Thesis topics
These ideas are intended to inspire your own thesis. They are starting points that you can adapt, combine, or develop further in discussion with me. Choose a theme and explore the briefs; the final research question, scope, and methods will be agreed together.
AI-001Writing objectives for AI agentsControlled task experiment
a) Background and research problem
AI delegation requires people to express objectives, constraints, and success criteria. A promising extension of research on specification costs is to examine whether structured instructions reduce the total effort needed to obtain an acceptable result. The relevant outcome includes preparation and correction time, rather than output speed alone.
b) Research question
Does a structured objective template reduce total delegation effort while maintaining task quality?
c) Data, methods and approach
Compare free-form and template-based instructions on two bounded business tasks. Use the same model version and settings, counterbalance task order, and record preparation time, revisions, and completion quality against a predefined rubric. Use independent assessors where possible and plan participant recruitment and sample size before collection.
d) Expected contribution
Evidence on the trade-off between specifying a task and correcting its output, with a tested instruction template for business use.
AI-002AI disclosure and trust in startup pitchesVignette experiment
a) Background and research problem
Founders can use AI to prepare investor-facing communication. Disclosure of this assistance may alter perceptions of competence and authenticity even when the pitch content is identical. Signaling and perceived authenticity offer possible theoretical perspectives. The opportunity is to isolate the disclosure effect in a financing context.
b) Research question
How does disclosure of AI assistance affect perceived founder credibility and willingness to consider an investment?
c) Data, methods and approach
Randomly show participants the same fictional pitch with either an AI-assistance disclosure or no disclosure. Measure credibility, authenticity, and stated investment interest using established measures where suitable. A narrower study can focus on one pitch; an extended study can add a second context or investor-experience comparison. Interpret hypothetical interest separately from actual investment.
d) Expected contribution
An estimate of the disclosure effect in a controlled pitch setting and implications for how founders communicate AI use.
AI-003Substantive and generic AI language in annual reportsContent analysis
a) Background and research problem
Annual reports may describe concrete AI applications or use general statements about technological opportunity. A focused extension of AI-engagement measurement would examine the specificity of disclosure. Mention frequency alone may not distinguish operational use from aspiration.
b) Research question
How does the specificity of AI disclosure vary across firms within one sector?
c) Data, methods and approach
Select 30–50 firms in one sector and a single reporting year. Obtain annual reports through SEC EDGAR or issuer websites. Code AI-related passages for named applications, deployment stage, resources, and risks. Double-code a subset. An extension can compare two years or test agreement between a dictionary and manual coding.
d) Expected contribution
A transparent coding framework and an assessment of what AI-related language reveals about reported engagement.
AI-004Financial explanations and confidence calibrationRandomized experiment
a) Background and research problem
An explanation can make a financial answer seem convincing without improving its correctness. Confidence calibration, meaning the alignment of confidence with accuracy, provides a measurable outcome for studying AI assistance. The thesis would examine this relationship in a bounded educational decision task.
b) Research question
Does an AI-generated explanation improve the alignment between confidence and accuracy in financial decisions?
c) Data, methods and approach
Create financial tasks with objectively correct answers. Compare information alone with the same information plus fixed AI-generated explanations. Include correct and deliberately flawed explanations in a balanced design, subject to ethics requirements and debriefing. Measure accuracy and confidence, and account for repeated answers within participants. Use simulation or power analysis to set recruitment targets.
d) Expected contribution
Evidence on when explanations support understanding and when they encourage unwarranted confidence.
AI-005Autonomy and human oversight in financial AI productsComparative case analysis
a) Background and research problem
Products described as financial AI agents can differ in what they actually execute, the permissions they hold, and when human approval is required. A comparative study could clarify these distinctions using an explicit classification framework rather than accepting product labels as evidence of autonomy.
b) Research question
How do financial AI products differ in execution autonomy and human oversight?
c) Data, methods and approach
Select 10–15 products using documented inclusion criteria. Code public technical documentation, demonstrations, and terms for execution authority, approval steps, reversibility, and user responsibility. Date all observations and distinguish documented capabilities from promotional claims. An extension can include expert interviews or test a small set of accessible products in simulated conditions.
d) Expected contribution
A documented typology of autonomy and oversight that helps compare products and identify gaps in disclosure.
EF-001How startups explain the use of investment fundsContent analysis
a) Background and research problem
Fundraising materials differ in how clearly they explain what additional capital will finance. Information asymmetry and signaling can help frame a comparison of spending plans, milestones, and contingencies. The empirical opportunity is a transparent assessment within one financing setting.
b) Research question
How specific are startups about the intended use of funds in public fundraising materials?
c) Data, methods and approach
Select 30–50 equity-crowdfunding campaigns from one platform and period. Code spending categories, milestones, timelines, and risk contingencies. Preserve campaign dates and archived versions where possible. An extension can relate disclosure to funding outcomes if comparable outcome data exist, treating the relationship as observational and controlling for obvious campaign differences.
d) Expected contribution
A typology of funding-use disclosure and evidence on the information available to prospective investors.
EF-002Investor rights in tokenized startup financingComparative document analysis
a) Background and research problem
Tokenization can change the representation and transfer of an investment, while the underlying rights depend on offering terms. A financing comparison should therefore examine the documented rights rather than assume that a token creates ownership, liquidity, or governance rights.
b) Research question
How do investor rights and exit arrangements differ between tokenized offerings and conventional equity crowdfunding?
c) Data, methods and approach
Compare 10–15 offerings in one jurisdiction, using investor documents and platform terms. Code cash-flow rights, voting, information rights, transfer restrictions, and exit mechanisms. Match offerings by instrument type where possible and separate legal rights from technical features. Explain the limitations of interpreting documents without observing enforcement in practice.
d) Expected contribution
A comparison matrix that clarifies the economic rights associated with different financing formats.
EF-003Managing financial and sustainability tensions in startupsQualitative interviews
a) Background and research problem
Young firms may face decisions in which financial viability and sustainability ambitions conflict. An extension of research on entrepreneurial learning could examine one specific decision domain, such as supplier selection or product design, and the practices founders use to resolve tensions.
b) Research question
How do founders manage financial and sustainability tensions in supplier selection?
c) Data, methods and approach
Conduct approximately 12–18 interviews with founders in a defined sector, adjusting the final sample to information richness. Ask about concrete decisions, alternatives, and outcomes rather than general attitudes. Use thematic analysis informed by paradox or organizational-learning perspectives. Seek documentary corroboration and contrasting cases where feasible.
d) Expected contribution
A process account of how founders negotiate competing objectives, including conditions under which particular practices are useful.
EF-004Investor feedback and changes to startup pitchesLongitudinal qualitative cases
a) Background and research problem
Pitch materials evolve as founders receive feedback, but the changes may involve presentation, business assumptions, or financing strategy. Studying successive versions offers a way to observe how feedback is interpreted and translated into action, using entrepreneurial learning as an optional lens.
b) Research question
How does investor feedback change the content and underlying assumptions of startup pitches?
c) Data, methods and approach
Recruit four to six startups willing to share at least two pitch versions and associated feedback. Conduct founder interviews and code changes in market claims, financial assumptions, use of funds, and risk discussion. Build within-case timelines before comparing cases. Incubator cooperation may facilitate recruitment but is not assumed.
d) Expected contribution
A grounded account of feedback-driven revision and a distinction between cosmetic changes and substantive learning.
EF-005Local investment opportunities and willingness to investVignette experiment
a) Background and research problem
Local investment opportunities may appeal through geographic proximity and perceived community benefits. These considerations can be examined separately from expected returns and risk. Familiarity or place attachment provides a possible theoretical perspective, linked to research on local retail investment.
b) Research question
Does geographic proximity affect stated investment willingness when financial characteristics are held constant?
c) Data, methods and approach
Randomize otherwise identical fictional investment descriptions as local or distant. Measure allocation intentions, perceived risk, familiarity, and community benefit. Define localness explicitly and account for participants’ location. An extension can add a tokenized-versus-conventional format factor, provided recruitment supports the larger design.
d) Expected contribution
Evidence on geographic framing and investor preferences, with implications for presenting local financing opportunities.
DA-001Stablecoin resilience across market venuesEvent study and time series
a) Background and research problem
A stablecoin can trade at different prices across venues during stress. An extension of empirical stablecoin research would compare the depth and persistence of deviations across a small number of markets. The design should distinguish market-price behavior from issuer redemption conditions.
b) Research question
How do the size and recovery time of stablecoin price deviations differ across venues during selected stress episodes?
c) Data, methods and approach
Select two stablecoins, two or three venues, and two documented episodes. Obtain sufficiently frequent historical prices and volumes; verify coverage before fixing the question. Calculate peg deviations, duration, and recovery measures with consistent quote currencies and timestamps. Compare alternative sampling intervals and document missing or stale observations.
d) Expected contribution
A venue-level comparison of observed resilience and evidence on the sensitivity of stability measures to data choices.
DA-002Concentration of voting in decentralized organizationsGovernance data analysis
a) Background and research problem
Formal voting rights and observed participation can diverge in token-based organizations. A manageable thesis can distinguish concentration among active voters from concentration of all eligible voting power. Comparing proposal types may reveal where participation differs within the same governance system.
b) Research question
How concentrated is observed voting power, and does concentration vary by proposal type?
c) Data, methods and approach
Collect one year of proposal and vote records for two or three Snapshot spaces after testing export access. Classify proposals and compute concentration and participation counts. A narrower study can focus on one space; an extended study can compare governance designs. Calculate turnout only where a defensible eligible-power denominator is available. Wallets do not necessarily represent distinct people.
d) Expected contribution
A reproducible assessment of voting concentration and a comparison of participation across decisions.
DA-003Liquidity around crypto asset exchange listingsEvent study
a) Background and research problem
An exchange listing may change trading access, but listing-related price movements do not necessarily imply sustained improvements in liquidity. An extension of cross-listing research can focus on liquidity persistence and the difference between announcement and trading dates.
b) Research question
How does observed liquidity evolve before and after additional exchange listings?
c) Data, methods and approach
Select 20–40 listings from one exchange and period with verifiable timestamps. Use daily prices and volumes for a longer window, or spreads where accessible. Choose measures supported by the data and control for market movements. Exclude or flag overlapping announcements and compare short and longer windows. Listing selection prevents a simple causal interpretation.
d) Expected contribution
Evidence on the persistence of liquidity changes and a replicable event dataset.
DA-004Fan token promises and delivered participationComparative case analysis
a) Background and research problem
Fan tokens may be promoted through participation, exclusive benefits, or financial appeal. A useful extension of fan-token research is to compare publicly stated benefits with documented activities and their continuity. The thesis would examine a defined period rather than infer fan satisfaction from activity counts.
b) Research question
How closely do documented fan-token activities correspond to the benefits promised by issuing clubs?
c) Data, methods and approach
Select six to ten clubs and a fixed 12-month period. Code launch materials, benefit descriptions, published polls, and documented events. Record missing archives and distinguish consequential decisions from symbolic activities. An extension can interview fans or club staff if access is secured.
d) Expected contribution
A framework for assessing promise delivery and evidence on the forms of participation offered to supporters.
DA-005Transaction fees and activity across Ethereum applicationsReplication and time series
a) Background and research problem
Research on Ethereum fees and application activity provides a basis for replication in a later period or a narrower application group. An updated study could examine whether relationships are stable when transaction activity shifts between environments. The comparison must account for changes in application classification.
b) Research question
How stable is the relationship between Ethereum fees and activity in a selected application category across two periods?
c) Data, methods and approach
Select one category, such as stablecoins or decentralized exchanges, and reproduce defensible activity and fee measures from documented blockchain queries. Compare two pre-specified periods with regression or time-series methods appropriate to stationarity and lag structure. Address common trends and activity migration. Distinguish predictive relationships from structural causality.
d) Expected contribution
A transparent replication and evidence on the temporal stability of the selected relationship.
