BUS708 Statistical Modelling For Historical Market Settings And Expectations Essay

Questions:

1. Give a brief introduction about the assignment and search related article and write a paragraph of summary which supports your assignment. You need to give the full citation of the article.
2.Give a short description about this dataset. Is this primary or secondary data? What are types of variablesinvolved? Explain briefly what are the possible cases used in this study.
3. Explain how you collect the data and discuss its limitation (e.g. whether your sample is biased). Is this primary or secondary data? What is/are the type(s) of variable(s) involved? Give a description of cases you consider for this data set.

Answer:

Handling and transporting services, goods, and materials by road, air, rail, sea or air has to turn out to be a complicated endeavor. As the world gets more connected, it has never been so possibly tricky to move someone or something from one point to point another so fast, efficiently and safely. Well-developed logistics and freight businesses are facing exertions satisfying their clients’ full range of new preferences. Their skills, networks, services, and physical assets are an outcome of historical market settings and expectations (Tseng et al., 2005). A new breed of rivals is carving off bits and pieces of the industry, offering intensive services that some look upon as offering more value and innovation than the more extensive, but less targeted, provisions of the developed models of business (Christopher and Peck, 2004).

Australia’s logistics and freight sector include postal services, transport, warehousing, and other transport related services provided to customers across all industries in Australia (Council, 2006). The correct Logistics sector is projected to represent 8.7% of the state’s GDP, it is a substantial cost for Australia’s bulk export businesses, and Australia’s substantial import of manufactured goods implies that effective supply chains from ports to customers are indispensable for warranting that consumers of imported commodities are getting the merchandises at the lowest probable prices (Anderson et al. 2005).

Australia’s far-flung but exceedingly urbanized population of nearly 25 million people presents an obstacle to planners and policy makers of the state’s transport infrastructure (Christopher and Peck, 2004). The structure and design of transportation to and from key population hubs will be fundamental to the growth of trades, to residential expansion, and to people’s aptitude to commute not only to and from municipalities but from business hub to business hub and suburb to suburb (Ghaderi et al., 2015). For Infrastructure in Australia, the main factor to managing infrastructure development is realizing consensus on the major priorities for nationally substantial investment.

1 (b) This dataset presents counts of tap ons and tap offs created on the Opal-ticketing system for the duration of two non-consecutive weeks in 2016. This data is secondary data as it was obtained from the online site of Opal-ticketing system. As said by Nguyen and Tongzon (2010), secondary data constitutes data that has been gathered by somebody other than the user. Conventional sources of secondary data include censuses, government publications, peer-reviewed journals, newspapers, magazines etc. The advantages of secondary data are several. First, they are economical in terms of time and resources. Second, secondary data offer a basis for comparison for the data that is collected by other investigators. Third, secondary data sources are readily available (Harrell and Bradley, 2009). Finally, secondary data helps improve the understanding of the problem being investigated. On the other side, secondary data has been faulted for various reasons including providing inappropriate data as the data is collected by other people, lack of control by the researcher over data quality and quality issues.

The dataset, in this case, is constituted by both categorical and numeric variables. Categorical variables as stated by Hox and Boeije (2005), are variables that someone can assign categories, but the groups have no natural order. In this case, the categorical variables in dataset 1 are the mode of transport (train, bus and light rail), location, tap, and date. On the other side, the values of a numerical variable are numbers (Neuman, 2013). The numeric variable in this data set is time (which is a continuous variable) and count (which is a desecrate variable). Discrete can be further categorized into continuous or discrete variables (Yin, 2009). The discrete variable only take on a finite number of values while continuous variable has an infinite number of possible values (Saunders et al., 2016).

1 (c) The dataset 2 is primary data as I collected it personally for a particular reason. On the word of Black et al. (2002), primary data is an original data that is collected first-hand by the investigator in a particular research project or project. Primary data has several advantages according to researchers. First, primary data is very reliable as an investigator can duplicate the procedure to check the validity of the results, as they understand how the data was gathered and analysed (Neuman, 2013). Second, primary data collection provides the latest data as data obtained from previous years is less likely to answer the questions that a researcher wants to address consistently. Lastly, primary data allow researchers to be subjective in types of data they are gathering in line with the hypothesis they are trying to test. Regardless of the advantages of primary data, this method of data collection is faulted for being expensive regarding resources and time consuming (Yin, 2009).


The main methods of collecting primary data include direct observations, survey questionnaires, and conducting interviews (oral or phone interviews) (Neuman, 2013). In our contest, an online survey questionnaire was randomly distributed to the targeted respondents and the responses recorded for analysis. Simple random sampling is a research technique where each sample element of a given size has an equal chance of being selected (Neuman, 2013). The use of online questionnaire was preferred in this survey as it is less costly regarding administration and is convenience as it enables respondents to participate in any study at any place provided they are connected to the internet. Gadgets like mobile phones, tablets, pcs and desktops usually allow participation or respondents in online surveys.

The two variables of interest under this study were gender and mode of transport as the investigator sought to investigate the relationship between the two. Since this was a small scale survey, a total of 30 responses were recorded out of the 50 targeted responses, which represents a 60% response rate. Under this case, both gender (male and female) and mode of transport (train, light rail and buses) are categorical variables.

2(a) Table 1: Summary statistics

Mode of Transport used by NSW people from 8th to 14th of August 2016

No. of people

Train

81061

Bus

42186

Ferry

1318

Lightrail

1028

Total

125593

Figure 1: Mode of transport used by NSW people from 8th to 14th of August 2016

From Figure 1, it is evident that train was the most used mode of transportation by NSW people from 8th to 14th of August 2016 with 81,061 people affirming this followed by bus with 42,186 people. On the other side, light rail was the least used mode of transport with only 1,028 acknowledging that they used it during 8th to 14th of August 2016.


2(b) ince the total population of travellers is 125593, then 50% is 62,797 (the 0.5 has been rounded off as we cannot have a half person). The null hypothesis of this model is stated as there are more than 50% of public transport users in NSW users (62,797) of the particular mode of transport (train, ferry, bus and light rail). The alternative reads that there are no more than 50% of public transport users in NSW users (62,797) of a particular mode of transport (train, ferry, bus and light rail). Since the percentage representation of the NSW people who used train is 0.6454 (64%), we can reject the alternative hypothesis and accept the null to conclude that there are more than 50% of public transport users in NSW users of particular mode train.

3 (a) Table 2: No. of persons who used Parramatta, Bankstown and Gosford towns

Town

No. of persons

Parramatta

4087

Bankstown

446

Gosford

75

Figure 2: No. of persons who used Parramatta, Bankstown and Gosford towns

As evidenced by Table 2 and Figure 2 as well, considering the three municipalities, we can conclude that the majority of the people (4087 or 89%) visited Parramatta Town to access train services.

3(b) The null hypothesis is stated as there is difference a between mean counts of taps on and off whereas the alternative hypothesis is stated as there is no difference between mean counts of taps on and off.

Table 3: t-Test: Two-Sample Assuming Equal Variances

On

Off

Mean

130.8378

120.614

Variance

140872.8

96601.74

Observations

487

513

Pooled Variance

118160.6

Hypothesized Mean Difference

0

df

998

t Stat

0.470107

P(T<=t) one-tail

0.319191

t Critical one-tail

1.646382

P(T<=t) two-tail

0.638381

t Critical two-tail

1.962344

From Table 3, we can observe that P(T<=t) two-tail 0.638381 is more than the p-value which is 0.05 at 95% confidence level. Thus we reject the alternative hypothesis and accept the null hypothesis. We can thus conclude that there is difference a between mean counts of taps on and off.

3(c) Based on the observations from (b) and (a) above, this paper concludes that the government the best place for the government to build an underground Railway line to central would be from Parramatta.

4(a) Table 4: Summary statistics

Gender

Ferry

Bus

Train

Lightrail

Total

Male

1

5

7

2

15

Female

3

2

6

4

15

Total

4

7

13

6

30

Figure 3: Preferred Mode of Transport

From Figure 3, it can be observed that train was the most preferred mode of transport for the sampled group with 13 respondents confirming this, followed by buses with a representation 0f seven persons. However, the ferry was the least preferred mode of transport for the studied group, i.e. only four persons out of 30 confirmed they travel by ferry.

