Entrepreneurial domains, theories, and new ideas: evidence from a lab experiment

Last registered on March 19, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Entrepreneurial domains, theories, and new ideas: evidence from a lab experiment

RCT ID

AEARCTR-0013189

Initial registration date

March 18, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

March 19, 2024, 5:26 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Italy

Region

Lombardy

Primary Investigator

Name

Claudia Frosi

Affiliation

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

George Chondrakis

PI Affiliation

ESADE

Contact Investigator

PI Name

Luisa Gagliardi

PI Affiliation

Bocconi University

Contact Investigator

PI Name

Myriam Mariani

PI Affiliation

Bocconi University

Contact Investigator

Additional Trial Information

Status

In development

Start date

2024-03-19

End date

2024-12-31

Keywords

Behavior, Education, Firms & Productivity

Additional Keywords

Entrepreneurial Domains, Entrepreneurial Cognition

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Entrepreneurs are often depicted as boundedly rational and myopic decision-makers who rely on local search and readily available data (Moore, Oesch, and Zietsma, 2007; Simon, 1957). When searching and attempting to define or assess market opportunities, entrepreneurs are guided by their pre-existing experiences. Their knowledge sets coming from individual experiences, backgrounds, values, and beliefs, or domains, define what is salient when framing strategic and entrepreneurial problems (Camuffo, Gambardella, and Pignataro, 2023).

While there is a growing body of literature arguing that theory framing is essential in strategic and entrepreneurial decision making (Felin and Zenger, 2017), little is known about the relationship between entrepreneurial domains, theory framing, and entrepreneurial idea (and solution) generation.

The goal of this project is to understand whether making entrepreneurs aware of the impact exerted by their domains on their entrepreneurial ideas triggers a change in experimentation and search, beyond initial domains.

Since exploration beyond what is known can expand the opportunity set at the expense of increased uncertainty and monetary and non-monetary (i.e., time, effort) costs, it is unclear whether exploration beyond initial domains can be beneficial or not. Given this trade-off, we argue that entrepreneurs who engage in “scientific entrepreneurship” (Zellweger and Zenger, 2023), adopting a theory-based experimentation approach, explore in a more structured way and are able to reduce the burden of expanded search (Camuffo, Cordova, Gambardella, and Spina, 2020).

To this aim, we investigate the impact that domain awareness (which shall trigger increased exploration) has in addition to scientific or theory-based experimentation (which shall reduce the cost of exploration). A companion study finds that domain awareness and theory-based experimentation have a complementary effect on the number of ideas generated by entrepreneurs (Frosi, Chondrakis, Gagliardi, and Mariani, 2024), but it remains agnostic as to what mechanism underlies this increase in idea generation. This study adopts a similar design, and administers a similar manipulation, but aims to uncover how the domains intervention, conditional on the scientific (or theory-based) intervention, impacts theory framing and decision-making. Moreover, this study aims to uncover the mechanisms underlying the main effects of receiving the domains intervention and the scientific (or theory-based) intervention (i.e., more ideas), attempting to understand what drives the increased number of ideas found in the field experiment with entrepreneurs. Lastly, this study aims to replicate the results of the companion study (Frosi, Chondrakis, Gagliardi, and Mariani, 2024) that finds that domain awareness and theory-based experimentation have a complementary effect on the number of ideas generated by entrepreneurs.

This project designs and implements a laboratory experiment following up on a field experiment with early-stage entrepreneurs carried out at ESADE University in Spain (previously pre-registered on AEA under the code AEARCTR-0009325).

In this laboratory experiment, we leverage on a sample of 140 entrepreneurship MSc students (enrolled in the entrepreneurship MSc program at Bocconi University) who have been trained on the theory-based approach to decision-making (the “scientific” or “theory” intervention), and we expose them to a second intervention, the “domains” intervention. In subsequent waves, we plan to involve more students from different courses (up to ~150 more students).

Participants to the study, who have already received the theory intervention, are randomly allocated to two groups for the domains intervention (treated, untreated). We adopt stratified randomization, using the two student cohorts that we leverage on as strata. This way, we ensure that each cohort has a similar split of treated and untreated participants.

