Frequently Asked Questions FAQ

Scale and Scoring
Validity and Reliability
Definition of Job Strain
National Means and International Comparison
 Translation

 

Scale and Scoring

Q. How many questions will you recommend if I only want to do a pilot study for my MSc dissertation?

    A: It depends on your research questions and your research objectives. However, we recommend the standard 49 questions of the JCQ which include the following questions: 6 items for skill discretion, 3 items for decision latitude, 8 items for macro-level decision latitude, 5 items for psychological demands, 5 items for physical demands, 6 items for job insecurity, 5 items for supervisor support, and 6 items for coworker support respectively (for detail, read the JCQ user's manual).  

Q. Except for the recommended questions, what other scales does the JCQ have?  

    A: The JCQ has many more questions which you might be interested in: scales measuring psychological strain and job dissatisfaction (26 items), scales for physical hazard and exposure (9 items), new scales for work social identity (5 items), customer contact problem(7 items), and global competitive effects (5 items).

Q. How can I compute the scale scores for the questionnaire scales?

    A: Please see the scoring formulae for main scales of the JCQ. Please refer to the JCQ user's manual for other formulae for supplementary and new scales to be tested.

Q. I have heard that you dropped the 9-item psychological demands scale and instead now recommend the researchers to use 5-item psychological demand. Is this true? In addition, can you let me know the reasons?

    A: Yes, it is true. We recommended in March 2002 at the ICOH 3rd CVD Conference that only the 5-item version of psychological demands be used. The 9-item version was no longer recommended by the JCQ Center. The reasons were two: 1) 9-items did not loaded on a factor (psychological demands) in factor analysis with a Canadian study and an European study, and 2) inversed associations of the 9-item psychological demands with psychological strain outcomes were reported in a Danish study and the JACE study, as opposed to the first 5-item psychological demands.

Q. Please verify the Job dissatisfaction formula. There are what appear to be "dots" between some numbers and it is known if this is a "power" or is to imply that it is a multiplication as the multiplication appears as a X in some places and an * others.

    A: In this case, it is a multiplication. From the next print of the JCQ manual and on this website, this is fixed.

Q. Can I add the Q 69 (additional supervisor support item) and Q 71 (additional coworker support item) of the JCQ into the supervisor support and coworker support scale scoring process respectively? What do you think about this idea?

    A: In fact, conceptually those questions can be included into the supervisor and coworker support scales. However, unfortunately, at this time, we have no enough information to answer to your question. We recommend you to test with your own data reliability and factor loadings of those questions and ask you to report those results to us or the international JCQ user's group. Q 50 ("hostile supervisor") and Q 55 ("hostile coworkers") have been recommended, however, it has not loaded consistently on the respective social support factors.

Q. I used the Job Content Questionnaire to collect some information for my master's thesis. According to your JCQ Guide, some of the variables are calculated using the "z-scored addition of VAR1 VAR2, etc."  (Specifically, this is applied to Total Psychological Stressors, Total Physical Hazards, Total Physical Stressors, and the Composite Psychological Strain variables). Is there any way you might explain this calculation in greater detail to me? 

    A: Here is an example(for calculating the total physical hazards scale score).  

    Step 1. Calculate Z score of Hazardous Condition (z1) = (individual score - m1)/sd1
    Step 2. Calculate Z score of Toxic Exposure (z2) = (individual score - m2)/sd2
               Mean (m1), Standard deviation (sd1) of the Hazard Cond scores in your sample (or population)
               Mean (m2), Standard deviation (sd2) of the Tox Expos scores in your sample (or population)
    Step 3. Sum up the two z-scores: Total Physical Hazard = z1 + z2.


Validity and Reliability

Q. Where can I get the validity and reliability information on the JCQ scales? Which articles do you recommend for me to read?

    A: We recommend you to read the following articles: 1. Karasek R and Theorell T. (1990). Healthy Work: Stress, Productivity, and the Reconstruction of Working Life (Appendix I). New York: Basic Books. 2. Karasek R, Schwarz J, Theorell T. Final NIOSH Report (1982). 3. Karasek R, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B. (1998) The Job Content Questionnaire (JCQ): An instrument for internationally comparative assessment of psychosocial job characteristics. J Occup Health Psychol, 3:322-355. 4. Karasek R, Choi B, Ostergren P, Ferrario M for the JACE study group. (2003) Comparative international reliability and validity study of JACE-Job Content Questionnaire (JCQ): European worker's psychosocial job characteristics (submitted).

