Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The ideal of a time-allocation study is a "god's-eye view" of everything everyone in the research population does. That this is completely unfeasible should not deflect us from recognizing that it is the ideal. When we compromise this ideal for practical reasons, we should do so reluctantly and self-reflectively, with the knowledge that too many compromises may result in collecting masses of misleading or useless data. In designing a time-allocation study, evaluation of alternative methods should incorporate the following four rules of thumb.
Coding Rules Must Be Explicit
At the present stage of research, when standards for describing human behaviour are neither rigorously developed nor widely shared, the best that researchers can hope for is to develop explicit descriptions of how they recorded particular behaviours and how they coded them into higher-order categories. Each study should, as a matter of course, produce a code-book that other researchers may obtain and use when comparing one study with another. Then it will not matter so much if one baseball-playing father is coded as "child-caretaking" while in another study such a father is coded as enjoying "recreation," because, by referring to the separate code-books for each study, comparative researchers can recode the behaviours to correspond.
This problem is among the most serious confronting time-allocation research. It was the single most time-consuming and difficult aspect of developing a standardized coding scheme for cross-cultural comparisons in the UCLA Time Allocation Project. In a coordinated field study, such as that of Gross and his co-workers in Central Brazil (1979), a team of researchers can achieve intercoder reliability through pre-field-work training. But such projects are rare, and the prevailing pattern of individualistic, and to a degree idiosyncratic, coding schemes means that some ambiguity in exactly how observed behaviour comes to be coded as this or that "activity" is inevitable whenever we are trying to interpret another researcher's time-allocation data.
For those responsible for making and implementing policy, this implies both a need for caution and an opportunity. Time-allocation data, being numerical, can be presented in decimal-point precision, but, given the limitations of data-collection techniques, must not be regarded as more than estimates of the general pattern of time allocation in a community. On the other hand, by designing their own time-allocation component into new research, planners can carefully define coding rules, train for intercoder reliability, and feel suitably confident of the accuracy of their findings (see, for example, Neumann and Bwibo, 1987). This is a case where careful planning well in advance of field research can pay large dividends in the quality of data obtained.
Samples Must Be Representative
One of the compromises with the god's-eye view ideal is that we can only observe or interview a fraction of the population a fraction of the time. Thus, all time-allocation research is done on samples. This means that the rules of statistical inference apply, and the basic requirement for representative sampling must be met. To the greatest possible extent, the individual subjects should be randomly drawn from the population, and the times for which their behaviour is recorded should be random moments drawn from the stream of time. When this is impossible, a researcher must pause and give careful thought to what is being lost in terms of the generalizability of the data, and should devise compensatory methods wherever possible. For example, Baksh and Paolisso (1986), in collecting a very large set of time-allocation observations among the Embu of Kenya, found that observing individuals at random throughout a large region of scattered homesteads was prohibitively costly in researchers' travel time. Instead, they devised a roundabout route through the territory that passed each homestead, and proceeded to move along that route making spot checks at each house along the way, resuming each day where the previous day's observations left off (a team of researchers took turns). Since each household visit involved certain unpredictable delays, the day's observations never ended at the same house, and over time an essentially random pattern of household visits resulted. This compromise allowed for a highly efficient use of researchers' time without seriously compromising representativeness.
Short-term Memory Is More Reliable than Long-term Memory
It has been demonstrated that reports of behaviour made immediately (within 2-3 minutes) after an observed event are much more accurate than reports made after several minutes have elapsed. when short-term memory retention gives way to the distortion of information in long-term memory, which is dominated by pattern-seeking. Thus, long-term memory conforms much more to "cognitive maps" of what behaviour should be like than does short-term memory. While fresh, it retains a fairly clear and objective image of what actually happened. These cognitive maps are partly cultural, so that long-term recall tends to conform to implicit, shared ethnocentric preconceptions, and partly personal, which may lead to the recollection of selected behaviours that confirm biases and support personal theories while failing to capture those that do not. It is especially important to keep in mind that this is done completely unconsciously. In fact, people are strongly inclined to doubt that they even do this, though the experimental evidence is overwhelming (Bernard et al., 1984).
