Performance Assessment of Faculties of Management Discipline From Student Perspective Using Statistical and Mining Methodologies

Chandrani Singh,Arpita Gopal,Santosh Mishra
International Journal of Data Engineering (IJDE) Volume (1): Issue (5) 63
Performance Assessment of Faculties of Management
Discipline From Student Perspective Using Statistical and
Mining Methodologies
Chandrani Singh singh.chandrani@gmail.com
Associate Professor, MCA Department
Sinhgad Institute of Business Administration and Research
Pune, Maharashtra - 411048
Arpita Gopal arpita.gopal@gmail.com
Director,MCA Department
Sinhgad Institute of Business Administration and Research
Pune, Maharashtra-411048
Santosh Mishra ssantosh.k.mishra@gmail.com
Sinhgad Institute of Business Administration
and Research
Pune,Maharashtra-411048
Abstract
This paper deals with Faculty Performance Assessment from student perspective using Statistical
Analysis and Mining techniques .Performance of a faculty depends on a number of parameters
(77 parameters as identified) and the performance assessment of a faculty/faculties are broadly
carried out by the Management Body ,the Student Community ,Self and Peer faculties of the
organization .The parameters act as performance indicators for an individual and group and
subsequently can impact on the decision making of the stakeholders. The idea proposed in this
research is to perform an analysis of faculty performance considering student feedback which can
directly or indirectly impact management’s decision, teaching standards and norms set by the
educational institute, understand certain patterns of faculty motivation, satisfaction, growth and
decline in future. The analysis depends on many factors, encompassing student’s feedback,
organizational feedback, institutional support in terms of finance, administration, research activity
etc. The statistical analysis and mining methodology used for extracting useful patterns from the
institutional database has been used to extract certain trends in faculty performance when
assessed on student feedback. The paper compares first the traditional approach with the
statistical approach and then justifies the usage of data mining classification technique for
deriving the results.
Keywords: Data Analysis, Mining, Clustering, Trend Extraction, Performance Prediction
1. INTRODUCTION
The applications of Data Mining in the field of higher education can truly be justified because
typical type of data mining questions used in the business world has counter part questions
relevant to higher education [2]. The need of Data Analysis and Mining in higher education is to
mine faculty and students data from various stakeholders’ perspective [7]. The methodology
adapted to design the system is dealt extensively in the previous paper [16].Initially 77
parameters were considered,50 faculties performance was assessed based on the feedback
obtained from various segments and averaged out to show the mean performance of Faculties
using traditional approach. The ongoing research on Faculty Assessment has enabled us to
increase our data size and implement segment slicing. The result generated in this paper is

strictly from student feedback. Around 3000 student records were taken into consideration. The
data was smoothened and profiled, inconsistent data was removed and the operational data
included two consecutive years student feedback from two institute’s of management discipline.
This data was then analyzed using conventional MS-Excel and the following pattern was derived
as shown in Figure 1.
Faculty Performance-Traditional
Approach
14%
11%
72%
3%
very good
good
satisfactory
poor
FIGURE 1: Faculty Performance – Traditional Approach
The accuracy of the result then was taken into rigorous consideration because the influence of
the other parameters on the faculty performance was missing considerably. The justification of
implementing statistical analysis and mining algorithms was required to extract intelligent
information and to perform complex calculations, trend analysis and sophisticated data modeling,
and reporting. The need was to identify critical information on the not so obvious data and extract
mission critical information and intelligence that would enable better decision by the academia.
This is an ongoing research work so comparative evaluation is behind the scope of this paper
since similar work has been performed only on monitoring student academic performance using
data mining technique.
2. DATA ANALYSIS AND MINING RATIONALE
The goal of higher education is to continually maintain quality and standards with the most
efficient procedures implemented for growth and the degree of quality teaching involves the
pertinent issues of how to enhance and evaluate it through overt and covert processes. Hence
the Data Mining processes for knowledge discovery is to subject various classification and
prediction procedures on the data. This helps institutes to predict certain trends of faculties in
terms of intellectual contribution, administrative services, and standards followed which cannot be
meted out using traditional approach.
3. CLASSIFICATION AND CLUSTERING
The classifier model used was the full training set and ZeroR algorithm was used to predict the
classified instances. The results of classification are as shown in Table 2. Initially incorrectly
classified instances was found to be around 28 % hence the data was again profiled to increase
the percentage of correctly classified instances.

