Citation
Fiet, J., Nixon, R., Gupta, M., & Patel, P. (2006). Entrepreneurial discovery by the working poor. Journal of Developmental Entrepreneurship, 11(3), 255-273.
Highlights
- The study’s objective was to examine the impact of systematic search training on the number of business ideas participants generated, particularly business ideas inspired by occupational experience and ideas likely to create wealth.
- The authors randomly assigned sample members into a treatment group, which could participate in a series of training sessions on systematic search, and a control group, which could not participate in that training. They then compared the number of ideas each group produced, overall and in categories defined by type of idea.
- The study found that members of the treatment group produced significantly more wealth-creating ideas and developed more ideas based on occupational experience than members of the control group.
- The quality of causal evidence presented in this report is low because it was based on a randomized controlled trial with high attrition and the authors did not adequately control for differences in background characteristics between the groups. This means we are not confident that the estimated effects are attributable to systematic search training; other factors are likely to have contributed.
Features of the Study
The authors invited low-income individuals identified by a civil rights organization in a single city to participate in the study and randomly assigned 110 people into a treatment or control group. Participants in both groups were offered a series of seven 45- to 60-minute sessions over a period of eight weeks. The first six sessions, held once per week in the first six weeks, focused on training. Participants in the treatment group received training in systematic search that directed them to use constrained and focused thinking in developing business ideas. Participants in the control group were directed to remain alert to business ideas that might emerge over the course of their customary activities, but did not receive systematic search training. The two training courses followed similar structures and differed only in their content.
Throughout the program, participants in both groups recorded business ideas in journals that the authors used as data in the analysis, including a description of the idea, how it was developed and in what context, and its value. The final session served as a debriefing in which the authors discussed the data they had collected over the previous six sessions.
Findings
- The study found that the treatment group generated 25 percent more wealth-creating ideas than the control group, a statistically significant finding.
- The study also found that the ideas generated by treatment group members were significantly more likely to spring from occupational experience, 36 percent compared with 15 percent.
Considerations for Interpreting the Findings
Although the authors randomly assigned participants to the treatment or control condition, only 15 of the 110 randomly assigned participants completed the training and were included in the authors’ analysis. When studies have attrition this high, they are not eligible for a high causal evidence rating, and CLEAR reviews them as nonexperimental studies. To receive a moderate causal evidence rating, studies must either establish the equivalence of the two groups being analyzed on their background characteristics or control for those characteristics in the analysis. In this study, the authors neither investigated differences between the treatment and control groups’ racial compositions nor controlled for race in their analysis of the number of wealth-creating ideas. The analysis of ideas emerging from occupational experience included no statistical controls. Therefore, in both analyses, differences in the underlying characteristics of the groups, rather than the training, might explain observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because it was based on a randomized controlled trial with high attrition and the authors did not adequately control for differences in background characteristics between the groups. This means we are not confident that the estimated effects are attributable to systematic search training; other factors are likely to have contributed.