Methodologically, this implies that direct observations will be most effective when the observer immediately records the behaviour as it happens. Similarly, for informant selfreports, greater accuracy will be achieved if the informant immediately records his or her behaviour. How to do this without totally disrupting the flow of behaviour will require some ingenuity. For example, researchers are beginning to experiment with beeper-type systems, whereby the subject is "beeped" at a random moment, and complies by quickly noting in a pocket notebook the time and a description of the activity in which he or she is engaged and then proceeding with the activity. This method, if it is successful, will combine some of the strongest features of both the direct observation and informant selfreport methods.
Direct Observation Generates Unexpected New Information
Researchers who specialize in direct observations rather than informant self-reports often comment on the serendipity of unexpected insights provided by the first-hand observation (Johnson, 1978b; Gillespie, 1979). Direct observation of behaviour on a random schedule forces the researcher out of routines, landing him or her in unexpected places at unexpected times. It enhances the participant observer aspect of the fieldwork, and generally strengthens and extends the field-worker's personal relations among the study population. It can help the investigator overcome some timidity or reluctance about being in certain situations, and, just as a picture is worth a thousand words, it gives the researcher the wealth of data and sense impressions which only direct apprehension of an event can provide, but which cannot be achieved by asking an informant to tell about it.
SUMMARY
Time-allocation data are valuable means for describing the behaviour of people in the full range of their activities. Incorporated during the planning stages, studying time allocation in the proposed project area can provide a wide range of information for project development. Integrated into project implementation, it can help monitor changes as they occur. For example, are members of a targeted group increasing their participation in income-generating activities? All methods of time-allocation research, however, involve compromises with the ideal of obtaining complete and accurate information. Researchers must select from a variety of possible approaches those which achieve their specific research goals most fully.
It is important to maintain the clear distinction between methods based on direct observation by a trained researcher and those based on informant self-report. Direct observation tends to be expensive of a researcher's time and may be perceived as intrusive by the subjects, whereas informant self-reports depend on the documented frailty of human memory and may suffer greatly if informants are indifferent or secretly hostile to the research project and its personnel.
Four rules of thumb can help strengthen any time-allocation project: coding rules and categories must be made explicit and available to other researchers; observations must be representative of the population to which they will be generalized; observations recorded immediately after the event has occurred (short-term memory) are much more reliable than observations recorded more than a few minutes after the event (long-term memory); and direct observation of the event, by placing the researcher in the fullness of the context of behaviour, generates unexpected new data and insights beyond the mere quantitative description of time use.
Obviously, descriptions of patterns of time allocation can be useful in a wide variety of ways, complementing (not replacing) data on economic behaviour, household structure, socio-economic status, ethnicity, and so on. Perhaps the most significant limitation from the policy standpoint is that, like many essentially ethnographic methods, the data are most complete when collected over a period of at least a year, in a fairly small community. How can such a method be of use in a real world where a research team may have only two weeks to survey a region and develop policy recommendations?
In cases of extreme urgency, collection of valid time-allocation data is probably impractical. But, with some foresight, several possibilities suggest themselves on how to include such data in many projects. First, as mentioned above, if one time-allocation study can be done in a region, future short-term research projects can make inferences of time use based on that study. Such projects might include household surveys measuring useful predictors of time-use patterns (as established in the study). Second, in many world areas it is practical to train local researchers to conduct time-allocation studies over the long term; an expert team of researchers may actually need to visit the area only for a short period. If the study is done in advance of the team visit, the team can use the data to develop more precise questions and to formulate their recommendations.
Finally, even if time-allocation data are only collected over a short period, they can be very useful. True, we cannot use such data to construct a model of time use for an entire year. But they can give a rich picture of the daily round, of where and how people spend their time, and in whose company. Particularly when the preferred methods (spot checks, informant diaries, and 24-hour recall) are used, it is much better to have such data than not to have it, for it provides insights and opens avenues of inquiry that no other kinds of information can offer.
NOTE
1. Support for this seven-year project, Standardized Cross-cultural Time Allocation Database, comes from National Science Foundation (NSF) grants BN84-19033 and BN87-04604.
REFERENCES
As, D. 1978. Studies of Time-use: Problems and Prospects. Acta Sociologica, 21(2): 125-141.
Andorka, R., 1. Harcsa, and 1. Niemi. 1983. Use of Time in Hungary and in Finland Central Statistical Office of Finland, Helsinki.
