Marmot et al (1997)

·         To investigate the association between workplace stress and stress-related illness in male and female civil servants.

·         This was part of the Whitehall studies, where a number of different psychosocial characteristics of work were investigated to test their association with illness.

·         This particular investigation focused on the negative correlation between job control and stress-related illness.

·         A sample of 10,308 civil servants aged 35-55 (6895 men – 67%, and 3413 women – 33%) were investigated in a longitudinal study over 3 years.

·         Research methods included questionnaires and observation.

·         Job control (an aspect of workplace stress) was measured through both a self-report survey and by independent assessments of the work environment by personnel managers. 

·         Job control was assessed on two occasions, three years apart.

·         Records were also kept of stress-related illness.

·         A correlational analysis was carried out to test the association between job control and stress-related illness.

·         Participant's with low job control were four times more likely to die of a heart attack than those with high job control.

·         They were also more likely to suffer from other stress-related disorders such as cancers, strokes, and gastrointestinal disorders.

·         These findings were consistent on both occasions that job control was measured and the association was still significant after other factors, such as employment grade, negative attitude to employment, job demands, social support and risk factors for CHD had been accounted for.

·         The findings seem to show that low job control is associated with high stress, as indicated by the number of stress-related illness.

·         There is an inverse social gradient in stress-related illness among British Civil Servants; as job control decreases, illness increases. In other words, the variables are negatively correlated.

·         Implications include the responsibility of employers to address job control as a source of stress and illness. Giving employees more freedom and control may decrease stress-related illness which would increase efficiency in workplace.

·         The self-report method is vulnerable to investigator effects and participants reactivity bias. The P’s may have guessed that an association between job control and stress-related illness was being looked for, and so reported low job control if suffering from illness and high job control if not.

·         Weaknesses of the correlational method mean that there is no control over the variable (job control), which makes interpretations difficult as cause and effect cannot be established (causation can only be established when an IV is directly manipulated). Therefore, it cannot be said that low job control causes stress-related illness; only that an association can be established. This lack of control also means that other factors such as personality and coping skills may be involved in the association.

·         The jobs performed by those high and low in control differed in several ways other than simply control, i.e. those having high levels of job control generally earn more money, have more interesting jobs, have more opportunity for interpersonal contact, than those having low levels of job control. We do not know which of these factors is most closely associated with heart disease.