Java程序辅导
C C++ Java Python Processing编程在线培训 程序编写 软件开发 视频讲解
C C++ Java Python Processing编程在线培训 程序编写 软件开发 视频讲解
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 1 of 14
RMIT University
Learning and Teaching Investment Fund 2008
Final Report
Due date is February 20, 2009 to your LTIF College Coordinator
Project title New ways of Learning - Teaching and assessment of large classes
Project leader Mali Abdollahian
Team members Assoc Prof Cliff Da Costa, Dr Yousong Luo, Dr Claude Zorzan, Dr Ian,
Grundy,Anthony Bedford,
Funds approved $45,433
Funds acquitted (attach
financial statement)
Introduction
This project is a service teaching innovation which will build on the current RMIT
leadership and student feedback project. It is designed to address key educational
problems and issues experienced by lecturers and those identified as the top three
issues by students in their course experience feedback. Service teaching is an integral
aspect of learning and teaching in Higher education and is acknowledged by
universities nationally and internationally as an important element of cross-disciplinary
study. Service teaching has many models. In a traditional service teaching model the
discipline expertise is recognized and valued and is utilized across courses, programs
and discipline areas within the university. Educational quality of courses and programs
are enhanced when the students are taught by discipline experts with the appropriate
expertise, skills and knowledge base rather than home program lecturers.
Student learning outcomes and experiences should be satisfying and memorable
experiences for the students. However; this is not currently the case for some courses
in several programs. Effective service teaching and best practices still remain
undefined. This project will develop, implement and evaluate three classroom
curriculum initiatives each addressing a specific educational problem in providing
maths and stats service teaching to large classes.
Detailed project
description and outline
of what was done
The initiatives to be addressed in this project are:
Innovation 1 – Addressing student diversity in multi-discipline large classes
by providing relevant disciplinary context-related exemplars in teaching
service classes
Innovation 2 – Addressing tutor teaching practices by trialling innovative
ways of providing and supporting tutors in large service classes
Innovation 3 - Exploring and implementing effective ways to provide feedback
to students in large service classes
The proposed integrated learning system will utilise existing IT infrastructure currently
available at RMIT and will blend with what is available via resources such as
WebLearn.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 2 of 14
Description - Innovation 1
This innovation led by Dr Mali Abdollahian and Dr Yousong Luo will use the course
Maths 2123 to develop, pilot and evaluate relevant discipline specific examples for
teaching the course. During semester 1, 2008, 3 3rd and 4th year students from the
relevant disciplines will be selected to develop relevant data and examples for
teaching statistics and mathematics in the first year multi-discipline large class. These
examples will be reviewed and validated by the chief investigators, then uploaded into
the course material on the Distributed Learning Management system (DLS). The
redesign of the online component of the course will include changing the resources
and exemplars into sub-categories related to the multiple disciplines and student
cohort taught in the service teaching classes.
These exemplars will be used for teaching and evaluated in terms of their
effectiveness in student learning experiences and outcomes.
Description - Innovation 2
This innovation led by Dr Ian Grundy and Dr Claude Zorzan will institute a pilot of
using the Queensland University model of volunteer tutoring cited in the Australian
University’s Educational quality best-practice here at RMIT in the course Maths 2117.
Two lecturers are exploring the possibilities and ways of implementing it here and
would apply what they have learnt into RMIT L&T practices. The initiative uses 3rd and
4th year students to peer tutor lower years of the program on a voluntary basis. All
tutors are mentored and supported by academics and only the good committed tutors
are provided formal training in learning and teaching and as accredited tutors.
In the first semester 2008 the investigators will set up the scheme, select tutors,
provide support and mentor them when they are conducting tutorials. In 2nd semester
this will continue but there will be interviews to find out what are the key success
factors and barriers to implementing such a scheme. The interviews will also identify
the key elements that tutors will need on their training. The project will be evaluated
and refined for use in 2009 with focus of improving teaching practices and student
learning outcomes.
Description - Innovation 3
Innovation 3 led by A/P Cliff da Costa and Dr Anthony Bedford will expand on their
current practices of teaching online by exploring effective ways of providing feedback
to the students. Currently students use Weblearn applications to self assess their
learning and are provided hurdles so they achieve mastery of the topics and their
learning objectives. What is lacking in this approach is providing immediate, prompt
and appropriate feedback for learning.
