Survey Response Classification for Non-Random Noise Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for quantifying drug use and related behaviors in online survey data suffer from inaccuracies due to non-random noise, which current statistical techniques struggle to effectively identify and remove, leading to unreliable estimates.
Innovation Solution
A method involving a survey questionnaire with non-existent products, second-generation interval null hypothesis, confidence intervals, and second-generation p-values to classify responses as noise, signal, or indeterminate, followed by computer simulations to refine estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional statistical techniques (outlier removal, smoothing, careless response removal) are used to handle noise in survey data, then the processing is simple and quick, but the accuracy and reliability of statistical estimates deteriorate due to inability to effectively identify non-random noise
Solution Approach 1:
The patent segments the survey questionnaire into multiple sections with different types of questions (attention-grabbing items, filler questions, scale questions) and segments responses into categories (inattentive, careless, random noise, signal) using different statistical tests for each segment, allowing targeted noise removal while preserving valid data
Solution Approach 2:
The patent introduces second-generation p-values as an intermediary statistical measure that bridges traditional hypothesis testing and effect size estimation, providing a more nuanced classification of responses that improves accuracy without requiring completely new methodologies
2Ease of operation
If attention-grabbing items and simple statistics (addition, standard deviations, correlation calculations) are used to classify inattentive response patterns, then the method is easy to implement, but the reliability of noise identification deteriorates due to oversimplification
Solution Approach 1:
The patent changes the statistical parameters used for classification from simple descriptive statistics to second-generation p-values that incorporate effect size, variability, and hypothesis testing, improving reliability while maintaining computational feasibility through automated procedures
3Productivity
If binary classification (noise or signal) is used for survey responses, then the classification is simple and quick, but the accuracy deteriorates due to inability to capture indeterminate or ambiguous responses
Solution Approach 1:
The patent segments the classification system into three distinct categories (noise, signal, indeterminate) based on second-generation p-value thresholds, allowing ambiguous responses to be identified separately rather than forced into binary categories, improving precision without significantly increasing processing time
Solution Approach 2:
The patent applies different classification criteria and thresholds to different types of questions and response patterns, allowing locally optimized classification that improves precision for each question type while maintaining overall processing efficiency
Data Source
AI summary
The invention outlined here is an innovative approach to increasing the accuracy of survey responses by combining novel classification of inaccurate survey responses as noise with state-of-the-art statistical techniques. This invention innovatively combines 1) a novel method to quantify inaccurate survey responses, with 2) statistical distribution assessment of variability to quantify bounds of classification, and 3) statistical classification of responses into at least 3 categories of inaccuracy. This invention is implemented by a computer and will generate estimates of variability, which are subsequently utilized in classification. These estimates can be effectively used to classify field responses as either signal, noise, or indeterminate and be used to probabilistically adjust numerical calculations of field response in surveys.