Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Methods of data collection
Processing and interpreting data
Procedures for collecting and analysing time-activity data efficiently and effectively
Appendix 1. Messer, mitla field notes, 4 September 1980
World Hunger Program, Brown University, Providence, Rhode Island, USA
The purpose of this paper is to indicate how studies of time allocation can be combined with nutritional studies to help illuminate household and general socio-economic conditions that impinge on nutrition and health.
The organization of time is a significant feature of any life-style. The ways people classify intervals in time, schedule activities according to these intervals, and fit such activity schedules into time comprise aspects of their time sense, which varies from culture to culture.
What people do in time and how they time what they do are aspects of cultural identity. Usually a distinction is made between Western industrialized societies in which "time is of the essence" and non-industrialized societies in which "productivity" is evaluated in terms of social relations and cultural values in addition to material output "per unit of time." Yet even in nonindustrialized societies, work demands at particular times may create time conflicts as different activities, such as food production and household maintenance, compete for time and create bottlenecks in individual and household schedules. The priorities people set in expending time, the urgency with which people complete tasks, and the ways in which people perceive their seasonal cycles of production and consumption are all features of their strategies for survival.
Time budgets and time-activity patterns are useful analytical units for understanding how people produce income and meet household needs in any socio-cultural group. Time and activity data can, for example, provide significant measures showing how the organization of individual or household activities, along with other aspects of the environment, contribute to the food provisioning and the nutritional and health maintenance of individuals, households, and communities. Studies of the organization of time, from scientific and folk perspectives, can identify many of the factors that affect what, when, and how often people eat. Additional comparisons of activity patterns among individuals (categorized by age, sex, or occupation) and households (classified according to occupational type, income level, numbers and ages of children, or family structure) can also further our understanding of why some households or cultural patterns succeed and others fail nutritionally under apparently the same environmental conditions.
Some Applications of Time - activity Data for Nutritional Studies
Anthropologists, psychologists, and economists have been collecting time-activity data for years to answer questions within their disciplines. Time-activity data have become increasingly important for nutritional studies as (a) functional measures have been increasingly emphasized in nutrition and (b) as the research perspective has broadened beyond studies of the individual to comparisons of the effects of different individuals on the nutritional status of others in their households. Nutritionists as well as anthropologists are now looking at nutrition in its entire social context to understand the important linkages between social and household variables and nutritional well-being. These questions, which form the basis for an ongoing study of the relationship between nutrient intake and function (Calloway et al., 1979), are essential if we are to understand the relationships between nutrition, socio-cultural organization, and various categories of work and leisure performance. Projects have been organized in terms of the relationships between nutrient intake and: disease-immune response; reproductive function; work capacity, i.e. physical function; psychological (cognitive) function; and social function (Calloway et al., 1979). To understand these relationships, one needs data on nutrient intake as well as (a) typical household tasks, household organization of tasks, typical times it takes to complete certain sets of tasks, and activity schedules of the focal female and others around her; (b) meal scheduling and the manner of acquiring, preparing, and serving food; (c) sources of infection; and (d) patterns of child care in households. Additionally, one needs to know where to look for linkages between social organization, time use, and biological well-being.
Some of the approaches anthropologists and other social scientists have used to study the relationships between time allocation and nutritional status include careful assessment of seasonal and household differences in time use. Sharman's (1970) investigation of the relationships between food beliefs and practices, income, and nutritional status in rural Uganda, for example, shed light on several pertinent issues. She concluded that social organization of household labour, including how women expended time and other resources, appeared to be a significant factor in predicting the nutritional outcome of children. In this society, management of resources, not just their availability soon after the harvest, determined whether there would be sufficient food for household members throughout the year.
Sharman found it was crucial to examine co-operation within and between household units. She noted that the generosity, wastefulness, and frugality of the various household managers were critical factors, as some women spent much of their time and substantial food resources early in the post-harvest season preparing beer for sociability groups. The expenditures depended on their desire for this form of sociability and on the amount of household labour available for their organization. Also, Sharman found that men and women who spent most of their time working other people's fields ate fewer meals per day, and used nutritionally poorer ingredients that were quicker to prepare. Moreover, their children tended to scavenge more. The nutritionally poorest staple and relishes (vegetable-protein complement) were, furthermore, those that required the least time to prepare.
