Presentation
Debunking the ”5-8” and “15”: a Proposed Framework for Sample Size Selection
Event Type
Oral Presentations
TimeWednesday, June 8th2:00pm - 2:30pm EDT
Location
DescriptionAre sample sizes for Formative and Summative studies rules, a best guess, or guidelines that can be bent? Are you curious? Then this presentation will be a great discussion starter. Elise Rukavina, Riley Gervais, and Adam Czibur will discuss the background of the very familiar 5 – 8 (Formative) and 15+ (Summative) sample sizes, explore aspects and considerations that are lacking in current methodology, and provide recommendations for sample size determination that is based on product risk.
The investigation of the historical background on where the standard “5 – 8” and “15+” sample sizes and their justification came from has revealed gaps in the real-world application of these standards, often leaving Human Factors Professionals using these exact sample sizes without further considerations.
This current thinking on sample sizes will be challenged to broaden its perspective through posing thought-provoking questions and issues related to the current standards and expectations, with a goal of opening up the conversation and dialogue around process improvements and providing the audience with content to reflect on.
With the impossible nature of quantifying or predicting human error, conducting iterative product design and evaluation along with analyzing critical use-related errors and residual risk trends can shine light on the journey of use-related risk-reduction in the product. The value of utilizing these powerful tools to drive selection of appropriate formative and summative testing sample sizes will be discussed in this presentation.
To support the discussion, the similarities between software and human factors risk-reduction will also be covered to convey a nondeterministic understanding and approach to identifying and addressing use-related risk to improve safe and effective use.
Concluding this presentation, a consolidated approach for deriving sample sizes based on product risks will be discussed, including key points relating to extensive risk reduction activities and their impact on Formative and Summative sample sizes.
The investigation of the historical background on where the standard “5 – 8” and “15+” sample sizes and their justification came from has revealed gaps in the real-world application of these standards, often leaving Human Factors Professionals using these exact sample sizes without further considerations.
This current thinking on sample sizes will be challenged to broaden its perspective through posing thought-provoking questions and issues related to the current standards and expectations, with a goal of opening up the conversation and dialogue around process improvements and providing the audience with content to reflect on.
With the impossible nature of quantifying or predicting human error, conducting iterative product design and evaluation along with analyzing critical use-related errors and residual risk trends can shine light on the journey of use-related risk-reduction in the product. The value of utilizing these powerful tools to drive selection of appropriate formative and summative testing sample sizes will be discussed in this presentation.
To support the discussion, the similarities between software and human factors risk-reduction will also be covered to convey a nondeterministic understanding and approach to identifying and addressing use-related risk to improve safe and effective use.
Concluding this presentation, a consolidated approach for deriving sample sizes based on product risks will be discussed, including key points relating to extensive risk reduction activities and their impact on Formative and Summative sample sizes.
Authors
