Download presentation
Presentation is loading. Please wait.
Published byἹππολύτη Ζωγράφος Modified over 6 years ago
1
Business Statistics and Operations Management: Teaching Strategies and Assessment of Student Learning in Traditional vs. Online Classes presented by Dr. Burcu Adivar joint work with Dr. Li Chen Fayetteville State University College of Business and Economics Department of Management, Marketing & Entrepreneurship 2017 DSI Annual Conference Nov 18, 2017 Washington, DC
2
Outline Objective Description of the study
Effects of age, gender and race on learning outcomes of the quantitative courses Comparative analysis of online vs. face-to-face teaching Correlation analysis Multivariate analysis Conclusions
3
Objective To better understand student learning experience in quantitative courses To measure the effect of quantitative courses’ learning outcomes on the CLA exam performance To use analytics before proposing curriculum changes and redesigning assessment of learning outcomes To explain the determinant of CLA scores and use predictive models to increase student and faculty awareness for better preparedness
4
Description of the Data
MGMT335 Operations Management is a core course for students in College of Business and Economics BADM 216 Business Statistics is the prerequisite course Face-to-face and online sections are offered every Fall, Spring and summer We are using Canvas system and same syllabus, same text for all sections in our teaching. Canvas system provides number of page views number of participation in the course website number of messages sent to the instructor online amount of time he has spent in the course website course grade, student pictures (gender, age, race) Number of ontime, late and missed assignments Banner system provides student transcripts statistics grade Dean’s Office provides CLA score cards for students Total score 154 Student records from 5 sections of OM: Fall 2016 (F2F, ONL), Spring 2017(F2F, ONL), Summer 2016 (F2F)
5
Age Factor STAT_Grade OM_Grade CLA_Score Age N Mean Min Max % of Total
Variance Y 110 2.94 1 4 71.89% 0.86 78.8 3 103 72.06% 282 34 1029 746 1347 77.84% 20663 A 44 2.88 28.11% 1.06 76.4 3.15 102 27.94% 526 10 996 782 1187 22.16% 19688 All 154 2.92 100.00% 0.91 78.1 350 1021 20171
6
Gender Factor STAT_Grade OM_Grade CLA_Score Gender N % of Total Mean
Std Dev Min Max F 94 61.33% 2.94 0.95 1 4 61.40% 78.6 17.9 3.15 103 29 66.57% 1032 134 782 1281 M 60 38.67% 2.9 0.96 38.60% 77.4 20.1 3 102 15 33.43% 1002 159 746 1347 All 154 100.00% 2.92 78.1 18.7 44 1021 142
7
Race Factor STAT_Grade OM_Grade CLA_Score Race N Mean Min Max
% of Total Variance A 95 2.71 1 4 68.72% 0.83 73.7 3 102 70.21% 400 28 1003 746 1347 66.35% 17383 W 3.57 26.74% 0.25 91.4 70 103 25.65% 69.3 11 1104 908 1281 28.71% 20579 H 5 3.4 2 4.55% 0.8 82.7 98 4.14% 184 1044 901 1187 4.94% 40898 All 128 2.92 100.00% 78 370 41 1032 19951
8
Online vs. Face to face STAT_Grade OM_Grade CLA_Score Type N Mean Min
Max % of Total Variance F2F 65 2.77 1 4 40.00% 0.77 78.7 3.15 99.7 42.51% 260 27 1015 746 1347 60.99% 20432 ONL 89 3.03 60.00% 0.99 77.7 3 103 57.49% 419 17 1031 842 1271 39.01% 20835 All 154 2.92 100.00% 0.91 78.1 350 44 1021 20171
9
Page Views vs. F/O Graph Builder
10
Participations vs. F/O Graph Builder
13
Correlations Multivariate Page Views Participations On Time Missing
Activity time STAT_Grade CLA_Score OM_Grade 1.0000 0.5276 0.4661 0.0720 0.3519 0.4189 0.8060 0.2647 0.1174 0.5238 0.2587 0.1098 0.4344 0.0135 0.0078 0.2108 0.0348 0.3668 0.5257 0.4521 Multivariate
14
Correlation Probability
Page Views Participations On Time Missing Activity time STAT_Grade CLA_Score OM_Grade <.0001 0.3749 0.7078 0.8504 0.0009 0.4480 0.0716 0.0012 0.4779 0.8678 0.0002 0.1024 0.9236 0.1697 0.6679 0.0143 0.0021 Multivariate
15
Boosted Tree for CLA_Score
RSquare RMSE N 0.792 44 Term Number of Splits SS Portion STAT_Grade 12 0.2642 OM_Grade 14 0.1643 Participations 0.1555 Page Views 16 0.0981 Number of Messages 10 0.0726 Activity time 0.0567 communication 0.0507 On Time 11 0.0420 Age 24 0.0410 Race 0.0188 Type 0.0168 Late 3 0.0123 Attendance 1 0.0036 Gender 2 0.0018 Missing 0.0015
16
Prediction Profiler Boosted Tree for CLA_Score
17
Effect Summary Summary of Fit Fit Group > Response OM_Grade
Source LogWorth PValue Missing 25.537 STAT_Grade 11.221 Type 8.976 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) 154 Source DF Sum of Squares Mean Square F Ratio Lack Of Fit 52 2.4602 Pure Error 98 79.778 Prob > F Total Error 150 <.0001* Max RSq 0.8540 Fit Group > Response OM_Grade Fit Group > Response OM_Grade
18
Parameter Estimates Analysis of Variance Source DF Sum of Squares
Mean Square F Ratio Model 3 Error 150 120.2 Prob > F C. Total 153 <.0001* Parameter Estimates Term Estimate Std Error t Ratio Prob>|t| Intercept 20.88 <.0001* Missing -12.97 STAT_Grade 7.47 Type[F2F] 6.51 Fit Group > Response OM_Grade
20
Boosted Tree for OM_Grade
RSquare RMSE N 0.858 154 Term Number of Splits SS Portion Missing 24 0.4247 STAT_Grade 29 0.2307 Page Views 0.1060 On Time 19 0.0965 Participations 6 0.0690 communication 10 0.0243 Attendance 9 0.0213 Type 11 0.0102 Activity time 0.0055 Late 5 0.0052 Age 0.0034 Number of Messages 2 Gender 0.0000 Boosted Tree for OM_Grade
21
Prediction Profiler Boosted Tree for OM_Grade
22
Regression results
23
Conclusions Our data shows wide difference among students (for example, variance is large among students' OM and CLA grade, for page views, participation and communication, online students have larger range) Our data helps us identify key factors for students performance. (for example, the correlation analysis shows that page views, participation and on time submission have strong correlation with OM grade) Our analysis helps instructors. (for example, instructors can remind students early when they observe students' performance on those key factors are not good)
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.