Methodological Issues

Many studies have been published since the mid-1970s on the effectiveness of bereavement intervention but there are major methodological flaws with most of them (Allumbaugh & Hoyt, 1999; Dowdney, 2000; Kato & Mann, 1999). Stroebe, Stroebe & Schut (2003) argue it is unethical to carry out poorly designed bereavement research given the sensitivity of the subject. The following aspects have been criticised frequently in study design.

Recruitment and Attrition

Recruiting and maintaining subjects in large studies is problematic. Bereavement research is no exception. Parkes & Weiss (1983) reported that of 274 potential subjects 'nearly three-fifths agreed to be interviewed' and Schut et al. (1997) report that only just over one fifth (22 %) of 358 eligible subjects agreed to participate in their study of gender differences and counselling programmes. Gallagher-Thompson et al. (1993) mailed 2450 bereaved individuals of whom 735 (30 %) replied and only 212 met the criteria for age and proximity for domiciliary interviews. Large numbers of potential subjects need to be identified before adequate numbers of suitable subjects can be found to evaluate the effect of different variables.

It is tempting to speculate on the mental health of those who decline to participate in research. Some would argue that these individuals have coped well and are reluctant to invest time and energy into something for which they feel no need. Others would argue an opposing view and suggest these individuals are not coping well and feel too distressed. The situation is further complicated by the suggestion that the decision whether or not to participate in bereavement research may be gender related (Stroebe & Stroebe, 1989). A different rate of recruitment or attrition between groups adds to the dilemma in identifying the effects of different interventions. A third of participants in one study dropped out after finding out to which one of four groups they had been allocated (Barrett, 1978, cited in Kato & Mann, 1999). People may choose the intervention with which they feel most comfortable rather the one that is more effective for their needs (Parkes, 2000) but it is arguable that the client's perception of the credibility of an intervention may enhance its effectiveness.

The social context of the recruitment invitation can influence uptake. Couples who had experienced perinatal loss were more likely to accept the offer of counselling support if the therapist was present with the clinician at the time of the interview (Lilford et al., 1994).

Bereaved children can be difficult to access because carers are protective and concerned that questioning would distress their children further. If adequate numbers of children are recruited, assessing the effectiveness of a programme continues to present challenges (Stokes, Wyer & Crossley, 1997)


Most participants in bereavement research are widows. A meta-analysis of 35 studies reported that 84 % of the participants were female with the modal person being a widow of 52 years (Allumbaugh & Hoyt, 1999). It has been suggested by Stroebe (1998) that within Western culture men and women respond differently to bereavement and this effect may confound results in studies in which gender has not been controlled. Earlier work by Stroebe & Stroebe (1989) found that widowers who agreed to participate in research were less depressed than those who declined whereas widows who agreed to participate were more depressed than those who declined. Bereaved women are more likely than bereaved men to be members of support groups and this is relevant if a sample is recruited in this way. In addition, it is arguable that there are differences between joiners and non-joiners of social groups regardless of their gender.

Nature of the Death

Some deaths are sudden but even where there has been a period of forewarning it may be seen as sudden by those bereaved. Anticipatory grief was described by Lindemann (1944) and the concept was refined by Rando (1988). The term originated to describe the experience of grieving so fully before an expected death that emotional ties are relinquished - the missing soldier returns but finds no welcome. A different view is that attachment strengthens in the face of loss and anticipatory grief is not the same as the grief felt subsequently. A preparatory period can offer advantages but if the final illness is of a protracted and uncertain duration then this brings different stressors. A study of 73 caregivers bereaved by the death of an elderly relative identified that quality of support prior to the death was more important in adjustment to bereavement than subsequent support (Bass, Bowman & Noelke, 1991).

Sudden, unexpected death is considered more problematic for the bereaved person and the research of Parkes & Weiss (1983) supports this, but there are confounding variables. Sudden death is more likely to be the cause of death in a younger population and so those bereaved may be younger and possible have dependent children. The death is more likely to be violent, whether this is caused by an industrial accident or major incident, and in these circumstances the death will involve lengthy legal processes and publicity. These issues are discussed later.

The Type and Nature of the Relationship

For a parent there is an untimeliness about the death of a child whether that child is in its early years or has become an adult. A source of practical and emotional support may be lost with the death of an adult child. If the dead person was not valued by their community then the bereaved survivor will feel isolated in grief as can happen with the bereaved parents of children with learning disabilities (Cathcart, 1996).

Parkes (1972) suggested that an ambivalent relationship is more complicated to grieve as unresolved tensions are left with the survivor. Pincus (1974) explored how the psy-chodynamics of marital relationships influence living with subsequent bereavement. The apparently stronger partner may have gained self-esteem by caring for the apparently less confident but this need is unmet by the death of the latter.

Researchers have varied in their definition of 'elderly'. Gallagher-Thompson et al. (1993) set the limit at 55 years and over whereas Strobe & Stroebe (1993) used a sample with a mean age of 53 years in a study of younger widows. Lund, Caserta & Dimond (1993) examined bereavement in 'later life' and surveyed a bereaved population over the age of 50 years. This lack of agreement adds to the difficulty in drawing conclusions about the impact of bereavement on an elderly population. The participants in the Lund, Caserta & Dimond (1993) study ranged from 50 years to 93 years but the authors note with surprise that age was not an important predictor of outcome in spousal bereavement but personal resources such as flexibility and a sense of self-competence were more significant.

Culture and Ethnicity

Rosenblatt has been exploring the universality of grief for nearly three decades (Rosenblatt, 1996; Rosenblatt, Walsh & Jackson, 1976) and reflects how his own perception has shifted from believing that social scientists can understand another culture by careful observation and recording to an acknowledgement that these observations are shaped by the culture which is an integral part of the observer. Practices that may seem callous to one observer such as a certain way of disposing of a body may be seen as commendable by another because this preserves the future wellbeing of the whole community.

There will be individual variation within a culture and there can be conflict within a family over the 'proper' way to show respect for the dead when practice is changing within a society. Two separate traditions may come together. Viking graves have been found in Scotland carved with Christian symbols as well as containing the material goods associated with pagan burial.

Social Desirability

This factor may be relevant if answering questionnaires or being interviewed in bereavement when there are strong expectations from the social group. It may be acceptable that there is a sense of relief at a death that follows a lengthy period of poor quality of life but it may be less easy to acknowledge a sense of relief if the relationship had private tensions. Bereaved relatives or families in the public eye will find their reactions scrutinised and be offered sympathy or criticism depending on whether their behaviour is considered 'appropriate'.

Length of Follow-up

Some studies report a follow-up period of a few months only, whereas others report a year or more. Yet during the first year of bereavement there are many 'firsts', such as the first holiday, birthday or anniversary, significant events that bereaved people may use as milestones. Lake et al. could trace only half their 78 participants at six months. Parkes & Weiss (1983) were unusual in tracking some participants for two to four years. The authors of one major meta-analysis criticised the lack of long-term follow-up in many of the studies they surveyed (Allumbaugh & Hoyt, 1999).

Self Improvement Fast Track

Self Improvement Fast Track

Surefire Ways To Put Your Self Improvement On The Fast Track. This Book Is One Of The Most Valuable Resources In The World When It Comes To Accelerated Learning Techniques For People New To Personal Development.

Get My Free Ebook

Post a comment