In this study of the sequelae of conjugal bereavement it was hypothesized that widows who resolved the bereavement crisis in a healthy manner would differ from those with a bad outcome in their perception of the supportiveness of interpersonal relationships during the crisis. In order to obtain matched groups of good outcome and bad outcome subjects, information was obtained by questionnaire from an unselected sample of 132 widows concerning their physical and mental health during the 13 months following bereavement. On the basis of this information 28 subjects (21.2 per cent.) were rated "bad outcome", 57 (43.2 per cent.) were rated "good outcome" and 47 (35.6 per cent.) "intermediate", and from the two extreme groups 20 good outcome subjects were matched with 20 bad outcome subjects for later interview. Of the various social and personal characteristics of the 132 widows recorded on the health questionnaires, only age of widow and age of husband were found to have a statistically significant relationship to illness score, with younger widows and/or widows of younger husbands reporting greater deterioration in health following bereavement.
In the main study each of the 20 good outcome subjects and 20 bad outcome subjects was interviewed. Information was elicited firstly about specific persons available to the widow during her bereavement crisis and about her perception of them as helpful, unhelpful or neither; because of small numbers only qualitative discussion of this set of data was possible. Secondly, the widows were asked about specific forms of interpersonal exchange and whether these had been perceived as helpful, unhelpful, indifferent, absent or needed, and in statistical analysis significant and meaningful differences were found between good and bad outcome groups. Over the total 59 items employed in this second part of the interview there was a gross difference between the groups in the frequency of "needed" responses, bad outcome subjects tending to perceive the environment as failing to meet their needs during the bereavement crisis. While there were no overall differences in "helpful" responses, there were significant differences between the groups on subsets of items, good outcome subjects tending to perceive permissive support as helpful and bad outcome subjects appreciating more active encouragement from the environment. Bad outcome subjects also tended to perceive the environment as actively unhelpful, the relevant interchanges usually involving either the blocking of a widow's expression of affect, or overt or covert hostility directed towards her. Attempts to focus her attention on the future, and to discourage her thinking of the past, tended also to be found unhelpful by bad outcome subjects. The results were considered to be of theoretical and practical interest.