IB-001Investment communities and perceived credibilityVignette experiment
a) Background and research problem
Investors encounter recommendations accompanied by visible community approval. Social proof can influence credibility even when the information itself is unchanged. A controlled design can isolate the role of displayed approval from recommendation content and observed market performance.
b) Research question
How does displayed community approval affect the credibility of an investment recommendation?
c) Data, methods and approach
Randomize a fictional investment post to display high, low, or no visible approval. Hold text and author details constant. Measure perceived credibility, willingness to investigate, and stated investment interest. Use neutral assets and a manipulation check. An extension can examine financial literacy or compare community contexts with sufficient statistical power.
d) Expected contribution
Evidence on visible approval as an information cue and its relationship with perceived credibility.
IB-002Risk disclosure readability and investor comprehensionRandomized comprehension experiment
a) Background and research problem
Risk statements can differ in linguistic complexity even when they communicate the same facts. A thesis can examine whether simpler wording improves comprehension and whether perceived risk changes alongside understanding. Financial literacy provides a possible moderator rather than a substitute for measuring comprehension.
b) Research question
Does plain-language risk disclosure improve comprehension of a digital investment product?
c) Data, methods and approach
Create two versions of a fictional product disclosure with equivalent facts and different readability. Randomly assign participants and ask objectively scored comprehension questions. Measure perceived risk and literacy using suitable items. Pilot semantic equivalence and difficulty. An extension can test one interaction or a delayed comprehension measure.
d) Expected contribution
Evidence on disclosure wording and an evaluated example of clearer investor communication.
IB-003Financial literacy and stablecoin risk understandingSurvey
a) Background and research problem
Understanding a stablecoin requires distinguishing price stability, issuer exposure, redemption arrangements, and custody. An extension of adoption research can focus on objective risk knowledge and its relationship with financial and digital literacy, rather than self-reported familiarity alone.
b) Research question
How are financial and digital literacy associated with understanding stablecoin risks?
c) Data, methods and approach
Conduct a bounded survey of adults with and without digital-asset experience. Develop neutral scenario questions covering distinct risk dimensions and pilot them with experts. Analyse knowledge scores and associations with literacy and experience. Report recruitment limitations and avoid population-wide prevalence claims from a convenience sample.
d) Expected contribution
A tested knowledge instrument and evidence on which risk dimensions participants understand least well.
IB-004Utility and speculation in digital collectible ownershipInterviews or survey
a) Background and research problem
Digital collectibles may appeal through practical access, identity, community, and anticipated resale value. A focused extension of ownership-motivation research could examine one collection category and how participants explain changes in their motivations over time.
b) Research question
How do owners describe changes in the balance between utility and speculation in their digital collectible holdings?
c) Data, methods and approach
Recruit 12–20 current and former owners within one defined category for interviews. Use concrete purchase, holding, and exit episodes to support thematic analysis. Include contrasting experiences and acknowledge recall and selection bias. A survey alternative requires a defensible motivation scale and a feasible recruitment route agreed in advance.
d) Expected contribution
An account of changing ownership motivations and a refinement of motivation categories within a specific setting.
IB-005Experience and reactions to investment lossesVignette survey
a) Background and research problem
People may respond differently to losses depending on experience, prior expectations, and willingness to accept risk. Standardized scenarios offer a manageable way to compare reported responses while holding the loss and product information constant. Loss aversion or the disposition effect can guide the interpretation.
b) Research question
How is investment experience associated with intended responses to a standardized investment loss?
c) Data, methods and approach
Present identical hypothetical loss scenarios to participants with different experience levels. Record intended selling, holding, or additional investment, alongside risk tolerance and literacy. Use clear recruitment criteria and multivariable analysis appropriate to the outcome. An extension can vary the product label experimentally. Experience itself remains observational.
d) Expected contribution
Evidence on reported loss responses and a distinction between product-framing effects and experience-related associations.
TS-001Specificity of corporate biodiversity commitmentsContent analysis
a) Background and research problem
Corporate biodiversity statements can vary in the precision of their targets, implementation plans, and progress indicators. A thesis can assess this variation in a defined sector without assuming that reporting specificity proves ecological performance. Accountability provides a possible organizing perspective.
b) Research question
How specific and measurable are biodiversity commitments reported by firms in one sector?
c) Data, methods and approach
Select 25–40 firms and one reporting year. Collect sustainability reports and code baselines, target dates, geographic scope, indicators, and responsibilities. Double-code a subset and document absent disclosures. An extension can compare years or sectors. Use a consistent reporting boundary and identify changes in report availability.
d) Expected contribution
A reproducible disclosure framework and evidence on the transparency of corporate commitments.
TS-002Evidence of impact in blockchain energy projectsStructured evidence review
a) Background and research problem
Energy projects can describe expected blockchain benefits without providing comparable outcome evidence. An extension of blockchain-and-energy research could examine the evidence supporting one claimed benefit, such as lower transaction costs in peer-to-peer trading, and the quality of comparison baselines.
b) Research question
What empirical evidence supports claims that blockchain reduces transaction costs in peer-to-peer energy trading?
c) Data, methods and approach
Conduct a structured search with explicit dates, sources, and inclusion rules. Extract project scale, operating context, cost definition, comparator, and study design. Separate simulations, pilots, and operational evaluations. Assess evidence quality and synthesize findings narratively; use quantitative pooling only if measures and contexts are sufficiently comparable.
d) Expected contribution
An evidence map that distinguishes demonstrated outcomes from projected benefits and identifies measurement needs.
TS-003Digital identity and willingness to share personal dataVignette experiment
a) Background and research problem
Digital identity services differ in the information requested and the control users are offered. Privacy calculus can frame how people weigh convenience against data-sharing concerns. A narrow experiment can assess preferences for one service context rather than compare entire identity architectures.
b) Research question
How does offering user control over data sharing affect willingness to use a digital identity service?
c) Data, methods and approach
Use a fictional municipal mobility service with identical convenience and security descriptions. Randomize granular sharing control versus a bundled sharing arrangement. Measure adoption intention, perceived control, and privacy concern. An extension can vary the amount of requested data if sample size permits. Measure understanding of the scenario.
d) Expected contribution
Evidence on the role of perceived control in stated adoption preferences for digital identity services.
TS-004Research connections between AI and entrepreneurial financeBibliometric mapping and focused review
a) Background and research problem
AI and entrepreneurial finance span several research communities. A thesis can map how studies connect financing decisions, startup activity, and AI, then use close reading to interpret those connections. Bibliometric clusters should be treated as analytical results requiring validation, rather than self-explanatory research gaps.
b) Research question
Which research streams connect AI and entrepreneurial finance, and how much do they draw on shared literature?
c) Data, methods and approach
Build a documented corpus using OpenAlex or an institutionally available database. Pilot search precision and recall, clean records, and apply bibliographic coupling or co-citation where coverage supports it. Validate clusters through reading and sensitivity checks. Set a publication cutoff and account for citation age and database coverage.
d) Expected contribution
A reproducible research map and a focused agenda grounded in the content of the identified streams.
TS-005How consumers identify regulated online gambling providersInterface content analysis
a) Background and research problem
Consumers may need to determine whether an online gambling provider is licensed in their jurisdiction. A bounded study can assess the visibility and consistency of licensing information without assuming that displayed claims are accurate or that visibility alone changes consumer choices.
b) Research question
How clearly do licensed providers communicate licensing status and consumer-protection information?
c) Data, methods and approach
Select one jurisdiction and a dated sample drawn from its official licensed-provider list. Review publicly accessible pages without registering, depositing, or gambling. Code visibility, links to official verification, and consumer-protection information. An extension can add a non-gambling identification task with adults, subject to ethics review. Recheck the official list at collection.
d) Expected contribution
A transparent communication audit and practical suggestions for making licensing information easier to verify.
AI-006Market reactions to corporate AI partnership announcementsEvent study
a) Background and research problem
An AI partnership announcement may convey new capabilities or simply associate a firm with a prominent technology provider. Signaling theory motivates examining which disclosed commitments distinguish announcements. A focused study can compare reactions to concrete implementation plans and general expressions of intent without assuming that an announcement proves operational improvement.
b) Research question
Do stock-market reactions differ between AI partnership announcements with concrete commitments and those with general statements?
c) Data, methods and approach
Assemble dated announcements by listed firms in one market and period. Code implementation scope before inspecting returns. Estimate abnormal returns using a pre-event market model and short event windows. Align announcements with trading hours, flag concurrent earnings or acquisitions, and adjust inference for clustered dates. Pilot historical price access and the number of usable events.
d) Expected contribution
A reproducible announcement dataset and evidence on how markets respond to different forms of AI commitment.
DA-006Security incidents and spillovers across crypto assetsEvent study
a) Background and research problem
A publicly disclosed protocol exploit can affect the associated token and other assets perceived as connected. A bounded study can distinguish direct exposure from broad sector repricing. The key challenge is defining connections before an incident, rather than interpreting whichever assets subsequently fall as exposed.
b) Research question
Are abnormal returns around protocol exploits larger for assets with documented pre-incident connections?
c) Data, methods and approach
Select a consistent class of incidents with verifiable first-public timestamps and tradable tokens. Define exposure using pre-event protocol links or categories. Estimate abnormal returns against a crypto-market benchmark, excluding mechanically affected benchmark components where material. Compare alternative windows and account for events that share dates or assets. Document missing and delisted tokens.
d) Expected contribution
Evidence on the reach of incident-related repricing and a transparent framework for measuring exposure.
DA-007Token unlocks and anticipated selling pressureEvent study
a) Background and research problem
Token releases can be scheduled well before they occur, so the release date need not represent new information. Studying returns, turnover, and liquidity before and after unlocks can reveal the timing of adjustment. The research opportunity is to separate scheduled supply changes from unexpected revisions to the schedule.
b) Research question
How do returns and trading activity evolve around pre-announced token unlocks, and does unlock size matter?
c) Data, methods and approach
Build a sample with archived schedules, actual release dates, beneficiary categories, and circulating supply measured before each event. Analyse market-adjusted returns and turnover across anticipation and release windows. Distinguish schedule changes, exclude overlapping token events, and cluster repeated observations by token and date where feasible. Confirm archival access before sampling.
d) Expected contribution
A comparison of anticipation and release patterns that clarifies how event timing changes the interpretation of token-market evidence.
EN-006Corporate venture investments as signals to public marketsEvent study
a) Background and research problem
A corporate venture investment can communicate access to new capabilities, strategic experimentation, or a costly distraction. The public investor is the observable firm, while the startup is usually private. Strategic fit provides a testable explanation for variation in announcement responses without requiring private startup valuations.
b) Research question
How is strategic fit between a listed corporate investor and a startup associated with the investor’s announcement return?
c) Data, methods and approach
Collect corporate venture investment announcements from investor releases over a bounded period. Code technological or market fit using descriptions available at announcement. Estimate abnormal returns for the listed parent and compare fit categories, investment stages, and disclosed commitments. Screen simultaneous material news and repeated investments; use conservative inference if the sample is small.
d) Expected contribution
Evidence on how investors evaluate corporate experimentation and a replicable measure of disclosed strategic fit.
TS-006Cybersecurity disclosures and the value of accountabilityEvent study and disclosure analysis
a) Background and research problem
Cybersecurity incidents differ in severity and in how clearly firms describe responsibility, remediation, and affected operations. A thesis can examine whether disclosure characteristics accompany different market responses. Such an association cannot by itself establish that communication caused the response, because severity also shapes what firms disclose.
b) Research question
How are accountability and remediation details in initial incident disclosures associated with abnormal stock returns?
c) Data, methods and approach
Select listed firms within one disclosure regime and period. Verify initial disclosure times from filings and releases. Independently code incident scope, responsibility language, and remediation detail, then estimate short-window abnormal returns. Control for observable severity, flag unrelated news, and test alternative coding rules. Treat unreported severity as an interpretation limit.
d) Expected contribution
A disclosure coding framework and evidence connecting incident communication with market valuation responses.
SI-001Green bond announcements and the credibility of environmental commitmentsEvent study
a) Background and research problem
A green bond announcement combines financing news with environmental claims. External review and specific use-of-proceeds commitments may shape credibility, but issuers choosing these features may already differ. A focused event study can examine valuation responses while separating new announcements from subsequent issuance dates.
b) Research question
Do equity-market reactions to first green bond announcements vary with external review and use-of-proceeds specificity?
c) Data, methods and approach
Identify first announcements by listed issuers in one market, record external-review status at announcement, and code planned uses of proceeds. Estimate market-model abnormal equity returns. Compare issuers cautiously, accounting for financing size, firm characteristics, concurrent news, and clustered dates. Pilot document and price coverage before defining the full sample.
d) Expected contribution
Evidence on which observable financing commitments accompany stronger market responses, with limits on causal interpretation made explicit.
SI-002Greenwashing allegations and reputational spilloversEvent study
a) Background and research problem
A documented greenwashing allegation may challenge one firm’s credibility and influence perceptions of similar firms. Legitimacy theory suggests examining whether scrutiny remains firm-specific or reaches an industry. A thesis should distinguish an allegation, a regulatory finding, and a company response rather than pool them as equivalent events.
b) Research question
Do public greenwashing allegations generate abnormal returns for industry peers as well as the targeted firm?
c) Data, methods and approach
Choose one event type and use dated regulator releases or a consistently defined source of allegations. Define peers using pre-event industry information. Estimate short-window abnormal returns for targets and peer portfolios, checking overlapping events, common announcements, and alternative peer definitions. Report allegations as allegations and record later outcomes separately.
d) Expected contribution
Evidence on the scope of reputational spillovers and a transparent distinction between event types.
EN-007Which early milestones predict startup survival?Longitudinal cohort and survival analysis
a) Background and research problem
Startup success is often represented by funding even though survival, revenue growth, and acquisition capture different outcomes. A cohort study can use a clearly defined endpoint and information observable early in a venture’s life. Survival analysis allows firms still operating at the observation cutoff to remain in the analysis.
b) Research question
Which milestones observed in a startup’s first year are associated with subsequent survival?
c) Data, methods and approach
Define a founding cohort in one ecosystem using a registry or accessible venture database. Verify operating status and dates from multiple sources. Code milestones from information available during the first year, then model later closure with right censoring. Distinguish acquisition from failure and test how alternative status definitions affect results. Report missing histories and selection coverage.
d) Expected contribution
A time-aware assessment of early indicators and a clearer operational definition of startup survival.
EN-008Founder experience and different forms of startup successArchival cohort comparison
a) Background and research problem
Prior founding experience may bring knowledge and networks, but repeated entrepreneurship also reflects selection into another venture. A useful study can compare several observable outcomes rather than treat fundraising as universal success. Human capital theory provides a basis for distinguishing industry experience from experience of founding itself.
b) Research question
How are prior founding and industry experience associated with survival and follow-on funding in a defined startup cohort?
c) Data, methods and approach
Build a cohort with founder biographies recorded near entry. Define experience categories before observing outcomes and use a common follow-up period. Model survival and follow-on funding separately with sector, cohort, and initial-resource controls. Audit biography missingness and run sensitivity checks. Matching may improve comparability but does not remove unobserved founder selection.
d) Expected contribution
Evidence on whether experience relates similarly to continuity and external financing, with transparent limits on causal interpretation.
EN-009Founding-team complementarity and venture progressTeam-level archival study
a) Background and research problem
Diverse founding teams may combine complementary skills while facing coordination costs. Broad demographic diversity measures do not directly capture functional complementarity. A focused study can examine whether combinations of technical, commercial, and sector experience relate to achieving a specified operational milestone.
b) Research question
Are complementary functional backgrounds among founders associated with reaching a first commercial product milestone?
c) Data, methods and approach
Select startups founded in a narrow sector and period. Code founder experience from dated biographies and define complementarity before analysing outcomes. Verify a first commercial launch independently of funding announcements. Compare teams while accounting for team size, founding year, and initial resources; test alternative experience coding. Avoid inferring individual traits from names or images.
d) Expected contribution
A reproducible functional-complementarity measure and evidence on its association with early venture execution.
EN-010Accelerator participation and subsequent venture developmentMatched cohort study
a) Background and research problem
Accelerators offer mentoring, visibility, and contacts, but admitted startups may already have stronger prospects. Comparing participants with all other startups would mix programme selection with possible programme benefits. A focused study can assess observable differences and clearly define what the available comparison group supports.
b) Research question
How do follow-on funding and survival differ between accelerator participants and observably similar non-participants?
c) Data, methods and approach
Study one programme and admission cohort with a common follow-up horizon. Prefer eligible unsuccessful applicants if access is agreed; otherwise construct a matched comparison using pre-programme attributes. Document balance, exclusions, and missing outcomes. Use a credible admission threshold only if its assignment rule and data support such a design. State remaining selection limits.
d) Expected contribution
A programme-specific outcome comparison and an assessment of how comparison-group choice changes conclusions.
EN-011When founders decide to pivotInterview study
a) Background and research problem
Pivot decisions combine market feedback, resource constraints, and founder commitments. Retrospective success stories can make these decisions look more deliberate than they were. Interviewing founders about concrete decision episodes, including unsuccessful changes, can reveal how they interpret feedback and decide whether to persist.
b) Research question
How do founders distinguish feedback that calls for a pivot from feedback that supports continued experimentation?
c) Data, methods and approach
Recruit founders from one sector with contrasting pivot outcomes. Conduct semi-structured interviews around dated decisions and, with consent, triangulate accounts with pitch versions or product records. Use abductive thematic analysis and actively seek contradictory cases. Begin with a feasible recruitment target and adjust for information richness; protect commercially sensitive accounts.
d) Expected contribution
A process account of pivot decisions and a framework linking feedback interpretation to changes in venture direction.
EN-012Learning from startup shutdownsComparative case study
a) Background and research problem
Public shutdown accounts can emphasize external conditions or founder decisions and may serve reputational purposes. Combining these accounts with independent evidence can reveal recurring explanations without treating founder narratives as objective diagnoses. The study can also examine what founders say they would change in a later venture.
b) Research question
How do founders explain startup closure, and which explanations are supported or complicated by contemporaneous evidence?
c) Data, methods and approach
Select a bounded set of closures within one sector. Collect founder postmortems, dated product and funding records, and optional follow-up interviews. Code internal and external attributions, compare narratives across cases, and identify missing evidence. Include less-publicized closures where feasible and avoid estimating population failure causes from a voluntary narrative sample.
d) Expected contribution
A typology of closure explanations and insight into the relationship between entrepreneurial learning and reputation management.
EN-013Bootstrapping and venture growth ambitionsComparative interview study
a) Background and research problem
Founders may choose external finance to accelerate growth or avoid it to retain autonomy. Observed funding status alone does not reveal whether a venture is constrained or deliberately self-financed. A comparative qualitative study can examine financing choices in relation to founders’ definitions of success.
b) Research question
How do founders reconcile growth ambitions, autonomy, and financing constraints when choosing to bootstrap?
c) Data, methods and approach
Interview founders of self-financed and externally financed ventures in one industry and similar age range. Reconstruct concrete financing decisions, alternatives considered, and changes in goals. Analyse within-case timelines and cross-case contrasts; include founders who switched financing strategies. Treat reported motivations as accounts to triangulate, not definitive evidence of optimal choices.
d) Expected contribution
A grounded account of financing choices that distinguishes voluntary bootstrapping from limited access to capital.
EN-014Can simple models predict follow-on startup funding?Predictive modelling
a) Background and research problem
Predictions of venture success can be inflated by using information recorded after the predicted outcome or by testing on similar firms from the same period. A bounded prediction study can focus on a verifiable funding milestone and evaluate whether complex models improve on simple, interpretable baselines.
b) Research question
How accurately can information available at a startup’s first funding round predict a subsequent round within a fixed horizon?
c) Data, methods and approach
Use an accessible or licensed venture dataset with dated attributes and funding records. Set a common prediction date, exclude future information, and retain firms with sufficient follow-up. Split training and testing chronologically; compare logistic regression with one more flexible model. Report calibration, precision-recall performance, class imbalance, and missing-data handling alongside baseline predictions.
d) Expected contribution
A reproducible forecasting benchmark and evidence on the incremental value of model complexity for a specific funding outcome.
PP-001Risk allocation and renegotiation in public-private partnershipsComparative contract and case analysis
a) Background and research problem
Long-term public-private partnerships must allocate construction, demand, and operating risks while allowing for changing conditions. Contractual allocation may differ from who ultimately absorbs losses. A bounded comparison can examine how initially specified risks feature in later renegotiation without assuming that every amendment represents project failure.
b) Research question
How do initial risk-allocation clauses relate to the issues addressed in later PPP renegotiations?
c) Data, methods and approach
Select a small set of comparable transport or municipal-service PPPs in one jurisdiction with accessible contracts and amendments. Code risk ownership, adjustment triggers, and documented changes; triangulate with audit reports and council records. Build case timelines and compare alternative explanations such as shocks and procurement conditions. Separate original commitments from ex-post assessments.
d) Expected contribution
A structured comparison of contractual risk allocation and subsequent adaptation, with implications for contract transparency and design.
PP-002Citizen acceptance of public-private service deliveryFactorial vignette experiment
a) Background and research problem
A service delivered through a PPP can be judged on price and quality as well as accountability and private profit. A controlled scenario can examine whether oversight arrangements change acceptance when service performance is held constant. The proposal focuses on perceived legitimacy rather than the actual superiority of an ownership model.
b) Research question
Does independent oversight increase citizen acceptance of a PPP, and does acceptance depend on the service’s public importance?
c) Data, methods and approach
Present a fictional long-term municipal service partnership with identical cost, quality, and risk information. Randomize oversight arrangements and, if recruitment permits, one service context. Measure acceptance, accountability perceptions, and understanding of the scenario. Pilot realistic wording, predefine the primary comparison, and choose sample size for the intended effects.
d) Expected contribution
Evidence on which governance features influence acceptance and a distinction between performance expectations and procedural concerns.
PP-003Why smaller firms struggle to participate in PPP consortiaInterview study
a) Background and research problem
PPP procurement can require financial capacity, long commitments, and coordination among multiple firms. These demands may shape how smaller suppliers enter a consortium or remain subcontractors. A focused interview study can examine barriers and negotiation practices from both public-procurement and business perspectives.
b) Research question
Which procurement and consortium practices enable or constrain smaller firms’ participation in PPP projects?
c) Data, methods and approach
Choose one infrastructure or public-service sector. Interview procurement officers, consortium leads, and smaller firms, including unsuccessful applicants where accessible. Use a common guide around a recent procurement episode and triangulate with tender requirements. Compare explanations across roles through thematic analysis; distinguish formal eligibility rules from informal partner-selection practices.
d) Expected contribution
An account of participation barriers and practical options for improving access while preserving project-delivery requirements.
PP-004Do PPP performance metrics capture public value?Systematic literature review and document comparison
a) Background and research problem
PPP performance can be assessed through delivery time, fiscal cost, service quality, accessibility, or distributional outcomes. Different measures can produce different judgments of success. A focused review can examine which dimensions are actually evaluated and whether project reporting reflects the outcomes emphasized in research.
b) Research question
How is public value measured in empirical PPP evaluations, and which dimensions are absent from project performance reports?
c) Data, methods and approach
Limit the review to one service sector and a defined period. Document databases, searches, screening decisions, and quality appraisal; extract outcome definitions, comparators, and stakeholder perspectives. Compare the resulting framework with a small purposive set of accessible audit or performance reports. Keep academic evidence and illustrative project documents analytically distinct.
d) Expected contribution
A measurement framework showing how PPP evaluations represent public value and where reporting could be strengthened.
ML-001Moral framing and acceptance of carbon-offset servicesVignette experiment
a) Background and research problem
Carbon-offset services can be presented as responsible action or criticized as permission to pollute. This makes them a useful setting for studying how moral interpretations shape market acceptance. A controlled study can hold the service and environmental evidence constant while varying the meaning attached to purchase.
b) Research question
How do responsibility and compensation frames affect the perceived acceptability of purchasing carbon offsets?
c) Data, methods and approach
Randomly assign adults to equivalent fictional service descriptions with different moral frames and a neutral comparison. Measure perceived legitimacy and purchase intention, with prior environmental attitudes collected before exposure. Pilot semantic equivalence and distinguish moral approval from perceived environmental effectiveness. Any subgroup comparison should be planned and adequately powered.
d) Expected contribution
Evidence on how moral framing changes evaluations of an otherwise identical market offering.