In semester 1 2008 the investigators will explore common mistakes students make,
common misconceptions and develop appropriate feedback for the databank. This
information will be gleaned from the multiple choice tests that the students will have
done during semester 1 and in 2007. They will input this on the Weblearn database so
that it feeds into the teaching and learning modules on the DLS In semester 2 first
year students will use it. The feedback would be refined as students use the learning
assessment tasks. The project will be evaluated and refined for use in 2009 with the
focus of improving teaching practices and student learning outcomes.
Attach the full and
detailed report and
evaluation of your
project outcomes
including evidence of
Innovation 1
Due to the fact that grant money was finalized at the end of March, 2008 we could not
get enough students on time to complete the project for the first semester of 2008,
however we are ready to implement it in the first semester of 2009. During semester
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 3 of 14
the impact the project
has had. Also make
reference to how the
outcomes address the
five key objectives:
Improved student
learning experiences,
outcomes and
employment
opportunities
Innovation
Strategic alignment
University wide
application
Value for money
1, 2008, three 3rd and 4th year students from the relevant disciplines were selected to
develop discipline related data and examples for teaching statistics in the first year
multi-discipline large class. Lecture notes were prepared in three disciplines by
changing the exemplars to applications in Food science, Environmental Science and
Biomedicine. The redesign of the online component of the course involved a
separation of resources and exemplars into sub-categories related to the multiple
disciplines and student cohort taught in the service teaching classes. These examples
were reviewed and validated by the chief investigators, then uploaded into the course
material on the Distributed Learning Management system (DLS). Although the lecture
material covered in class was common students were able to refer to their discipline
oriented lecture notes with relevant examples in the DLS. Because the examples are
related to the disciplinary context, students find it more interesting and easy to
understand.
Apart from disciplinary related lecture notes, the step by step graphical Learning
Guide of MINITAB (statistical package that is used heavily in the course) for two
mathematics courses (both courses have more than 220 students) was prepared with
8 web pages to help students learn to use MINITAB to perform statistical analysis on
the topics covered in their lecture notes. In order to make the MINITAB Guide an
effective self teaching tool, we have presented step by step graphical designed
examples for each topic that enables students to use the package for their required
statistical analysis without getting help from any one.
We have also generated a smart weblearn question bank to address the diversity in
the academic background of students in the course. In the weblearn test questions a
help link was inserted for each question that required use of the MINITAB statistical
package. The link displays step by step graphical application of the MINITAB for the
related topic. Therefore students can use the MINITAB Learning Guide for any
difficulties they come across in solving problems.
Another important change done in weblearn tests is the introduction of statistical
definitions help link. In the questions of weblearn modules. For each question in the
weblearn question bank, if students click on a statistical word they are not familiar with
they will be directed to a webpage of definitions and related examples.
These changes would enable student to get instant feedback on each question.
This is a significant step that helps them learn independently at their own pace about
the topics taught during the lecture.
The Minitab help link was implemented in 2008 and the feedback was very positive
since the students who needed extra help could get it instantly on line. They also
mentioned that the link made it easy for them to complete their weblearn tests with
additional lab assistance.
However, the definition help link will be implemented for the first time during the first
semester of 2009.
These changes were trialled on two mathematics course each of which has a statistics
and a mathematics component. While the Minitab learning guide help link and the
definition help link were added to the statistics component, other changes were
introduced to the weblearn tests within mathematics component. Under the topics
covered in the two courses, new question banks are created. The majority of the
questions in those question banks are new with only a few questions modified from
the existing questions. The new questions with built-in intelligence were created with
the help of a maths honours student and a computer science master student who had
knowledge of Maple, LaTex, HTML and Graphic editing. There are two major
innovations to the new questions. The first is the feedback mechanism. When
students answered a question wrongly they will be provided with a detailed solution to
their individualized question. This will help them to achieve better in their next
attempt. The second is the editing tool for entering Maple codes. Most questions
require students to enter a symbolic mathematical expression as their answer and this
is done by using Maple code. Previous CES shows that students have a great
difficulty with entering Maple code. Now the new questions employ an external Java
Applet: DragMath to assist students to edit mathematical expressions graphically.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 4 of 14
mathematics component will also be implemented in the first semester of 2009.