"Mothercraft" - time management and household child management - has also been suggested as a significant variable in other nutritional studies. Evaluating effects of a mother-child health programme in Cali, Colombia, Franklin and Valdés (1979) quantified the relationship between mother's time and abilities and children's nutritional status. They considered mothers' use of time and food (measured in calories) as complementary scarce resources, which together could produce more or less acceptable levels of child nutrition. Employing a regression model, their data support the hypothesis that a mother's time and abilities, not just food availability, significantly influence child nutrition.
Time allocation of focal females may also affect nutritional well-being via an interaction between nutrition and infection. Kumar (1977), in an investigation of the relationships between occupations, incomes of women, and children's nutritional status, studied how seasonal conflicts in time allocation between agricultural tasks and child care could periodically subject children to infection and nutritional stress. She noted how, during peaks in seasonal work in the fields, women were inhibited from spending adequate time providing food and health care for children, who were, in turn, cared for by others, usually older siblings. Infants and very young children subjected to this periodic neglect while their mothers joined harvesting teams in the fields might not be adequately fed or protected from the sun, and often became ill. As a result, children of those mothers who spent the most time working in the fields and, therefore, earned higher incomes were often worse off nutritionally than those of mothers working shorter hours and earning less. The results of time-allocation observations could document, then, how women's work, given the social organization of child care, is related to child nutrition and health maintenance. Such studies might explain the seemingly skewed negative relationship between mother's income and children's nutritional and health status in some socio-cultural situations.
Studies of household labour organization and time allocation by all members have also furthered understanding of the variables intervening between levels of income and nutritional outcomes of a household (e.g. Nag, White, and Peet. 1978; Quizon-King, 1978). Such timeallocation records can indicate (a) how the labour of all household members, including children, contributes to household maintenance, income, and nutritional and health status, and (b) how time allocation by women for work outside and inside the home interferes with child feeding and child care. Time-allocation records can be particularly useful in examining the health and nutritional consequences of women's seasonal labour patterns and surrogate-mother child care. Women working outside the home have less time for breast-feeding and child care but more income to buy food: are their households better or worse off?
In a study by Tripp (1981) of Nigerian women, households of women working outside the home were better off. Those women who worked as traders had children who showed fewer signs of malnutrition than women not so employed. He concluded that women who earned their own income invested it mainly in food for their children, which accounted for their children's superior nutritional levels.
In contrast, Popkin and co-workers (Popkin and Solon, 1976; Popkin, 1980), in a series of studies, suggested that women who work as traders in the Philippines have children who are less well-nourished in terms of vitamin A status than their nonworking counterparts. They suggest that the reason for the vitamin A deficit is that working mothers have less time to prepare leafy green vegetable sauces that are rich in vitamin A to go with the dietary staple, rice, than those who spend their time gathering such ingredients for subsistence. Thus, as women enter the cash economy, their purchased diet may be inferior to the home-cooked meals they no longer have the time to prepare. However, none of the studies by Popkin and co-workers provides dietary and time data sufficient to support such a view. Such data would include when work took place and when income was a available over periods of a day, a season, and a year and would provide more adequate information than occupation and income categories.
An important use of time studies is in evaluating the actual or potential nutritional impact of socio-economic development policies. For example, some policies can be anticipated to have adverse effects on women by increasing their workloads, which may already average more than ten hours a day (Scrimshaw, 1982). Where men are lured into non-agricultural occupations, women may be left to run household farms completely alone. Superbreeds of cows, pigs, and poultry, which require more forage and water, make more work for women who often must carry these items. Fertilized crops grow more weeds, and women must pull them. Insecticides require water, and women must draw it (Fortmann, 1981, pp. 211-212). Precise calculations of such increases in "drudgery" and their impact on women's ability to care for and feed their households can only be obtained through time-allocation studies.
Water and food-processing technologies can also be evaluated in relation to their effects on women's time allocations. In what seasons would such labour-saving devices as mills be most desired? How have traditional social arrangements managed to decrease the time needed for food preparation in certain high labour-demand seasons? What are the effects of technological changes on such social arrangements?
Hemmings-Gapihan (1981), in her study of the potential impact of a solar energy unit in Upper Volta, used small samples of women to observe the amount of time they spent drawing water and pounding grain in different seasons - tasks that would be facilitated by the introduction of a solar energy cell. Time allocation by women to household and food-production activities, as well as leisure, was recorded to determine the effects of modernization on household structure, the social organization of labour, and women's work. Such examples of time-allocation analysis are important for understanding the relationships between household organization and nutrient intakes of members.