ML-002How gambling entrepreneurs justify their businessInterview study
a) Background and research problem
Entrepreneurs in gambling-related businesses may face conflicting expectations from customers, employees, regulators, and communities. Interviews can investigate how founders address these audiences and where they draw boundaries around acceptable activity. The aim is to understand justification practices, without assuming that a participant’s account demonstrates social responsibility.
b) Research question
How do gambling-sector entrepreneurs explain and defend the social acceptability of their business to different audiences?
c) Data, methods and approach
Recruit adult founders or senior managers within one legally defined market, including varied business models. Conduct interviews around specific stakeholder challenges and compare accounts with public communications. Use abductive coding for justifications, audience differences, and tensions. Protect identities where disclosure could harm participants and distinguish formal compliance claims from broader moral arguments.
d) Expected contribution
A grounded account of moral boundary work and audience-specific legitimacy strategies in a contested market.
ML-003Dual-use startups and contested investor identitiesComparative interview study
a) Background and research problem
Technologies with civilian and defense applications can attract audiences with opposing views of the same use. Founders may need to explain how these uses fit their venture’s identity. A qualitative study can examine how financing conversations influence public positioning and internal boundaries around customers or applications.
b) Research question
How do dual-use startup founders negotiate identity and acceptable applications when engaging different investors?
c) Data, methods and approach
Interview founders and investors in a defined technology field, using recent financing or customer-selection episodes. Compare civilian, defense-oriented, and mixed positioning with public pitch materials. Analyse recurring tensions and contrasting cases; avoid collecting classified or commercially restricted information. Include ventures that declined an application or investor where accessible.
d) Expected contribution
Insight into how moral disagreement shapes venture positioning, financing relationships, and self-imposed market boundaries.
ML-004When fintech ventures do not fit a familiar categoryFactorial experiment
a) Background and research problem
A venture combining familiar categories may appear innovative but also be harder to understand. Category incongruence concerns a mismatch with audience expectations; it is not equivalent to poor quality. A pitch experiment can test whether an explanation connecting the categories helps audiences evaluate an otherwise identical offering.
b) Research question
Does explaining the connection between unfamiliar category combinations improve venture comprehensibility and investment interest?
c) Data, methods and approach
Create fictional fintech pitches with pretested category combinations and matched functional information. Randomize category fit and the presence of an integrating explanation. Measure comprehension, perceived novelty, and investment interest separately. Balance pitch length, predefine the key interaction, and recruit an audience suited to the claim; general-public responses do not represent professional investment decisions.
d) Expected contribution
Evidence on when explanatory framing helps ventures whose positioning crosses category expectations.
ML-005Category labels and the evaluation of hybrid venturesConjoint experiment
a) Background and research problem
A venture pursuing commercial and social aims may be evaluated differently when described as a technology startup, social enterprise, or combination of both. A conjoint design can isolate label effects alongside explicit product attributes. The study should distinguish the category label from evidence of actual social impact.
b) Research question
How do category labels change preferences for ventures that combine commercial growth with a social mission?
c) Data, methods and approach
Present repeated choices between fictional venture profiles varying category label, revenue evidence, impact verification, and growth plans. Use realistic attribute combinations, randomize profile order, and account for repeated choices within respondents. Measure respondent priorities before tasks and report attribute effects with uncertainty. Pilot comprehension and limit the number of attributes.
d) Expected contribution
A comparison of category cues and substantive evidence in evaluations of hybrid ventures.
ML-006Brand-category incongruence in entry into digital assetsVignette experiment
a) Background and research problem
An established organization entering digital assets may be seen as extending its capabilities or departing from its identity. A controlled design can study perceived fit while keeping the proposed product constant. This helps distinguish responses to the offering from responses to the organization associated with it.
b) Research question
How does perceived fit between an established brand’s category and a digital-asset offering affect credibility and adoption interest?
c) Data, methods and approach
Use fictional established brands with pretested industry identities and the same digital-asset product description. Randomize brand category and measure perceived fit, credibility, and interest. Hold product benefits and risks constant, assess prior asset familiarity, and avoid real-brand reputational confounds. Treat adoption intention as a stated preference.
d) Expected contribution
Evidence on category fit as a boundary condition for diversification into digital assets.
ML-007How platforms try to repair legitimacy after a crisisLongitudinal content analysis
a) Background and research problem
A platform facing a public failure may issue apologies, change procedures, or seek external assurance. These actions address different audience concerns and can be traced over time. A thesis can analyse repair attempts without presuming that visible communication restored stakeholder approval.
b) Research question
How do digital-asset platforms combine explanation, corrective action, and external assurance when attempting to repair legitimacy?
c) Data, methods and approach
Select several well-documented crises with accessible statements before and after the event. Construct timelines and code communication alongside independently documented organizational changes. Compare sequences, audience references, and implementation evidence. Add stakeholder responses only through a clearly defined sample and avoid equating posting volume or favorable comments with legitimacy restoration.
d) Expected contribution
A process typology of legitimacy-repair attempts and a distinction between communicated commitments and documented action.
ML-008One platform, different standards of legitimacyStakeholder interview study
a) Background and research problem
A digital platform may be useful to customers while raising concerns among workers, local authorities, or community groups. Legitimacy theory distinguishes audience evaluations rather than assuming a single shared judgment. A focused case can examine which expectations conflict and how the platform responds to them.
b) Research question
How do different stakeholder groups judge the legitimacy of the same digital platform?
c) Data, methods and approach
Choose one platform and local setting with feasible access. Interview representatives of three stakeholder groups using comparable prompts about concrete practices. Analyse criteria of approval, perceived benefits, and points of conflict; triangulate with platform policies and public records. Use theoretical concepts as sensitizing ideas while retaining categories that emerge from the interviews.
d) Expected contribution
An audience-specific map of legitimacy judgments and an explanation of tensions that aggregate reputation measures may miss.
ML-009Verified traction versus founder claims in startup pitchesFactorial experiment
a) Background and research problem
Startup pitches contain claims about users, revenue, and growth whose credibility may depend on verifiability. Signaling theory directs attention to what audiences can observe and whether a claim is costly to imitate. An experiment can hold the claimed traction constant while changing its verification and the founder’s commitment.
b) Research question
How does independent verification of a traction claim affect investor evaluations relative to an unverified founder statement?
c) Data, methods and approach
Randomize fictional pitches with identical traction figures and either documented third-party verification or a self-reported claim. A second factor can vary a clearly explained founder commitment if sample size permits. Measure credibility and willingness to investigate; assess whether participants understood the verification. Use suitable investor recruitment or bound claims to the sampled audience.
d) Expected contribution
Evidence on the credibility of verifiable signals and guidance on distinguishing signal content from signal assurance.
ML-010When startup signals contradict one anotherFactorial vignette experiment
a) Background and research problem
A venture may display prestigious affiliations while providing weak operating evidence, or strong customer evidence with few endorsements. Studying such combinations can reveal how audiences integrate signals rather than evaluate each cue alone. A controlled design can test whether operating evidence changes the value attributed to an affiliation.
b) Research question
Does customer-traction evidence moderate the effect of a prestigious affiliation on startup credibility?
c) Data, methods and approach
Create matched pitches varying affiliation and customer evidence independently, using pretested fictional institutions or carefully controlled descriptions. Measure credibility, perceived quality, and willingness to request more information. Predefine the interaction and verify that respondents understand both cues. Account for repeated pitch evaluations if the design presents more than one venture per participant.
d) Expected contribution
Evidence on signal combinations and the conditions under which affiliation adds to, or fails to compensate for, operating evidence.
AI-007What does AI change in entrepreneurial decision-making?Systematic literature review
a) Background and research problem
AI can enter opportunity assessment, forecasting, and resource allocation through different tasks and forms of human involvement. Treating all AI use as one intervention can obscure these differences. A focused systematic review can examine how empirical studies measure decision quality and distinguish assistance from delegation.
b) Research question
Under what task and human-involvement conditions is AI use associated with better entrepreneurial decision quality?
c) Data, methods and approach
Define eligible entrepreneurial decisions and empirical study designs before searching scholarly databases. Record the protocol, screening flow, and exclusion reasons. Extract task, AI role, comparison condition, outcome measures, and study quality. Synthesize results by design and context; use meta-analysis only for sufficiently comparable outcomes. Check recent reviews before finalizing the scope.
d) Expected contribution
An evidence-based account of where AI-assisted decisions have been evaluated and which claims remain insufficiently tested.
DA-008The intellectual structure of decentralized-finance researchBibliometric review
a) Background and research problem
Decentralized-finance research spans financial economics, information systems, and computer science. A bibliometric review can examine how these communities connect, while close reading explains what the mapped groups actually study. The contribution depends on a bounded question and robust corpus construction rather than a list of prominent authors.
b) Research question
Which research communities shape decentralized-finance scholarship, and where do their knowledge bases overlap?
c) Data, methods and approach
Build and document a corpus from OpenAlex or an institutionally available citation database. Validate search precision, merge versions, and audit reference completeness. Use bibliographic coupling or co-citation where supported, with sensitivity checks for thresholds and database coverage. Read representative and boundary papers to interpret clusters; normalize comparisons for publication age where appropriate.
d) Expected contribution
A reproducible map of research communities and substantive connections that can motivate focused future studies.
EN-015What counts as startup success in empirical research?Systematic literature review
a) Background and research problem
Startup studies use survival, funding, growth, profitability, and exits as outcomes, but these measures are not interchangeable. A review can examine whether reported success factors depend on how success is defined. The initial focus should be narrow enough to compare designs meaningfully, such as one venture stage or technology sector.
b) Research question
How do definitions and measurement horizons of startup success shape the predictors reported in empirical studies?
c) Data, methods and approach
Search selected scholarly databases using a preregistered or dated protocol. Extract outcome definitions, observation periods, predictor timing, sample selection, and research design. Appraise risks such as survivorship and information leakage. Compare findings within outcome categories and avoid counting statistical significance as an effect-size synthesis. Assess recent reviews before setting the final scope.
d) Expected contribution
A measurement framework and a synthesis clarifying which success claims concern financing, continuity, or operating performance.
SI-003Does sustainable venture finance create additional impact?Systematic literature review
a) Background and research problem
Funding a sustainable venture does not automatically show that the financing produced environmental or social outcomes beyond what would otherwise have occurred. A systematic review can distinguish financial additionality, changed business activity, and measured impact. This creates a focused contribution around evidence quality rather than broad claims about sustainable finance.
b) Research question
How do empirical studies establish additionality in sustainable venture finance?
c) Data, methods and approach
Define a financing segment and include studies that evaluate a stated counterfactual or comparison. Document searches and screening, then extract financing mechanism, outcome level, comparator, timing, and identification strategy. Appraise selection and measurement limitations and synthesize by evidential strength. Record intended impact separately from observed outcomes; do not pool unlike indicators.
d) Expected contribution
A framework for assessing additionality claims and a research agenda tied to specific measurement and identification gaps.
ML-011How research connects morality and market organizationBibliometric review and interpretive synthesis
a) Background and research problem
Research on morally contested exchange appears under several labels, including stigma, valuation, legitimacy, and moral markets. A bibliometric study can examine whether these conversations share an intellectual base. The review should define its conceptual boundaries carefully rather than treat every ethics-related market paper as equivalent.
b) Research question
How are research streams on morality and market organization connected, and which concepts link them?
c) Data, methods and approach
Pilot several concept-based search strings and manually assess inclusion boundaries. Build a deduplicated citation corpus, map bibliographic coupling, and test whether database and keyword choices change clusters. Read representative papers to compare meanings of morality, contestation, and market evaluation. Use the network as a guide to interpretation, not proof that an unconnected area is a research gap.
d) Expected contribution
An integrated conceptual map showing shared foundations and differences among research streams on contested exchange.
ML-012When category incongruence helps or hurts organizationsSystematic literature review
a) Background and research problem
Crossing category expectations may attract interest or reduce comprehensibility, depending on audiences and contexts. Studies of category spanning, atypicality, and incongruence may operationalize related but distinct constructs. A systematic review can clarify these differences before comparing evidence about evaluation outcomes.
b) Research question
Which audience and context conditions explain positive versus negative evaluations of category-incongruent organizations or offerings?
c) Data, methods and approach
Specify construct definitions and eligible empirical studies before searching. Code how category fit is measured or manipulated, which audience evaluates it, and the evaluation outcome. Separate experimental and observational evidence and assess design quality. Compare proposed mechanisms and boundary conditions through structured synthesis; pool effects only where constructs and metrics align.
d) Expected contribution
A clarified construct framework and evidence map of the conditions under which category incongruence is rewarded or penalized.