Innovation 2
The large class considered for this innovation was a mathematics course servicing on
Engineering department. The course had a mathematics and statistics component and
an enrolment of over 220 students. Tutors were selected from the relevant
Engineering department and also from the Maths department.
For the mathematics component, students were divided into two groups. Three tutors
from the Engineering department were assigned to one group and 3 tutors from the
Mathematics department were assigned to the other group. At the end of semester 2
2008 the Mathematics practice class marks of each group was then compared using a
two-sample t-test to decide whether the home department tutors are more effective in
the learning process or not. The results show that there was no significant difference
in marks between the two groups. In the same two groups, students who scored exam
marks above 40% were also compared using a two sample t-test and again there was
no significant differences between the two groups.
For the statistics component, students were again divided into two groups. A tutor
from the Engineering department was assigned to one group and a tutor from the
Mathematics department was assigned to the other group. Statistics Weblearn test
marks for each group were then compared using a two-sample t-test. The results
showed a significant difference in marks between the two groups. i.e. students who
had a tutor from their home department achieved higher marks compared with
students who were tutored by statistics student. It should be mentioned that students
could do the test at any time within the opening period of each test ( this was at least a
two week period), therefore we have no prior- knowledge of whether they did the test
while the tutor was present in the lab or not. For these two groups students with exam
marks above 40% were compared using a two-sample t-test. The result showed that
there was no significant difference in the exam marks between the two groups.
The graphical and statistical results for this innovation are attached in Appendix A.
Innovation 3
The service courses considered for this innovation were the two compulsory first-year
statistics courses included in the Bachelor of Applied Science (Psychology) degree
program. Typically these courses have around 90 students enrolled at both the
Bundoora and City campuses. These students usually have a strong dislike for
mathematics, including statistics, and consider it onerous being forced to undertake
these courses despite the clear need for the objective skill set in their studies.
Assessment in these service courses comprises mainly of multiple-choice question
based paper tests. Initial analysis of the student performance on Semester 1 2008
assessment tasks identified some unexpected, potentially confounding factors, viz: (a)
the number of response items; (b) the placement of the correct response; and (c)
previous student performance. The project was then modified for these students in
Semester 2 2008 to allow further identification and analysis of these potentially
confounding factors.
Analysis of assessment tasks undertaking in Semester 1 2008 indicated that students,
when presented with a four response item question, were more likely to select an
incorrect response than expected (p <0.001). Further to this, the placement of the
correct response in these four response questions was also found to be a
considerable influence on the student’s performance (with p<0.001). Given these
results, the first assessment task presented to the students in the Semester 2 unit was
adapted to contain only 4 response type questions and considerable care was taken
to ensure the wording and style of each question were consistent, i.e. only positively
phrased questions. Great care was also taken to ensure that the correct responses
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 5 of 14
were randomly placed to avoid creating predictable patterns in the correct responses.
Following this assessment task, students were given a thorough feedback session,
covering identified areas of weakness and ensuring that students understood why a
particular response was the most correct. Students were also encouraged to discuss
why they had selected a particular incorrect response to assist in the identification of
areas of weakness and/or confusion. Analysis of their performance on this task
demonstrated that the correct response positioning effect had disappeared (with
p=0.309), however, a new unexpected result arose from the analysis. The
performance of these students was found to be highly predictable (with R2=0.98 and
p<0.001) perhaps suggesting that whilst students may learn new material, they master
this new material to the same level as previous material. Subsequent assessment
tasks demonstrated similar results in terms of predicting student performance. (See
Appendix B)
Despite these potentially confounding factors, we identified four main areas of
weakness for students in these service courses, viz: (1) students miss vital steps in
calculations; (2) students confuse technical terms (e.g. thinking bimodal is the same
as bivariate); (3) students misread the question or response text; and (4) students
misread data tables. Following the work undertaken for the first assessment task in
Semester 2 and the subsequent feedback session, we determined that these factors
had been reduced to just (3) and (4) on the remaining assessment tasks. This
outcome strongly suggests that the bi-directional feedback between the students and
the lecturers allows both parties to better identify areas of weakness and implement
appropriate intervention strategies. However, further research involving a larger
sample size and more teaching staff is required to validate this finding.
