GLOBAL TRADE AND MARKETS - Assessment Description and Guide

For both assessment 1 (presentation) and assessment 2 (written research essay) you will be required to work in a group on the same research topic. Oncampus groups should consist of 3 to 4 individuals. Offcampus students should consist of 2 to 3 individuals. You are free to  form your own groups, whether through face-to-face contact or by using the ‘group formation’ forum on CloudDeakin. Ensure once your group is formed that you go to CloudDeakin and get all group members to register under an agreed upon group number that currently has no other people registered to it. You are required to finalise your group formation by the second week of trimester. Individuals who have problems forming groups must inform me by the third week of trimester. Failure to do so may result in no marks for the first assessment item. Datelines for assessments can be found in your unit guide.

Your first priority after forming a group is to come up with a research topic – the same topic will be usef for both assessment items. The research topic will be based on a hypothesised relationship between two variables of some economic interest. Group members should independently think of several possible research topics, and subsequently, come together and choose the most appropriate one. The best research essays from this unit will be published online at the website Deakin Papers on International Business Economics (DPIBE) subject to revisions. Students are highly encouraged to refer to these past essays to get a good idea of the standard required, as well as a source of ideas for topics. 

 Assessment 1 (presentation)

Oncampus groups will be given 12 minutes to present on their research topic. Presentations will occur during the second half of seminar time slots. Some groups will have to present outside of seminar time slots. However, in the interest of fairness, the specific timeslot associated with any particular group will only be announced during the first day of presentations. Any group may be chosen to present on the first day; failure to present on the

day will result in no marks being awarded for the first assessment item. This ensures that all groups must have their presentations ready during the first week of presentations.

During the presentation all group members are expected to be present. At the end of the presentation all students are free to ask questions and feedback will be given immediately; therefore, prepare to take notes.

Off campus groups do not need to present. Instead, they submit their slides directly to me at aaron.nicholas@deakin.edu.au by the due date stipulated in the unit guide. Offcampus groups should bear in mind that their slides will be read as if it is a document, therefore groups are encouraged to include more details (whether in the slides or in a separate document). The  total word count of the submitted documents should not exceed 750.

The purpose of this first assessment is for groups to:

1.   Define a clear and testable research hypothesis
2.   Identify the type of data and theories the group will focus on
3.  Obtain early feedback on how to construct the research essay . The presentation should contain the following:
4.  Hypothesis: state your hypothesis clearly. Also, state any compelling reasons for the alternate hypothesis that you plan to focus on.
5.  Justification: state the theories and evidence from other research papers that lend support to your hypothesis and alternate hypothesis. Also, highlight how your study will be different from the existing literature.
6.  Data and analysis: what variables/data will you need for your analysis? Where will you get the data from? How can you use your data to confirm/reject your hypothesis? (E.g. give an example of how patterns in the data should look like if your hypothesis is supported).

You do not have to actually complete the research analysis for the presentation. The point of the presentation is to show me how you will do the analysis. Consequently, I also do not expect groups to form conclusions at this point.

Assessment 2 (research essay)

The research essay has a strict 2000 word limit. This limit includes in-text references but not the reference list at the end and not the abstract. Use either the Harvard or APA style of referencing:

Submit your essay online using CloudDeakin. Prior to submission ensure the following:

1)                Your group has completed the cover sheet (only 1 sheet per group is required).

2)    Each group member must complete the submission quiz found online on CloudDeakin. Failure to complete this results in an incomplete submission.

3)    When you submit (only one person needs to submit), there should be a Turnitin bar informing you of the percentage of similarity your document has with other sources on the internet. If this is high, please ensure you have referenced and cited sources  you have used accurately as failing to do so amounts to plagiarism. You can resubmit your assignment over a previous submission – it will automatically overwrite.

The research paper should contain the following:

  Abstract: a short paragraph (around 150-200 words) about your hypothesis, your data source, and your main result. The abstract does not count towards your word limit.

Introduction and hypothesis: state your hypothesis clearly. Also, state the  alternative hypothesis. State how you will test your hypothesis and the type of  patterns in the data you will need in order to accept or reject your hypothesis.

 Theory and literature review: what have other studies done that are related to yours, and how does your study differ from those studies? These should ideally include both theoretical and empirical studies. Some of these studies should support your hypothesis, and other studies should support your alternate hypothesis.

 Analysis: this is where you use graphs, tables, figures and summary statistics to support or reject your hypothesis. You should show as far as you can whether the data corresponds to the particular theories and reasons behind the hypothesis/alternate hypothesis (whichever one is found to be supported in the data).

Conclusion and limitations: what are the implications of your findings and what were the limitations of your study?

Marks will be based primarily on sections (2), (3) and (4). Additionally, marks will be awarded based on the overall clarity and coherence among all five sections of the paper. The marking rubric will be available online prior to submission.

 Group Disputes and Peer Assessment

While disputes are not common, they do happen. As we shall discuss in weeks 3 and 4, when there are groups, there is a tendency for ‘free-riding’. To avoid disputes, your group should stipulate a schedule for meetings and tasks, and ensure each group member commits to the schedule. Where possible each group member should keep track of their own personal contribution to the group. This is easiest by keeping different electronic versions of the presentation and research paper with your changes and contributions tracked. E-mails also leave automatic trails. You do not have to go overboard with this: the main purpose of this is to allow easy answering of the question: “what was your contribution to the group?”.

