Decision Making
Project Funding Task
Also called: escalation of commitment task, commitment escalation, entrapment, Staw's escalation paradigm
Measure escalation of commitment: participants fund one of two divisions of an invented company, see it do poorly or well, then split a second budget of points, with responsibility and results set by your study.
3-5 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Try it here first, with no account.
- Papers, 2021 to 2025
- 162
- Evidence grade
- C
- Registered clinical studies
- 0
What it measures
Measures escalation of commitment. Participants give 10 points of an invented company's development budget to one of two divisions, or read that a predecessor did, and see five years of results. Then they split 20 more points between the divisions. The main score is the points put back into the division funded first.
Constructs
- Escalation of commitment
- Self-justification
- Decisions after negative feedback
- Resource allocation
Research uses
- Organizational behavior and management research
- Judgment and decision-making research
- Behavioral economics
- Comparison of conditions within a study
How the paradigm works
The project funding task measures escalation of commitment, the tendency to keep putting resources into a losing course of action, above all one the decision maker chose (Staw, 1976). It follows Staw's design with a company and numbers written for this task, and with points, never money. The participant takes the role of planning director at a company with two divisions of about the same size, Lighting and Furniture, whose yearly results have fallen for five years. The board gives 10 points of extra development budget to one division only: in the standard version the participant chooses the division (personal responsibility); in a setting, the participant reads that a predecessor chose it, a division drawn from the session's seed. Five more years of results follow. In the standard version the funded division keeps falling while the other recovers; in a setting, the funded division recovers while the other keeps falling. A check question on splitting a budget comes next, and then the board gives 20 more points, which the participant splits between the two divisions in any way on a bar. Either setting can be drawn at random for each session, so one study can spread its participants over the four conditions. There is no time limit.
More about the paradigm
The main score is the number of points, out of 20, put back into the division funded first. In Staw's study, business students who had chosen a division that then declined allocated an average of 13.07 million of 20 million dollars to it, more than in any of the other three conditions. A meta-analysis of 166 samples estimated the correlation of personal responsibility for the initial decision with escalation at .258 (Sleesman and colleagues, 2012). The responsibility effect replicated in two experiments of Schultze, Pfeiffer and Schulz-Hardt (2012) but not in an online replication of Staw's case with 365 participants, who simply gave more to the division that did well (Białek, Węgrzyn and Meyers, 2021). The condition is saved and scored with each session, so studies compare conditions across participants. Scores describe allocations in an invented case; they are research measures and do not judge a person's decisions at work.
Scores
- Points Reinvestedpoints
- Points of the second budget of 20 given to the division funded first, whether the participant or the predecessor funded it.
- Higher means more commitment to the first choice; compare conditions
- Personal Responsibility0/1
- 1 when the participant made the first decision, 0 when a predecessor did, as the study set it or the session drew it.
- Informational: the condition
- Poor Results0/1
- 1 when the funded division kept falling and the other recovered, 0 when the funded division recovered.
- Informational: the condition
- Check Correct0/1
- 1 when the check question on splitting a budget was answered correctly, 0 when not.
- 1 expected
- First Decision Timems
- Time from the case screen to the choice of a division; empty when a predecessor decided.
- Informational
- Split Timems
- Time from the split screen to confirming the split.
- Informational
- Decisions Under 100 mscount
- Decisions made less than 100 ms after their screen appeared.
- 0 expected
- Decisionscount
- Decisions made: the first decision when the participant made it, and the split.
- Informational
- Planned Decisionscount
- 2 when the participant makes the first decision, 1 when a predecessor did.
- Informational
Use and evidence
162 papers from 2021 to 2025 name this task in their title or abstract, down 28% compared with 2016 to 2020. That ranks it 186 of 632 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 41 |
| 2024 | 30 |
| 2023 | 30 |
| 2022 | 24 |
| 2021 | 37 |
| 2020 | 45 |
| 2019 | 43 |
| 2018 | 49 |
| 2017 | 46 |
| 2016 | 42 |
| 2015 | 39 |
| 2014 | 62 |
| 2013 | 50 |
| 2012 | 37 |
| 2011 | 30 |
| 2010 | 30 |
| 2009 | 26 |
| 2008 | 24 |
| 2007 | 33 |
| 2006 | 17 |
| 2005 | 23 |
| 2004 | 15 |
| 2003 | 15 |
| 2002 | 7 |
| 2001 | 11 |
| 2000 | 10 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 2
- Papers on reliability or validity
- 33
- Registered clinical studies using it as an outcome
- 0
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
Note on the counts: includes field studies of managers and investors.
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.44.0, counted 12 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Case
- An invented company with two divisions of about the same size, Lighting and Furniture, whose yearly results fell for five years. Budgets and results are in points; no screen mentions money.
- First decision
- 10 points of development budget go to one division only. Standard: the participant chooses the division (personal responsibility). Setting: a predecessor chose it, the division drawn from the session's seed.
- Results
- Five more years. Standard: the funded division keeps falling while the other recovers. Setting: the funded division recovers while the other keeps falling.
- Second decision
- 20 more points, split between the two divisions in any way on a bar, with buttons that move 1 point either way. One check question with feedback comes first.
- Conditions
- Either setting can be drawn at random for each session, so one study spreads its participants over the four conditions of Staw's design. The condition is saved on every row and scored.
- Order
- Which division is on the left in the tables, buttons and bar is drawn from the session's seed. No time limit.
- Scores
- Points out of 20 put back into the division funded first, with the two condition codes, the check answer and the decision times.
The company and its numbers are written for this library; Staw's case is not used. In Staw (1976), people who had chosen a division that then declined put the most back into it. Later studies found the responsibility effect in some samples and not in others, so compare conditions within a study. The scores describe allocations in an invented case and do not judge a person's decisions at work.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: the first decision, when the participant made it, and the split of the second budget, in the selected run. No split is right or wrong. A missing split, a wrong check answer and decisions under 200 ms are flagged; with two decisions at most, speed never invalidates under the shared rule.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| First decision | The participant chooses (standard) | The participant chooses (standard), A predecessor chose, Random for each session |
| Results after the first decision | Poor for the funded division (standard) | Poor for the funded division (standard), Good for the funded division, Random for each session |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Project Funding Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Staw BM, 1976), with the standard settings (first decision: The participant chooses; results after the first decision: Poor for the funded division). The main score was points reinvested (points of the second budget of 20 given to the division funded first, whether the participant or the predecessor funded it).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Staw, B. M. (1976). Knee-deep in the big muddy: A study of escalating commitment to a chosen course of action. Organizational Behavior and Human Performance, 16(1), 27-44. https://doi.org/10.1016/0030-5073(76)90005-2The design: a first allocation to one of two divisions, by the participant or another officer, five years of good or poor results, and a second allocation split in any way.
- Sleesman, D. J., Conlon, D. E., McNamara, G., & Miles, J. E. (2012). Cleaning up the big muddy: A meta-analytic review of the determinants of escalation of commitment. Academy of Management Journal, 55(3), 541-562. https://doi.org/10.5465/amj.2010.0696Meta-analysis of the determinants of escalation, personal responsibility among them.
- Schultze, T., Pfeiffer, F., & Schulz-Hardt, S. (2012). Biased information processing in the escalation paradigm: Information search and information evaluation as potential mediators of escalating commitment. Journal of Applied Psychology, 97(1), 16-32. https://doi.org/10.1037/a0024739Replications of the responsibility effect in the two-division paradigm.
- Białek, M., Węgrzyn, M., & Meyers, E. A. (2021). Escalation of commitment is independent of numeracy and cognitive reflection. Failed replication and extension of Staw (1976). Economics and Business Review, 7(2), 5-16. https://doi.org/10.18559/ebr.2021.2.2An online replication of Staw's design with a slider for the second split.
Source papers
- Staw (1976): 240 business students. Mean allocation to the previously chosen division, out of 20 million dollars: 13.07 million with personal responsibility and poor results; 11.08 million with personal responsibility and 8.89 million without it, across results.Paradigm source, Adults (business students)
Questions
How long does the Project Funding Task take?
About 3 to 5 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Points Reinvested, Personal Responsibility, Poor Results, Check Correct, First Decision Time, Split Time, Decisions Under 100 ms, Decisions and Planned Decisions. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Project Funding Task used in research?
162 papers from 2021 to 2025 name it in their title or abstract, down 28% compared with 2016 to 2020. That ranks it 186 of 632 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.44.0, counted 12 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Project Funding Task have?
In the published research, 2 papers report its test-retest reliability and 0 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
First decision (The participant chooses (standard), A predecessor chose, Random for each session) and results after the first decision (Poor for the funded division (standard), Good for the funded division, Random for each session). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese, with its words in the same language. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.