b. Intervention studies; c. It should try to capture new labor market change effects. A parsimonious core was needed. Should there be a "module" on labor market flexibility?
2. Karasek raised the issue of the "power of the JCQ" being lost without a short - but very consistently used instrument - thus the need for a compact "core" to maintain the generalizable validity of its scales.
3. A subgroup to work further was defined: RK plus the JCQ Board, and T. Theorell, J. Johnson and A. Ostry.
B. Detailed Discussions: Psychological Demands
1. The instrument was too generic. How specific should it be? Who does this research (for example T, Kristiansen) How homogeneous was the impact on health outcomes. Decision we must keep a core, but it was claimed that the current scale works best in large population studies, but for many practical work-reorganization focused studies more specific lists of stressors were needed.
2. Multiple parsimonious demand exposure modules - for specific situations were needed. Johnson suggested than rather than a very large number of these (one per company type) a short list of " say 5" basic modules by industry and occupation should be defined: capturing much of the variance within "five types of work", but not an occupation-specific list.
3. For intervention goals, the issue of "vertical" (by scale) vs. "horizontal" (by industry) differentiation of instrument components was discussed. How to draw in an employer - requires company specificity, measuring "concrete" risks in each area, and effects of complex organizational dynamics.
4. A sub-group to work further was defined: N. Kawakami [not presnt], A. Tsutsumi, R Ruggules, [T. Kristiansen - not present], M. Kerr.
C. Detailed Discussions: Decision Latitiude.
1. The future importance of a macro-decision latitude expansion, in the context of neo-liberalism and job insecurity issues was universally agreed upon. Macro changes have outflanked the micro job control assessed by the current JCQ. However, the question of what is macro-level decision latitude was not so easy to resolve, with clear overlaps coming with job insecurity, and even some forms of social trust and support. Should it be measured at the individual level or the organization level. Should register data be included? Could it be a measure of organizational power (?). Changes in the labor market structure were crucial to assess forms of the economist's market "flexibility" so different from Industrial Democracy based worker control and skill development conceptions.
2. Theorell suggested that organization justice conceptions tested in Finnish studies were important, and also the question of how conflicts were resolved at a company.
3. A subgroup to work further was: defined as [ ---- ].
D. Detailed Discussions: Social Support and related Topics
1 A very broad discussion occurred, with topics raised as important that were far more extensive than the original JCQ Social Support concepts. Social capital was discussed (social trust, social coherence, and the meaning of work) The conflicting ideological bases were noted by Jeff Johnson: conflict theories not agreeing with social capital conceptualizations since they often overlooked alienation and exploitation problems - which the Demand/control model inherently addressed. The importance of worker collectivity concepts was noted. Karasek mentioned the conception of creating social capital in conducive production - an issues related to active work in the demand/control model. A danger was noted is scanning too broad a range of diffuse topics in the JCQ.
2. In a later sub-group meting, an even broader range of linkages between the broad concept of "social support" and other topics was noted: the decline of macro-level trust in the context of neo-liberalism, undermining workplace social support and social capital development by workers in their jobs. The development of dramatic ideological contractions between workforce and management goals in some situations was noted.
3. A subgroup to work further was defined: P-O. Ostergren, J. Johnson, R. Ruggules, and L, McDonald, and R. Karasek
E. Detailed Discussions: Link to Effort/Reward? - Job Insecurity?
1. The question of whether the JCQ should (a) incorporate some Effort/Reward questions, or (b) develop a collaborative questionnaire section with E/R was raised. Karasek explained that these issues had been broached with Johannes Siegrist, a as a way of "uniting" psychosocial forces.
2.Fairness and respect were though to be important extensions for the JCQ. The questions of the meaning of "status control" and it relation to the JCQ dimensions was raised. The major question was whether the JCQ would loose its identity or theoretical coherence. No clear decision was reached in a rather short discussion.
3. Job Insecurity issues came up again as a part of this discussion, which also above had overlapss with the macro-level decision latitude discussion. Karasek noted the idea of adding psychological demands and job insecurity for some "simple-format" analyses to extend the Demand/Control model use.
4. A subgroup to work further was defined: T. Theorell, R. Karasek, A. Ostry, M. Kerr, I. Houtman, N. von Vechel
F. Other Brief Discussions
1. Validation processes/Objective assessment. A subgroup to work further was defined: R. Ruggules, L. McDonald, A de Lange, [T Ktistensen - not present].
