Fall

08

Data Analysis Report

Assessment #2

Tayla Wearne
Student No.: 9499920

Tutor: Jenny Houtsma

Date: 3rd September 2015

Word Count: 1,290

Q u e e n s l a n d U n i v e r s i t y o f Te c h n o l o g y
Table of Contents
1.0 Introduction................................................................3
2.0 Outliers....................................................................... 3
3.0 Distribution of Exam Results........................................4
4.0 Possible Influences on Exam Results.............................5
4.1 Influence of Gender...........................................................5
4.2 Influence of Single or Double Degree.................................6
4.3 Influence of Quiz Result.....................................................7
4.4 Influence of Report Result.................................................8
4.5 Influence of Report by Gender...........................................9
5.0 Conclusion................................................................10
Appendix........................................................................11

Page 2
1.0 Introduction
As part of the universitys goal to provide the best education
experience possible, universities are always looking for ways to
better understand the learning of students. This report aims to
summarise, analyse and interpret student performance data and to
identify the influences that contribute to a students final exam
score. The information obtained in this report will be used in
universities, so that they can better assist the students to achieve
their full potential. The data being analysed contains numerical and
categorical data. The numerical data includes the scores each
student attained for 3 pieces of assessment, the quizzes, the report
and the final exam. The categorical data included was the students
gender and whether the students were studying a single or double
degree at university. The analytical agenda for this report includes
analysing the distribution of the final exam results and what
influences these exam results.

2.0 Outliers
An outlier can be identified as a data point that lies an abnormal
distance from other values in a certain sample of data. It is
important to consider these observations to understand why they
appeared and what it means for the data being observed. When an
outlier is observed it is up to the person doing to analysis to decide
weather or not it is appropriate to exclude the data point. In this
data set there were no Z scores below -3 or above 3, therefore there
are no outliers.

Page 3
3.0 Distribution of Exam Results
The histogram below represents the distribution of the final exam
results. The results have been arranged into numerically ordered
classes. The number of observations in each class represents the
corresponding frequency of that class.

Exam Results Distribution

250

200

150
Frequency
100

50

Figure 1 Exam Results Distribution

The key features of the histogram in figure 1 tell us important

information regarding the exam results. The mean tells us that the
average mark the students got on the final exam was 28.52. This is
quite poor and just below half of the highest possible score of 60.
We can also see that more than half of students got less than half of
the exam correct. The standard deviation is 12.40, which means
that the data is relatively close to the median. The median (26.5) is
less than the mean (28.52), which tells us the data is right skewed.
We can also see that the lowest value a student scored was 5 and
the highest value a student scored was 59 out of a possible 60. It
also tells us that variation between each value in the data set and
the mean is equal to 153.67. The modal class is the 20-30 class,
with 232 students.

Page 4
4.0 Possible Influences on Exam Results

There are possible factors that could explain a students final exam
result. It is important to investigate these influences to better
understand a students performance. The possible influences that
affect students results in final exam include; gender, whether they
are studying single or double degree, their quiz result, their report
result and finally the report results by gender. In order to investigate
these influences, the data must be analysed in tables or charts and
summarised to better interpret the data.

4.1 Influence of Gender

The following graph shows the final exam results by gender and
depicts numerical and categorical data. The average and the
standard deviation represent the numerical data in the chart, whilst
the gender represents the categorical data.

Exam Results and Gender

St Dev
Female
Male

Average

Figure 2- Exam Results and Gender

Figure 2 indicates that on average females tend to do better in final

exam than males do. A females average result is 29.25, whereas a
males average is equal to 27.74. However they appear have the
same standard deviation of 12.4, which means that males and

Page 5
females results varied the same amount from their average exam
results.

4.2 Influence of Single or Double Degree

The figure below illustrates the effect that studying a single or

double degree has on the students exam result. This figure again
includes both categorical (single or double degree) and numerical
(exam result).