The domains intervention, which is at the core of this pre-analysis plan, makes participants aware of their domains (the set of knowledge they derive from their past experiences and background) and of the impact that domains exert on the framing of their entrepreneurial theories.

The experiment is based on a business simulation where participants will be provided with a problem identified by an early-stage start-up, and will be asked to frame a “theory of value” (Agarwal et al., 2023) that maps the identified problem to an entrepreneurial solution. Following the first theory framing attempt, treated and untreated participants are shown different videos:

1) Treated participants will receive an 8.10-minutes training video emphasizing the existence of domain-defining factors, their impact on theory formulation, and the advantages and disadvantages derived from exploring beyond them. Moreover, they will be asked to reflect on their own individual domains and how these may be influencing the theory framing that they are coming up with for the business simulation; the “treatment video” will also talk about scaling a business, and focusing on how start-ups move from theory formulation to execution.

2) Non-treated participants will receive an 8.10-minutes training (placebo) video talking about scaling a business, and focusing on how start-ups move from theory formulation to execution.

After the video, we will collect a number of self-reported and objective measures of performance, and we will compare them across different experimental cells to investigate the interaction between the domains and the theory-based interventions on exploration, theory formulation, and entrepreneurial solution generation.

External Link(s)

Registration Citation

Citation

Chondrakis, George et al. 2024. "Entrepreneurial domains, theories, and new ideas: evidence from a lab experiment." AEA RCT Registry. March 19. https://doi.org/10.1257/rct.13189-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

In terms of experimental flow, participants will first be assessed based on their understanding of theory-based experimentation. At this stage, we will collect some questions on the familiarity of decision and / or industry for the participant.

Next, they will be exposed to the case of an early-stage startup in the process of defining its first idea configuration. We will ask participants to propose a first idea configuration in the form of a theory, indicating what the key elements of this theory are (i.e., attributes) and how they are connected together through causal relationships (i.e., links) to reach an end solution.

After this first attempt, we will ask students to indicate their confidence (from 0 to 100) that the theory they have developed is true (i.e., the likelihood that such configuration will be successful) and their perception on the completeness (from 0 to 100) of the configuration (i.e., whether the configuration includes all relevant elements).

Then, we will ask them to watch an 8.10-minutes training video (different for control and treatment conditions), followed by a manipulation check.

At this point, we will ask them to decide whether they want to update the initial configuration by exploring (and “unlocking”) new attributes or not. Participants will then be shown more attributes (if they have decided to “explore”) or the same attributes as in the first attempt (if they have decided to not to “explore”) and they will create a second theory causal map attempt. In this second attempt, they will be free to include the new attributes or not, and to connect initial and new attributes as they wish. Similarly to attempt 1, participants will be asked to report their level of confidence and of perceived completeness of the theory configuration. We will run this exploration-theory formulation cycle one more time.

Finally, participants will be asked to answer to a final survey, collecting the key outcome variables of this study.

Theory configuration tasks within the experimental flow will be timed to grasp a measure of cognitive effort. In total, the activity will take approximately 1 hour to complete.

Participants will receive the training and solve the provided prompt in Bocconi facilities. We will collect their informed consents prior to the experiment.

Intervention (Hidden)

Intervention Start Date

2024-03-19

Intervention End Date

2024-03-20

Primary Outcomes

Primary Outcomes (end points)

The primary outcomes that we analyse in this study include:

• Level of exploration effort: a variable indicating whether individuals decide to explore (and “unlock” more attributes or not)
• Theory causal map components:
o Number of different theories generated (3 attempts)
o Number and type of attributes used
o Number and type of attribute categories used
o Number of links among attributes
o Breadth of theories (number of vertical links)
o Depth of theories (number of horizontal links)
o Confidence and completeness measures of theory causal maps: 0-100 indicators
o Quality of theory causal maps (vs. benchmark, vs. “true” causal map, external evaluation)
o Level of cognitive effort: self-reported measure (1-7 Likert scale) and platform-derived measure (time spent on theory configuration)

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

In this laboratory experiment, we leverage on a sample of 140 entrepreneurship MSc students (enrolled at the entrepreneurship MSc program at Bocconi University) who have been trained on the theory-based approach to decision-making (the “scientific” or “theory” intervention), and we expose them to a second intervention, the “domains” intervention. In subsequent waves, we plan to involve more students from different courses (up to ~150 more students).

The domains intervention, which is at the core of this pre-registration, makes participants aware of their domains (the set of knowledge they derive from their past experiences and background) and of the impact that domains exert on the framing of their entrepreneurial theories.
Participants to the study, who have already received the theory intervention, are randomly allocated to two groups for the domains intervention (treated, untreated). We adopt stratified randomization, using the two student cohorts that we leverage on as strata. This way, we ensure that each cohort has a similar split of treated and untreated participants.

The activity is presented as one of the requirements of a curricular course on strategic decision-making, and will take place during one of the course’s sessions (after participants have received the “theory” intervention).

The experiment is based on a business simulation where participants will be provided with a problem identified by an early-stage start-up, and will be asked to frame a “theory of value” (Agarwal et al., 2023) that maps the identified problem to an entrepreneurial solution. Following the first theory framing attempt, treated and untreated participants are shown different videos:

1) Treated participants will receive an 8.10-minutes training video emphasizing the existence of domain-defining factors, their impact on theory formulation, and the advantages and disadvantages derived from exploring beyond them. Moreover, they will be asked to reflect on their own individual domains and how these may be influencing the theory framing that they are coming up with for the business simulation;

2) Non-treated entrepreneurs will receive an 8.10-minutes training (placebo) video talking about scaling a business, and focusing on how start-ups move from theory formulation to execution.

After the video, we will run a manipulation check, and we will ask participants in both groups whether they would like to engage in exploration, gaining access to new problem elements (i.e., attributes – Camuffo, Gambardella, and Pignataro, 2023) at the cost of time, or whether they would like to keep relying on initial elements. Participants in both groups will be asked to come up with two more theory attempts, regardless of whether they decided to explore in exploration or not.

Throughout the simulation, we will collect a number of self-reported measures, including participants’ beliefs about their theories (confidence, completeness), whether they have thought of alternative solutions for the problem provided, whether they would have considered additional attributes or not, and whether the activity entails high cognitive effort or not. Moreover, we will collect all theory causal map attempts made by participants, measuring changes from one attempt to the next, as well as measuring a number of indicators on the theory causal maps’ composition (in terms of number of attributes, number of attribute categories, number of links among attributes), on their configuration (breadth and depth of causal maps), as well as on their quality (deviation from an aggregated benchmark, deviation from a “true” solution, external evaluation). Lastly, we will measure exploration decisions made by participants throughout the simulation (e.g., engage in exploration vs. not) as well as the time spent in each theory framing step.

Experimental Design Details

Randomization Method

We adopt stratified randomization, using the two student cohorts that we leverage on as strata.

We run randomization procedures in the office. We first split the complete sample into the 2 cohorts. Then, we perform randomization by cohort. We start by assigning a random number (from 0 to 1) that follows a uniform distribution to each student. We then rank students by their random numbers, and assign the first half of the ranking to the control and the second half to the treatment.

We run this procedure 10 times, using 10 different seeds, and select one of the seeds. We then run balance checks on the selected seed and find no imbalances across the groups.

Randomization Unit

Individual.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

140 individuals.

Sample size: planned number of observations

140 individuals. For some variables (i.e., theory causal maps), 3 data collection points (pre- and post-treatment), resulting in 420 observations. For other variables, 1 data collection point (post-treatment), resulting in 140 observations.

Sample size (or number of clusters) by treatment arms

140 individuals, 50-50 split: 70 individuals in the domains treated group, 70 individuals in the control group.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We justify the size of the sample based on experimental power calculations conducted using G*Power 3.1 (command F-tests, ANOVA: repeated measures, within-between interactions). We assume we detect a small effect size (f=0.15). We assume standard type I and type II errors (α err prob = 0.05 Power (1-β err prob) = 0.95). Given the 2 experimental cells, about 3 observations per individual on key outcome variables (theory causal maps), and making standard assumptions about correlation among repeated measures (0.5) and correction for non-sphericity (ε = 1), we obtain a required total sample size of approximately 116 participants. Given a targeted sample size of 140 students, we aim at a panel dataset composed of 420 observations.

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB