Jin, Q., & Imbrie, P., & Lin, J. J., & Chen, X. (2011, June), A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering  Paper presented at 2011 ASEE Annual Conference & Exposition, Vancouver, BC. 10.18260/1-2--17352

\bibitem{asee_peer_17352}
  Jin, Q., \& Imbrie, P., \& Lin, J. J., \& Chen, X. (2011, June), \emph{A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering}
  Paper presented at 2011 ASEE Annual Conference \& Exposition, Vancouver, BC.
  10.18260/1-2--17352

Qu Jin, P.K. Imbrie, Joe J.J. Lin, and Xingyu Chen.  "A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering".  2011 ASEE Annual Conference & Exposition, Vancouver, BC, 2011, June.  ASEE Conferences, 2011.  https://peer.asee.org/17352 Internet.  08 Jan, 2025

\bibitem{asee_peer_17352}
  Qu Jin, P.K. Imbrie, Joe J.J. Lin, and Xingyu Chen.
  "A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering".
  \emph{2011 ASEE Annual Conference \& Exposition, Vancouver, BC, 2011, June}.
  ASEE Conferences, 2011.
  https://peer.asee.org/17352 Internet.  08 Jan, 2025

@INPROCEEDINGS{asee_peer_17352
author = "Qu Jin, P.K. Imbrie, Joe J.J. Lin, and Xingyu Chen"
title = "A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering"
booktitle = "2011 ASEE Annual Conference \& Exposition"
year = "2011"
month = "June"
address = "Vancouver, BC"
publisher = "ASEE Conferences"
note = {https://peer.asee.org/17352}
number = {10.18260/1-2--17352}
}

TY  - CPAPER
AB  -                    A Mutli-Outcome Hybrid Model for Predicting                          Student Success in EngineeringMany statistical modeling methods have been studied in predicting success outcomes of collegestudents, such as linear regression, logistic regression, and discriminate analysis. These methodsare principally used to predict single outcomes (e.g. retention in engineering). Recently, neuralnetwork models have been developed to predict student’s outcomes and they are shown to havebetter prediction performances over more traditional methods (e.g., SEM, Logistic Regression).One unique feature of the neural networks is they can predict multiple outcomes in the samemodel.First year retention and GPA are two of the most widely used success outcomes of engineeringstudents. The important factors identified using a model with these two outcomes may betterexplain what success means to engineering students. The purpose of the study is to design neuralnetwork models predicting first year retention and GPA of engineering students in the samemodels, to evaluate the predicting performances of the multi-outcome models, and to comparethe important factors identified by the models with those identified by single-outcome models.Participants in the study included 1500 first year engineering students enrolled in a largeMidwestern university. Cognitive input variables included high school GPAs, standardized testscores, and the grades and number of semesters in math, science and English courses in highschool. Affective input variables were measured using the Student Attitudinal SuccessInstrument (SASI), which include: nine major factors: leadership, deep learning, surface learning,teamwork, academic self-efficacy, motivation, metacognition, expectancy-value, and majordecision.The research questions to be answered are:   1) To what extent do the multi-outcome models improve student success predictions?   2) How are the factors identified by the multi-outcome models for success compared with       the factors identified by single-outcome models?To be presented are:   1) The prediction performances of the multi-outcome neural network models.   2) Comparison of the factors identified by the multi-outcome and single-outcome models.
AU  - Qu Jin
AU  - P.K. Imbrie
AU  - Joe J.J. Lin
AU  - Xingyu Chen
CY  - Vancouver, BC
DA  - 2011/06/26
PB  - ASEE Conferences
TI  - A Multi-Outcome Hybrid Model for Predicting Student Success in Engineering
UR  - https://peer.asee.org/17352
DO  - 10.18260/1-2--17352
ER  -