Work with regard to predicting student performance and refining assessment tasks is
ongoing. We are attempting to establish a tool based upon Rasch models (Item
Response Theory) to quantify the fairness and reliability of a question and thereby an
assessment task. This work is intended to inform future assessment development and
to aid in providing prompt feedback to both the students and the lecturers involved.
Dissemination of
project outcomes both
completed and planned.
This should include
both within RMIT and
externally.
The major benefits that would accrue from innovation 1 are:
a. A Student learns the statistics and mathematical concept via an examplar and
sees its immediate application in their respective discipline
b. Student can simulate the data analysis employed in the discipline which
enhances understanding of the concept
c. Student-learning occurs more efficiently and is integrated within their own
discipline areas
Innovation 2 indicated that if we train the tutors from the service department they can
be as effective as tutors from the mathematics department in tutoring mathematics
and statistics courses, despite the fact that their mathematical or statistical knowledge
would not be at the same level of Maths and stats students.
Since the major part of innovation 1 has just been completed and is going to be
implemented in the first semester of 2009 we will be publishing the final guidelines
after analysing the CES score in 2009. A draft paper is already prepared to be
submitted in an Education conference in 2009.
We have already achieved improved CES score of 10 points on one of the pilot course
where the improved statistics weblearn test and innovation 2 were implemented in
semester 2 of 2008.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 6 of 14
Summary of the project,
outcomes, impacts and
dissemination
The Main objective of this project was to develop an integrated learning system that
would utilise existing IT infrastructure currently available at RMIT and would blend with
what is available via resources such as WebLearn. The project was designed to:
Address student diversity in multi-discipline large classes by providing relevant
disciplinary context-related exemplars in teaching service classes
Address tutor teaching practices by trialling innovative ways of providing and
supporting tutors in large service classes
Exploring and implementing effective ways to provide feedback to students in
large service classes
Lecture notes are prepared in three disciplines by changing the exemplars to
applications in Food science, Environmental Science and Biomedicine. Students in
multi-discipline class will be able to refer to their discipline oriented lecture notes with
relevant examples. This part of the project will be implemented in the first semester of
2009.
A smart weblearn question bank has been generated to address the diversity in the
academic background of students in the course. This will ultimately help students
learn independently at their own pace about the topics taught during the lecture.
Within the statistical weblearn question bank a help link was inserted for each
question that required use of the MINITAB statistical package. The link displayed step
by step graphical applications of the MINITAB package for the related topic.
Another significant change undertaken in the weblearn tests is the introduction of a
statistical definitions ‘help link’. For each question in the weblearn question bank, if
students were unfamiliar with a specific statistical term they would be directed to a
webpage of definition and related examples which further enhanced their
understanding of the term and its application.
For the mathematical component of these courses new questions with built-in
intelligence were created. There are two major innovations to the new questions bank.
The first of which is the feedback mechanism which provides detailed solution to any
incorrect answers and the second is the editing tool for entering Maple codes. Most
questions require students to enter a symbolic mathematical expression as their
answer and this is done by using Maple code. Previous CES shows that students
have great difficulty with entering Maple code. As a result of this, new questions now
employ an external Java Applet: DragMath to assist students to edit mathematical
expressions graphically.
Tutors from the Engineering department were trained to tutor one group of
engineering students for the maths and statistics components of the course. The other
group’s tutors were composed from the Maths department. The statistical analysis of
the exam marks showed that there is no significant difference in the exam marks
between the groups.
We have identified the following four main areas of weakness for students in the
service courses:
(1) students miss vital steps in calculations;
(2) students confuse technical terms (e.g. thinking bimodal is the same as bivariate);
(3) students misread the question or response text; and
(4) students misread data tables.
Following the work undertaken for the first assessment task in Semester 2 of 2008
and the subsequent feedback sessions, we determined that these factors had been
reduced to just (3) and (4) on the remaining assessment tasks. This outcome strongly
suggests that the bi-directional feedback between the students and the lecturers
allows both parties to better identify areas of weakness and implement appropriate
intervention strategies. However, further research involving a larger sample size and
more teaching staff is required to validate this finding.
