Can undergraduates learn programming with a “Virtual Professor”? Findings from a pilot implementation of a blended instructional strategy

Dan Cernusca; Clayton E Price

Download Paper | Permalink

Conference: 2013 ASEE Annual Conference & Exposition
Location: Atlanta, Georgia
Publication Date: June 23, 2013
Start Date: June 23, 2013
End Date: June 26, 2013
ISSN: 2153-5965
Conference Session: Computer Based Grading and Learning Styles
Tagged Division: Computers in Education
Page Count: 14
Page Numbers: 23.268.1 - 23.268.14
DOI: 10.18260/1-2--19282
Permanent URL: https://peer.asee.org/19282
Download Count: 296

Request a correction

Paper Authors

biography

Dan Cernusca Missouri University of Science & Technology

visit author page

Dr. Dan Cernusca is Instructional Design specialist in the Department of Global Learning at the Missouri University of Science and Technology. He received his Ph.D. in Information Science and Learning Technologies in 2007 from University of Missouri – Columbia. He also holds a B.S. and a Ph.D. from the University of Sibiu, Romania with a specialization in manufacturing technologies and respectively cutting-tools design. His research interests include Design-Based Research in technology-enabled learning contexts, technology-mediated problem solving, applications of dynamic modeling for learning of complex topics, and the impact of epistemic beliefs on learning with technology.

visit author page

biography

Clayton E Price Missouri University of Science and Technology

visit author page

Professor Price has varied interests in the sciences, having earned degrees in geology/geophysics, mathematics, and computer science. He has taught at S&T for 32 years, currently in the computer science department. He teaches introductory programming classes in C++ and the numerical analysis courses. As assistant to the chairman, he advises freshmen and transfer students. Price's interests center on his teaching and investigating new paradigms for delivering a university education. He has recently developed on-line material for his core programming course, and hopes to expand this effort in the future.

visit author page

Download Paper | Permalink

Abstract

Can undergraduates learn programming with a “Virtual Professor”? Findings from a pilot implementation of a blended instructional strategyThis paper presents the main findings from the pilot implementation of a blendedinstructional strategy in one section of a multi-section course of introduction toprogramming with C++. The implemented strategy blended pre-recorded online lecturesand homework assignments, with one weekly optional face-to-face meeting dedicated toanswering students’ questions related to the assigned instructional task for that period.The same instructor taught both the blended instruction section (treatment group) and onesection using traditional face-to-face lecture approach (control group). In addition, boththe blended and the traditional sections were involved in face-to-face laboratoryactivities.The focus of this study was two-fold: a) to determine whether the blended formatnegatively impacted students’ final performance, and b) to identify the major predictorsof final performance in this course so that we will be able to use them in the future tomonitor course effectiveness throughout the semester. The data was collected usingonline surveys, one at the entry point and second during the last week of the course. Thevariables measured in the entry survey (self-efficacy, perceived engagement and difficult)were used to test the homogeneity of control and treatment group. We found nostatistically significant differences between the control and treatment groups at the entrypoint. A one-way ANOVA analysis indicated no statistically significant differences infinal course performance (final percentage score) between the control (traditionallecture) and the treatment (blended lecture) groups. Overall, the “virtual professor”and the “live” professor performed at a comparable level.To move beyond the overall analysis we proposed and tested a path analysis model that reflects the major predictors of instructional performance as proposed by the educational research literature. The proposed predictors shown inthe path analysis model were measured in the exit survey using validated scales reportedin previous educational research studies. The minimum discrepancy measured by chi-square was not significant (χ2 (3) = .26, p = .97) which indicates that there is an adequateclose fit between the hypothesized model and the perfect fit model.The adequacy of fit is also strengthened by the value of the ratio of the minimumdiscrepancy to the degrees of freedom, CDMIN/DF = .09, which is significantly smallerthan 2.0 as recommended in the literature. The major goodness-of-fit measures supporteda good fit: CFI = .99, RMSEA = .001 and Holter (p = .05) = 1320. The analysis of pathcoefficients indicated several expected findings. First, perceived engagement was asignificant positive predictor of self-efficacy, and self-efficacy was a significant positivepredictor for the final grade. Second, perceived difficulty was a significant negativepredictor for the final grade, and this impact was partially mediated by self-efficacy. Asfor unexpected findings, the enrollment section (1-traditional or 2-blended lecture) had astatistically significant negative impact on the final grade. This finding suggests that theuse of blended lecture has the potential to generate lower grades in the course in a fullimplementation strategy.