Q. Please let me know the general reliability statistics of the JCQ scales?

    A: We present below the average Cronbach's alphas of the main JCQ scales from the two studies (Karasek et al., 1998; Karasek et al.,2003) mentioned above. The 1998 study was a cross-national validity and reliability study of the JCQ scales that examined six broadly representative populations from four advanced industrial societies from the United States, Canada, the Netherlands, and Japan. The 2003 study examined the cross-national validity and reliability of the JCQ scales in five European countries: Belgium, France, Italy, the Netherland, and Sweden. For more details, refer to the two papers. Many researchers pointed out the relatively weak construct validity of psychological demands scale (see Karasek et al., 1998).

Main JCQ scales

Average Cronbach's alphas

Karasek et al. (1998)

Karasek et al.(2003)

Men

Women

Men

Women

Skill discretion

0.73

0.75

0.73

0.72

Decision authority

0.68

0.68

0.63

0.66

Psychological demands(5-items)

0.63

0.63

0.59

0.61

Supervisor support

0.84

0.84

0.85

0.86

Coworker support

0.75

0.77

0.79

0.80

Physical demands

0.86

0.79

0.86

0.84

Job insecurity

0.61(3-items)

0.58(3-tems)

0.48(4-items)

0.47(4-items)

Q. Could you explain why the reliability of the job insecurity scale is low?

    A: The Cronbach's alphas of job insecurity scale were very low in both studies mentioned above, but it should be considered that the job insecurity scale includes two types of information: (a) overall assessments of job insecurity and future career prospects and (b) specific data about layoff and work instability history. Although a statistically more homogenous scale could be achieved by dropping some questions, the robustness of the scale’s interpretability would suffer.


Q. Do you have a list of reliability and validity studies of the JCQ by country? What's the fastest way to get detailed information about the reliability and validity study of the JCQ in a specific country's context?

    A: Please refer to the following studies (not exhaustive, we keep updating the list of studies) or contact the researchers who translated the JCQ into non-English version and do the reliability and validity study for the translated JCQs.

Q. Do you have and are you willing to provide a list of the articles that have examined your job control/job demand theory of job strain. I have found and used many supportive articles, but I am now looking for a comprehensive list and thought you might have one.

     A: Please refer to the following studies (not exhaustive, we keep updating the list of studies).

     Mental Health

     Cardiovascular Disease   

     Musculoskeletal Disorders

     Critiques on the Demand-Control Model



Definition of High Job Strain

Q. How can I define the "high job strain" group according to the Karasek's Demand and Control Model?
   A: Please refer to the study: Landsbergis PA, Schnall PL, Warren K, Pickering TG, Schwarz JE. (1994). Association between ambulatory blood pressure and alternative formulations of job strain. Scand J Work Environ Health, 20, 314-28. In the paper, you will see four different methods (two "quadrant definition" and two "quotients definition" of high strain groups) that are also described on the website http://www.workhealth.org
   The first quadrant definition is "below the population mean on decision latitude and above the mean on psychlogical demands". The second quadrant definition is based on the "national-level" mean values on decision latitude and psychological demands. The first quotient definition is based on a continuous score of job strain by computing "psyhcological demands divided by decision latitude". This quotient term is nonlinear and tends to give more weight to decision latitude (denominator) than psychological demands. The second quotient definition is based on the equal weight of both decision latitude and psychological demands. For both quotient definitions it is absolutely essential that the resulting values be truncated to eliminate the extreme values that develop when dividing by very low/high values of decision latitude - which can easily distort regression findings.
   Karasek, Choi, et al (Comparative scale scores between the JCQ and JCQ-like questionaires in the JACE study (manuscript), 2004) observe that the classification error in identification of the job strain population is reduced by exclusion of that segment of the strain population that is closest to the population mean (the consistency among different instruments all measuring strain improves). A practical method for accomplishing this goal begins by dividing both the psychological demands and the decision latitude scales into quartiles. This yields a bivariate distribution of 16 cells.  This distribution can generate a seven-level scale by stepping incrementally along the diagonal from low to high strain (see Figure 2).  For dichotomization to identify a "job strain" population, it is recommended that the three corner cells (levels 6 and 7) be defined as "job strain", and the remainder as non strain.
    Also, a multiple level active-passive diagonal can be defined in an analogous manner, but following orthogonal diagonals to the low-high strain diagonlas. This would yield another seven-level scale which we would then recommend cutting at (6, 7) vs. (1 to 5) to generate a "passive job sub-population." A more elaborate version of the this technique can be used to modify the often-used four-part demand control model diagram utilizing four quadrants of: strain job, low strain job, active job and passive job."  A new "mid-population" category can be generated by  consolidating the closest-to-the-bivariate-mean quartile-by-quartile cells from each of the four quadrants. This yields a new five part version of the diagram (see Figure 2B), which may be relevant to some analytic purposes.