For advancing the productivity of women, preliminary studies of women's time allocation and patterns of household decision-making regarding all aspects of food provisioning can yield significant information about potential food increases that could result from introducing new forms of income production. If the labour time to clean, water, and feed household animals, as well as the daily rations of time and cash to supply them with food, could be shown not to interfere with women's time allocated to feeding and providing health care for human beings (including their own health and well-being), then development policies that included animal husbandry might be beneficial. Such policies might be particularly important in instances in which women's income seems to go directly and almost exclusively into providing more food for their households.
Time data on scheduling work and household tasks can also help in the development of programmes, including mother-child health and nutrition projects. Such projects might schedule activities to least interfere with household production routines. Preliminary background data for programme development involve observation of whole days and timed observation of selected activities in a limited number of households over short periods of time in different seasons.
As Nevin Scrimshaw, among other perceptive nutritionists, has noted, it is unwise to design health and nutrition programmes that will add yet another burden in time to the already overworked female running a household. He has suggested, as an example, that national and international agencies planning nutrition interventions in the form of supplementary foods consider a low-cost processed food. Such a food would not require elaborate food preparation that would compete with other activities in a mother's already full day (Scrimshaw, 1982). The appeal to health and nutrition planners to pay more attention to the value of women's time in planning their programmes and evaluating the reasons for low participation rates is customary in the anthropological literature. Again, selective observational methods, perhaps combined with selective interviews focusing on time constraints, could aid in the wise design of nutritional studies.
Collection of time-use-activity data is not new to anthropology, economics, socioeconomic planning, or developmental psychology. Within ecological anthropology, for example, human ecologists, focusing on the human population as their unit of analysis, have timed activities as an integral part of their investigations tracing the flow of energy and other nutrients through local and regional ecosystems. Using time-motion studies to estimate human energy expenditure in food production (Rappaport, 1967), or estimating time and energy spent in production relative to consumption of food (Lee, 1969), anthropologists have calculated the relative effort expended in the food quest under varying conditions of resource availability and social organization.
Socio-ecological anthropologists, to construct or reconstruct "optimal foraging strategies" for human social groups in particular environments, have used the method of linear programming to calculate the hypothetical time and effort (energy) expenditure required to obtain resources of specific nutritional value (see Winterhalder and Smith, 1981). In explaining deviations of actual from expected values, they have often suggested that factors other than efficiency or ease of energy-calorie acquisition are valued by societies in their food quests. Tastes, especially for the fat in food, or political factors, such as the relative safety in times of social upheaval of sowing and harvesting crops with a given growing time and product return, are dimensions that can be useful in predicting and comparing food choices within and across cultures (Jochim, 1981).
Primatologists, focusing on the social group rather than the ecological population as their unit of analysis, have used time-frames to compare and contrast the frequency of feeding of adult males versus females or immature members or either sex in primate societies. Following single individuals, or censusing at regular intervals whole social groups, they have analysed how relative ranking affects the efficiency of feeding and the nutritive value of the intake of individuals or classes of individuals within primate groups. For a review of the methods of observation used in primate studies, their comparability, and relative merits, see especially Altmann (1974) and essays and appendices in Clutton-Brock (1977).
Similar methods have been used in socio-cultural studies. Sharman (1980), in a research design reminiscent of methods used in primate studies, analysed the "frequency and timing of eating events, the extent to which households eat together and eat the same thing; and the general format and content of meals." She examined the relationships between dietary choices and allocation of resources, both income and time, within households, with special attention to the nutrition of children under five, the effect of low income, and the impact of a women-infant-child feeding programme (WIC). Her preliminary study resulted in a series of research questions directed at improving the quality of reporting about the social context of nutrition and nutrition interventions: What are the complete eating patterns of household and individual members? What is the relationship of work schedule to eating patterns? What is the relationship of income, time, and shopping/food expenditures? Also, she suggests a route to dietary analysis in modern complex societies that would reflect factors such as the shortage of time, the dislike of cooking, the timing and frequency of meals, as dependent on domestic chores, visiting, and cash-work patterns on food choices, expenditures, and consumption patterns.
Economic studies by anthropologists and other social scientists have also involved research on time use. Particularly those studies investigating how people's time schedules affect their productivity, health, and quality of life, or, reciprocally, how productivity, health, and quality of life affect people's performance of various types of activities, have contributed data and methodologies. The pre-Second World War social anthropological study of the Bemba of Northern Rhodesia by Audrey Richards (1939), for example, was designed to answer the British colonial economist's questions about why natives did not work harder and whether it was possible to increase local food production.