=== Classifier model (full training set) ===
ZeroR predicts class value: satisfactory
Time taken to build model: 0.02 seconds
=== Evaluation on training set ===
=== Summary ===
Correctly Classified Instances 2890
Kappa statistic 0
Mean absolute error 0.0254
Root mean squared error 0.112
Relative absolute error 100%
Root relative squared error 100 %
Total Number of Instances 2890
TABLE 2: Classifier model (full training set)
Then clustering of the correctly classified data was performed using EM algorithm where clusters
were generated based on the parameter value and for every parameter cluster the percentage
of ratings were found out as shown in the table 4 and then the cumulative value was averaged
out to find out the mean and the ratings were represented using percentages. A snapshot of the
cluster formation which is an intermediate process is also shown in the table 3.
Row
Id.
Cluster
id
Dist
clust-
1
Dist
clust-2
Subject_
Knowledge
Teachin
g Ability
_with_u
se_of
_new_Te
aching_
Aids
Motivation
_Self_Stud
ents
Aggre_per
84 1 1.1136 4.3082 8 8 4 40
85 1 1.9408 2.024 9 9 5 43
86 2 3.5222 0.27612 10 10 5 48
87 2 3.5222 0.27612 10 10 5 48
88 2 3.2658 0.81656 10 10 5 45
89 2 2.9407 1.901 10 8 5 47
90 2 3.7414 0.3378 10 10 5 50
91 2 3.7414 0.3378 10 10 5 50
92 2 3.7414 0.3378 10 10 5 50
TABLE 3: A snapshot of the intermediate cluster generation process
Cluster
1
Cluster
2
Cluster
3
Cluster
4
Cluster
5
Cluster
6
Cluster
7
Cluster
8
very good 2.839 6.0199 59.6441 218.529 50.278
9
16.6368 2.0933 45.884
2
Good 67.632
6
11.8359 17.4102 2.0227 45.918
4
37.7739 10.66 56.701
3
Satisfactory 636.33 4.9919 115.724 1.7551 195.54
4
312.472 148.40
7
223.75
3
Poor 7.6643 6.0078 1.0183 1.3386 3.4087 8.7138 14.303 3.3781

Cluster 9 Cluster 10 Cluster 11 Cluster 12 Cluster 13 Cluster 14 Average
1 1.9775 40.0564 10.0671 1.0022 5.9714
32.99999
3.0049 4.0798 39.2667 40.6924 4.997 1.0043
24.50001
44.8842 8.9885 180.93 183.253 6.017 6.9517
147.8572
9.3508 2.663 1.0011 10.1516 1.0006 1.0005
5.071443
58.2399 17.7088 261.254 244.164 13.0168 14.9279
[total] 714.466 28.8554 193.796 223.646 295.15 375.597 175.463 329.717
Table 4: Clustered data of Faculty Performance based on student feedback
The ratings were then represented using pie chart which is shown in the figure below and the
representation reveals more accuracy than the traditional approach because the clusters
generated are influenced heavily by all the attributes which had been taken into consideration.
FIGURE 2: Faculty Average Performance Using Data Mining Technique
Faculty Average Performance Using
Data Mining Technique
16%
12%
70%
2% Very good
Good
Satisfactory
Poor

4. FACULTY PERFORMANCE TREND EXTRACTION USING OLAP
STATISTICAL TOOL
TABLE 5: Performance trend- management faculty
In the above section the overall performance of the management faculties based on student
feedback have been shown first using traditional approach and then by using mining methodology
to provide a more accurate result. In this section we have used OLAP statistical tool to extract
certain trends in faculty performance and also to assess individual faculty performance across
several parameters represented in the cube form and extracted it in to the grid and synchronized
with the chart. The patterns as identified are as follows:

- The consistency in performance of faculties in the Associate Professor level was found to
be more leveled than the faculties at the lecturer level.
- Also the dip in the performance when analyzed across the lecturer level was found to be
more than that at the Associate Professor Level.
- The performance of the visiting faculties showed a subsequent drop in spite of them having
considerable industry and teaching experience.
5. CONCLUSION AND FUTURE WORK
The future work will contain association rule mining on the student feedback database and
dependencies will be analyzed to draw some meaningful conclusions.
6. REFERENCES
1. Breiman, L., Friedman, J.H., Olshen, R., and Stone, C.J., 1984. Classification and
RegressionTree Wadsworth & Brooks/Cole Advanced Books & Software, Pacific California.
2. A.K. Jain and R. C. Dubes. [1988]. Algorithms for Clustering Data. Prentice Hall.
3. R Agrawal, R Srikant Fast Algorithms for Mining Association rules in Large Databases (1994)
by Proceedings of the VLDB.
4. Ganti, V., Gehrke, J. and Ramakrishnan, R. 1999a. CACTUS-Clustering Categorical Data
Using Summaries. In Proceedings of the 5th ACM SIGKDD, 73-83, San Diego, CA.
5. GUHA, S., RASTOGI, R., and SHIM, K. 1999. ROCK: A robust clustering algorithm for
categorical attributes. In Proceedings of the 15th ICDE, 512-521, Sydney, Australia.
6. Zaki, M.J. Scalable algorithms for association mining Knowledge and Data Engineering, IEEE
Transactions on Volume 12, Issue 3, May/Jun 2000 Page(s):372 390 Digital Object Identifier
10.1109/69.846291
7. Chiu, T., Fang, D., Chen, J., and Wang, Y. 2001. A Robust and scalable clustering algorithm for
mixed type attributes in large database environments. In Proceedings of the 7th ACM SIGKDD,
263-268, San Francisco, CA.
8. Luan J. [2002] “Data Mining and Knowledge Management in higher Education” Presentation at
AIR Forum, Toronto, Canada.
9. Fathi Elloumi, Ph.D., David Annand. [2002] Integrating Faculty Research Performance
Evaluation and the Balanced Scorecard in AU Strategic Planning: A Collaborative Model.
10. Raoul A. Arreola, Michael Theall, and Lawrence M. Aleamoni [2003] Beyond Scholarship:
Recognizing the Multiple Roles of the Professoriate." Paper presented at the Annual Meeting of
the American Educational Research Association (Chicago, IL, April 21-25, 2003).
11. M.R.K. Krishna Rao. [2004] Faculty and Student Motivation: KFUPM Faculty Perspectives
12. Karin Sixl-Daniell, Amy Wong, and Jeremy B. Williams.[2004] The virtual university and the
quality assurance process: Recruiting and retaining the right faculty. Proceedings of the 21st
ASCILITE Conference.
13. Emmanuel N. Ogor.[2007] Student Academic Performance Monitoring and Evaluation Using
Data Mining Techniques. Electronics, Robotics and Automotive Mechanics Conference, 2007.
CERMA 2007 Volume, Issue, 25-28 Sept. 2007 Page(s): 354 – 359 Digital Object Identifier
10.1109/CERMA.2007.4367712.

14. Amy Wong and Jason Fitzsimmons [2008] Student Evaluation of Faculty: An Analysis of Survey
Results. U21GlobalWorking Paper Series, No. 003/2008.
15. Cristóbal Romero, Sebastián Ventura, Pedro G. Espejo and César Hervás.[2008], Data Mining
Algorithms to Classify Students. The 1st International Conference on Educational Data Mining
Montréal, Québec, Canada, June 20-21, 2008 Proceedings.
16. Chandrani Singh ,Dr. Arpita Gopal Performance Analysis of Faculty Using Data Mining
Techniques,IJFCSA-2010,1st edition

Performance Assessment of Faculties of Management Discipline From Student Perspective Using Statistical and Mining Methodologies

More Related Content

What's hot (18)

Viewers also liked (9)

Similar to Performance Assessment of Faculties of Management Discipline From Student Perspective Using Statistical and Mining Methodologies (20)

More from Waqas Tariq (20)

Recently uploaded (20)

Performance Assessment of Faculties of Management Discipline From Student Perspective Using Statistical and Mining Methodologies