National Means and International Comparison

Q. I have heard that the means and standard deviations for decision authority and decision latitude scale scores of the US QES data were incorrect in Table 3 of the study (Karasek et al. 1998). Is it true?  

   A: Yes, it is true. U.S. QES Decision Authority means (Table 3, Karasek et al., J Occup Heath Psychol, 1998) are incorrect as printed and must be multiplied by a factor of 1.20. This also changes Decision Latitude, since it is the sum of Skill Discretion and Decision Authority. This error also was found in the basic JCQ standardization scores used for the US. This only effects U.S. scores.   

Revised Decision Authority

 

 

 

 

U.S. QES means:

For Men

37.7, SD 9.6

For Women

33.8, SD 10.3

As published (error)

 

31.3,SD 8.0

 

28.1,SD 8.5

Revised Decision Latitude

 

 

 

 

U.S. QES means:

For Men

72.6, SD 15.4

For Women

65.7, SD 15.8

As published (error)

 

66.2,SD 13.9

 

60.0,SD 14.2

    For the JOHP article, this change has the effect of bringing the means of the QES sample closer to the means of the other sample, and thus strengthens the articles primary conclusion: that the JCQ means are relatively similar across full occupation spectrum samples in developed countries. The primary use of JCQ materials which is affected by this is to construct a job strain term (or other statistical references) that use the U.S. population means. In general, a “job strain” term defined by scale cut points based on the old (error) scores would define a more limited population (i.e. a higher strain group) than is identified by the revised score. In general U.S. JCQ scale scores are not published but can be obtained for occupations by request to JCQ Center.

Q. We collected occupational category information. I have been hoping to classify our job categories for men and women to reflect degrees of job strain that could lead to various cardiac states. Do you have a list of occupational categories by male and female and the ratings of decision latitude and psychological demands in a tabular format?  

   A:.Yes, we do have such information and have used it for years. We have JCQ national standard scores(for U.S.). JCQ scale scores are available for the 31 scales listed in the JCQ User's Guide, broken down into the following subgroups (listed in the Guide):

Q.  I want to compare my country's JCQ scale scores with those of other countries. Does the JCQ Center have a huge JCQ database from many countries? Is it arranged by sex, occupation, and industry? And if so, can I have access to the database?   

   A:.There is no such system at present. However, the JCQ Center may try in the future to build such comprehensive JCQ database from many countries for the international comparison of psychosocial job characteristics. Furthermore, the JCQ Center has indeed required the following; " our permission to use the instrument on larger studies is contingent on users providing a copy of basic statistics and the job data file based on our questions. This requirement might conceivably allow us to update the national and international scores in the future"(in the JCQ Users' Guide). In the future, the JCQ Center may begin this task encouraging a collaborative international project by putting together the isolated JCQ databases from the developing countries(South Korea, Brazil, Mexico, Taiwan, etc.) as well as developed countries (U.S., Canada, Japan, the Netherlands, Belgium, France, Italy, Sweden, etc) in close cooperation with the international JCQ users' groups. The eligibility for having access to the system and requirements for publishing the results by using the system will be discussed as this project develops.

Translation

Q. So far, how many language versions of the JCQ does the JCQ Center have?

   A: So far, the JCQ was translated into nineteeen countries' languages (Belgium(Flemish), Bulgarian, Chinese, Dutch, French, Greek, Iceland, Italian, Japanese, Korean (South Korea), Mexican (Spanish), Norwegian, Portuguese(Brazil), Puerto Rico, Russian, Spanish, Swedish, Taiwanese, Thai). See Translations.

Q. I would like to translate the JCQ into my country's language. Please let me know the procedure of the translation.

   A: Step I: translate the JCQ questions into the languages of the site countries
        Step II: please back-translate the translated questions into English
        Step III: please send both translated and back-translated versions to the JCQ Center.
   For details, please contact the JCQ Center. For the quality of translation, please use professional translators (Not students). Finally approved translated version will be distributed by the JCQ Center in the future to other users under the same conditions as the English version.