Australian Rental Market Analysis

This assignment analyzes rental market data in Australia, focusing on weekly rent prices for international students and trends across different suburbs. It utilizes statistical techniques like ANOVA to compare average rent across locations and explores the relationship between bond amounts and weekly rents using scatter plots. The analysis concludes with insights into rental variations based on location and bedroom type, advising clients on budget considerations when choosing a rental property.

STATISTICS AND DATA ANALYSIS
Weekly Rent Analysis
STUDENT ID:
[Pick the date]

STATISTICS AND DATA ANALYSIS
Introduction
a) The given case highlights a fictional situation involving a service company which caters to
those people who have come to Australia either temporarily or permanently. A pivotal service
that the company provides constitutes accommodation which is difficult to find in Australia
especially if there is a budget constraint. For providing suggestion to the customers in relation
to weekly rent in various suburbs and the underlying patterns according to dwelling, a sample
consisting of data on 500 properties have been provided which span across four suburbs i.e.
Randwick, Sydney, Auburn & Parramatta. Also, for catering to the student community, a
statistical analysis of the rent paid on the weekly basis has been carried out so as to provide
prudent suggestions.
b) The data obtained for international students weekly rent is primary data as the same has been
collected directly from the students and forms dataset 1. It is quite possible that this data is
biased as underlying sampling is not random and also the sample size taken is 20 which is
quite small. In such a situation, there could be potential under or over representation of key
attributes which can distort the validity of the results obtained through statistical analysis. If
the given data had been collected from any source which had already collected this data, then
it would have been termed as secondary data which is not the case here. The key variable of
interest is the weekly rent which essentially represents quantitative type of data and tends to
highlight a ratio measurement scale. The quantitative data type is apparent considering the
numerical values of rent.
c) There is another dataset 2 which highlights the corresponding values of 500 observations
which in turn is secondary data obtained from the “Rental Bond Board Property Data” which
is published by the ‘Department of Finance, Services and Innovation’. There are various
variables in this dataset. The quantitative or numerical data are the bedroom number, bond
amount along with weekly rent which are all measured using the ratio measurement scale.
Besides, there are various qualitative or non-numeric data in the form of suburb and dwelling
type which are represented in the form of nominal measurement scale. The given dataset’s
first five cases have been illustrated below.
International Students’ Weekly Rent
1

STATISTICS AND DATA ANALYSIS
(a) The descriptive analysis for weekly rent amount paid by International students is
highlighted below:
The below highlighted histogram represents value of weekly rent amounts paid by international
students
0 to 356 356 to
411 411 to
467 467 to
522 522 to
578 578 to
633 633 to
689 689 to
744 More
than 744
0
1
2
3
4
5
6
Histogram - Weekly Rent (International
students (AUD))
Weekly rents of international students (AUD)
FRequency
(b) It is apparent from the descriptive analysis of weekly rent paid by the international
students that the value of measures of central tendency (mean, median and mode) is not
2

STATISTICS AND DATA ANALYSIS
same. Therefore, it can be concluded that data distribution is not normal. Further, the
value of skew is positive, which indicates that data has rightwards tail and presence of
outliers at the higher range of data. The positive skew indicates that median value of the
weekly rent paid by international student is lesser than the mean value of weekly rent.
Further, the value of standard deviation is significantly lower than the mean value of
weekly rent and hence, low to moderate dispersion is present in the data set.
Rental Bond Board Property: Dwelling type
a) The below highlighted pivot table shows the total number of flat and houses (dwelling type)
for the given data.
The below highlighted bar chart is the graphical representation of dwelling type distribution in
the given data.
It is apparent from the above highlighted distribution table and bar chart that approximately 96.6
% of the type of dwelling is of flat type that means about 483 dwelling out of 500 dwelling is of
flat type. Further, only 3.4% of the type of dwelling is of house type.
3

STATISTICS AND DATA ANALYSIS
b) The central objective is to check the validation of the claim “house dwelling type has
proportion of lower than 10%.”
Null hypothesis H0 : p=10 %Proportion of house dwelling type has the proportion same as 10%.
Alternative hypothesis H1 : p<10 %Proportion of house dwelling type has the proportion lower
than 10%.
Probability = 17 /500 = 0.034
Number of observation = 500
Standard deviation error = √ 0.1 ( 1−0.1 )
500 =0.01342
Value of z statistics = ¿ ( 0.034−0.1 )
0.01342 =−4.918
The p value corresponding to the z statistics is nearly 0.00.
Let the significance level = 5%
It can be seen that p value is lower than significance level and therefore, the conclusion can be
made that null hypothesis would be rejected and alternative hypothesis would be accepted.
Hence, the claim is right that proportion of house dwelling type has the proportion lower than
10%.
(c) Numerical summary of dwelling type for the given four suburbs are determined with the help
of pivot table and is highlighted below:
The graphical representation is show below in the form of bar chart of the dwelling type for the
given suburbs.
4