2. The Validation/Objectivity topic was also linked to Individual vs. Organization level data and Register data. A subgroup to work further was defined: L. McDonald, I Houtman, A. Ostry, A Mikkensen, M. Dollard..
3. Job Insecurity was linked to the topic Downsizing. A subgroup to work further was J. Johnson and P. Landsbergis.
4. Active Work and active work outcomes. A subgroup to work further was defined: A. de Lange, M. Dollard. A Mikkensen,
G. Format to be Followed in Four Major Decision Areas by JCQ User Groups
(Goals, Decision latitude, Psychosocial Demands, Social Support and Related)
1. Conceptual summary
2. Evidence summary
3. Example of possible questions
4. Trade-off implications
5. Decision Tree
_____End of Karasek Minutes_____
KERR's Minutes from the JCQ Workshop, Toronto, March 19, 2003
•Minutes Plenary, Five High Priority Topics (Task 5)
Prepared by Mickey Kerr, IWH, Toronto, (4-25-2003)
NOTE: This section of the meeting discussed the 16 points from the file “Today’s Tasks” also found in longer version on the main JCQ meeting agenda file. There were two parts to the session: 1) some background information and discussion about the main points of interest; 2) prioritizing down to 5-6 key areas for future work using the list of 19 “Revised” Discussion Topics handed out at the meeting (see file named IWH-JCQrevdisctopics.doc)
Part 1: Background informationPrior to discussing the research priorities for the future the group heard a report back from the last JCQ meeting (Dusseldorf) for the 3 main groups of the latitude scale (PL), the demands scale (IH) and the support scale (PO). In addition to these more focused discussions, there was also some broader discussion about methodological issues and goals issues for the JCQ.
Minutes from discussion
Latitude
– PL reviewed main discussion points from his pre-workshop document
– Need to address control at work group, unit and organizational levels
Demands
- main point for demands scale may relate to issue of homogeneity versus heterogeneity of samples, especially where physical demands may overlap in some settings
- also an issue with use of quantitative version of response scale
Support
- did not receive much attention at last meeting in Dusseldorf
- PO suggested that social support at the organizational level needs consideration
- Also had long discussion about social capital, including bonding in small workplaces, bridging, conducive value, and linking vertically in organizations
- social capital in relation to control rather than support also discussed
- concern raised though about use of term “social capital” given its very broad meaning
- led to discussion about use of “integral variables” (e.g. programs) versus derived variables (aggregated) like percent newspaper readership
- should this type of data be included in the JCQ or shoulod a list be made of what to collect outside of the JCQ?
Note: lack of discussion about the JCQ job insecurity scale raised at this point
Methodological Issues
- use of agreement versus frequency response set discussed
- should deleting items be done on basis of psychometric properties only or should policy goals of JCQ also be brought into decisions?
- Issue of coverage raised including notion of generic versus (occupation) specific items and suggestion that more items may be needed or useful for interventions (despite psychometric properties)
- JCQ will not measure everything so possibly can recommend other things to collect or have modules, such as core (required) and optional?
Other Issues
- should we focus on refining the current instrument to improve specific components of existing JCQ?
- perhaps can create modules that can be optional, such as social capital
- need to link surveys to economic indicators but this is very complex and needs to have input from economists (not represented at meeting)
- concerns about how well JCQ captures the reality of today’s workplace environments
- issue of whether or not to include the ERI items within the JCQ as an ERI subscale
- if ERI is included would the intrinsic components also be included or only the items used to determine ERI ratio itself?
- this raised issue as to whether or not other research models might also be “adopted” by JCQ
- issue raised about concerns with inclusion of self-reflexive questions or negative mood questions
- measuring outcomes for “active behaviour” was also raised as something currently lacking in existing JCQ format
Initials:AO = Aleck OstryPart 2: Prioritizing Future Work
Based on the above discussions and the documents circulated prior to the meeting, research priorities for the future were identified by a “show of hands” process, using the list of 19 “Revised” Discussion Topics handed out at the meeting (see file named IWH-JCQrevdisctopics.doc). Six priority areas, ranked here by order of priority (highest first) were identified: 1) goals of the JCQ; 2) refining the psychological demands scale; 3) use of a macro-level latitude scale; 4) revisions to the social support scale including the issue of social capital; 5) inclusion of the ERI and respect items; 6) objective validation of JCQ constructs is needed. Working groups with a leader were identified for each priority area.
