Financial Diaries are not journals that families keep or self-reported data of any kind. Rather, researchers visit the families every two weeks to interview them about all financial activity and events in their lives. In these interviews, trained enumerators ask about all income, expenses, and transactions in financial instruments (e.g., saving at home, borrowing from a bank, an installment loan, and so on) held by household members.
In the most recent Diaries projects, researchers have logged information using custom database software applied on tablet computers. Before delving into the details of how much each member spent in the preceding two weeks and on what (i.e., the cash flows interviews), we collect information on household demographics and family history, physical assets, and income sources. The detailed cash flow interviews are guided by this initial information which is used to generate a cash flows interview template unique to each family.
The high-frequency nature of the study and the fact that the researchers must visit families in a relatively close geographic area constrains the sample size to about thirty families per researcher. Additionally, due to the intensive time commitment of the project, participating in an hour-long interview every two weeks, and concerns about attrition, Diaries households are not selected randomly. Households are also selected purposively to provide breadth to the sample. In the Smallholder Financial Diaries, for example, households were selected to gain samples with crop or irrigation diversity.
While the sample size and selection are qualitative, the data at the household level are quantitative in nature and generate thousands of cash flow data points. The Mexico Financial Diaries dataset, for example, contains over 220,000 individual cash flow data points.