STATISTICS AND DATA ANALYSIS
From the above shown bar chart and table, it can be said that the most prominent type dwelling
type is flat which is having highest rank among all the given suburbs. From the last row of the
table, it can be seen that in Sydney suburb, all dwelling is of flat type and no houses are present
in the given sample data.
d) The above analysis in terms of houses clearly indicates that house representation is highest in
Auburn when considered in percentage and also on absolute number basis. Thus, taking the
assumption that the given sample faithfully mirrors the underlying population, it would be
fruitful for any client who wants a house to rent should make a suburb choice optimally as house
availability is quite limited. Besides, since only very few houses are available, the rent asked for
by the owner could be on the higher side owing to limited supply. As a result, it would be
advisable that the company provides guidance to the concerned client in relation to the relative
merits and demerits of both dwelling choices as per the customer and thereby allow a prudent
choice to be made.
Rental Bond Board Property Data – Weekly Rent
a) The task aims to illustrate the trends in relation to ongoing weekly rent for 2 bedroom
dwellings in various suburbs present in the sample data. The summary is presented as follows.
5

STATISTICS AND DATA ANALYSIS
The above summary in the graphical manner is showcased as follows.
The above summary apparently reflects that weekly rent for a two bedroom dwelling seems to be
highest in Sydney and lowest in Auburn. But it would be naïve to reach such a summary based
on total rent collections as the two bedroom dwellings available in the sample in each suburb is
not the same. Hence, for comparison of rent it makes more sense to compute the average weekly
rent in each suburb for a 2 bedroom property.
b) In the given case, the number of suburbs across which mean comparison needs to be
performed in higher than 2, therefore t test would be found lacking and suitable alternative in
the form of one factor ANOVA test would be used for comparison of weekly rents across the
four suburbs.
.
H0: There is no significant statistical difference between the average weekly rent of two bedroom
dwellings in different suburbs
H1: Atleast one of the suburb has an average weekly rent of two bedroom dwellings which does
not match with the other suburbs.
6

STATISTICS AND DATA ANALYSIS
The relevant excel output from ANOVA is illustrated as follows.
The above computation hints at the F value being computed as 185.999 with the p value as 0.00.
Taking the level of significance as 5%, it may be concluded that the computed p value from the
ANOVA test does not exceed the significance level, hence there is availability of significant
statistical evidence for null hypothesis rejection and alternative hypothesis acceptance. Hence, it
would be concluded that the average weekly rent of two bedroom dwellings tends to vary across
the four suburbs and cannot be assumed to be same.
c) On the basis of the above hypothesis test, it would be fair to conclude that average rent is not
the same for 2 bedroom dwellings across suburbs. Hence, the client needs to be mindful of
this fact and thereby choose a desirable locality based on the available budget. For instance,
Auburn is expected to have the lowest average weekly rent while it is expected to be
maximum for Sydney. Thus, the client needs to be considerate of this fact while exhibiting
his choice.
Bond Amount
a) The requisite scatter plot for highlighting the underlying relationship between the two
variables of interest i.e. bond amount and weekly rent is illustrated as follows.
.
7

STATISTICS AND DATA ANALYSIS
The distinct linear trend highlighted in the graph above clearly reflects the existence of a high
strength positive linear relationship between the bond amount and weekly rent. Although there
are a few aberrations in this regard, but they are very few and the existence of a near perfect
linear trend cannot be denied. But existence of this trend is quite expected since the bond
amount given to the property owner is usually the one month or four weeks rent and this
convention is quite common as the linear trend also hints.
b) The correlation coefficient which highlights the strength and direction of the linear
relationship has been computed and indicates strong positive linear association as indicated by
the graph above. The correlation coefficient value for the given data is 0.889. Despite the
linear trend in scatter plot, the coefficient is quite less than the maximum possible value of 1
as there are some outliers which tend to lower the value. However, it cannot be denied that the
general trend is that the bond amount to be expressed as four times the weekly rent. Hence,
any deviation from the same should be analysed with suspicion.
Conclusion
a) In the wake of discussion carried above, conclusion may be drawn that the weekly rent data in
case of international students was not normal in distribution as there was presence of skew.
Also, outliers on the higher side were present which lead to mean distortion but the overall
dispersion in data was moderate only. For the weekly rent data on suburbs, it is apparent that
houses are very scarce in relation to flat dwelling type indicating that the latter is the preferred
mode. Hence, a client while looking to rent a house should keep the availability in mind.
8

STATISTICS AND DATA ANALYSIS
Also, the mean rent across the suburbs varies for dwelling with two bedrooms. Hence, this
difference should be given consideration by the client who needs to narrow down on locality
based on the underlying budget available for rent. Besides, the bond amount has a strong and
linear relationship with weekly rent which is quite expected as the bond amount asked by the
owner of the dwelling is usually the equivalent of four week rent.
b) Keeping in mind the future research, it makes sense to analyse the weekly rent across the
suburbs for different bedrooms and hence the findings could be compared across bedroom to
bring to light any interesting finding that may seem strange. Further, more suburbs can be
added to the list based on their difference in attributes which would allow the researchers to
understand as to why a particular suburb is expensive while the other one is cheaper. Thus,
with better understanding of the attributes contributing to rent in case of locality, individual
dwelling and bedrooms more precise estimates of the intrinsic rent value can be obtained from
a futuristic point of view.
9