Frameworks and strategic decision-making

The experiment randomly assigns participants to one of four conditions that vary in what supplementary material, if any, they receive alongside the business case before generating strategic options.
(1) Control. Participants receive the case study describing the company's situation and a placebo visual. They then proceed directly to the option-generation task with standard instructions to identify the strategic options available to the company. No framework or supplementary material is provided.
(2) Integrated framework. In addition to the case and the visual placebo, participants receive a visual framework that organizes strategic options along three interdependent dimensions. The framework makes all choice dimensions simultaneously visible and organizes them into an integrated structure that supports the assembly of distinct strategic alternatives.
(3) Words-only. In addition to the case and the visual placebo, participants receive a flat, ungrouped list of every term that appears in the integrated framework. This condition holds informational content constant while removing all structural features, isolating the contribution of making interdependent choice dimensions simultaneously visible.
(4) Single-framework (Ansoff). In addition to the case and the visual placebo, participants receive Ansoff's growth matrix, which organizes strategic options along two interacting dimensions (products and markets). This is one of the three component frameworks of the integrated framework.

2026-04-23

2026-06-30

We will observe the following outcomes:
(1) the number of total options that the student brainstormed
(2) the number/share of strategic (rather than operational) options that the student brainstormed
(3) a binary measure of whether the options were mutually exclusive
(4) the number/share of options that suggest that the company continue its current strategy (vs. exit or expand)
(5) binary measures of whether the best-chosen option is to "continue", "exit", or "strategic"

Some of the measures will be coded by two independent coders (or a fine-tuned LLM) using a rubric who are blind to the random assignment, e.g.:
(1) whether the options are strategic or operational
(2) whether the options are mutually exclusive
(3) whether the options suggest that the company continue its current strategy vs. exit or expand

All participants will be asked to articulate a problem statement and their thought process in developing the problem statement.
Survey questions to all participants: (1) Describe how you developed your options; (2) How confident are you in the options you have developed? (3) How would you describe the difficulty of the task? (4) Which specific data from the case did you rely on the most to complete this exercise? (5) Beyond the data provided in the case, what additional data would be helpful for you to complete this exercise?
Additional survey questions to non-control arms: (1) Did you find the framework provided in the instructions helpful in developing your options? (2) Please elaborate on why it was helpful or not helpful.

This project aims to understand how using a framework affects the crafting of strategic options. This project is an extension of prior works (Kim, Hyunjin and Nety Wu. 2022. "Frameworks and strategic decision-making." AEA RCT Registry. October 03. https://doi.org/10.1257/rct.10146-1.0. Kim, Hyunjin and Nety Wu. 2023. "Frameworks and strategic decision-making." AEA RCT Registry. February 21. https://doi.org/10.1257/rct.10961-1.0. Kim, Hyunjin and Nety Wu. 2023. "Frameworks and strategic decision-making." AEA RCT Registry. September 20. https://doi.org/10.1257/rct.12114-1.0. Kim, Hyunjin and Nety Wu. 2024. "Frameworks and strategic decision-making." AEA RCT Registry. February 14. https://doi.org/10.1257/rct.12963-1.0. Kim, Hyunjin and Nety Wu. 2024. "Frameworks and strategic decision-making." AEA RCT Registry. April 02. https://doi.org/10.1257/rct.13245-1.0).

We conduct a four-arm randomized controlled experiment in which participants generate strategic options for a firm described in a business case. The experiment is administered online via Qualtrics, with participants recruited through Prolific. Eligible participants are aged 24–60, hold a master's degree (MA, MSc, or MPhil), are fluent in English, have at least three years of work experience, and currently hold a managerial or senior role.
Upon entering the survey, participants provide informed consent and complete screening questions to verify eligibility. They then read a case study about Rated, an online review platform facing strategic challenges. A minimum reading time of one minute is enforced, followed by comprehension checks to confirm understanding of the case material.
After the case and briefing, participants are randomly assigned to one of four conditions — control, integrated framework, words-only, or Ansoff — which vary the supplementary material provided before the option-generation task. Participants are then asked to generate all strategic options they see as available to the firm, and indicate which option they consider best. They will then engage in reflection questions on how they generate these options, what they see as the main problems, and what data they used. Treatment-group participants additionally answer a reflection question about how the material they received influenced their thinking.
The survey includes explicit instructions not to use AI tools, and copy-paste functionality is disabled in both directions. Prolific's built-in LLM detection and bot-check features are enabled throughout.

Individuals are randomized by Qualtrics randomizer.

Simple randomization at the individual level.

No

N/A

1000 participants

250 participants control, 250 participants with integrated framework, 250 participants with words-only, 250 participants with Ansoff framework

With 250 participants per arm, simple randomization, no clustering, the minimum detectable effect sizes at 80% power are: - Continuous DV (number of strategic options): Cohen's d = 0.31 - Binary DV (mutual exclusivity): a difference of 11.36 percentage points Calculation is based on parameters from previously run experiments. The confirmatory family comprises 5 pre-registered pairwise contrasts: integrated vs. control, integrated vs. words-only, integrated vs. single-framework (Ansoff), control vs. words-only, and control vs. single-framework. The remaining pairwise contrast (words-only vs. single-framework) is exploratory.