AI-008AI in venture-capital due diligence: evidence and practiceMultivocal literature review
a) Background and research problem
Academic studies and practitioner reports may emphasize different benefits and limitations of AI in venture due diligence. A multivocal review combines scholarly work with systematically selected grey literature while assessing each source’s credibility. The opportunity is to distinguish demonstrated decision improvements from implementation advice and vendor claims.
b) Research question
Where do academic evidence and practitioner accounts agree or diverge on AI-supported venture due diligence?
c) Data, methods and approach
Define due-diligence tasks and search scholarly databases plus a documented set of investor, professional, and technology sources. Set dates and stopping rules for grey-literature searches. Appraise authorship, evidence, commercial interests, and traceability; deduplicate repeated claims. Compare benefits, risks, and evidence strength by source type, retaining disagreement rather than treating all sources as equally persuasive.
d) Expected contribution
A research-practice comparison and a transparent inventory of tested capabilities, reported experiences, and unsubstantiated claims.
DA-009Asset tokenization: promised efficiencies and observed resultsMultivocal literature review
a) Background and research problem
Tokenization projects may claim faster settlement, lower costs, or wider access, but reports differ in baselines and maturity. A multivocal review can compare academic evaluations with project and industry accounts in one asset class. The main analytical task is to identify which benefits are measured and under what conditions.
b) Research question
Which claimed efficiencies of asset tokenization are supported by observed evidence in a selected asset class?
c) Data, methods and approach
Choose one asset class and search scholarly studies alongside a documented sample of project reports, regulator publications, and industry analyses. Extract operating stage, comparator, cost boundary, outcome, and evidence source. Rate transparency and commercial interests, trace repeated claims to originals, and distinguish pilots or simulations from operational results.
d) Expected contribution
An evidence matrix connecting efficiency claims to measurement quality and project conditions.
SI-004What makes a corporate carbon claim credible?Multivocal literature review
a) Background and research problem
Academic research, standards bodies, NGOs, and industry guides may use different criteria to assess corporate carbon claims. A multivocal review can compare these criteria and identify which are supported by empirical validation. The project should specify a claim type, such as product-level carbon neutrality, and a dated scope.
b) Research question
Where do scholarly and practitioner sources converge or disagree on the credibility of a defined type of corporate carbon claim?
c) Data, methods and approach
Search academic databases and preselected standards, regulator, NGO, and industry repositories. Record versions and access dates, since guidance changes. Code claim boundaries, verification, accounting assumptions, and supporting evidence. Appraise source interests and authority separately from empirical quality, then compare criteria across source types without treating guidance as evidence of actual effectiveness.
d) Expected contribution
A transparent assessment framework and an account of unresolved differences in how carbon-claim credibility is judged.
TS-007Digital sovereignty across disciplinesBibliometric review
a) Background and research problem
Digital sovereignty is discussed in relation to public infrastructure, organizational dependence, and control over data or technology. The same term may refer to different levels of analysis. A bibliometric review combined with close reading can map these meanings and examine whether research communities exchange ideas.
b) Research question
How do disciplines define digital sovereignty, and how connected are their research conversations?
c) Data, methods and approach
Construct a bounded scholarly corpus with documented language and date limits. Clean author keywords and references, then map co-word patterns and bibliographic coupling where coverage permits. Read papers from each cluster to identify the actor, resource, and type of control being discussed. Test sensitivity to alternative search terms and cluster thresholds.
d) Expected contribution
A conceptually interpreted research map that distinguishes levels of sovereignty and opportunities for cross-disciplinary comparison.
TS-008Data trusts and collective control over dataScoping review
a) Background and research problem
Data trusts and related collective-governance arrangements can differ in legal form, decision rights, and participant representation. An initial scoping review can establish what arrangements have been studied and what outcomes are reported. The contribution is a structured map of the field rather than an estimate of effectiveness across incomparable models.
b) Research question
Which governance arrangements are described as data trusts, and what evidence exists about participant control in practice?
c) Data, methods and approach
Define inclusion criteria distinguishing data trusts from adjacent arrangements. Search academic and selected institutional sources using documented procedures. Chart legal or organizational form, decision rights, implementation stage, stakeholder representation, and evaluated outcomes. Separate conceptual proposals from operational cases and use a transparent screening flow. Narrow geography or application if the corpus is too broad.
d) Expected contribution
A taxonomy of governance arrangements and an evidence map showing where participant control has actually been examined.
AI-009When should an AI assistant hand a task back to a human?Behavioral experiment
a) Background and research problem
An AI assistant may continue confidently or request human review when uncertain. A controlled task can examine whether the timing and explanation of a handoff improve decisions or simply increase interruption costs. The research opportunity is to measure both decision quality and the human effort needed to achieve it.
b) Research question
Do uncertainty-based handoff prompts improve human-AI task performance relative to uninterrupted AI assistance?
c) Data, methods and approach
Use bounded business-analysis tasks with known evaluation criteria and standardized assistant outputs. Randomize handoff prompts while holding underlying information constant. Measure error detection, final decision quality, completion time, and unnecessary interventions. Counterbalance task order, pilot difficulty, and keep model outputs reproducible. Plan sample size around a primary outcome and include appropriate debriefing.
d) Expected contribution
Evidence on the trade-off between oversight and interruption, with practical implications for allocating human attention.
AI-010How entrepreneurs actually integrate AI into daily workDiary study and follow-up interviews
a) Background and research problem
One-off surveys can obscure how AI use changes across tasks and over time. A short diary study can capture adoption, abandonment, and workarounds close to when they happen. This offers a way to examine the difference between intended uses and routines that become part of everyday entrepreneurial work.
b) Research question
Which task characteristics lead entrepreneurs to retain, revise, or abandon AI-assisted work routines?
c) Data, methods and approach
Recruit a small, varied group of founders for several weeks of brief task diaries, followed by interviews. Record task purpose, tool role, revisions, perceived usefulness, and reasons for discontinuing use without collecting confidential prompts by default. Analyse changes within participants and contrasts across tasks. Track missing diary entries and the possibility that observation changes behavior.
d) Expected contribution
A process account of AI appropriation grounded in repeated work episodes rather than retrospective adoption intentions.
EN-016Investor-network position and follow-on financingNetwork analysis and cohort modelling
a) Background and research problem
Investors’ prior co-investment relationships may provide startups with access to further financing. However, network position can also reflect the kinds of ventures an investor selects. A longitudinal network study can construct relationships before the focal investment and examine their association with a defined funding milestone.
b) Research question
Is an early investor’s pre-existing network position associated with a startup obtaining follow-on financing?
c) Data, methods and approach
Use an accessible or licensed dataset of dated financing rounds. Build investor networks only from rounds before each startup’s focal round, resolving investor identities and fund-parent relationships. Compare follow-on outcomes over a common horizon with venture-stage, sector, and cohort controls. Test alternative network definitions and account for multiple startups sharing investors.
d) Expected contribution
Evidence on the association between financing networks and venture progression, with explicit safeguards against using future information.
EN-017Different routes to early startup growthFuzzy-set qualitative comparative analysis
a) Background and research problem
Startups may reach early growth through different combinations of resources, customer access, and founder experience. A configurational approach can explore these combinations instead of assuming one best predictor. The design requires conceptually grounded conditions and transparent calibration, especially with a modest number of ventures.
b) Research question
Which combinations of founder experience, customer access, and financing conditions are associated with early revenue growth?
c) Data, methods and approach
Assemble a purposive, sufficiently varied set of ventures in one sector with comparable revenue measures. Select a small number of theoretically justified conditions and calibrate sets using substantive thresholds. Analyse necessary and sufficient configurations, inspect contradictory cases, and test alternative calibrations. Use case knowledge to interpret solutions and report limited diversity.
d) Expected contribution
A bounded account of alternative growth configurations and hypotheses that can be examined in larger samples.
DA-010Delegation networks in decentralized governanceNetwork analysis
a) Background and research problem
Voting delegation can concentrate decision influence without changing the formal distribution of token ownership. A network approach can examine who receives delegated power and how this structure changes around major decisions. The study must distinguish wallet-level observations from claims about distinct individuals or coordinated groups.
b) Research question
How does the structure of voting delegation change around major governance proposals?
c) Data, methods and approach
Select one governance system with accessible historical delegation records and proposal dates. Build time-specific directed networks and measure concentration, turnover, and persistence of delegates. Compare ordinary and pre-defined major proposals, documenting voting-rule changes and off-chain delegation gaps. Use sensitivity checks for inactive accounts and avoid attributing identities or intent from network position alone.
d) Expected contribution
A reproducible view of delegated influence and its stability across governance decisions.
DA-011How sensitive are stablecoin risk rankings to measurement choices?Replication and robustness study
a) Background and research problem
Stablecoins may appear more or less stable depending on the venue, sampling interval, and metric used. A replication study can examine whether published or commonly used rankings survive reasonable measurement changes. This contributes methodological clarity without requiring a claim that one design measures every dimension of stablecoin risk.
b) Research question
How robust are stablecoin stability rankings to alternative price sources, sampling intervals, and deviation metrics?
c) Data, methods and approach
Select a small set of assets with overlapping historical coverage. Reproduce one documented stability measure, then vary venue aggregation, sampling frequency, and treatment of missing observations. Compare ranking agreement and identify episodes driving changes. Separate market-price stability from reserve quality and redemption risk, which require different evidence. Publish code and retrieval details where licences permit.
d) Expected contribution
A sensitivity analysis showing which stability comparisons are robust and which depend strongly on data choices.
IB-006Do small monetary incentives change investment-risk choices?Incentivized experiment
a) Background and research problem
People may report risk preferences differently when a decision has a real, bounded financial consequence. Comparing hypothetical and incentivized tasks can test how much inference depends on the elicitation format. A controlled study should keep probabilities and information constant and avoid equating laboratory choices with all real-world investing.
b) Research question
Do hypothetical and modestly incentivized investment tasks yield different risk choices and confidence ratings?
c) Data, methods and approach
Randomly assign adults to matched allocation tasks with hypothetical outcomes or clearly explained, capped bonus payments. Use researcher-funded rewards without requiring participant deposits. Predefine the payment mechanism and primary risk measure, check probability comprehension, and analyse choice and confidence separately. Arrange participant recruitment, budget, and ethics review before running the experiment.
d) Expected contribution
Evidence on how elicitation format affects observed risk preferences and the interpretation of survey-based investment measures.
IB-007When an investment conflicts with an investor’s valuesInterview study
a) Background and research problem
Investors may face tension between financial returns and objections to a firm’s products or practices. Interviews can examine how they justify participation, draw exclusions, or revise judgments after investing. The focus is on decision processes and moral boundaries rather than estimating the prevalence of ethical investing.
b) Research question
How do individual investors resolve perceived conflicts between expected returns and personal moral commitments?
c) Data, methods and approach
Recruit adults with contrasting investment approaches and ask about concrete inclusion, exclusion, or divestment episodes. Use participant-selected examples without requesting account access or identifying financial records. Analyse justifications and changes over time, seek disconfirming cases, and consider impression-management and recall effects. Keep claims bounded to the recruited sample.
d) Expected contribution
A typology of moral-financial trade-offs and insight into how investors maintain or revise market boundaries.
TS-009Platform-rule changes and the quality of financial discussionQuasi-experimental panel study
a) Background and research problem
A platform rule change may alter which financial discussions remain visible and how users participate. A before-and-after comparison alone may confuse the rule with market conditions or broader platform changes. A focused study can assess whether comparable communities provide a defensible counterfactual.
b) Research question
How does a clearly dated moderation-rule change relate to the prevalence of substantiated claims in financial discussions?
c) Data, methods and approach
Identify a documented rule change with permitted access to archived posts and a plausible unaffected comparison group. Develop and validate a claim-quality codebook, then analyse a panel using difference-in-differences only if pre-trends, timing, and spillover checks support it. Otherwise use a clearly descriptive interrupted series. Minimize personal data and follow platform access conditions.
d) Expected contribution
Evidence on discussion-quality changes and an explicit assessment of how credible the comparison design is.
