Methodology

Abstract

We took a set of the psychometric instruments reviewed in Chap. 2 and gave them to a sample of 1,580 small business owners. The majority of businesses had 0–5 employees, had been in operation for more than 3 years, and had $10,000 USD or less in monthly sales. The sample had nearly an equal mix of males and females, typically between the ages of 25 and 54, and over three quarters had at least completed high-school studies. These entrepreneurs were selected because they had loans at one of six participating banks and microfinance institutions across Peru, Kenya, Colombia, and South Africa. Most loans were between $800 and $3,000, though the banks in Kenya & South Africa included clients with loans from $20,000 up to $100,000. Each client was given the series of assessments by representatives of the financial institution. The clients already had their loans for at least six months, and therefore the clients knew that their answers would not directly impact their loan (i.e. a low-stakes setting). This is useful for research, though not ideal for implementation as psychometric tools for credit scoring would be used in practice in a high-stakes setting, where applicants will try to manipulate their responses to get approval of their loan application. In evaluating the contribution of psychometric instruments to better identify high-potential entrepreneurs and direct finance to their ventures, there are two principal outcomes of interest: Business performance (best represented with company profits); and loan repayment (i.e., did the person default). We collected loan repayment history from the financial institution, and profit levels as reported by the entrepreneurs, to compare to responses on the psychometric assessments.