Introduction

In this chapter I deal with methods for collecting behavioral and economic data on productive inputs and outputs. Any attempt at the collection of quantitative data requires that the researcher should ideally have prior knowledge of the full range of economic activities and perform preliminary evaluations of the accuracy of data collection procedures and coding schemes. This will prevent false starts, increase cross-cultural comparability, and lead to a more systematic account of activities. Whenever possible, I encourage researchers to rely on observational data as a kind of gold standard: it produces data amenable to sophisticated quantitative analysis, is crucial for theory testing, is more easily used for cross-cultural comparison than qualitative observations, and reduces the known errors in recall data (see Stange et al. 1998 for an illuminating account of recall errors compared to direct observation). Nevertheless, because of intrusiveness, labor intensiveness, and cultural sensitivities in direct observation, recall data are oftentimes required, but may be integrated into behavioral records. Techniques for reducing recall error (e.g., short time frames) and cross-checking are recommended in such cases.

Conventionally, economic activities may be defined as behaviors whose end result is the production of a material good (e.g., food or artifact), maintenance of an object (e.g., tool repair), or provisioning of a service (e.g., assisting a neighbor building a house). Production activities can be easily characterized along a set of input measures (time, energy, or even risk) but output measures are more complex. In food production outputs are conventionally weight, kilocalories, or macronutrients (carbohydrates, fats, and proteins) and represent common denominators for comparative and analytic purposes. Which of these currencies is appropriate will depend on the research question posed.