Domains and frames of reference in entrepreneurial decision-making

How to cite

Frosi, Claudia; Chondrakis, George; Gambardella, Alfonso; Camuffo, Arnaldo; Mariani, Myriam; Gagliardi, Luisa. 2022. "Domains and frames of reference in entrepreneurial decision-making." AEA RCT Registry. AEARCTR-0009325. https://doi.org/10.1257/rct.9325-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Claudia Frosi* , George Chondrakis , Alfonso Gambardella , Arnaldo Camuffo , Myriam Mariani , Luisa Gagliardi

* Primary investigator

Registry ID
AEARCTR-0009325
DOI
https://doi.org/10.1257/rct.9325-1.0
First registered
Last updated
Status
In development
Trial dates
2021-12-21 – 2023-03-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/9325

Abstract

Entrepreneurial ventures are essential in driving economic growth through the introduction of innovation via new technologies, products, or services. In the key decision-making process of idea generation, entrepreneurs come up with their own theories about market phenomena and business opportunities. In doing so, they leverage on their own endowed domains, the sum of their experiences, knowledge, social and demographic backgrounds, values, beliefs, and tastes, which act as frames of reference.

While there is a growing body of literature arguing that theory framing is essential in strategic and entrepreneurial decision making, no account studies where entrepreneurial theories come from and in which way they evolve from initial domains.

The goal of this project is to understand the implications of acknowledging the existence of domains and their anchoring and constraining effect on entrepreneurial idea generation. We want to understand under which circumstances exploration past endowed domains can be beneficial in terms of key entrepreneurial outcomes such as idea breadth and performance.

To this aim, the project designs and implements an experiment within an already existing randomized control trial (RCT) taking place at ESADE University in Spain (previously pre-registered on AEA under the code AEARCTR-0008945).

Participants are first randomly allocated to groups for the first intervention (focused on the application of the Scientific Approach to entrepreneurial decision making) and then re-randomized into additional groups for the second intervention (the domain intervention we refer to in this pre-registration).

Intervention

We will conduct an experiment nested within another Randomized Control Trial ("A Scientific Approach to Innovation Management" - RCT ID: AEARCTR-0008945). The main manipulation consists of 6x4h sessions of training including interactive lectures and coaching by qualified mentors/instructors - each working with a subset of the start-up sample. Entrepreneurs are randomly assigned to three different groups. Two groups receive training on two popular approaches to entrepreneurial experimentation: the lean startup method and the scientific approach to decision-making. The last sub-sample is a pure control group, made of entrepreneurs who do not take any training course.

We will implement a 2x3 design, thus running the second manipulation with a randomly allocated subset of entrepreneurs belonging to each of the 3 groups of the above described trial. Treated entrepreneurs will receive a 1-hour training emphasizing the existence of domain-defining factors, their impact on business idea formulation, and the advantages and disadvantages derived from exploring beyond them. Moreover, they will receive additional periodic nudges emphasizing the same topics for the whole duration of the experiment. Non-treated entrepreneurs will receive a 1-hour placebo training consisting of the same talk by another successful entrepreneur.

Participants are first assigned to the 3 different treatment groups pertaining to the primary trial by stratified randomization (since we have English-only and Spanish-only speakers, we randomize within the stratum of language). We then run a second randomization, assigning participants through simple randomization into the 2 groups pertaining to the second manipulation.

Experimental design

We will conduct an experiment nested within the main manipulation, following a 2x3 design, thus running the second manipulation with a randomly allocated subset of entrepreneurs belonging to each of the 3 groups of the main experiment (i.e., treated with the lean startup method, treated with the scientific approach to decision-making, pure control):

1) Treated entrepreneurs will receive a 1-hour training emphasizing the existence of domain-defining factors, their impact on business idea formulation, and the advantages and disadvantages deriving from exploring beyond them. Moreover, they will receive additional periodic nudges emphasizing the same topics for the whole duration of the experiment.

2) Non-treated entrepreneurs will receive a 1-hour placebo training consisting of the same talk by another successful entrepreneur.

In order to avoid contamination across groups receiving different “main treatments”, we will set up independent cohorts and run independent classes. We will have a total of 5 treated (2 lean, 2 scientific, 1 pure control) and of 5 non treated classes (2 lean, 2 scientific, 1 pure control). In order to control in for instructor fixed effects, we will also randomize the allocation of the 2 instructors to each class.

We will observe the impact of the second manipulation over time, as the data collection process will be aligned to that of the main treatment and take place over the course of one year.

Design details

Randomization method
Participants are first assigned to the 3 different treatment groups pertaining to the primary trial by stratified randomization (stratum: language, as some participants speak only Spanish and some speak only English). We then run a second randomization, assigning participants into the 2 groups pertaining to the second manipulation. Everything is performed in the office using STATA.
Randomization unit
Individual
Clusters
150 individuals
Observations
150 individual-month observations
Arms
Primary experiment: 40% scientific treatment, 40% lean treatment, 20% pure control Add-on experiment: 50% treated, 50% non treated
Primary outcomes
Idea breadth (level of applicability of an idea across contexts, number of contexts, distance across contexts) Performance (revenues) Scale (number of employees)
Secondary outcomes
Exploration effort (level of effort incurred to collect feedback on a business idea)

Cited by

No citation data available for this registration yet.

Keywords