2.1. Goals of the JCQ
Working group – BK + JJ, AO, and JCQ Board
- background discussion about how JCQ was started and its context then and now including issues of Fordism ,Taylorism possibly less relevant now than when JCQ was developed so “old” JCQ can remain but a new module may be needed to capture current “flexible” work arrangements
- a need to broaden conceptual range of how we understand work and thus broaden JCQ to reflect this
- perhaps this might be best done outside of existing JCQ though, as discussion ensued about “scientist” versus “practitioner” goals for JCQ
- also a policy level or goals consideration and it was noted that some countries were already working on this but more work needed
- for example, can JCQ use already invested research capital to convince large scale international surveys to use a common JCQ format
2.2. Refining the psychological demands (PD) scale
Working group – IL + MK, RR, and current PD group members from JCQ Board
- perhaps more basic work needs to be done on this scale than on latitude
- what (other) databases are available to test some hypotheses generated in group? For example, could determine extent of different kinds of demands (physical, emotional, psychological, time, etc.) in different populations
- lack empirical evidence to determine when something is really a demand
- could hold some core set of items for all versions but then expand for specific situations (e.g. emotional demands)
- some work being done on this by Tage Kristensen’s group in Copenhagen and it was suggested that he be asked about this
- JCQ Center no longer recommends the 9-items demands scale, rather the 5-item one now being used
- some concern current JCQ (and other) demands scales may not be sensitive to change
- some concern also about how items may be filtered in certain settings, such that psychological demands items (e.g. work hard, work fast) may be interpreted as physical demands items if work physically demands
2.3. Macro-level latitude scale
Working group – PL + AO, POS, and current latitude group members from JCQ Board
- some suggestions for items can be found in Table 12 of the paper provided prior to workshop by Landsbergis et al.
- need some clarity on what the “macro” level really is
- is it organization level or group level or both?
- organizational justice questions from Finland might provide some guidance on this
- asking workers and aggregating up may not give adequate data
- some of the items from Table 12 mentioned above have already been asked in some Swedish studies
2.4. Social support scale including the issue of social capital
Working group – PO + JJ and current support group members from JCQ Board
This issue combined points 3, 8 and 9 from the revised topic list (of 19)
- should social capital be included or not?
- some concern over use of term
- is it better to use specific terms for aspects that might be important such as social coherence, trust, etc.?
- critique of social capital by C. Mutaner was raised as being useful here
- gender, race and class sectors exist in society and this impacts how you are mobilized collectively but this is very different than social capital
- JCQ is more in line with alienation, exploitation but traditionally ignores class, race and some other relevant social capital issues so it can be very dangerous to bring it in to JCQ as the conceptual model may not be well suited for it
- some discussion that support questions were originally developed and conceived of as a buffer but are not an “exchange value”
- need to think about how these topics can be advanced since email does not seem to be a very productive forum for this, as we are still at a conceptualization phase
2.5. Inclusion of the ERI and respect items
Working group – TT, AO, IH, MK, Natasja van Vegchel and members from JCQ Board
- some (e.g. AO) indicated they had data from both models and that they could use guidance from group for addressing specific questions about how to best analyze it
- should specific components of ERI be pulled out rather than using full ERI model
- what does addition of ERI mean conceptually for the JCQ?
- how would ERI items be included into the model/analysis?
- job insecurity already in JCQ needs to be better refined to help address this issue, and this was done to some extent by combining points 15 and 19 from the revised topic list and identifying a possible working group to look at this issue, including JJ, MD and ??
- is the ERI a measure of the worker or the work environment, as there is some debate about the full model, but especially the intrinsic components
2.6. Objective validation of JCQ constructs / gathering data at different levels
Working group – RR + LM, AO, IH, Maureen Dollard, Auslag Mikkelsen, Annette de Lange and members from JCQ Board
This issue was a combination of points 13 and 14 from the revised list of 19 points
- not much time was available for discussion of this point
- LM mentioned that NIOSH has a project going already that is directly relevant to this topic
Possible Sites for Next JCQ Meeting:
ICOH WOPS– Aug 2005
ICOH-CVD – March 2005
ICBM – Mainz, Germany Aug 2004
_____End of Kerr Minutes ____
LANSDSBERGIS' Minutes from the JCQ Workshop, Toronto, March 19, 2003
•Minutes of Decision Latitude Scale" Plenary Discussion, and Subgroup
Prepared by Paul Landsbergis (5-27-03), with input from Irene Houtman, Jan de Jonge, Tsutsumi Akizumi, Reiner Ruggules, Aleck Ostry and Per Oystein Saksvik.