Richards investigated how Bemba men, women, and children scheduled their activities, the time required for particular activities, and the relationships between social organization, work performance, and the local diet. On the basis of her observations of task performance by different women over several seasons, she was able to suggest how insufficient nutrient intake by adult women during the lean season, which was also the planting season, affected their ability to perform both productive tasks and household maintenance. She was also able to identify ways in which the nutritional deficiencies of adult women may have interfered with current and prospective production and nutrition of all household members. She blamed the situation on the British colonial interests, which had drawn men away from their home villages to the mines, and left the women with the entire burden of clearing new fields in addition to their usual gardening tasks.
Following Richards, more recent studies have addressed questions concerning the household factors, including organization of female labour, that affect production and distribution of household resources and, by extension, the nutrition and health of household members. Timing the performance of common subsistence activities across households forms a part of investigations (e.g. Sharman, 1970). Also, social anthropologists have tried to develop appropriate units in work and time by which to calculate the costs to various types of production, for example, how many persons with what technological instruments are needed to plough a certain class of field (Wallman, 1969).
Other anthropological studies drawing on time-activity data have addressed the questions whether people work "harder" (or, at least, longer) in some cultures (Gaulin and Konner, 1977) and how the industrialization of production lightens or increases workloads (Minge-Klevana, 1980). In related studies, time-allocation profiles of men, women, and children have also contributed to economists' understandings of socio-economic factors operative in women's fertility decisions and their decisions to enter the paid labour force (summarized in Evenson, 1981). Time use has also been a dimension in general social comparisons between societies (Szalai, 1972), and time-activity data have been collected to further recreation facility planning by showing how people prefer to allocate leisure time. The studies of Chapin (1974) on leisure-time activities in the greater Washington, D.C. area have added to the corpus of interview methods used in time-activity studies.
Psychological studies, in contrast, have employed time-frames to observe how parents interact with children, or, alternatively, how children "interact" with their environments as part of more general studies of child development (Whiting and Whiting, 1975). A few of these studies have investigated, in addition, child behavioural development in relation to nutritional supplementation of the mother and/or the child (see Herrera et al., 1980).
More sociologically oriented studies have dealt with related historical and comparative cultural questions of the changing perception and quality of time in different historical contexts (Moore, 1963; Thompson, 1967). The notion of time as an economic commodity to be rationed, spent, or wasted and the temporal dimensions of social and economic life are important qualifying factors for any quantitative study of time use. "Non-economic" allocations of resources, including time and energy allotted to ritual, social, or other economically "nonessential" activities must also be accounted for in arriving at an accurate depiction of the interrelationships between "work" and nutrient intake and nutritional status.
The differing time senses of industrial versus non-industrial societies have been dealt with in a number of historical studies (Thompson, 1967) and ethnographic reviews of societies and cultures undergoing technological transformations (Mead, 1955). These suggest that, outside Western industrial societies, one should expect no simple relationship between time allocated to work and household productive income, or, more to the point, no simple way to sample and measure "time use" to predict economic and nutritional well-being.
Pacing of activities over daily, weekly, seasonal, and yearly cycles may be determined by competing internal (household) and external (economic) demands, both natural and sociocultural in origin. For example, demands on women's time may be seasonally shaped not only by the need for female labour in planting and harvesting, but also by the cultural rules for the social organization of labour allocated to each task. How much effort should be expended by particular classes of individual - i.e. how hard and how long people work to meet culturally defined goals - is also a product of cultural training, expectations, and the cultural food system. These factors affect measurable physiological potentials of individuals to perform work, as well as standard tasks and standards for their performance.
Studies of ritual time are also important. For example, the ritual observance of one day of rest and heartier food consumption in every seven in many Jewish, Christian, and Moslem societies which are otherwise marginally nourished, and periods of ritual preparation and consumption in various societies, suggest that simple averaging of time (and activities) over a number of days will not yield a meaningful picture of how time-use patterns are related to the nutrition of individuals, households, and communities. Any study which proposed to use timeallocation analysis as a means to describe and compare social and economic conditions must, therefore, begin by collecting preliminary data on the "social value" of time and its implications before proceeding further.
With these caveats on qualitative time factors in mind, I will proceed in the next section to examine several possible ways to study time allocation quantitatively.
Contents - Previous - Next