Where The Evidence Comes From — Meta-Analysis And The Research Substrate

The distribution of correlations or d-values as reported across a research literature prior to artifact correction — its observed mean and observed variance — confounding true and artifactual components and appearing to show conflicting findings. (Building trustworthy cumulative knowledge that moves the world.)

started 0 · finished 0 (claimed) · not yet measured (verified) · no data (n<5)

Derived from: Where The Evidence Comes From — Meta-Analysis And The Research SubstrateT1

Ordered tasks (16) — this is what a project auto-creates

#1Observed Effect Size Distribution (Mean and Variance)HighTo Do
Describes what you actually confront in the raw literature: a distribution of correlations or d-values that mixes true effect, sampling error, and artifacts into an apparently contradictory mess. You learn to read it as a confounded signal, not a verdict.
Assignee: Unassigned · due 0 days after project start · takes 14 days
#2Substantive Domain Antecedents (Person/Context/Leadership)LowTo Do
Identifies the driver variables — person, context, and leadership characteristics — that a meta-analysis treats as the causal starting point, and shows why their construct definitions govern what you can legitimately pool.
Assignee: Unassigned · due 14 days after project start · takes 3 days
#3Substantive Mediating MechanismsLowTo Do
The psychological and relational variables — satisfaction, commitment, trust, LMX, strategy use — that carry effects from antecedents to outcomes, and how meta-analysis reconstructs those pathways from primary correlations.
Assignee: Unassigned · due 17 days after project start · takes 3 days
#4Substantive Domain OutcomesLowTo Do
Address the dependent variables a meta-analysis ultimately estimates — performance, engagement, strain, citizenship behavior — and how outcome definition and measurement source shape the pooled effect.
Assignee: Unassigned · due 20 days after project start · takes 3 days
#5True Effect Size Distribution (Mean and Variance)HighTo Do
Define the target every meta-analysis aims at: the real distribution of effects that would emerge from flawless, infinite-sample studies. You learn to think of an effect not as a single number but as a mean with genuine spread.
Assignee: Unassigned · due 23 days after project start · takes 14 days
#6Study Design ArtifactsHighTo Do
Catalogs the specific imperfections in primary studies — unreliable measures, restricted ranges, dichotomized variables, finite samples — that systematically distort what any single study reports. You learn to see each observed correlation as already corrupted before you touch it.
Assignee: Unassigned · due 37 days after project start · takes 14 days
#7Sampling ErrorHighTo Do
Isolates the single largest source of apparent disagreement across studies: the random scatter introduced by finite samples. You learn why small-sample studies bounce around a true value without any real effect differing between them.
Assignee: Unassigned · due 51 days after project start · takes 14 days
#8Systematic Attenuation / Underestimation of True RelationshipMediumTo Do
The directional bias — always downward — that measurement error, dichotomization, and range restriction impose on the average observed effect. You learn why the literature systematically understates how strong relationships really are.
Assignee: Unassigned · due 65 days after project start · takes 7 days
#9Artifactual / Spurious VarianceMediumTo Do
Address the fraction of between-study variance that comes from artifacts rather than real heterogeneity. You learn to decompose total observed variance and ask how much is left once the artifacts are removed.
Assignee: Unassigned · due 72 days after project start · takes 7 days
#10Artifact Correction Procedure (Meta-Analysis)MediumTo Do
The analytic engine of psychometric meta-analysis: the procedures that strip artifactual distortion from observed statistics to recover true parameters. You learn what the corrections do and what they require of your data.
Assignee: Unassigned · due 79 days after project start · takes 7 days
#11Valid Moderator DetectionLowTo Do
Focus on about finding the real variables — populations, settings, treatments — that make an effect genuinely differ, and separating them from statistical mirages. You learn when a moderator search is warranted and how to conduct it honestly.
Assignee: Unassigned · due 86 days after project start · takes 3 days
#12Capitalization on ChanceLowTo Do
Names the specific statistical trap that undermines moderator analysis: testing many study-characteristic hypotheses against few studies inflates the odds of spurious 'findings.' You learn to recognize and quantify the risk.
Assignee: Unassigned · due 89 days after project start · takes 3 days
#13Cumulative Scientific KnowledgeMediumTo Do
The ultimate payoff: stable, generalizable facts assembled across many studies into coherent theory. You learn what distinguishes genuine accumulation from a pile of disconnected results.
Assignee: Unassigned · due 92 days after project start · takes 7 days
#14Validity of Generalized InferencesLowTo Do
Concerns whether the conclusions you draw from a synthesis are actually trustworthy and appropriately scoped. You learn to distinguish a defensible generalization from an overreach.
Assignee: Unassigned · due 99 days after project start · takes 3 days
#15Rigor of Synthesis ProcessLowTo Do
The workflow that makes a synthesis credible: problem formulation, comprehensive search, systematic coding, sound analysis, and transparent reporting. You learn that method quality upstream determines trust downstream.
Assignee: Unassigned · due 102 days after project start · takes 3 days
#16Utility for Policy and PracticeLowTo Do
Tells you what separates a synthesis that decision-makers actually use from one that sits uncited in a journal archive. It covers relevance, accessibility, and the pathway from pooled estimate to defensible action.
Assignee: Unassigned · due 105 days after project start · takes 3 days