SI-005What investors value in biodiversity financeDiscrete-choice experiment
a) Background and research problem
Biodiversity finance combines uncertain returns with impact claims that may be difficult to compare. A choice experiment can examine whether investors value independently verified outcomes, local benefits, or financial attributes. It should use a clearly explained hypothetical product rather than imply that a real product offers the stated returns.
b) Research question
How do impact verification and project location influence preferences for hypothetical biodiversity investments relative to risk and return?
c) Data, methods and approach
Design a small set of credible product attributes with expert input and pilot comprehension. Recruit adults with relevant investment interest for repeated choices including an opt-out. Estimate attribute effects while accounting for repeated responses, and examine preference heterogeneity only where sample size supports it. Keep hypothetical willingness to invest separate from actual purchase behavior.
d) Expected contribution
Evidence on the relative importance of impact signals and financial terms, with implications for clearer product communication.
SI-006From TNFD disclosure to financially useful nature-risk informationComparative disclosure analysis
a) Background and research problem
The Taskforce on Nature-related Financial Disclosures (TNFD) offers a framework for reporting dependencies, impacts, risks, and opportunities related to nature. A focused research problem is whether company reports connect local ecological exposure to identifiable business activities and financial consequences, rather than stopping at general commitments.
b) Research question
How clearly do TNFD-aligned reports connect location-specific nature dependencies to business risks and financial decisions?
c) Data, methods and approach
Select 15–25 firms in one nature-dependent industry with publicly available reports. Develop a coding rubric covering locations, dependencies, time horizons, financial transmission channels, and management responses. Double-code a subset and distinguish missing disclosure from missing internal assessment. Compare firms of similar size and reporting maturity; pilot whether enough comparable reports exist.
d) Expected contribution
A reproducible assessment of disclosure usefulness and examples of how nature-related reporting could better inform financial analysis.
SI-007How financial institutions use TNFD in investment and lending decisionsPractitioner interview study
a) Background and research problem
Nature-related frameworks can support financial institutions in identifying dependencies and exposures. Adopting a reporting framework does not itself demonstrate a change in portfolio selection, credit assessment, or client engagement. Organizational routines offer a useful lens for examining how assessment results enter actual decisions.
b) Research question
Under what conditions does TNFD-related assessment influence investment or lending decisions?
c) Data, methods and approach
Interview approximately 10–15 analysts, risk professionals, or sustainability managers within either banking or asset management. Ask for concrete decision episodes, including cases where assessment results had little influence. Compare interviews with public methodology documents and code data constraints, internal responsibilities, incentives, and decision rules. Secure initial recruitment access before fixing the scope and anonymize commercially sensitive examples.
d) Expected contribution
An account of the organizational steps and barriers between nature-risk assessment and financial decision-making.
SI-008What makes a biodiversity credit claim credible?Multivocal literature review
a) Background and research problem
Biodiversity credit proposals raise questions about what is measured, which improvements are additional, and how long outcomes persist. Standards, scientific studies, and market reports provide different forms of evidence. Legitimacy theory can help explain how verification arrangements support credibility without treating certification as proof of ecological success.
b) Research question
Which verification arrangements support credible biodiversity credit claims, and where does outcome evidence remain limited?
c) Data, methods and approach
Systematically search academic studies alongside a bounded sample of credit standards, project documents, independent evaluations, and critical reports. Code baselines, additionality, measurement uncertainty, durability, and stakeholder participation. Appraise sources according to their purpose and independence; distinguish design requirements from observed results. Narrow to one habitat or credit approach if the literature is too heterogeneous.
d) Expected contribution
An evidence map separating proposed safeguards from demonstrated outcomes, with a practical checklist for evaluating biodiversity finance claims.
SI-009Do biodiversity-linked financing targets reward meaningful ecological progress?Financing-document case comparison
a) Background and research problem
A loan or bond can link financial terms to environmental performance indicators. For biodiversity, the choice of baseline, target, and verification process is especially consequential: a measurable activity may be easier to reward than a meaningful ecological outcome. Contract design provides a focused way to study this tension.
b) Research question
How do biodiversity-related financing indicators connect financial incentives to measurable ecological outcomes?
c) Data, methods and approach
Pilot a search for 6–10 publicly documented financing cases with explicit biodiversity targets. Compare financing frameworks, target definitions, external reviews, and subsequent progress reports. Code ambition, controllability, independent verification, and the size and timing of financial incentives. Separate missing contract information from weak design and reported target achievement from independently demonstrated ecological improvement.
d) Expected contribution
A comparison of biodiversity-linked incentive designs and recommendations for making their claimed environmental contribution more assessable.
SI-010How location-specific biodiversity risk affects investment judgmentsInformation experiment
a) Background and research problem
Financial exposure to nature can depend on where a company operates and how its activities rely on local ecosystems. Aggregate sustainability ratings may communicate different information from a concrete explanation of a production site's dependence on water or pollination. The research problem concerns risk interpretation rather than preference for a green product label.
b) Research question
Does location-specific nature-risk information change perceived financial risk and investment allocation compared with an aggregate rating?
c) Data, methods and approach
Randomly assign participants to matched fictional company profiles with generic sustainability information, a location-specific dependency explanation, or both. Keep financial figures constant and measure risk comprehension, allocation, and confidence. Pretest the materials for equal readability, justify sample size, and distinguish hypothetical allocation from actual investment behavior.
d) Expected contribution
Evidence on which forms of nature-risk information help investors connect ecological dependencies to financial exposure.
DA-012Why merchants adopt—or abandon—stablecoin paymentsMerchant interview study
a) Background and research problem
The availability of stablecoin checkout technology does not establish that merchants find it useful. Adoption may depend on customer demand, accounting, refunds, conversion costs, and the ability to receive local currency. Technology adoption theory can organize a comparison of merchants with different implementation experiences.
b) Research question
Which operational and commercial factors explain continued use, rejection, or abandonment of stablecoin payments by merchants?
c) Data, methods and approach
Recruit approximately 12–18 merchants in one sector or country, including adopters, former users, and non-adopters who considered the option. Reconstruct the decision and payment workflow through interviews and public checkout documentation. Where available, request anonymized usage ranges and cost examples. Distinguish offering stablecoin payment from customers actually using it, and document recruitment selection.
d) Expected contribution
A grounded account of stablecoin payment adoption and the frictions providers must address to sustain merchant use.
DA-013Stablecoin checkout: fees, refunds, and willingness to payConsumer choice experiment
a) Background and research problem
Stablecoin payment adoption involves more than familiarity with cryptoassets. At checkout, consumers may trade lower fees against uncertainty about refunds, exchange-rate conversion, and who resolves a dispute. A controlled choice setting can separate these payment features from general enthusiasm for the technology.
b) Research question
How do price savings, refund arrangements, and automatic currency conversion affect consumers' choice of stablecoin payment?
c) Data, methods and approach
Create repeated choices between conventional and stablecoin payment options for a fictional purchase. Vary a small set of clearly explained attributes using a balanced experimental design. Measure choices, comprehension, and previous payment experience; estimate attribute effects with uncertainty intervals. Pretest terminology, plan recruitment and sample size, and conduct the study without transferring funds or requiring a wallet.
d) Expected contribution
An estimate of the payment features associated with stated adoption, with clear limits on extrapolating to real purchases.
DA-014The full cost of a stablecoin cross-border paymentPayment-route cost comparison
a) Background and research problem
A low blockchain transaction fee does not establish that a cross-border payment is cheap from sender to recipient. Conversion spreads, funding fees, withdrawal charges, and access restrictions can change the comparison. A useful research gap is a transparent, reproducible comparison for a narrowly defined payment corridor.
b) Research question
When is a stablecoin payment route cheaper and faster than conventional alternatives after accounting for the complete payment journey?
c) Data, methods and approach
Select one currency corridor and several transfer sizes. Collect time-stamped public quotes and fee schedules for accessible stablecoin routes and conventional providers over a fixed period. Include on- and off-ramp costs, exchange-rate benchmarks, and eligibility conditions. Separate quoted from observed completion times and use sensitivity analysis for uncertain costs; no live transfers are required.
d) Expected contribution
A replicable route-comparison framework identifying the assumptions under which stablecoins offer a payment advantage.
DA-015Who can actually redeem a stablecoin at face value?Issuer-policy and disclosure analysis
a) Background and research problem
A stablecoin holder's practical access to redemption can differ from the headline promise of price stability. Direct redemption may depend on customer eligibility, minimum amounts, fees, intermediaries, and operating hours. These arrangements matter independently of the market price and reserve composition.
b) Research question
How do redemption arrangements differ across stablecoin issuers, and what do those differences imply for different types of holder?
c) Data, methods and approach
Compare official terms, reserve disclosures, redemption instructions, and dated policy updates for 6–10 fiat-backed stablecoins. Construct common retail and institutional holder scenarios within one jurisdiction. Record who can redeem, through which route, at what disclosed cost, and under which restrictions. Archive document dates and distinguish contractual statements from observed redemption performance.
d) Expected contribution
A structured map of redemption access that clarifies why holding a stablecoin and having a direct claim on its issuer are different economic positions.
AI-011Do teams of AI agents improve startup due diligence?Controlled task benchmark
a) Background and research problem
Assigning different analytical roles to multiple AI agents may improve coverage, but it also creates coordination costs and can reproduce the same errors across roles. The relevant comparison is whether an agent team produces better investment analysis at a comparable resource budget.
b) Research question
Do multiple cooperating AI agents improve the quality of startup due diligence relative to a single agent under comparable budgets?
c) Data, methods and approach
Build a small benchmark of fictional startup dossiers with known inconsistencies and missing information. Compare single-agent and multi-agent workflows using the same underlying model, source material, tool access, and total spending limit. Repeat runs and have blinded assessors score evidence accuracy, missed issues, and unsupported conclusions. Record latency and coordination overhead; treat cases, rather than repeated runs alone, as the basis for generalization.
d) Expected contribution
Evidence on when agent specialization improves due diligence and when coordination consumes resources without improving decisions.
AI-012Responsibility when an AI agent makes a business mistakeFactorial vignette experiment
a) Background and research problem
Delegating work to an AI agent can distribute control across a manager, an employee, and a software provider. After a mistake, perceived responsibility may depend on who approved the task and how much discretion the agent had. Attribution theory provides a basis for studying these judgments.
b) Research question
How do approval arrangements and delegated discretion affect responsibility judgments after an AI agent's business error?
c) Data, methods and approach
Randomize adult participants across fictional procurement scenarios that vary prior human approval and the agent's permitted discretion while holding the loss constant. Measure attributed responsibility, perceived fairness, and preferred corrective action. Pretest whether participants understand the control arrangements and plan sample size around the main effects and interaction. If recruiting managers, report their experience rather than assuming professional representativeness.
d) Expected contribution
Evidence on how organizations can make accountability understandable when business tasks are delegated to AI agents.
AI-013Selling AI agents: subscriptions, usage fees, or payment for outcomes?Founder and customer interviews
a) Background and research problem
AI agent ventures can price access, resource consumption, or completed outcomes. Each model allocates uncertainty differently when task difficulty and output quality vary. Transaction-cost and business-model perspectives can help examine how founders define an outcome customers are willing to pay for.
b) Research question
How do AI agent startups choose pricing models and allocate the risk of unsuccessful task completion?
c) Data, methods and approach
Select one application area, such as customer support or sales administration. Compare public pricing and service terms for 10–15 providers, then interview a feasible subset of founders and business customers. Code outcome definitions, exceptions, monitoring costs, and reasons for changing pricing. Distinguish published prices from negotiated contracts and claims of performance from customer evidence.
d) Expected contribution
A typology of AI agent revenue models and the conditions under which outcome-based pricing is commercially credible.
AI-014The hidden maintenance work of using AI agentsLongitudinal workflow study
a) Background and research problem
An AI agent may perform well in an initial demonstration but require ongoing monitoring, instruction updates, and recovery from changing inputs. This maintenance work can alter the economics of delegation. A longitudinal study can examine costs that a one-off productivity test misses.
b) Research question
How does maintaining an AI agent change the time savings and reliability of a recurring business workflow?
c) Data, methods and approach
Follow 4–6 small organizations using agents for one comparable, low-risk administrative task over 4–6 weeks, subject to recruitment access. Collect short activity diaries, anonymized task counts, correction time, and weekly interviews. Compare initial setup with later operation and distinguish successful completion from time merely shifted to supervision. Record changes in models or tools and avoid retaining confidential business inputs.
d) Expected contribution
An account of recurring delegation costs and practical measures for assessing whether agent adoption produces sustained efficiency gains.