If anyone did not receive a copy of these documents, contact paul.landsbergis@mssm.edu
II. General discussion of JCQ decision latitude scale
A number of issues were raised in discussion of the summary statements from the 2002 JCQ workshop and the proposed article on the JCQ decision latitude scale. The main focus of discussion was on the thoroughness of the literature review in the proposed scale article and the nature of “macro-level” decision latitude and how it should be measured. In addition, there was some discussion of trends over time in job decision latitude.
Several speakers commented that new data needed to be added to the current draft of the latitude scale article – that earlier studies (e.g., those examining cardiovascular disease, psychological strain and musculoskeletal disorders as health outcomes) needed to be reviewed and added to the article, along with any new unpublished data that is available.
In addition, it was suggested that we locate studies that compare the JCQ latitude scale with other latitude scales. (Can anyone suggest such studies?) Similarly, it was suggested that we examine studies that use “objective” measures of decision latitude, and see how those might compare to self-reported latitude in associations with health outcomes.
Several speakers recommended that we do not use the term “predictive validity” to describe associations between decision latitude and health outcomes in cross-sectional studies. Such a term should be reserved for prospective studies. The term “concurrent validity” would be appropriate for cross-sectional studies.
B. The nature of “macro-level” decision latitude and how it should be measured
Several speakers suggested examining downsizing and restructuring and their effect on decision latitude and on health outcomes. Another speaker raised the question of the association between “macro-level” decision latitude and the concept of “organizational justice”. For example, are there structures within an organization for resolving disputes between labor and management or between employees? If yes, this could be an example of both “macro-level” influence or control as well as “organizational justice”. For example, in Swedish national surveys, there are questions about one’s influence over one’s representatives at work or one’s managers. Related concepts would be those of “trust” within an organization, positive labor relations, or positive organizational “climate”.
Another question raised was whether the current decision latitude scale captures changes in worker control occurring due to changes in the global economy and the changing labor market?
Other speakers raised the question of how best to measure “macro-level” influence or control. Can it be best measured by asking individuals within an organization, or, in addition, should it be measured by organizational-level data that may be available? Should individual-level responses be aggregated by group or organization?
Other speakers raised questions about sampling strategies. During periods of downsizing or restructuring, it is important to also sample those people who are no longer employed at the organization, and not just those who remain on the job.
Finally, it was suggested that a term such as “organizational-level” latitude or control might be a clearer and more meaningful term than “macro-level” latitude or control, which is a little more ambiguous.
C. Trends over time in job decision latitude
It was reported that, in Denmark, trends over time in job decision latitude (and job demands) were based solely on changes in the distribution of various industries in Denmark, and not on changes within industries. The current draft of the scale article does not include a section on trends in decision latitude. However, that is an important issue that needs to be updated, perhaps in a separate paper.
III. Decision latitude scale “break-out” group meeting
In the afternoon, a smaller group of seven researchers interested in the decision latitude scale met informally. It was decided that we needed to finish the scale article, and to prepare a more detailed proposal for a vote next year on choices for the scale.
A. Scale article
To finish the scale article, brief but more thorough literature reviews need to be added to each section of the article. The article had originally been intended as an update of the review paper by Karasek et al., in the Journal of Occupational Health Psychology (JOHP), 1998. Our goal had been to summarize some new unpublished data, which had become available after 1998. However, it has become clear in discussions over the past two years, including the discussion in Toronto, that, for the article to allow us to make informed decisions about the scale, the article needs to take into account past research as well as new research.
Therefore, we need the assistance from all members of the scale sub-committee in completing the article. Listed below is a table listing each section of paper and a column for “previously published studies” and “new unpublished data”. Please help in this process by volunteering to prepare a brief review of one of the topics listed below. Some of this information could be abstracted from the 1998 paper by Karasek et al in JOHP. For some topics there may be very little data available in the literature aside from what already is included in the draft of our article. Paul Landsbergis volunteered to summarize studies on the association between decision latitude and cardiovascular disease (CVD). Aleck Ostry has volunteered to summarize research using questions on “macro-level” decision latitude. Please let Paul Landsbergis know which topics you can help to review. Any person providing substantive material can be added as a co-author of the article.