Appendix A: Innovation 2
Group 1 – students who had tutors from engineering department for Mathematics
Group 2 – students who had who tutors from Mathematics department for Mathematics
Comparison of the Maths assignment marks
D
at
a
Group 2Group 1
20
15
10
5
0
Boxplot of Group 1, Group 2
Two-Sample T-Test and CI: Group 1, Group 2
Two-sample T for Group 1 vs Group 2
N Mean StDev SE Mean
Group 1 49 17.64 2.79 0.40
Group 2 119 17.48 3.76 0.34
Difference = mu (Group 1) - mu (Group 2)
Estimate for difference: 0.164346
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 7 of 14
95% CI for difference: (-0.879030, 1.207722)
T-Test of difference = 0 (vs not =): T-Value = 0.31 P-Value = 0.756 DF = 119
Conclusion: P-Value = 0.756 > 0.05
Do not reject H0. There is no significant difference in maths assignment marks between engineering department
tutor group and maths department tutor group.
Comparison of the students who have achieved Exam marks above 40% and had different tutors for the
maths component of the course
Group 1 – students who had tutors from engineering department for Mathematics
Group 2 – students who had who tutors from mathematics department for Mathematics
D
at
a
Group 2Group 1
100
90
80
70
60
50
40
Boxplot of Group 1, Group 2
Two-Sample T-Test and CI: Group 1, Group 2
Two-sample T for Group 1 vs Group 2
N Mean StDev SE Mean
Group 1 43 65.8 14.8 2.3
Group 2 105 64.1 12.0 1.2
Difference = mu (Group 1) - mu (Group 2)
Estimate for difference: 1.72292
95% CI for difference: (-3.35133, 6.79718)
T-Test of difference = 0 (vs not =): T-Value = 0.68 P-Value = 0.500 DF = 65
Conclusion: P-Value = 0.5 > 0.05
Do not reject H0. There is no significant difference in MATH2114 exam marks between engineering department
tutor group and maths department tutor group.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 8 of 14
Comparison of Statistics weblearn tests marks
Group 1 – students who had tutors from engineering department for Mathematics
Group 2 – students who had who tutors from mathematics department for Mathematics
D
at
a
Group 2group 1
30
25
20
15
10
5
0
Boxplot of group 1, Group 2
Two-Sample T-Test and CI: group 1, Group 2
Two-sample T for group 1 vs Group 2
N Mean StDev SE Mean
group 1 43 23.02 3.62 0.55
Group 2 126 21.17 6.49 0.58
Difference = mu (group 1) - mu (Group 2)
Estimate for difference: 1.84865
95% CI for difference: (0.26691, 3.43039)
T-Test of difference = 0 (vs not =): T-Value = 2.31 P-Value = 0.022 DF = 131
Conclusion: P-Value = 0.022 < 0.05
Reject H0. There is a significant difference in MATH2114 Statistics weblearn marks between engineering
department tutor group and maths department tutor group.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 9 of 14
After eliminating the outliers,
Two-Sample T-Test and CI: group 1, Group 2
Two-sample T for group 1 vs Group 2
N Mean StDev SE Mean
Group 1 42 23.24 3.38 0.52
Group 2 123 21.64 5.82 0.52
Difference = mu (group 1) - mu (group 2)
Estimate for difference: 1.59582
95% CI for difference: (0.13229, 3.05935)
T-Test of difference = 0 (vs not =): T-Value = 2.16 P-Value = 0.033 DF = 123
Conclusion: P-Value = 0.033 < 0.05
Reject H0. There is a significant difference in MATH2114 Statistics weblearn test marks between engineering
department tutor group and maths department tutor group.
Same conclusion as above.
Comparison of the students who have achieved Exam marks above 40% and had different tutors for the
statistics component of the course
Group 1 – students who had tutors from engineering department for Mathematics
Group 2 – students who had who tutors from mathematics department for Mathematics
D
at
a
Group 2Group 1
100
90
80
70
60
50
40
Boxplot of Group 1, Group 2
Two-Sample T-Test and CI: Group 1, Group 2
Two-sample T for Group 1 vs Group 2
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 10 of 14
N Mean StDev SE Mean
Group 1 40 64.1 12.9 2.0
Group 2 110 64.7 12.9 1.2
Difference = mu (Group 1) - mu (Group 2)
Estimate for difference: -0.615909
95% CI for difference: (-5.371550, 4.139732)
T-Test of difference = 0 (vs not =): T-Value = -0.26 P-Value = 0.797 DF = 69
Conclusion: P-Value = 0.797 > 0.05
Do not reject H0. There is no significant difference in MATH2114 exam marks between engineering department
tutor group and aths department tutor group.
