Can it help me structure survey attributes and levels?

Yes, it offers specific guidance on orthogonality, attribute density, and how to handle nested or constrained randomization without compromising causal inference.

How does this skill help with statistical power analysis?

It provides closed-form formulas and simulation logic for calculating standard errors and effective sample sizes for both main Average Marginal Component Effects (AMCE) and complex interaction effects.

What are Type S and Type M errors in this context?

It helps researchers avoid 'Sign' (Type S) and 'Magnitude' (Type M) errors, which are common in underpowered designs and can lead to incorrect effect directions or exaggerated effect sizes.

Does it support specific statistical tools or R packages?

The skill provides logic compatible with industry-standard tools like cjpowR, DeclareDesign, FindIt, and cjbart for simulation-based diagnostics and heterogeneity detection.

Conjoint Design Expert

Name: Conjoint Design Expert
Author: scdenney

byscdenney

•

데이터 과학 및 ML

Facilitates the design, power analysis, and statistical specification of conjoint and factorial vignette experiments based on social science standards.

This skill provides specialized logic and methodological guidance for researchers conducting conjoint and factorial experiments. It assists in architecting attribute structures, calculating statistical power using closed-form formulas and simulation-based tools, and drafting rigorous pre-analysis plans. By integrating best practices for Average Marginal Component Effects (AMCE) estimation, interaction power, and treatment validation, it ensures experimental designs are statistically robust, theoretically sound, and aligned with current social science literature.

주요 기능

0115 GitHub stars

02Validation Logic: Frameworks for auditing experimental realism, attention checks, and salience-induced bias.

03Statistical Specification: Assistance writing regression formulas for AMCE, Marginal Means (MM), and interaction effects.

04Power Analysis: Sophisticated calculators for Effective N, AMCE precision, and interaction effect budgeting.

05Attribute Architecture: Expert guidance on orthogonality, randomization order, and managing respondent cognitive load.

06PAP Drafting: Support for documenting attribute restrictions and multiple-testing corrections in pre-analysis plans.

사용 사례

01Conducting a power analysis to determine the required sample size for detecting specific AMCE magnitudes.

02Drafting the methodological and statistical sections of a pre-analysis plan for an experimental study.

03Structuring an attribute-level table for a political science, economics, or marketing survey.

주요 기능

0115 GitHub stars

02Validation Logic: Frameworks for auditing experimental realism, attention checks, and salience-induced bias.

03Statistical Specification: Assistance writing regression formulas for AMCE, Marginal Means (MM), and interaction effects.

04Power Analysis: Sophisticated calculators for Effective N, AMCE precision, and interaction effect budgeting.

05Attribute Architecture: Expert guidance on orthogonality, randomization order, and managing respondent cognitive load.

06PAP Drafting: Support for documenting attribute restrictions and multiple-testing corrections in pre-analysis plans.

사용 사례

01Conducting a power analysis to determine the required sample size for detecting specific AMCE magnitudes.

02Drafting the methodological and statistical sections of a pre-analysis plan for an experimental study.

03Structuring an attribute-level table for a political science, economics, or marketing survey.