Citation
Format

Cernusca, D., & Price, C. E. (2013, June), Can undergraduates learn programming with a “Virtual Professor”? Findings from a pilot implementation of a blended instructional strategy Paper presented at 2013 ASEE Annual Conference & Exposition, Atlanta, Georgia. 10.18260/1-2--19282

TY  - CPAPER
AB  -          Can undergraduates learn programming with a “Virtual          Professor”? Findings from a pilot implementation of a                     blended instructional strategyThis paper presents the main findings from the pilot implementation of a blendedinstructional strategy in one section of a multi-section course of introduction toprogramming with C++. The implemented strategy blended pre-recorded online lecturesand homework assignments, with one weekly optional face-to-face meeting dedicated toanswering students’ questions related to the assigned instructional task for that period.The same instructor taught both the blended instruction section (treatment group) and onesection using traditional face-to-face lecture approach (control group). In addition, boththe blended and the traditional sections were involved in face-to-face laboratoryactivities.The focus of this study was two-fold: a) to determine whether the blended formatnegatively impacted students’ final performance, and b) to identify the major predictorsof final performance in this course so that we will be able to use them in the future tomonitor course effectiveness throughout the semester. The data was collected usingonline surveys, one at the entry point and second during the last week of the course. Thevariables measured in the entry survey (self-efficacy, perceived engagement and difficult)were used to test the homogeneity of control and treatment group. We found nostatistically significant differences between the control and treatment groups at the entrypoint. A one-way ANOVA analysis indicated no statistically significant differences infinal course performance (final percentage score) between the control (traditionallecture) and the treatment (blended lecture) groups. Overall, the “virtual professor”and the “live” professor performed at a comparable level.To move beyond the overall analysis we proposed and tested a path analysis model                                                               that reflects the                                                               major predictors                                                               of instructional                                                               performance as                                                               proposed by the                                                               educational                                                               research                                                               literature.                                                                         The proposed                                                                         predictors shown inthe path analysis model were measured in the exit survey using validated scales reportedin previous educational research studies. The minimum discrepancy measured by chi-square was not significant (χ2 (3) = .26, p = .97) which indicates that there is an adequateclose fit between the hypothesized model and the perfect fit model.The adequacy of fit is also strengthened by the value of the ratio of the minimumdiscrepancy to the degrees of freedom, CDMIN/DF = .09, which is significantly smallerthan 2.0 as recommended in the literature. The major goodness-of-fit measures supporteda good fit: CFI = .99, RMSEA = .001 and Holter (p = .05) = 1320. The analysis of pathcoefficients indicated several expected findings. First, perceived engagement was asignificant positive predictor of self-efficacy, and self-efficacy was a significant positivepredictor for the final grade. Second, perceived difficulty was a significant negativepredictor for the final grade, and this impact was partially mediated by self-efficacy. Asfor unexpected findings, the enrollment section (1-traditional or 2-blended lecture) had astatistically significant negative impact on the final grade. This finding suggests that theuse of blended lecture has the potential to generate lower grades in the course in a fullimplementation strategy.
AU  - Dan Cernusca
AU  - Clayton E Price
CY  - Atlanta, Georgia
DA  - 2013/06/23
PB  - ASEE Conferences
TI  - Can undergraduates learn programming with a “Virtual Professor”? Findings from a pilot implementation of a blended instructional strategy
UR  - https://peer.asee.org/19282
DO  - 10.18260/1-2--19282
ER  -