Expert vs. novice: Problem decomposition/recomposition in engineering design

Abstract

The purpose of this research was to investigate the differences of using problem decomposition and problem recomposition among dyads of engineering experts, dyads of engineering seniors, and dyads of engineering freshmen. Fifty participants took part in this study. Ten were engineering design experts, 20 were engineering seniors, and 20 were engineering freshmen. Participants worked in dyads to complete an engineering design challenge within an hour. The entire design process was video and audio recorded.After the design session, members participated in a group interview.This study used protocol analysis as the methodology. Video and audio data were transcribed, segmented, and coded. Two coding systems including the FBS ontology and "levels of the problem" were used in this study. A series of statistical techniques were iv used to analyze data. Interview data and participants' design sketches also worked as supplemental data to help answer the research questions.By analyzing the quantitative and qualitative data, it was found that students used less problem decomposition and problem recomposoition than engineer experts in engineering design. This result implies that engineering education should place more importance on teaching problem decomposition and problem recomposition. Students were found to spend less cognitive effort when considering the problem as a whole and interactions between subsystems than engineer experts. In addition, students were also found to spend more cognitive effort when considering details of subsystems. These results showed that students tended to use dept-first decomposition and experts tended to use breadth-first decomposition in engineering design. The use of Function (F), Behavior (B), and Structure (S) among engineering experts, engineering seniors, and engineering freshmen was compared on three levels. Level 1 represents designers consider the problem as an integral whole, Level 2 represents designers consider interactions between subsystems, and Level 3 represents designers consider details of subsystems. The results showed that students used more S on Level 1 and 3 but they used less The purpose of this research was to investigate the design behavior. A commonly used design strategy called "problem decomposition and recomposition" was the focus of the study. Dyads of engineering experts, dyads of engineering seniors, and dyads of engineering freshmen completed an engineering design challenge within an hour. The entire design process was video and audio recorded. After the design session, participants participated in a group interview.The conversation and behavior from the design were transcribed, segmented, and coded. Two coding systems including the FBS ontology and "levels of the problem" were used in this study. A series of statistical techniques were used to analyze data. Interview data and participants' design sketches also worked as supplemental data to help answer the research questions.There were a series of interesting findings that emerged from analyzing the quantitat...

Protocol Analysis in Engineering Design Education Research: Observations, Limitations, and Opportunities

“…Participants are typically presented with an intersection they are familiar with, and they are asked to include any relevant diagrams and assumptions they made during their design (Adams et al, 2003;Atman and Bursic, 1996;Cardella et al, 2002;Cardella et al, 2008;Roberts et al, 2007). • In the window problem participants design a device which could provide assistance to a person with a disability to open a stuck, double-hung window without relying on electric power (Gero et al, 2013;Kannengiesser & Gero, 2017;Lammi & Gero, 2011, Lammi, 2011Song & Becker, 2014;Wells et al, 2016;Williams et al, 2012). This specific task can be attempted by fairly novice designers.…”

Quantifying Differences Between Professional Expert Engineers and Engineering Students Designing: Empirical Foundations for Improved Engineering Education

“…While each of these methods produced interesting results, the most promising method is protocol studies. It has become the basis of the current cognitive study of designers (Atman, et al 2008;Badke-Schaub et al 2007;Becker & Mentzer, 2012;Christensen & Schunn 2007;Gericke, et al 2007;Gero, Kan & Jiang, 2014;Kavakli & Gero, 2002;McDonnell & Lloyd, 2007;McNeill, et al, 1998;Song, 2014;Suwa, et al, 1998;Suwa, Gero & Purcell, 2000;Williams, et al, 2013).…”

“…FBS Processes This project builds on previous NSF-funded projects that looked at the longitudinal development of design cognition of undergraduate engineering students across two contiguous years (Williams, Lee, Gero & Paretti, 2013). Results from a pilot study at Utah State University show there is a significant gap between the cognitive behavior of novice and professional expert engineering designers (Song, 2014). This project makes use of those results and focuses on gaps.…”

Board 33: Empirical Foundations for Improved Engineering Education: Differences Between Engineering Students and Professional Expert Engineers while Designing

“…Building on previously NSF-funded projects that looked at the longitudinal development of design cognition of undergraduate engineering students across two contiguous years (Williams, Lee, Gero & Paretti, 2013), this research seeks to further understand undergraduate student design cognition, but more importantly, to establish an inventory of expert design cognition. Results from a pilot study at a western land-grant university show there is a significant gap between the cognitive behavior of novice and professional expert engineering designers (Song, 2014). Additionally, a meta-analysis of design cognition studies indicated that there are commonalities, as well as differences, between students and professional designers (Gero, Kannengiesser & Pourmohamadi, 2012).…”

“…Cognitive studies fall into five main methodological categories: protocol studies, interviews (Cross & Cross, 1998), input-output experiments; where the designer is treated as a black box which produces the behaviors in the outputs for changes in inputs (Purcell, Williams, et al, 1993), anthropological studies (Lopez-Mesa & Thompson, 2006), and questionnaires. While each of these methods produce intriguing results, protocol studies have become the basis of the current cognitive study of designers (Atman, et al 2008;Badke-Schaub et al 2007;Becker & Mentzer, 2012;Christensen & Schunn 2007;Gericke, et al 2007;Gero, Kan & Jiang, 2014;Kavakli & Gero, 2002;McDonnell & Lloyd, 2007;McNeill, et al, 1998;Song, 2014;Suwa, et al, 1998;Suwa, Gero & Purcell, 2000;Williams, et al, 2013).…”