Appendix B: Innovation 3
Analysis of four response questions (1) versus other questions (2) in Semester 1 course:
Chi-Square Test: Incorrect, Correct
Expected counts are printed below observed counts
Chi-Square contributions are printed below expected counts
Incorrect Correct Total
1 306 986 1292
379.46 912.54
14.220 5.913
2 393 695 1088
319.54 768.46
16.886 7.022
Total 699 1681 2380
Chi-Sq = 44.042, DF = 1, P-Value = 0.000
Note:
We reject the null hypothesis with p=3 x 10-11 and =0.05. The data provides significant evidence that
there is an association between the number of response items on a question and the performance of
students (i.e. Correct or Incorrect).
Analysis of four response questions correct answer placement in Semester 1 course:
Chi-Square Test: Incorrect, Correct
Expected counts are printed below observed counts
Chi-Square contributions are printed below expected counts
Incorrect Correct Total
1 66 240 306
72.47 233.53
0.578 0.179
2 68 306 374
88.58 285.42
4.781 1.484
3 89 319 408
96.63 311.37
0.603 0.187
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 11 of 14
4 83 121 204
48.32 155.68
24.899 7.727
Total 306 986 1292
Chi-Sq = 40.438, DF = 3, P-Value = 0.000
Note:
We reject the null hypothesis with p=8.9 x 10-9 and =0.05. The data provides significant evidence that
there is an association between the placement of the correct answer response item on a question and the
performance of students (i.e. Correct or Incorrect).
Analysis of four response questions correct answer placement in Semester 2 course:
Chi-Square Test: Incorrect, Correct
Expected counts are printed below observed counts
Chi-Square contributions are printed below expected counts
Incorrect Correct Total
1 98 210 308
100.98 207.02
0.088 0.043
2 129 235 364
119.34 244.66
0.782 0.381
3 107 257 364
119.34 244.66
1.276 0.622
4 125 239 364
119.34 244.66
0.268 0.131
Total 459 941 1400
Chi-Sq = 3.592, DF = 3, P-Value = 0.309
We do not reject the null hypothesis with p=0.309 and =0.05. The data does not provide significant
evidence that there is an association between the placement of the correct answer response item on a
question and the performance of students (i.e. Correct or Incorrect).
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 12 of 14
Analysis of student performance in Semester 2 course using Semester 1 final mark as a predictor:
Scatterplot showing possible linear relationship:
908070605040
45
40
35
30
25
20
Semester 1, 2008 course Final Mark
Se
m
es
te
r
2,
2
00
8
Ta
sk
1
m
ar
k
Scatterplot of S2 Task 1 mark vs S1 Final Mark
Regression analysis:
Regression Analysis: Task 1 mark versus S1, 2008 Final Mark
The regression equation is
Task 1 mark = 0.465 S1, 2008 Final Mark
Predictor Coef SE Coef T P
Noconstant
S1, 2008 Final Mark 0.46530 0.01402 33.19 0.000
S = 5.19493
Analysis of Variance
Source DF SS MS F P
Regression 1 29720 29720 1101.27 0.000
Residual Error 26 702 27
Total 27 30422
Unusual Observations
Obs S1, 2008 Final Mark Task 1 Fit SE Fit Residual St Resid
26 68.0 43.00 31.64 0.95 11.36 2.22R
29 63.0 18.00 29.31 0.88 -11.31 -2.21R
R denotes an observation with a large standardized residual.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 13 of 14
Lecturer notes about unusual observations:
Subject 26 had demonstrated a natural aptitude and curiosity for the material
Subject 29 had difficulties with the material and was identified as “at risk” following this
assessment. A successful intervention resulted in this student achieving a Pass grade in the
upper range (i.e. 55-59).
Note: R2 = 0.9769 and
We reject the null hypothesis (H0: =0) with p=3 x 10-22 and =0.05. The data provides significant
evidence that the slope of the population regression line is non-zero.
We also reject the null hypothesis (H0: =0) with p=3 x 10-22 and =0.05. The data provides significant
evidence that the population correlation coefficient is non-zero.
RMIT University
Document: SET LTIF Final Report Abdollahian.doc/Katrina Woodland
Save Date: 06-03-2009
Page 14 of 14