It is evident from Figure 4 that the majority of females (7) and males (6) prefer traveling by train. However, it is also evident that majority of males (5) love traveling by bus compared to their female equals (2). Whereas this is the case, the majority of females love traveling by lightrail (4) and ferry (3) than their male counterpart (2) and by ferry (1).

Q5 Discussion

From the analysis and the subsequent findings that were obtained above, this paper can conclude that the train is the most favored mode of transport by both genders, followed by buses. However, the use of light rail and ferry is not the best choice for NSW people. Another critical observation that was made is the majority of males love travelling by bus whereas the majority of females love travelling by light rail and ferry. Nevertheless, the government should seek to exploit railway transport as it is the most preferred mode of transportation. This was evidenced by the population of people that travelled by train between the duration 8th to 14th of August 2016. Regarding setting building an underground railway to central, Parramatta is preferred to Bankstown and Gosford. Thus, NSW should consider setting an underground railway line from Parramatta as it has the highest traffic flow compared to the other two towns.

Conclusion

International logistics and freight undertakings are questionably the most multifaceted and thought-provoking supply chains in existence. In one sense, it is a fantastic industry to service, as its growth will continue contributing to the growth of Australia’s economy and changing its shape. It is a tough industry as it is highly competitive and profit margins are way low, with lucrativeness depending so much on “savvy” business decisions and strong operational relationships. It is also an exceedingly monitored market, necessitating a constant balance between future innovation, productivity enhancements and safety compliance. As Australia’s logistics and freight activities grow, one of the biggest tests will be for the sector to work smarter. Thus, Australia need to use the most suitable vehicle types, including, streamline intermodal interfaces, higher productivity freight automobiles, reassure a competitive and dependable rail sector, deploy future technologies to enhance operational results and information visibility, and warrant that the labor force has the right knowledge training, skills, and qualities to succeed.

Recommendation for future Research

Aside from these key findings, there are other areas that scholars, economist, and other government stakeholders can explore to ensure there is maximization of revenue that is generate from the transport sector. These areas include:

  1. Studying the correlation between the location of railway, bus, ferry, or lightrail boarding places and the mode of transport that is chosen.
  2. Studying the cost-benefits implication of travelling by different modes of transport
  3. Studying the determinants of various modes of transport

Reference List

Anderson, S., Allen, J. and Browne, M., 2005. Urban logistics––how can it meet policy makers’ sustainability objectives?. Journal of transport geography, 13(1), pp.71-81.

Black, J.A., Paez, A. and Suthanaya, P.A., 2002. Sustainable urban transportation: performance indicators and some analytical approaches. Journal of urban planning and development, 128(4), pp.184-209.

Christopher, M. and Peck, H., 2004. Building the resilient supply chain. The international journal of logistics management, 15(2), pp.1-14.

Council, A.T., 2006. National guidelines for transport system management in Australia. Background material, 5.

Ghaderi, H., Fei, J. and Cahoon, S., 2015. The impediments to the competitiveness of the rail industry in Australia: the case of the non-bulk freight market. Asia Pacific Journal of Marketing and Logistics, 27(1), pp.127-145.

Harrell, M.C. and Bradley, M.A., 2009. Data collection methods. Semi-structured interviews and focus groups. Rand National Defense Research Inst santa monica ca.

Hox, J.J. and Boeije, H.R., 2005. Data collection, primary versus secondary.

Neuman, W.L., 2013. Social research methods: Qualitative and quantitative approaches. Pearson education.

Nguyen, H.O. and Tongzon, J., 2010. Causal nexus between the transport and logistics sector and trade: The case of Australia. Transport policy, 17(3), pp.135-146.

Tseng, Y.Y., Yue, W.L. and Taylor, M.A., 2005. The role of transportation in logistics chain. Eastern Asia Society for Transportation Studies.

Yin, R.K., 2009. Case study research: Design and methods (applied social research methods). London and Singapore: Sage.

How to cite this essay: