A paradigm for sequential experimentation

Abstract

m odel; H um an system interface.M ost human factors design problem s involve larg e data spaces with so man y factors that a single experim ent investigating all of these factors sim ultaneously is unreasonable. A n alternative is the use of sequential experim entation. In this paper, an integrated research paradigm for sequential experim entation is presented. Speci® c m ethods and procedures are described for the four steps: planning research strategy, selecting, describing and optim izing design varia bles. A hypothetical example is provided to illustrate the use of the research paradigm . The usefulness of this overall strategy for sequential experim entation in conducting ergonom ics research is discussed.

Human Factors Experimental Design and Analysis Reference

“…Sequential experimentation as described in Topic 24 is a critical part of the empirical model building repertoire of procedures. Han, Williges, and Williges (1997) provided an extensive flowchart paradigm accompanied by an example of using their paradigm for conducting sequential experimentation in a human factors problem related to a passenger seat design used in a transportation system. Their paradigm lists four major stages with major and minor considerations in each stage.…”

“…As shown on this slide, Han, Williges, and Williges (1997) recommend three major planning considerations dealing with defining the boundaries of the experimental space of interest, determining experimental procedural constraints and requirements, and developing a comprehensive recording tool that can be used across the series of experiments for investigating all the factors of interest. Stage 3 of the Han, Williges, and Williges (1997) paradigm is the major data collection and analysis phase of their sequential experimentation approach. Four of the eight major considerations in their paradigm flowchart are listed on this slide along with minor considerations within each major topic.…”

“…Comparability of data across experiments is assessed by evaluating common data points across experiments, and factor adjustments are made, if necessary. Han, Williges, and Williges (1997) paradigm is the final stage and deals with optimizing the interface design. Five major considerations are listed on this slide for interface optimization using empirical models resulting from the integrated database in Step 3.…”

“…The Han, Williges, and Williges (1997) paradigm should be used in conjunction with the Williges, Williges, and Han (1993) guidelines as discussed in Topic 24 for conducting sequential experimentation. As human factors and ergonomic researchers have more experience with using sequential research procedures, this initial paradigm and guidelines for sequential experimentation will need to be expanded.…”

“…Empirical Model Building. Han, Williges, and Williges (1997), Williges (1981), and Williges, Williges, and Han (1993) provide details on empirical model building in human factors research. Box and Draper (1987) and Myers and Montgomery (2002) Context/Purpose Determine if the median acceptability ratings given by non-professional photographers are higher than those of the professional photographers.…”