Note that once the presentation is complete, individuals will not be allowed to change groups. If there are any concerns about group member contributions, please speak to your group members first and try to resolve any differences. If no resolution can be achieved, then  inform me as soon as you can. If the problem persists, and all of the following criteria are satisfied, then members of a group identified as not contributing may suffer a penalty in their assignment to reflect this. The criteria for lodging a dispute are as follows:

• If at least two other group members stipulate that a member has contributed less than 20%. 
• The accused member cannot provide evidence to suggest otherwise
• The accused member’s lack of contribution was brought to his/her attention at least two weeks prior to submission of the assignment in question (please keep a record of this)
• The group demonstrates that they have given the accused member sufficient opportunity to contribute (the earlier the group informs me, the easier it is to keep track of this)
• The extent of this penalty is determined by degree of non-contribution of the accused member. Hence, to avoid being in a position where you are unable to provide evidence of your contribution, ensure you keep physical or electronic evidence of your contribution.
• Remember however your priority as a group is to encourage others to contribute, rather than accusing them of lack of contribution. While contributions of each member should ideally be equal, in effect they will not be – this alone is insufficient grounds for a dispute.

  Assessment Guide

 Working as a group

It is useful to have a leader who coordinates meetings and holds the group accountable to a schedule. To keep track of this, construct a schedule early on and e-mail it out to everyone. This e-mail could also stipulate the roles and tasks of each member. Keep records of the responsibilities and deadlines assigned to each group member to encourage accountability Remember that the group leader is a coordinator and not a dictator.

The group should discuss where and how they would like to contribute to the group. Avoid assigning the writing of different sections of the essay/presentation to different group members. This tends to result in an uncoordinated essay that reads very poorly. Rather, the group should come together after doing independent research and decide how to structure the entire essay. I.e. the group should only assign different roles at the early stages  where research and information is required. Once all the information has been gathered the group needs to sit down and work together.

 Forming a hypothesis

The key step in choosing a research topic is the formation of a specific research hypothesis. A hypothesis in applied economics requires a question of causality between two observable variables; i.e. does A cause B where ‘A’ and ‘B’ are variables which are observable. Furthermore the sign of causality should be stated; e.g. does A increase (or decrease) B?

Consider the following example of a hypothesis:

“Does economic growth increase income inequality within a country?”

Variable A is economic growth. Variable B is income inequality. The hypothesised relationship between A and B is positive. It is important to highlight that the causality flows from economic growth to income inequality, and not vice versa. Both economic growth and income inequality are, in principal, observable. It is useful at the outset to define the actual variable that will be used to represent/measure A and B; e.g. economic growth will be measured by annual changes in Gross National Income while income inequality will be measured by the Gini Index.

It is also useful to define the domain or scope of your hypothesis. In the example above, it is made clear that the comparison being made is within a country; i.e. if economic growth increases in Australia, does income inequality also increase? An alternative would be a cross- sectional study with the following hypothesis: "Do countries with higher economic growth have high income inequality"?

It is also important to recognise that the larger the scope of your study (e.g. studying 100 countries) the less likely you are to find coherent patterns. It is often beneficial to restrict your study to smaller groups such as certain states/cities or types of consumers. You may also choose to do your own small-scale survey, though if doing so, please consult with me first.

 Justifying the hypothesis

The hypothesis must be informed by some theory and/or evidence in the research literature. This can be a combination of research articles from academic journals, along with research reports from government organisations and other independent research bodies. There are several ways to go about this. The easiest way to locate articles and reports is by doing an online search. To restrict your search to academic journals, consider doing a search using Google Scholar. When you locate a journal article online, you will be able to access them directly if you are logged on to a Deakin University computer. Otherwise, take note of the journal and article number, and use the following library link: http://www.deakin.edu.au/library/a-z/journals.php

There you will be able to access most journals and their corresponding articles.

Note that while it is okay to quote or refer to other sources such as news articles, to demonstrate you have done sufficient research, the bulk of your reading should be from academic journals and research reports.

In the example hypothesis, there are several reasons one might believe that economic growth increases inequality. One of these reasons could be that if growth is driven by cheap labour owned by large monopolies, monopoly owners and shareholders will accrue all the increases in wealth that the country enjoys. The Kuznet’s Curve is a classic example of theory that tries to describe the relationship between growth and inequality.

While thinking about reasons that justify the hypothesis, it may be worthwhile considering if you should refine your hypothesis to make it more specific. For example, instead of talking about economic growth in general, the hypothesis could be changed to the following:

“Does economic growth increase income inequality within developing countries?”

Now the hypothesis is restricted to only developing countries. This may be the case because the theory stipulates that monopolies tend to have more power and be less regulated in developing countries; this may also simply be due to data limitations. Again, you may also want to focus your study on only a specific country, or region.