Influence of Degree on Exam Results

No

Double Degree?

Yes

0 5 10 15 20 25 30 35

Average result

Figure 3- Influence Degree on Exam Results

Figure 3 clearly illustrates that on average students who study a

double degree do better in the final exam than students who study a
single degree. The average result of a student that studies a double
degree is 31.55, which is a moderate difference to the average of
26.94 for a student that studies a single degree. Therefore it can be
said that that the type of degree you are studying (single or double)
can influence how well you do in the final exam.

Page 6
4.3 Influence of Quiz Result

The next influence to be considered is that of the quiz result. This

analysis will investigate if a students quiz result is linked to their
exam result. Both the data sets are numerical and we can therefore
measure the relationship between the two by measuring ones
increase against the change in the other variable. The figure below
is a scatterplot.

Relationship between Quiz and Exam Result

60

50

40
f(x) = 1.96x - 1.33
Exam Result 30

20

10

Figure 4- Relationship Between Quiz and Exam Result

Page 7
The correlation in figure 4 is 0.5364. This means that there is a
moderate positive relationship between the quiz and the exam
results. This suggests students should be encouraged to do the
quizzes because it can assist them in preparing for the final exam. It
also shows that the quiz results are a good indication of how the
students might perform in the exam. The correlation could be
stronger but there are reasons some students do better on different
pieces of assessment. Some students may have felt more
comfortable doing quizzes at home in their own time rather than
doing an exam.

4.4 Influence of Report Result

This section investigates the relationship between the students that

do well on the report and their final exam results. The two variables
will be compared by calculating the correlation of the coefficients;
this will reveal the type of relationship there is between the results.

Relationship between the Report and the Exam

60

50

40

Exam Result 30 f(x) = 1.11x + 13.2

20

10

Figure 5: Relationship between the Report and the Exam

Page 8
In figure 5 the correlation is 0.2928. This tells us that the
relationship between the variables is weak positive. This tells us
students that did well on the report didnt necessarily do well on the
exam. This can be because of a number of reasons. One reason is
that a report piece of assessment is something students can do at
home and receive help from different people, whereas when doing
an exam, students have set time and need to recall all the
information theyve learnt throughout the semester. However
students should still be encouraged to try their best on the report,
as it can be practice and equip them with knowledge they need for
the exam.

4.5 Influence of Report by Gender

This final segment will investigate whether or not there is evidence

to suggest that that a student (male of female) who does well in the
report will do well in the final examination.

Page 9
Relationship between Gender and the Report to the Final Exam Results

Female
Linear (Female)
Male
Linear (Male)

Figure 6- Relationship between gender and the Report to the

Final Exam Results

Figure 6 shows the correlation of both females and males report

results to their exam results. Firstly, females correlation was
calculated at 0.3072, this means there is a relationship between
females report and exam results, this relationship is moderate
positive. Secondly, males correlation was calculated at 0.2719, this
is classified as a weak positive relationship.
This data tells us that females who do well on the report also do well
on the exam, but that males who do well on the report may not do
well on the exam. The gender effect that can be seen is that
females tend to have a stronger relationship between their results,
whereas males barley have a relationship between results.

5.0 Conclusion

To conclude this report, there are two main effects that influence the
final exam results. The first influence on the final exam results is the
quiz results. The relationship between the quiz and exam results
was moderate strong. This tells us that the quiz is important and

Page 10
students should be encouraged to do the quizzes and to take them
seriously. Students that did well on the quizzes in turn did well on
the final exam. The second influence is the influence of females
report results; there is evidence (a moderate positive relationship)
that suggests females that did well in the report did well in the
exam.

Appendix
To find outliers, the Z score must be calculated; this is done
following the z score formula: mean-exam result
standard deviation.

Page 11
If any of the Z scores are below -3 or above 3, they are considered
an outlier. Some examples of the calculations can be seen below.

Page 12