Collecting time use data: Accelerometers or survey methods?
How people spend their time can influence their health, nutrition, and income. However, accurately measuring time use can be very challenging. In this blog, Vidya S.R. Vemireddy, a PhD candidate in applied economics at the Tata-Cornell Institute for Agriculture and Nutrition at Cornell University, compares two methods of measuring time use – 24-hour recall surveys and accelerometers – based on her dissertation field work in Maharashtra, India.
This post originally appeared on the Gender-Nutrition Idea Exchange, a blog maintained by the CGIAR Research Program on Agriculture for Nutrition and Health (A4NH).
Time is a critical input to do any daily activity, from earning income to caring for one’s family. In developing economies, women are often involved in domestic work, such as chores and providing care for their children and the elderly, and income-generating activities like agriculture. These commitments can lead to severe time constraints for women. For example, women may have to trade time for childcare with working an extra hour in agriculture. These constraints and tradeoffs may have negative consequences for women’s and children’s nutrition.
To better understand the relationship between agriculture, nutrition, and time use, we need accurate data on how people spend their time and energy. However, measuring time use is not easy and each of the methods for doing so has advantages and disadvantages. In this blog, based on focus groups and key informant interviews conducted in rural India, I compare the traditional 24-hour recall method with the newer method of using accelerometers to measure energy expenditure.
Traditional survey-based tools – such as the 24-hour recall – are extensively used in research. These techniques are especially suited for developing economy contexts where a majority of the population is poor, illiterate, and lives in remote areas. But, time use surveys alone cannot capture the energy expenditure, which is important for understanding the nutritional consequences of one’s daily activities. One alternative is using accelerometers such as Fitbits, which are gaining popularity as tools to capture energy expenditure and time use in field surveys (see this blog for more details).
Time use or energy expenditure?
Detailed time use surveys are suitable when the objective is to obtain a detailed catalogue time spent on each activity in the past 24 hours. Using these surveys, one can estimate energy expenditure using country guidelines of energy requirements for each activity. Accelerometers capture the intensity of activity in a given time period, from which one can calculate energy expenditure, but cannot be used to quantify time use burden.
Budget friendly or not?
While accelerometers are suitable for small samples in narrowed contexts, they are expensive and require additional team members and training. Time use surveys can be effectively implemented over large samples more easily, requiring only a survey team and training.
What does each method capture?
Time use surveys can capture multitasking through recall of secondary and tertiary activities. The same is not true for accelerometers, as they capture only energy expenditures. The accuracy of a time use survey is heavily dependent on the skill and expertise of the surveyor and the respondent’s ability to recall. On the other hand, accelerometers are technology driven and give inputs as they are received. Still, some still debate their reliability and accuracy, and they require charging and can be damaged in the field.
Which method is easier to implement?
Time use surveys require well trained enumerators, though monitoring these surveys is less cumbersome. Using tablets for collecting time use data makes the surveys more respondent-friendly, less error prone, and more interactive. In comparison, implementing and maintaining accelerometers can be challenging. Monitoring costs may be higher because the devices may get lost or subjects may remove them. The devices need to be charged, which can be complicated if the survey is being conducted in remote locations. Plus, they need to be durable, making them unsuitable for certain research contexts. Most accelerometers devices require Bluetooth-enabled base devices to store information and data services for transferring data.
In conclusion, accelerometers and time use surveys have different objectives, so the choice depends on a project’s purpose, sample size, context, and budget. In my fieldwork, high frequency time use surveys proved more useful to study the time tradeoffs that women face and the implications of agricultural work on nutrition. I found that designing and piloting an accurate and context-specific time use tool and training surveyors were the two main critical and challenging components in ensuring a successful data collection exercise.
Johnston D, Stevano S, Malapit HJ, Hull E, Kadiyala S. 2018. Time use as an explanation for the agri-nutrition disconnect? Evidence from rural areas in low and middle-income countries. Food Policy 76: 8-18.
Vemireddy V. Operationalizing panel data collection. Tata-Cornell Institute for Agriculture and Nutrition Blog.
Video: “Fieldnotes: Vidya Vemireddy”
Video: “Vidya Vemireddy on the Role of Women’s Time in Agriculture-Nutrition Linkages”
By Vidya Vemireddy
Vidya Vemireddy is a Tata-Cornell Scholar and PhD Candidate in Applied Economics and Management in the Dyson School at Cornell University. Her research focuses on women’s time use in agriculture activities and how that may influence nutrition at the household level.
Follower her on twitter @vvidyasachi.