Date of Award

August 2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Mechanical Engineering

Committee Member

Marissa L Shuffler

Committee Member

Cameron J Turner

Committee Member

Christiaan J Paredis


Functions are a critical concept in engineering that support problem clarification and early-stage conceptual design. Function modeling, like other early-stage design tools, relies on subjective inputs from designers and is influenced by individual differences in designers. While research on function modeling has investigated topics such as modeling representations, model construction techniques, model use for conceptual design, and modeling behaviors, the role of individual differences is largely unexplored. This research aims to investigate how cognitive attributes and other individual differences influence the function modeling process and outcomes. Due to limited insight available about the nature of the relationship between individual attributes and function modeling, a theory building approach is adopted. An input-process-output framework is developed to systematically identify measures that will represent different aspects of function modeling. Four cognitive attribute measures are selected for testing: (1) systemizing quotient, (2) risk propensity, (3) goal orientation, and (4) concept design thinking style inventory. A two-part protocol study in conducted. Participants are first asked to complete the set of surveys intended to capture the input measures in the form of individual differences. Following that, a protocol study session is scheduled where a video recording of the function modeling activity is collected. A protocol analysis is used to code videos into structured data, which are subsequently analyzed to generate process measures. The finalized Function Structure model is converted into a bipartite graph, which is then used to calculate graph complexity metrics. These along with a rubric-based evaluation of the model are used as output measures. The input, process, and output measures are then compared using an exhaustive pairwise regression analysis and a multiple regression analysis. Correlations highlighted from the regression analyses are discussed.

The learning goal orientation measure is found to be correlated with frequency of reading the problem statement, pointing towards a tendency to internalize the problem. Preference for different concept design thinking styles is found to correlate with different aspects of the modeling process. The “inquiring” thinking style is correlated with labeling flows, while the “exploring” thinking style is correlated with the number of modeling activities. Risk propensity is found to be inversely correlated with functions generated in the modeling process and directly correlated with the level of interconnection in the final model. Elements generated during the modeling process and the chaining methods used for introducing elements to the models are also correlated with the level of interconnection in the final model.

Following a discussion of potential relationships observed, a targeted experiment is designed and conducted to investigate the relationship between risk propensity and level of interconnection in the final model. Results shows that risk propensity does not correlate with function model size but does affect the level of interconnection in the model. To conclude, the correlations found are summarized and limitations of the study are discussed. A theoretical model of function modeling is proposed using the relationships discovered in this research. Finally, five future research questions are identified, corresponding hypotheses are formulated, and potential experiments are discussed. Applications of the research methods to other design tools are also explored as future work.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.