Analysis of International Flights in Australia

This report presents an analysis of the performance of key airports in Australia with regards to international flights. The analysis includes dataset 1 and dataset 2, examining trends and associations between variables. The findings suggest that the shape of the 'All Flights' variable is skewed and the average monthly flights do not exceed 30. The performance of airports is not significantly associated with the chosen airlines. Further research is recommended to explore differences in airport performance among user groups.

STATISTICS
STUDENT ID:
[Pick the date]

Table of Contents
Section 1: Introduction..........................................................................................................................2
Section 2: Analysis of single variable in Dataset 1.................................................................................3
Section 3: Analysis of two variables in Dataset 1...................................................................................5
Section 4: Collect and analysis Dataset2................................................................................................7
Section 5: Discussion and Conclusion....................................................................................................8
References.............................................................................................................................................9

Section 1: Introduction
a) Increasingly, the market share of international airlines and international flights in Australia is on
the upward trajectory. As a result, the given task aims to present a report highlighting the
performance of the various key airports in Australia with regards to international flights. This
becomes imperative in the backdrop of international airlines having 25% share in Australian flight
market. The growth rate witnessed by international operators at 5.4% is about three times that of
the regional and domestic operators. The thriving of international players to an extent may be
attributed to the cutting of capacity by Qantas Airways which is the largest airline based in Australia.
It is expected that this trend would continue going forward and requisite adjustments need to be
made by airlines to attract higher traffic (AnnaAero, 2018).
b) The given data (i.e. dataset 1) relates to information about selected international flights which
tend to flow out or in to Australian cities. This information includes the number of stops, date of
flight, stops, number of seats besides information about destination, route and start city. It is clearly
known that thee given data that has been provided has been obtained from Australian Government
Open Data and has not been collected by the University or myself. As a result, it would be wrong to
classify the given data as primary data and the given data would be termed as secondary data
(Medhi, 2016). There are a plethora of variables that have been used to capture the data and
majority of these variables are categorical with a nominal measurement scale. These variables are
non-numerical. However, there are certain variables such as stops, max seats, all flights which are
quantitative and are expressed by ratio scale of measurement (Eriksson and Kovalainen, 2015). The
possible cases used in the study relate to the various international flights which are either flying to
Australia or from Australia. Hence, the underlying population of interest would be all such
international flights which have Australia as the starting point or final destination.
c) The dataset 2 comprises of data about 29 flights being listed with focus on three variables namely
Australian City, International City and airline. This data has not been obtained from any other source
but rather has been obtained by conducted survey of 29 students at the KOI campus. As a result, it
would be fair to regard the given data as primary data and not secondary data. All the variables
included in the data are categorical in nature since they are represented by non-numerical data.
Further, the appropriate measurement scale for the data is nominal since the categorical data
cannot be arranged in any natural order (Flick, 2015). A key issue with regards to the data is that the
underlying sampling of KOI students is not based on random sampling but rather on convenience
sampling. As a result, it is possible that the responses collected from these students with regards to

the Australian city may be biased. It would have been preferable if the respondents would have
been collected through random sampling (Hillier, 2016). The cases that have been considered for
this are essentially all students who have taken any international flight from either of the three
airports namely Melbourne, Sydney or Brisbane.
Section 2: Analysis of single variable in
Dataset 1
a) The numerical summary of the variable “All Flights” is indicated below.
From the above summary statistics, it is apparent that the given data has a very high positive skew
which implies that there are certain flights which tend to have very high frequency during the
month. This can also be supported from the high range with regards to the given variable as
compared to the mean value of about 24. This clearly indicates presence of outliers on the positive
side of the mean which is leading to mean value being skewed higher than the median value. Based
on the above, it seems quite apparent that the shape of the given variable distribution would be
asymmetric thereby implying that the “All Flights” variable distribution is non-normal (Hair et. al.,
2015).
The histogram has also been obtained for the “All Flights” variable which is summarised below.

From the above histogram, it is apparent that the shape of the given variable is asymmetric as the
tail on the right hand of the mean is significantly longer than the corresponding tail on the left side
of the mean. Further, it is apparent that majority of the 1000 flights included in the sample tend to
have monthly frequency of lower than 40. However, there are certain flights which have very high
frequency and exceed 150 also. Clearly, such flights would be outliers and thereby it would be
appropriate to term the given distribution as non-normal (Medhi, 2016).
b) The objective is to highlight if the average monthly flight coming in and going out of Australia
exceed 30 or not using the inferential statistical technique named hypothesis testing. This inferential
statistical technique is being used as the underlying objective is to estimate population parameter
based on sample statistic. The requisite hypotheses for this test are stated below.
Null Hypothesis (H0): μ ≤ 30 i.e. the average monthly flights coming in and going out of Australia do
not exceed 30.
Alternative Hypothesis (H1): μ>30 i.e. the average monthly flights coming in and going out of
Australia does exceed 30.
The level of significance for this hypothesis test has been taken as 5%.
It is apparent from the given data that population standard deviation is not known and hence the
suitable test statistics for performing the hypothesis test would be t. Further, the given hypothesis
test would be a right tailed test which is essentially a one tail test which is apparent from the
alternative hypothesis (Flick, 2015). Using the sample information of 1000 flight, the hypothesis test

has been conducted using Excel functions and the following output has been obtained in this
regards.
The p value approach has been used for the given hypothesis testing. The decision rule under this
approach is that the null hypothesis would be rejected if the p value is lesser than the significance
level for the test. In the given case, the p value comes out as 1 which implies that it is greater than
the significance level of 5%. As a result, the evidence at hand is not sufficient to cause rejection of
null hypothesis and acceptance of alternate hypothesis (Eriksson and Kovalainen, 2015). Thus, it
would be fair to conclude that average monthly flights coming in and going out of Australia during
the period between September 2003 and September 2018 does not exceed 30.
Section 3: Analysis of two variables in
Dataset 1
a) The requisite numerical summary with regards to comparison of the three Australian cities and
selected airlines is rendered below.

The graphical summary in this regards is indicated as follows.
It is apparent from the above numerical summary and graphical summary that Melbourne tends to
have the maximum representation with 26 flights while Sydney has the lowest representation with
20 flights from the three selected airlines. Also, it regards to Air New Zealand, it is apparent that the
number of flights tend to exceed the other two (i.e. Cathay Pacific and Singapore Airlines). This is not
surprising considering the fact that New Zealand is neighbouring country to Australia and hence has
higher frequency of flights in comparison to the far flung destinations served by the other two
airlines. With regards to Air New Zealand, the maximum representation is from Sydney which is in
contrast with the other two airlines that tend to have lower flights to and from Sydney as compared
to the other two cities.
b) The objective is to conduct a suitable hypothesis test in order to highlight if there is an association
between city and airlines in the context of the underlying data taken into consideration for part a.
The relevant hypotheses for this test are as indicated below.
Null Hypothesis (H0): There is no significant association between Australia city (Melbourne, Sydney,
Brisbane) and airline (Cathay Pacific, Air New Zealand and Singapore Airlines).

Alternative Hypothesis (H1): There is significant association between Australia city (Melbourne,
Sydney and Brisbane) and airline (Cathay Pacific, Air New Zealand and Singapore Airlines).
Considering that both the given variables are categorical in nature, hence the Chi-square test of
independence is a suitable choice to test the hypothesis (Hillier, 2016). The level of significance for
the given test is given as 5%. The relevant output for the hypothesis test is obtained from Excel and
pasted below.
The p value approach has been used for the given hypothesis testing. The decision rule under this
approach is that the null hypothesis would be rejected if the p value is lesser than the significance
level for the test. In the given case, the p value comes out as 0.2986 which implies that it is greater
than the significance level of 5%. As a result, the evidence at hand is not sufficient to cause rejection
of null hypothesis and acceptance of alternate hypothesis (Hair et. al., 2015). Thus, it may be
concluded with 95% confidence that there is no significant association between Australia city
(Melbourne, Sydney, Brisbane) and airline (Cathay Pacific, Air New Zealand and Singapore Airlines).

c) There does not seem to any significant relationship between the airlines and the airport city. As a
result, to determine which airport performed the best, it would be prudent to rely on the actual
number of flight traffic that has been handled by the three airports in the context of the given
airlines. Considering that maximum number of flights has been handled by the Melbourne airport,
hence it would be appropriate to conclude that the best performance is of Melbourne airport
amongst the three choices available in this regards.
Section 4: Collect and analysis Dataset2
A summary of the performance of the three airports based on the sample data of 29 observations is
indicated below.

From the above graphical summary, it is apparent that more than 50% of the sample flights were
either taken from Sydney or were heading for Sydney. 34% of the sample flights were coming in or
heading out from Melbourne while the lowest share at 14% belongs to Brisbane. As a result, the
above data potentially indicates that the superior performance of Sydney. However, this contradicts
the conclusion derived from a larger sample data where the Melbourne was the best performing
airport and Sydney the least. This anomaly may be attributed to the non-random nature of sample
data owing to which it would be worthwhile to consider the conclusion drawn by using a larger
sample of 1000 flights over the 15 year period rather than the primary sample collected from 29 KOI
students.
Section 5: Discussion and Conclusion
Based on the above analysis, it is fair to conclude that the shape of “All Flights” variable is skewed.
This is not surprising considering the fact that underlying demand on different sectors would differ
owing to which the underlying frequency would show higher variation. Also, the hypothesis test
highlighted that average monthly flights from and to Australian cities did not exceed 30 during the
time from September 2003 to September 2018. With regards to the performance of airports, the
sample data with regards to the three chosen airlines supported Melbourne having highest number
of flights with lowest number of flights being witnessed at Sydney. The hypothesis test failed to find
any significant relationship between the three airports and the chosen airlines implying the
performance of airports is not dependent on the underlying airlines. In relation to the Dataset 2
(primary data), the best performance was seen by Sydney which accounted for more than 50%
flights. This may be attributed to the use of non-probability sampling technique owing to which more
preference would be given to Dataset 1.
In wake of the above analysis, further research need to be carried out by considering various user
groups such as KOI students and see if there is any difference in the performance of the different
airports. Additionally, the underlying reasons for difference in preferences of the various
international airlines also need to be considered.

References
AnnaAero (2018) International airlines now carry 25% of Australian traffic; Sunshine Coast is flying in
terms of passenger growth; preparing for non-stop UK, [online] Available at
https://www.anna.aero/2018/02/27/international-airlines-now-carry-25-australian-traffic/
[Assessed January 11, 2019] pp. 7-10.
Eriksson, P. & Kovalainen, A. (2015). Quantitative methods in business research (3rd ed.). London:
Sage Publications pp. 45-49.
Flick, U. (2015). Introducing research methodology: A beginner's guide to doing a research project
(4th ed.). New York: Sage Publications pp. 24-31.
Hair, J. F., Wolfinbarger, M., Money, A. H., Samouel, P., & Page, M. J. (2015). Essentials of business
research methods (2nd ed.). New York: Routledge pp. 36-38.
Hillier, F. (2016). Introduction to Operations Research. (6th ed.). New York: McGraw Hill Publications
pp. 68-72.
Medhi, J. (2016) Statistical Methods: An Introductory Text. (4th ed.). Sydney: New Age International
pp. 96-102.