Important variables: results of the positive-deviance mail survey

Between December 1983 and May 1984, a mail survey on the topic of positive deviance was sent to almost 700 nutrition and health professionals. The survey questionnaire listed variables that were known or hypothesized to contribute to positive deviance and asked the respondents to rate their importance. The researchers requested survey responses from those health professionals with an active interest in this topic. Responses were received from 68 persons in 36 countries in time to be included in the analysis (and from 25 more thereafter). This response rate of about 13 per cent indicated to the research group that a relatively large proportion of professionals in the nutrition field have an active interest in this topic.

Table 8. Topics ranked by percentage of "important", items

Results

The results of the survey, including the regions and countries of the respondents and descriptive statistics by rural and regional location, are presented in Appendix I in tables A to F. Table G summarizes the open-ended observations, suggestions, and comments written onto the questionnaires. The names and addresses of the respondents are listed in Appendix 2 in order to provide an informal reference group of professionals interested in the topic of positive deviance.

It is important not to overanalyse these data, which are subjective in nature and based overwhelmingly on personal observations by a selected group. One of the primary purposes of presenting the results is to provide researchers with a comprehensive list of variables and to enable them to refer to the survey responses on an item-by-item basis when designing their own field-studies. Items ranked as important by the survey respondents should clearly receive serious consideration in both research and programme design. It is interesting to note that 44 per cent of the questionnaire items received average rankings greater than 3 on a scale ranging from 0 to 4.

By topic area, in descending order of importance, the percentage of items under each topic with rank averages above 3 is shown in table 8.

Table 9 presents all items ranked as "very important" (score of 4) by more than 50 per cent of respondents in the total sample or in either the rural of urban subsamples. The writers believe these items should serve as a useful "shortlist" of factors that should be taken into consideration when studying positive deviance.

Both tables show that our experts gave highest importance to the role played by infection in positive deviance.

Some rural/urban differences emerged from the results. In general, variables reflecting modernization were ranked as being more important in the urban setting. Greater contact with the outside world, fewer visits to traditional healers, and less use of home remedies were ranked as significantly more important for urban than for rural mothers. Characteristics of other caretakers also ranked higher (quite possibly because urban employment tended to require separation of mother and child). In the rural areas, presence of siblings old enough to help the mother ranked higher. With respect to attitudes, mother's satisfaction with her life was ranked more important in

Table 9. Survey of expert knowledge and opinion on positive deviance in nutrition of young children (items agreed to be very important by 50 per cent or more of respondents in overall, rural, or urban categories)

^a. Not all responses could be included in this calculation since some were respondents who wrote out their information in longhand rather than answering the items. urban areas. Sex of the child was ranked less important, and timing of the birth more important in the urban setting.

The numbers representing the different regions are too few to permit statistical come parisons, although the regional values may be worth reviewing for individuals interested in specific items. For example, the practice of discrimination against female children was ranked as more important in Middle South Asia (Bangladesh, India, Nepal, Sri Lanka, and Turkey) than in other regions.

Discussion

In summary, the survey results underscore the importance of nutrition-infection interactions for the study of positive deviance. These high ratings given to health may in part reflect the fact that many of the respondents were clinicians who encountered malnutrition in sick children attending health facilities. The results confirm that many nutrition and health professionals acknowledge the importance of psychosocial factors contributing to child growth and particularly to the ability to thrive under conditions of adversity. They also illustrate the fact that conditions contributing to positive deviance differ significantly from one setting to another.

An interesting example of this difference is that sex of the child was ranked more important in the rural areas, while timing of the birth was more important in the urban areas. In many rural areas, particularly outside of Africa, land is passed down from father to son and the multigenerational patrilocal family is the production unit. Under these circumstances, a primary parent-son emotional bond may be required to ensure intergenerational commitment to the economic unit. The need for such preferential bonding would diminish with urbanization. However, timing of births increases in importance as couples begin to limit their fertility and mothers enter paid employment.