Baksh, M. 1984. Cultural Ecology and Change of the Machinguenga Indians of the Peruvian Amazon. Unpublished doctoral dissertation. Department of Anthropology, University of California, Los Angeles, Calif.
-. 1985. Faunal Food as a "Limiting Factor" on Amazonian Cultural Behavior: A Machi guenga Example. Res. Econ. Anthropol., 7: 145-175.
Baksh, M., and M. Paolisso. 1986. Time Allocation Methods. Unpublished MS. School of Public Health, University of California, Los Angeles, Calif.
Becker, G.S. 1965. A Theory of the Allocation of Time. Econ. 1., 75(2): 493-518.
-.1981. A Treatise on the Family. Harvard University Press, Cambridge, Mass.
Behrens, C. 1981. Time Allocation and Meat Procurement among the Shipibo Indians of Eastern Peru. Human Ecol., 92(2): 189-220.
-. 1984. Shipibo Ecology and Economy: A Mathematical Approach to Understanding Human Adaptation. Unpublished doctoral dissertation. Department of Anthropology, University of California, Los Angeles, Calif.
Bergman, R. 1980. Amazonian Economics: The Simplicity of Shipibo Wealth. Dellplain/University Microfilms, Syracuse, N.Y.
Bernard, H.R., P. Killworth, D. Kronenfeld, L. Sailer. 1984. The Problem of Informant Accuracy: The Validity of Retrospective Data. Ann. Rev. Anthropol., 13: 495-517.
Bevans, G. 1913. How Working Men Spend Their Time. Columbia University Press, New York.
Boserup, E. 1965. The Conditions of Agricultural Growth: Economics of Agrarian Change. Aldine, New York.
Carlstein, T. 1982. Time Resources, Society and Ecology. George Allen & Unwin, London.
Carneiro, R.L. 1979a. Forest Clearance among the Yanomamo: Observations and Implications. Anthropologica, 52: 39-76.
-. 1979b. Tree Felling with the Stone Ax: An Experiment Carried out among the Yanomamo Indians of Southern Venezuela. In: C. Kramer, ea., Ethnoarcheology, pp. 21-58. Columbia University Press, New York.
Clark, C., and H. Haswell. 1970. The Economics of Subsistence Agriculture. Macmillan, London.
D'Andrade, R.G. 1974. Memory and the Assessment of Behavior. In: H. Blalock, ea., Measurement in the Social Sciences. Aldine, Chicago.
Deere, C.D., and M.L. de Leal. 1979. Measuring Women's Work and Class Position. Stud. Fam. Plan., 10: 370-374.
Dufour, D. 1983. Nutrition in the Northwest: Household Dietary Intake and Time-energy Expenditure. In: R.B. Hames and W.T. Vickers, eds., Adaptive Responses of Native Amazonians, pp. 329-355. Academic Press, New York.
Edmundson, W. 1977. Individual Variations in Work Output per Unit Energy Intake in East Java. Ecol. Fd. Nutr., 6: 147-151.
Ember, C.R. 1983. The Relative Decline in Women's Contribution to Agriculture with Intensification. Am. Anthropologist, 85(2): 285-304.
Erasmus, C.J. 1955. Work Patterns in a Mayo Village. Am. Anthropologist, 57(2): 322-333.
Foster, G.M. 1948. Empire's Children: The People of Tzintzuntzan. Publication 6. Institute of Social Anthropology, Smithsonian Institution, Washington, D.C.
Fox, R.H. 1953. A Study of the Energy Expenditure of Africans Engaged in Various Rural Tasks. Unpublished doctoral dissertation. London University.
Gillespie, V.H. 1979. Rural Women's Time Use. Stud. Fam. Plan., 10: 383-384.
Gross, D.R. 1984. Time Allocation: A Tool for the Study of Cultural Behavior. Ann. Rev. Anthropol., 13: 519-558.
Gross, D.R., G. Eiten, N. Flowers, F. Leoi, M. Ritter, and D. Werner. 1979. Ecology and Acculturation among Native Peoples of Central Brazil. Science, 206(4422): 1043-1050.
Grossman, L.S. 1981. The Cultural Ecology of Economic Development. Ann. Assoc. Am. Geographers, 71(2): 220-236.
Grossman, L.S. 1984. Collecting Time-use Data in Third World Rural Communities. Department of Geography, Virginia Polytechnic Institute and State University.