EN-018Which milestones make a space startup investable?Investor vignette experiment
a) Background and research problem
Space ventures may demonstrate engineering progress before establishing repeat customers. Investors therefore face signals that differ in technical credibility and commercial relevance. Signaling theory provides a way to examine how independent technical validation and customer commitments jointly shape evaluations.
b) Research question
How do verified technical milestones and customer commitments affect perceived investability of a space startup?
c) Data, methods and approach
Develop fictional profiles within one segment, such as satellite components. Independently vary third-party technical validation and the strength of customer commitments while holding team and financing needs constant. Recruit participants with investment or entrepreneurial experience where feasible, pretest technical comprehensibility, and measure investment interest, uncertainty, and credibility. Plan sample size and report limitations if using a broader participant pool.
d) Expected contribution
Evidence on how space ventures can communicate progress and whether technical and commercial signals reinforce one another.
EN-019Public procurement as a stepping stone for space startupsComparative longitudinal case study
a) Background and research problem
Public organizations can become important early customers for space ventures. A contract may support validation and revenue while also creating customer dependence or influencing product priorities. Resource-dependence theory helps frame the trade-off between public demand and entry into commercial markets.
b) Research question
How do early public contracts shape space startups' subsequent financing and commercial customer development?
c) Data, methods and approach
Select 6–8 ventures within one space segment with verifiable public awards. Reconstruct timelines using procurement records, company filings, financing announcements, and founder interviews where available. Separate grants, procurement contracts, and formal partnerships. Compare cases with different later customer mixes and seek unsuccessful or stalled trajectories. Treat observed sequences as evidence about mechanisms, not proof that contracts caused funding success.
d) Expected contribution
A process model of how public demand can support or constrain the commercialization of space technologies.
EN-020Why firms adopt Earth-observation servicesCustomer and founder interview study
a) Background and research problem
Satellite data become commercially useful when customers can integrate the resulting information into decisions. High technical accuracy may be insufficient if a service is difficult to interpret, expensive to integrate, or poorly matched to an existing workflow. Entrepreneurship research can examine this transition from data to customer value.
b) Research question
What determines whether business customers move from piloting an Earth-observation service to paying for continued use?
c) Data, methods and approach
Focus on one use case, such as crop monitoring or infrastructure inspection. Interview approximately 10–15 providers and current or former pilot customers. Compare decision episodes, integration work, value metrics, and reasons for non-renewal. Where accessible, examine anonymized pilot evaluations. Avoid generalizing from provider testimonials alone and establish customer access before finalizing the study.
d) Expected contribution
An explanation of the barriers between technical demonstration and recurring revenue for ventures in the space economy.
EN-021Who pays for a cleaner orbital environment?Multivocal business-model review
a) Background and research problem
Orbital sustainability services, including debris removal, can create benefits extending beyond the organization paying for them. This creates a business-model problem involving incentives, responsibility, and public demand. Public-goods and entrepreneurship perspectives can help assess how proposed services acquire paying customers.
b) Research question
Which revenue and procurement models can support orbital sustainability ventures, and what evidence exists of customer commitment?
c) Data, methods and approach
Review academic studies, agency procurement documents, company disclosures, and independent evaluations for a bounded service category. Code the payer, beneficiary, pricing basis, risk allocation, and evidence of implementation. Distinguish research grants, demonstration awards, signed service contracts, and aspirations. Compare proposed models with documented transactions without attempting an engineering assessment of debris-removal technologies.
d) Expected contribution
An evidence-based map of commercialization models and the coordination problems they seek to solve in the space economy.
TS-010Business models and continuity promises in the digital afterlife industryPlatform and contract analysis
a) Background and research problem
Digital afterlife services can sell memorial archives, scheduled messages, or conversational representations of deceased people. A promise of long-term availability creates an economic problem when storage, model access, and support incur continuing costs. Business-model analysis can examine how providers explain service continuity.
b) Research question
How do digital afterlife providers align pricing and contractual commitments with the promise of long-term service availability?
c) Data, methods and approach
Sample 15–25 publicly accessible services and archive dated pricing pages, terms, and help materials. Code recurring versus one-off charges, continuity language, ownership changes, data export, and closure provisions. Distinguish explicit commitments from promotional wording and missing information. Use public documentation without uploading anyone's personal data or interacting with representations of real deceased individuals.
d) Expected contribution
A typology of digital afterlife business models and a framework for evaluating whether continuity promises are commercially and contractually clear.
TS-011Consent and control in digital afterlife servicesVignette experiment
a) Background and research problem
A digital representation may involve the wishes of the person represented, relatives, and later users. These interests need not coincide. A focused empirical question is how different consent and control arrangements affect the acceptability of a hypothetical service, without assuming that acceptance establishes ethical adequacy.
b) Research question
How do prior consent and family control over access or deletion affect acceptance of a digital afterlife service?
c) Data, methods and approach
Present adult participants with clearly fictional, neutrally worded service scenarios. Vary documented prior consent and the allocation of access and deletion rights; measure acceptability, trust, and perceived conflicts. Pretest comprehension, justify sample size, and complete applicable ethics review. Avoid targeting recently bereaved participants or using real personal representations, and allow participants to skip sensitive material.
d) Expected contribution
Evidence on stakeholder expectations that can inform service governance while preserving the distinction between consumer acceptance and normative justification.
TS-012Building legitimacy in the digital afterlife marketStakeholder interview study
a) Background and research problem
Commercial services surrounding memory and death can be evaluated through competing moral frames: care, preservation, exploitation, or intrusion. Legitimacy theory and research on moralized markets offer a basis for studying how providers justify their activities and how other stakeholders contest those justifications.
b) Research question
How do digital afterlife entrepreneurs and professional stakeholders negotiate the boundaries of acceptable commercialization?
c) Data, methods and approach
Interview approximately 12–15 founders, digital-estate professionals, ethicists, or relevant civil-society representatives. Compare their accounts with public service descriptions and debates. Code moral justifications, disputed practices, and boundary-setting strategies, deliberately seeking critical perspectives. Use professional accounts rather than recruiting bereaved service users; anonymize participants and distinguish stated principles from evidence of business practice.
d) Expected contribution
An explanation of legitimacy-building and moral disagreement in an emerging industry organized around identity, memory, and commercial value.
TS-013Can a digital afterlife service speak for someone in an advertisement?Factorial vignette experiment
a) Background and research problem
Using a posthumous digital representation in commercial communication can blur preservation, endorsement, and monetization. Even when prior permission is specified, the fit between the representation's memorial purpose and a commercial role may affect judgments. Category incongruence provides a focused theoretical lens.
b) Research question
How do commercial use and clearly documented prior permission affect the perceived legitimacy of a fictional posthumous representation?
c) Data, methods and approach
Use text-only scenarios about an entirely fictional adult and a fictional service. Vary noncommercial versus sponsored communication and the presence of explicit prior permission. Measure perceived appropriateness, trust, and willingness to use the service. Pretest the scenarios, plan sample size and ethics review, and avoid generating real people's likenesses or eliciting personal bereavement experiences.
d) Expected contribution
Evidence on the boundary between memorial services and commercial endorsement, with implications for business-model design and consent communication.
EN-022From quantum research to a first paying customerFounder interview and case study
a) Background and research problem
Quantum entrepreneurship involves translating scientific capabilities into a product that addresses a customer's problem. Technical progress and commercial readiness can follow different timelines. A commercialization perspective can examine how founders choose applications and demonstrate value without assuming that every quantum technology has the same maturity.
b) Research question
How do quantum startups select an initial market and move from a research demonstration to a paying customer?
c) Data, methods and approach
Focus on one segment, such as quantum sensing or quantum-enabling components. Interview 8–12 founders, commercialization specialists, or early customers and reconstruct a small set of venture timelines. Compare application selection, validation requirements, sales cycles, and changes in the value proposition. Triangulate interviews with public product and partnership records; assess commercial decisions rather than attempting to verify technical quantum advantage.
d) Expected contribution
A process account of early market development that identifies commercialization challenges specific to the selected quantum segment.
EN-023Financing milestones in quantum entrepreneurshipComparative financing case study
a) Background and research problem
Quantum ventures may combine research grants, strategic partnerships, and private equity while progressing through uncertain development stages. The research problem is how financing milestones connect scientific progress to commercial expectations, and whether different capital providers reward different forms of evidence.
b) Research question
How do public and private funding sources shape milestone choices and commercialization priorities in quantum startups?
c) Data, methods and approach
Study 6–8 ventures in one quantum segment and jurisdiction using public grant records, financing announcements, filings, and interviews where feasible. Reconstruct the sequence of funding and technical or customer milestones, including delays and changed targets. Distinguish funding announcements from completed rounds and grants from procurement. Compare mechanisms across cases without inferring that a funding type caused venture success.
d) Expected contribution
A grounded account of financing sequences and potential tensions between scientific development and commercial validation in quantum entrepreneurship.
EN-024Credible signals in quantum startup pitchesPitch evaluation experiment
a) Background and research problem
Potential investors may struggle to interpret technical claims made by quantum ventures. Signaling theory suggests that independently verifiable evidence and external affiliations may carry different weight. The question is whether evaluators distinguish evidence about a product from prestige associated with the team or its partners.
b) Research question
How do independent product validation and prestigious affiliations influence the perceived credibility of a quantum startup?
c) Data, methods and approach
Create matched fictional pitches in one accessible application area, varying independent validation and institutional affiliation. Hold the product claim and financial information constant. Recruit investors or experienced business participants where possible; measure credibility, perceived uncertainty, and willingness to seek more information. Pretest technical understanding, justify sample size, and avoid implying that a fictional validation establishes general quantum advantage.
d) Expected contribution
Evidence on how audiences evaluate scientific venture claims and when prestige may substitute for more directly relevant commercial evidence.
ML-013How well do prediction markets forecast resolved events?Forecast calibration study
a) Background and research problem
Prediction market prices are often interpreted as probabilities, but their forecasting usefulness should be evaluated against actual outcomes. Results may depend on forecast horizon, contract selection, liquidity, and the treatment of unresolved events. A transparent sampling design is central to a meaningful assessment.
b) Research question
How accurate and well calibrated are prediction market forecasts within one event category at different forecast horizons?
c) Data, methods and approach
Pilot historical price and resolution retrieval from a documented public API. Define a cohort of binary contracts before examining performance and collect forecasts at fixed horizons. Use Brier scores and calibration plots, with appropriate reference forecasts. Report missing, canceled, and unresolved contracts; distinguish last-trade prices from executable quotes and account for related contracts when estimating uncertainty.
d) Expected contribution
A reproducible evaluation of forecasting quality that shows where prediction market probabilities are informative and where measurement choices change the conclusions.
ML-014When prediction markets disagree about the same eventMatched-contract comparison
a) Background and research problem
Two prediction contracts can appear to concern the same event while differing in deadlines, outcome definitions, or resolution authorities. Price differences may therefore reflect different promises rather than conflicting beliefs. Contract design offers a practical way to study the limits of treating market probabilities as interchangeable.
b) Research question
How much cross-platform prediction market disagreement is associated with differences in contract wording and resolution rules?
c) Data, methods and approach
Identify a manageable set of apparently matched contracts in one event category on two accessible platforms. Archive their full rules and time-aligned price histories or quotes. Independently code deadline, outcome, and resolution differences, then compare price gaps while considering spreads and trading activity. Confirm data availability in a pilot and avoid labeling gaps as arbitrage without accounting for execution and settlement constraints.
d) Expected contribution
A contract-matching framework and evidence on when apparently comparable prediction market prices measure different underlying events.
ML-015Do prediction market probabilities improve business decisions?Decision experiment
a) Background and research problem
A probability forecast only creates business value if decision-makers use it appropriately. Labeling information as a prediction market forecast may change trust even when its numerical content is identical. Decision theory provides a basis for comparing choices with a benchmark derived from stated payoffs.
b) Research question
Does identifying a probability forecast as coming from a prediction market improve or distort its use in business decisions?
c) Data, methods and approach
Randomly present identical fictional demand forecasts with a prediction-market label, an analyst label, or no source label. Ask participants to choose inventory or project investments with explicit payoff tables. Measure expected-payoff performance, confidence, and understanding of probabilities. Pretest task difficulty, justify sample size, and use hypothetical business decisions rather than requiring betting or platform participation.
d) Expected contribution
Evidence on the practical value of forecast communication and on whether source labels encourage appropriate reliance or unwarranted confidence.
AI-015How much spending authority should an AI agent receive?Delegated-payment choice experiment
a) Background and research problem
Agentic payments allow software acting on a person's behalf to initiate purchases within some form of delegated authority. Adoption may depend on how spending limits, approval requirements, and revocation are expressed. Delegation theory helps frame the trade-off between convenience and retained control.
b) Research question
How do spending limits, approval rules, and the ability to revoke permission affect willingness to delegate purchases to an AI agent?
c) Data, methods and approach
Develop repeated choices between fictional agent-payment arrangements for a familiar shopping task. Vary total budget limits, approval thresholds, and revocation options while holding product quality and payment costs constant. Measure willingness to delegate and understanding of the agent's permissions. Pretest the scenarios, plan sample size, and conduct all decisions without real payments or account access.
d) Expected contribution
Evidence on understandable spending mandates and the control features associated with acceptance of agentic commerce.
AI-016Trust and responsibility in agentic payment protocolsComparative protocol-document study
a) Background and research problem
Agentic payment initiatives address different parts of a transaction, including the user's authorization, the merchant's payment request, and settlement. Treating these functions as interchangeable can obscure who relies on which evidence. A governance perspective can compare the allocation of trust and responsibility across a payment journey.
b) Research question
How do agentic payment designs document user intent, payment authorization, and responsibility when a transaction is disputed?
c) Data, methods and approach
Archive versioned specifications and official examples for initiatives such as AP2 and x402. Map their roles onto a common fictional purchase, distinguishing complementary layers from competing designs. Code identity assumptions, authorization evidence, settlement dependencies, and documented dispute mechanisms. Separate protocol capabilities from provider policies, implementation choices, and legal enforceability; validate interpretations against worked examples where available.
d) Expected contribution
A clear comparison of the governance questions addressed by different protocol layers and those left to surrounding institutions.
AI-017Would merchants accept purchases made by AI agents?Merchant and payment-provider interviews
a) Background and research problem
A merchant receiving an order from an AI agent must decide whether the transaction represents an authorized purchase and how to handle mistakes or returns. Faster checkout may offer limited value if evidence of authorization and responsibility is unclear. Merchant adoption is therefore a useful complement to consumer delegation research.
b) Research question
Which evidence and service guarantees would merchants require before accepting agent-initiated purchases?
c) Data, methods and approach
Interview approximately 10–15 e-commerce merchants and payment professionals using the same fictional purchase and dispute scenarios. Compare concerns about authorization, returns, identity, fees, and customer support. Include firms with different payment infrastructures and distinguish current operating experience from expectations about proposed systems. Confirm participant access early and keep technical claims tied to documented capabilities.
d) Expected contribution
A prioritized account of merchant requirements and unresolved coordination problems in the adoption of agentic payments.
AI-018AI washing: comparing corporate claims with disclosed evidenceClaim-evidence content analysis
a) Background and research problem
AI washing concerns potentially misleading representations of AI capabilities or use. However, a vague claim or limited public disclosure does not by itself establish deception. A useful research problem is how to assess the specificity and evidential support of claims without conflating promotional language, uncertainty, and demonstrable contradiction.
b) Research question
How consistently do firms support AI-related product claims with specific and publicly verifiable evidence?
c) Data, methods and approach
Select 25–40 firms in one product category and archive dated marketing pages, documentation, demonstrations, and relevant disclosures. Develop a rubric distinguishing specific supported claims, unclear claims, and claims contradicted by available evidence. Double-code a subset and use official enforcement findings only for cases where such findings exist. Treat undisclosed internal capabilities as unknown and document alternative explanations.
d) Expected contribution
A cautious, reproducible method for studying AI claim quality without presenting unverified allegations of misconduct.
AI-019Can a company regain trust after correcting an exaggerated AI claim?Trust-repair experiment
a) Background and research problem
An exaggerated AI claim may damage trust in a firm's technology and its honesty. Legitimacy repair research suggests that acknowledging an error and providing verifiable evidence may have different effects from simply replacing promotional wording. A controlled setting can examine the consequences of different corrections.
b) Research question
Which correction strategies restore trust after a fictional company acknowledges that it overstated its use of AI?
c) Data, methods and approach
Randomly assign participants to matched company scenarios with different responses: a brief correction, an explanation with responsibility-taking, or an explanation plus independent verification. Include a suitable comparison condition and measure credibility, product interest, and perceived accountability. Pretest the original claim and correction strength, justify sample size, and use invented firms to avoid unsupported reputational implications.
d) Expected contribution
Evidence on the limits and effectiveness of trust repair after misleading technology communication.
AI-020What counts as AI washing? An evidence-based reviewMultivocal literature review
a) Background and research problem
The term AI washing can describe inflated capability claims, misleading descriptions of human involvement, or superficial use of AI terminology. These meanings require different evidence. A focused review can clarify the concept and prevent research from classifying every promotional reference to AI as misconduct.
b) Research question
How is AI washing defined and evidenced across academic research, official enforcement material, and professional debate?
c) Data, methods and approach
Use a documented search strategy for scholarly literature and a bounded selection of regulator publications and professional sources. Code claim types, alleged versus established conduct, evidence standards, and proposed consequences. Assess source quality and independence, separate normative arguments from empirical findings, and track when several reports repeat the same underlying case.
d) Expected contribution
A working typology and measurement framework that supports more precise empirical research on AI washing and distinguishes suspicion from substantiated evidence.
AI-021Who stands behind an AI influencer's recommendation?Advertising disclosure experiment
a) Background and research problem
An AI influencer's recommendation can involve a platform, a brand, and human operators. Disclosing that the persona is artificial does not necessarily explain who approved the message. Accountability theory can guide a study of whether additional information about editorial responsibility changes audience judgments.
b) Research question
Does identifying the human or organizational party responsible for an AI influencer's recommendation affect trust and perceived accountability?
c) Data, methods and approach
Show participants matched fictional sponsored posts that all disclose the persona's AI status and sponsorship. Vary whether editorial responsibility is unspecified, assigned to a named role within an agency, or assigned to the sponsoring brand. Hold the message and visual style constant; measure credibility, responsibility judgments, and purchase interest. Pretest disclosure comprehension and plan sample size.
d) Expected contribution
Evidence on whether responsibility disclosure adds useful information beyond identifying a persona as AI-generated and an advertisement as sponsored.
AI-022Financial recommendations by AI influencersMatched content analysis
a) Background and research problem
AI personas can present financial commentary in a familiar influencer format. The relevant research question is whether the information supplies balanced, traceable claims and clear commercial disclosure. Comparing AI and human accounts requires care because account selection, audience, and content category may explain apparent differences.
b) Research question
How do financial claims, risk discussion, and sponsorship disclosures compare between disclosed AI influencers and matched human influencers?
c) Data, methods and approach
Pilot whether enough publicly documented AI accounts publish relevant content on one platform. Sample a fixed period and match human accounts by topic and approximate audience size. Code evidence links, return claims, risk statements, and commercial disclosures, with reliability checks. Define AI status from documented account information rather than appearance; if the sample is small, use a bounded comparative case study.
d) Expected contribution
A transparent assessment of financial communication quality that separates observable content characteristics from assumptions about the persona producing them.
AI-023When an AI influencer claims personal experienceCategory-congruence experiment
a) Background and research problem
Influencer endorsements often rely on claims of personal experience. Such claims may be interpreted differently when an openly artificial persona describes tasting food or using a physical product. Category incongruence offers a way to investigate whether audiences see these statements as acceptable fiction, confusing communication, or a credibility problem.
b) Research question
How does an AI influencer's claim of personal product experience affect credibility across different product categories?
c) Data, methods and approach
Create clearly fictional advertisements with a disclosed AI persona. Vary first-person experience claims versus factual product descriptions across one sensory and one functional product category. Measure perceived authenticity, appropriateness, comprehension, and purchase interest. Keep sponsorship disclosure constant, pretest stimulus plausibility, and justify sample size for the interaction between message and product category.
d) Expected contribution
Evidence on the commercial and legitimacy boundaries of first-person endorsement by artificial personas.
DA-016What does a tokenized real-world asset holder actually own?Offering-document comparison
a) Background and research problem
A token linked to a real-world asset can represent different economic arrangements, including a direct interest, a claim against an intermediary, or exposure through another instrument. Similar marketing labels can conceal differences in custody, redemption, and the link between the token and underlying asset.
b) Research question
How clearly do tokenized asset offerings explain holders' economic rights and the intermediaries on which those rights depend?
c) Data, methods and approach
Select 8–12 offerings within one asset class and jurisdiction, subject to document access. Compare offering documents, custody arrangements, redemption policies, and public risk disclosures. Map the chain between holder, issuer, custodian, and asset; code unclear or unavailable provisions. Distinguish documented contractual rights from their untested enforceability and seek specialist input for contested interpretations.
d) Expected contribution
A structured comparison of token-holder positions and a disclosure framework for understanding tokenized real-world asset products.
DA-017Does tokenization create round-the-clock liquidity?Market-data and redemption analysis
a) Background and research problem
A token may be transferable at any time even when its underlying asset, trading venue, or redemption facility operates on a narrower schedule. Transferability, secondary-market liquidity, and redemption are distinct properties. Tokenized money-market or bond products provide a setting for testing how these properties align.
b) Research question
How does usable liquidity in tokenized financial assets vary across trading hours, weekends, and redemption windows?
c) Data, methods and approach
Pilot data access for a small set of comparable products with public trading or quote data. Combine time-stamped spreads, executable depth where available, and transaction records with official transfer and redemption rules. Separate trades from transfers, avoid treating wallet counts as investor counts, and report missing liquidity measures. Compare time windows within products before making cross-product comparisons.
d) Expected contribution
Evidence on the conditions under which continuous token transfer translates into an investor's practical ability to buy, sell, or redeem.
DA-018Does fractional tokenization improve portfolio choices?Portfolio allocation experiment
a) Background and research problem
Tokenized real-world assets are often presented as a way to lower minimum investment sizes. Lower entry thresholds may enable diversification, but the tokenization label may also alter perceived novelty and risk. Separating the economic feature from its technological framing is necessary to assess investor behavior.
b) Research question
Do lower minimum investment sizes improve diversification, and does a tokenization label independently affect allocation choices?
c) Data, methods and approach
Use a factorial experiment with fictional asset portfolios, independently varying minimum investment size and tokenized versus conventional presentation. Keep expected returns, disclosed risks, and custody information constant. Measure concentration, understanding of ownership, and perceived risk under a fixed budget. Pretest the investment task, justify sample size, and distinguish hypothetical portfolio choices from subsequent investment performance.
d) Expected contribution
Evidence separating the behavioral effect of fractional access from the effect of presenting an investment as a tokenized asset.
DA-019Tokenized bond issuance: which efficiencies are demonstrated?Paired issuance case study
a) Background and research problem
Tokenization may change bond issuance and settlement processes, but a faster technical step does not necessarily reduce total issuance costs or the number of intermediaries. A focused comparison can distinguish implemented process improvements from projected benefits and costs shifted elsewhere in the transaction.
b) Research question
Which issuance and settlement efficiencies are supported by documented evidence from tokenized bond transactions?
c) Data, methods and approach
Compare 3–5 tokenized issuances with conventional transactions matched as closely as possible by issuer, jurisdiction, size, and complexity. Use prospectuses, issuer reports, infrastructure documentation, and practitioner interviews if accessible. Map process steps, settlement timing, intermediaries, and disclosed costs. Separate one-off pilot expenses from recurring costs and report unavailable information instead of assuming savings.
d) Expected contribution
A process-based evaluation of asset-tokenization benefits and a measurement framework for future comparisons of bond-market infrastructure.