When Better Questions Do Not Guarantee Better Answers: Discovery Plane Theory and Managerial Transparency in Earnings Calls

Strategic management increasingly recognizes that decision quality depends on how a problem is formulated before alternatives are selected. Yet empirical work rarely separates the architecture of a question from the organizational response that follows it. This study develops a representation-relative operationalization of Discovery Plane Theory (DPT) for strategic questioning and tests whether structurally stronger questions elicit more transparent managerial answers. The design separates coordinate construction from outcome testing. First, 744 answerable questions from ECT-QA are used to calibrate lexical models against archive-grounded proxies for Discovery Complexity (DC) and Question Compression (QC). Cross-validated performance is nontrivial for DC (R² = .331, $r = .576$) and stronger for QC (R² = .655, $r = .809$). These frozen models are then transferred to 2,747 human-annotated earnings-call question–answer pairs from SubjECTive-QA covering 115 firms and 2007–2021. The outcome is a transparency index derived from human ratings of clarity, specificity, and relevance. In company-, year-, and quarter-adjusted models with company-clustered standard errors, neither calibrated DC ($B = -.013$, $p = .646$), calibrated QC ($B = .007$, $p = .718$), nor their interaction ($B = .002$, $p = .969$) predicts transparency. Question length, by contrast, remains strongly positive ($B = .042$, $p < .001$). The result identifies a theoretically important question–response decoupling: a manager’s answer is jointly produced by the epistemic structure of the question and the strategic governance of disclosure. The paper contributes an auditable two-stage method for studying strategic questions, clarifies the boundary between question quality and organizational answerability, and shows why knowledge availability cannot be equated with transparent response.