Hames, R.B. 1979. A Comparison of the Efficiencies of the Shotgun and the Bow in Neotropical Forest Hunting. Human Ecol., 7(3): 219-252.
Hames, R.B., and W. Vickers. 1982. Optimal Foraging Theory as a Model to Explain Variability in Amazonian Hunting. Am. Ethnologist, 9(2): 358-378.
Hames, R.B., and W. Vickers, eds. 1983. Adaptive Responses to Native Amazonians. Academic Press, New York.
Hawkes, K., K. Hill, and J.F. O'Connell. 1982. Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay. Am. Ethnologist, 9: 379-398.
Hipsley, E.H., and N.E. Kirk. 1965. Studies of Dietary Intake and the Expenditure of Energy by New Guineans. South Pacific Commission Technical Paper 147. Noumea, New Caledonia.
Johnson, A. 1975. Time Allocation in a Machiguenga Community. Ethnology, 14(2): 301-310.
-. 1978a. In Search of the Affluent Society. Human Nature, 1(9): 50-59.
-. 1978b. Quantification in Cultural Anthropology. Stanford University Press, Stanford, Calif.
Johnson, A., and M. Baksh. 1987. Ecological and Structural Influences on the Proportions Wild Foods in the Diets of Two Machiguenga Communities. In: M. Harris and E.B. Ross, eds., Food and Evolution: Toward a Theory of Human Food Habits, pp. 387-405. Temple University Press, Philadelphia, PA.
Johnson, A., and C. Behrens. 1982. Nutritional Criteria in Machiguenga Food Production Decisions: A Linear Programming Analysis. Human Ecol., 10(2): 167-189.
Johnson, A., and O.R. Johnson. 1988. Time Allocation among the Machiguenga of Shimaa. In: A. Johnson, ea., Cross Cultural Studies in Time Allocation. Vol. I. Human Relations Area Files, New Haven, Conn.
Lee, R.B. 1969. ! Kung Bushmen Subsistence: An Input-Output Analysis. In: A.P. Vayda, ea., Environment and Cultural Behavior, pp. 41-79. Natural History Press, New York.
-. 1979. The ! Kung San. Cambridge University Press, Cambridge.
Lewis, O. 1951. Life in a Mexican Village: Tepotzlan Revisited. University of Illinois Press, Urbana, 111.
Linder, S.B. 1970. The Harried Leisure Class. Columbia University Press, New York.
Lizot, J. 1977. Population, Resources and Warfare among the Yanomami. Man, 12: 497-517.
Lundberg, G., M. Komarovsky, and M.A. McInery. 1934. Leisure: A Suburban Study. Columbia University Press, New York.
McCarthy, F., and M. McArthur. The Food Quest and Time Factor in Aboriginal Economic Life. Melbourne University Press, Melbourne.
McSweeney, B.G. 1979. Collection and Analysis of Data on Rural Women's Time Use. Stud. Fam. Plan., 10: 379-383.
Minge-Klevana, W. 1980. Does Labor Decrease with Industrialization? A Survey of Time Allocation Studies. Curr. Anthropol., 21(3): 279-298.
Montgomery, E., and A. Johnson. 1977. Machiguenga Energy Expenditure. Ecol. Fd. Nutr., 6: 97-105.
Munroe, R., and R.H. Munroe. 1971. Household Density and Infant Care in an East African Society. J. Soc. Psychol., 83: 3-13.
Munroe, R.L., R.H. Munroe, C. Michelson, A. Koel, R. Bolton, and C. Bolton. 1983. Time Allocation in Four Societies. Ethnology, 22(4): 355-370.
Nag, M., B. White, and R.C. Peet. 1978. Anthropological Approach to the Study of Economic Value of Children in Java and Nepal. Curr. Anthropol, 19(2): 293-306.
Neumann, C., and N. Bwibo. 1987. The Collaborative Research Support Program on Food Intake and Human Function, Kenya Project: Final Report. Report submitted to USAID under grant DAN-1309-G-1070-00. School of Public Health, University of California, Los Angeles, Calif.
Norgan, N.G., A. Ferro-Luzzi, and J.V.G.A. Durnin. 1974. The Nutrient and Energy Expenditure of 204 New Guinea Adults. Trans. R. Soc. (Ser. B), 268: 309-348.