Assistance is needed to complete the article. Please volunteer to review one section by adding your name:
Previously published studies New unpublished data
Reliability – test-retest _____________________ Tsutsumi __________
_____________________ Houtman-deJonge ___
internal consistency _____________________ Houtman-deJonge ___
internal consistency by SES _____________________ __________________
Validity – factorial stability over time _____________________ Houtman-deJonge___
one factor or two factors? _____________________ Houtman-deJonge___
concurrent validity (x-sectional) CVD: Landsbergis ______ Landsbergis ________
Psych strain: ___________ __________________
WMSDs: ______________ __________________
predictive validity (prospective) CVD: Landsbergis ______ Landsbergis ________
Psych strain: Rugulies?___ __________________
WMSDs: ______________ __________________
Shorter versions of scale _____________________ __________________
Macro-level latitude Ostry, Saksvik _________ Ostry, Saksvik ______
It was felt that we should continue to pursue the idea of a special journal issue that would contain, at the very least, the three scale papers, a paper on the JCQ and ERI, and a paper on the interaction between job demands and job decision latitude. Aleck Ostry has also volunteered to take the lead on a paper that will focus on “macro-level” latitude. Perhaps then the final product of our work will be two papers – one just focusing on the current JCQ task-level decision latitude scale and one just focusing on “macro-level” latitude.
A related issue that arose briefly in discussion is how to measure health at various levels, such as the individual, work group and organization.
A document on “debate/discussion questions” regarding the JCQ decision latitude scale, prepared with the help of Annet de Lange and Irene Houtman, was distributed at the workshop. It contains an overview of the issues regarding the decision latitude scale, as well as recommendations for scale items (e.g., “repetitive” work, “macro-level” items), and recommendations for analysis (e.g., are skill and authority two separate factors?, occupational specific control measures). This was a valuable document for workshop participants to consider. However, the group felt that this “debate/discussion” document would need to be expanded in order to prepare for a vote on scale issues at the next meeting. It was suggested that the expanded document have the following structure: conceptual summary, evidence summary, examples of questions to be added/deleted from the scale, implications (positive and negative) of the decisions to be made, and a “decision tree” with alternative choices.
Assistance (volunteers) is needed to help expand the document to prepare for a vote at the next meeting.
It was mentioned that there is little or no data comparing the associations of health outcomes with the decision latitude scale in situations where the scale includes “repetitive” work vs. those situations where the scale does not include “repetitive” work. Such information would be useful in making decisions about whether to include such an item in one’s analysis. If any member of this committee knows of such data or has the opportunity to reanalyse their own data sets to obtain this information, that would be very helpful.
Similarly, participants were interested in a more thorough review of studies that examined the associations between health outcomes and the “skill discretion” and “decision authority” sub-scales, as well as the “decision latitude” scale. Are there stronger associations between certain health outcomes and one of the sub-scales, or not? Such information would also be useful (in addition to data on factor structure), in deciding whether to analyze the sub-scales separately or not. Therefore, if any member of this committee knows of such data or has the opportunity to reanalyse their own data sets to obtain this information, that would be also very helpful.
Members of the JCQ Decision Latitude Scale Sub-committee
Bonkyoo Choi -- bongkyoo@hotmail.com
Manueal Cifuentes -- mailto:cifuentes_manuel@hotmail.com
Sean Collins -- mailto:Sean_Collins@uml.edu
Annet de Lange -- mailto:A.deLange@psych.kun.nl
Irene Houtman -- mailto:I.houtman@arbeid.tno.nl
Jan de Jonge -- j.dejonge@fss.uu.nl
Marcel Kornitzer -- mkornit@ulb.ac.be
Paul Landsbergis -- mailto:paul.landsbergis@mssm.edu
Carlos Lima -- mailto:C.Lima@iop.kcl.ac.uk
Edwin Pelfrene -- edwin.pelfrene@rup.ac.be
Hynek Pikhart -- hynekpikhart@netscape.net
Reiner Rugulies -- rer@ami.dk (new address)
Akizumi Tsutsumi -- akizumi@md.okayama-u.ac.jp
N. Van Vegchel -- n.vanvegchel@fss.uu.nl
(New members of committee)
Els Clays -- els.clays@rug.ac.be
Jeffrey Johnson -- jvjohnson@earthlink.net
Aslaug Mikkelsen -- Aslaug.Mikkelsen@rf.no
Aleck Ostry -- ostry@unixg.ubc.ca
Per Oystein Saksvik -- peros@svt.ntnu.no
Jamie Tessler -- jtessler@igc.apc.org
Heleen Van Mierlo -- mailto:h.v.mierlo@tm.tue.nl
____End of Landsbergis Minutes_____
HOUTMAN / TSUTSUMI's Minutes from the JCQ Workshop, Toronto, March 19, 2003
•Minutes of Psychological Demand Scale: Plenary Discussion, and Subgroup
Prepared by Irene Houtman and supplemented by Akizumi Tsutsumi (4-17-03)
Some registrations from the Workshop:
Related to the summary on decision latitude:
· Organizational justice/fairness should fit in the decision latitude discussion (Tores)
· Influence of downsizing missing/interesting (Aleck Ostry)
· Finnish studies should be included (Bob)
· Include studies by Jeff Johnson
· Include CVD studies
· Combine decision authority and skill discretion or not? Discussion on pro’s and con’s
· There are data sets available; deadline for providing analyses/results
· The internal structure is very central…. But is the strength of the model/concepts; given the way work is organized right now (lean production etc) provides a conceptual challenge for a more global approach. Is this captured enough? Broader domain NIOSH – data base
· Danish data sets are also much broader. Approach validations on other than self report of job control
Related to the summary of demands:
· Different partitioning on data sets; discussion on the stability and consistency of the scale…. Items measure different things (paper by Mickey Kerr, Leslie MacDonalds)
· Data on frequency versus degree of agreement; tested on Malmo sample. There was not so much of a difference
· Bob \: even in the early data there was evidence for separate concepts. At the conceptual level there is reason to take them into one scale. He links this to the physiological evidence on arousal. The label might, however, be addressed.
· For decision latitude, leaving out the CVD literature may lead to an underestimation of the effect. Here, leaving out this literature might lead to an overestimation of the effect of demands. Irene suggests that data of the Whitehall suggest that the effect of demands appeared to be underestimated by the concept of grade… correcting for grade resulted in plain effects of demands. Also the Jace data appear to support an independent effect of demands..
· Isostrain included?
Response and where to go:
Decision latitude:
è Paul (I did not record)
Demands:
· Bob: there is a general issue of lack of stability. Irene: in TNO database this is not the case. What seems to be the case is that blue/white collar might be responsive to different Demand-items, and that we also found in specific occupational groups –e.g. nurses- that quantitative demands could not be separated, but this is not the case in a general population. Another example is the finding that e.g. emotional demands in a general population are not related to outcome measures. But when analyzing occupational groups, it was associated to outcomes only in nurses and policemen. This touches to an issue that you might have to acknowledge the different behavior of demands in homogenous and heterogeneous populations. This may or may not be (partly) due to methodological restrictions like restrictions of range.
· Necessity to discriminate between different types of demands.
· Different (objective –versus- survey) methods are needed to study these issues… as an extension on ‘how to validate in the future’
· Trust relates to participation… this is the ‘vertical dimension/direction’ towards ‘social capital’…
· also the relation between social capital and control are relevant/should be understood/studied
· others indicate that there might be a danger of adopting the concept of social capital. It is concluded that social capital should be considered and discussed on the sublevel.. not as a macro concept.
· Remark: ‘we have to look at integral variables related to the multilevel concept. You should be careful since the history of the instrument is individual respondents. When you expand in the suggested direction you might come up with more than you can chew’.
· In order to expand on decision latitude (Tores) you must look in society, and link changes to economical, financial data/developments. We don’t have the competence to do that (interpret it all). You need input from economists.
· Result: new JCQ + paper on what cannot be measured with this instrument, e.g. At company level
General remarks:
1. Job insecurity seems to be a growing phenomenon. Does the JCQ scale need to be repaired, or does it fit into the decision latitude scale, more on macro level;
2. The issue on should we be heading for a general versus specific instrument is an extremely strategic one.
3. Are we capturing the labor market?… should we also include home-work?
Discussion on priorities:
1. Goal of the JCQ –core set only or modules, intervention aides etc. (17)
Responsible person Bob (together with Tores and JCQ board?)
Change or keep the same/Main alternatives
Þ objective worker characteristics
Þ labor market fluctuations
Þ subset of types of work
Scientific goal
Social analyst
Global economy goal
2. Psychological demands – generic versus specific /heterogeneous versus homogenous –coverage e.g. cognitive emotional
Trade off is important
Responsible person: demand group, with (in principle Norito Kawakami taking the lead)… ask if Tage/his group will join
3. Macro-level decision latitude @ individual versus collective; for collecting macro level information you need other sources of information.
Responsible person: decision latitude group, with Paul Landsbergis taking the lead.
4. (Social capital)/social support. Responsible person: Per Olov; he illustrates how to proceed in order to catch up with the other scales. Jeff will join in.
5. Fairness and respect. Not covered. New scale . Some collaboration with Johannes Siegrist about integration. What does it mean for the Job Strain Model. It was asked if ERI was part of the work environment… responsible person: Bob (same group as for 1)
6. how to validate in future—objective measures? (including the issue regarding gathering data only from individuals or also at organizational level)
The point is how we should expand and/or exclude the items (scales) for the future JCQ. And consensus should be reached on the criteria by taking into account not only the psychometric properties but also the predictivity and/or what we can say for policy making by the (new) JCQ. The numbers of the items are set around 45 (or some more), considering the feasibility. Combining database (current as well as new) across the study groups is another strategy.
The participants have belonged to one or few discussion groups and will start tasks. The plan is to reach a kind of consensus until 2005 when the next JCQ workshop will be held at the 4th international conference on the work environment and cardiovascular diseases (ICOH-CVD; March 10-12, UCLA) or The Second ICOH International Conference on Psychosocial Factors at Work—Job Stress - East meets West (August 23-26, Okayama, JAPAN).
Additonal options:
BMC, Mainz 2004
Exposures? July 2004
Aim was to create a format to deal with all issues raised
· Conceptual summary
· Evidence summary (literature + analyses/data bases)
· Examples of kinds of questions
· Discussion of implications, e.g. trade offs
· A clear discussion guide
Subgroup discussion : Demand scale
Aki Tsutsumi, Reiner Rugulies, and Irene Houtman present. (Aki Tsutsumi had discussions with Yawen Cheng later).
1. responsible : Norito, with support from others (Aki, Irene, ?Tage?/Reiner)
2. Demand scale: conceptual framework, tapping on questions like:
a. One or more scales (also question of how homogenous/heterogeneous is the sample)
b. Response differences, do they matter (associations),
c. Curve linear or U-shaped relationship to outcomes –too low demands may cause adverse health effects (for all types of demands?)
3. Evidence summary, data bases:
a. General ones:
i. JACE
ii. Japan
iii. Copsoq (different answering categories)
iv. Other Dutch (with different answering categories)
v. Taiwanese (Yawen Cheng)
ii. X plus SD
iii. Correlations among concepts (total and by gender)
iv. Factor analyses
v. Cronbach alpha’s
vi. Correlations with outcome measures (see Paul’s document about decision latitude on how)
The above should give answers to the following questions
vii. Does the same pattern of associations arise despite different answering categories?
viii. Other demand items more appropriate
ix. Prevalences of demands by occupation
x. Relation to outcomes curve linear?
b. Specific occupational groups:
there is a lot on nurses: (quantitative job demands = physical demands)
a. Jan/Natasja (Dutch)
b. Harriet Riese (Dutch)
c. Japan (Aki + Norito)
d. Europe (Hans Martin Hasselhorn)
e. Marie Soderfeld
Further:
f. Dairy farmers (Maureen Dollard)
g. Police (Irene Houtman)
Maybe homogenous groups within the larger samples:
a. Copsoq (Danish, Tage Kristensen)
b. TAS (Dutch, Smulders)
Actions to be taken:
a. Decide on a report/table structure, and the analyses to be performed on the databases
b. Contact owner of data base for availability, help, (Natasja offered help)
c. Make a feasible time plan.
d. Keep in touch, coordinate etc.
_____End of Houtman/Tsustsumi Minutes____
_____END OF ALL MINUTES _________