Paolisso, M., and R. Sackett. 1985. Traditional Meat Procurement Strategies among the Irapa-Yukpa of the Venezuelan-Colombian Border Area. Res. Econ. Anthropol., 7: 177-199.
Pimentel, D., and M. Pimentel. 1979. Food, Energy and Society. John Wiley & Sons, New York.
Pospisil, L. 1963. Kapauku Papuan Economy. Publications in Anthropology, 67. Yale University, New Haven, Conn.
Rappaport, R.A. 1971. Energy Flow in an Agricultural Society. Sci. Amer., 224(3): 116-132.
Reiss, A. 1959. Rural-Urban and Status Differences in Interpersonal Contacts. Am. J. Soc., 65(2): 182-195.
Richards, A. 1939. Land, Labour and Diet in Northern Rhodesia: An Economic Study of the Bemba Tribe. Oxford University Press, London.
Robinson, J.P. 1977. How Americans Use Time: Sociopsychological Analysis of Everyday Behavior. Praeger, New York.
Rogers B. 1984. Personal communication.
Rogoff B. 1978. Spot Observations: An Introduction and Examination. Quart. Newsl. Inst. Compar. Human Devel. (Rockefeller University), 2(2C): 21-26.
Ruthenberg, H. 1980. Farming Systems in the Tropics. Clarendon Press, Oxford.
Sahlins, M. 1972. Stone Age Economics. Tavistock, London.
Salisbury, R.F. 1962. From Stone to Steel: Economic Consequences of Technological Change in New Guinea. Melbourne University Press, Melbourne.
Sanjek, R. 1972. Ghanaian Networks. Unpublished doctoral dissertation. Columbia University, New York.
Schlossman, N.P. 1986. Work Resumption, Breast-feeding, and Time Allocation of Mothers in Boston, Massachusetts: The First Half-year Postpartum. Unpublished doctoral dissertation. Tufts University School of Nutrition, Medford, Mass.
Sillitoe, P. 1979. Stone versus Steel. Mankind, 12: 151-161.
Sorokin, P., and C. Berger. 1939. Time-budgets of Human Behavior. Harvard University Press, Cambridge, Mass.
Szalai A., P.E. Converse, P. Feldheim, E.K. Schevich, and P.E. Stone. 1972. The Use of Time: Daily Activities of Urban and Suburban Populations in Twelve Countries. Mouton, The Hague.
Tanaka, J. 1980. The San Hunter-gatherers of the Kalahari. Trans. D.W. Hughes. University of Tokyo Press, Tokyo.
Thomas. R.B. 1976. Energy Flow at High Altitude. In: P.T. Baker and M.A. Little, eds., Man in the Andes: A Multidisciplinary Study of High Altitude Quechua. Academy Press, Chicago, 111.
Townsend, W.H. 1969. Stone and Steel Tool Use in a New Guinea Society. Ethnology, 8: 199205.
Tripp, R.B. 1982. Time Allocation in Northern Ghana: An Example of the Random Visit Method. J. Devel. Areas, 16: 391-400.
Van Arsdale, P. 1978. Activity Patterns of Asmat Hunter-gatherers: A Time Budget Analysis. Mankind, 4: 453-460.
Waddell, E. 1972. The Mound Builders. University of Washington Press, Seattle, Wash.
White, B. 1976. Production and Reproduction in a Javanese Village. Agricultural Development Council, Bogor, Indonesia.
Whiting, B., and C.P. Edwards. 1974. A Cross-cultural Analysis of Sex Differences in the Behavior of Children Aged Three and Eleven. In: R.A. LeVine, ea., Culture and Personality, pp. 188-201. Aldine, Chicago.
Yost, J.A., and P.M. Kelley. 1983. Shotguns, Blowguns and Spears: The Analysis of Technological Efficiency. In: R.B. Hames and W.T. Vickers, eds., Adaptive Responses of Native Amazonians, pp. 189-224. Academic Press, New York.
Zeitlin, M.F., N.P. Schlossman, and A.S. Beiser. 1988. Validation of a Telephone Time Use Recall Instrument. In preparation.
Zelkind, I., and J. Sprug. 1974. Time Research: 1172 Studies. Scarecrow Press, Metchuen, N.J.