The alternate hypothesis

After constructing a well-defined hypothesis with the appropriate justification for it, it is important to ask yourself “is there an obvious answer to my hypothesis”? For example, the hypothesis “when the price of apples increase, the quantity demanded of apples decrease” is an obvious one, and is therefore not worth pursuing research on. One way to determine if the answer to your hypothesis is obvious is to consider the alternate hypotheses. The alternate hypotheses is simply whatever is implied when original hypothesis is rejection. For example:

Hypothesis:

“Does economic growth increase income inequality within a country?” Implied alternate hypothesis if hypothesis is rejected:

“Economic growth does not increase income inequality within a country”

Once the alternate hypothesis is identified it is important to research (from the literature) why such an alternate hypothesis might be true. In the growth and inequality example, it could be that economic growth in fact decreases inequality through an increase in competition, expansion of opportunities and development of welfare support. This type of justification for the alternate hypothesis is an inversion of the sign of the relationship. Another type of justification for the alternate hypothesis is a 'no relationship' one. However this is often very weak and un-compelling – many things have no relationships with each other. E.g. avoid the following justification for an alternate hypothesis:

"Economic growth has no relationship with income inequality within a country because there are many factors that influence income inequality within a country".

If anything the type of reasoning above simply demonstrates that you have picked a poor empirical strategy since you were unable to isolate/identify the effect of the variable you are ultimately interested in. There are always going to be multiple factors that affect one variable. Your job as the empirical researcher is to isolate the effect of your particular A variable. I.e. both your hypothesis and alternate hypothesis is about this isolated effect of your variable A on the variable B.

The alternate hypothesis is important both to highlight that the original hypothesis is not obvious (e.g. growth and inequality could have either a positive or negative relationship) as well as to highlight the neutrality of the researcher. That is, as a researcher you are spending equal time looking at both possibilities and are not biased at the outset towards one or the other.

Typically, many alternate hypotheses exist; i.e. there are many reasons why a hypothesis may not be true. You should pick the most compelling alternate hypothesis to contend with your original hypothesis.

Quick examples of some possible types of hypotheses 

1. Sign-focused

Hypothesis: A has a positive effect on B

Alt Hypothesis: A has a non-positive effect on B (or strictly negative if 'no effect' is not interesting)

2. Direction-focused

Hypothesis: A has a positive effect on B

Alt Hypothesis: A has no positive effect on B but B has a positive effect on A

3. Explanation-focused

Hypothesis: A has a positive effect on B because of C

Alt Hypothesis: A has a positive effect on B because of D

Data

You need to find data that closely represent your variables A and B. These can be usually found online on government websites or world organizations such as the United Nations and the World Bank. If looking for Australian data, a good source is the Australian Bureau of Statistics. The type of data available will usually restrict the scope of your analysis. For example, if you can find data specific only to the United States, it may be useful to think about how you can formulate your hypothesis to be specifically related to conditions in the United States.

Note that if you cannot find data that directly represents your variables, it may be useful to consider if there are proxy variables. For example, you may be interested in how technologically advanced a country is. However, technological advancement is not readily observable. You could instead use data on the number of patents registered in a year, or the amount of research publications produced by the country, or the amount of money spent on research and development. You should then state what you think are the limitations involved in using such a proxy.

Papers and reports often summarise their own data and findings from data. You can utilise this for your paper, but you must cite that you did not use a primary data source.

 The analysis

Prior to beginning the analysis you need to clearly formulate criteria/tests for your hypotheses; that is, you need to ask yourself “what pattern must the data exhibit for me to support/reject the original hypothesis?”. In the growth and inequality case, imagine that you had data for Australia from the period 1980-2010. Then the hypothesis is supported if GNI and the Gini coefficient have a positive relationship. The hypothesis is rejected if GNI and  the Gini coefficient have a negative (or no) relationship. The relationship between the two variables is usually most clearly depicted using a scatterplot. For those who are able, a regression line and test of statistical significance will also help. Note that you can choose to make the comparison across multiple countries at a specific period of time, rather than one country over time. You can also do both.

Do not use different data for your hypothesis and alternate hypothesis. The whole point of the research paper is to verify which hypothesis is true for a chosen set of data.

This is just the first stage of the analysis. What happens in the second stage will depend on what you find in the first stage. If in the first stage, you find evidence in support for your hypothesis, the next step is to verify that the reasons (justification) you gave for your hypothesis is really driving the relationship. For example, if the hypothesis is that inequality is higher as the country gets more developed because of monopoly power and low wages, then you need to find additional data to support that monopoly power has been increasing while wages remain below inflation during this period.

If on the other hand, the first stage rejects the hypothesis, you need to find additional data specific to the reasons and justification behind your alternate hypothesis. For example, if the alternate hypothesis is that inequality is lower as a country gets more developed because of higher taxes, then your next step would be to show that taxes have been increasing over the period. This is why good justifications for both the hypothesis and alternate hypothesis are crucial.

 Thinking about causality

Lastly, you should also think about control variables. While your primary goal is to find data on ‘A’ and ‘B’ you should also think about other factors that are correlated with A and B that could lead you to make misleading conclusions if not controlled for. For example, one may  be interested in whether there is gender discrimination in terms of wages in  Australia. Looking at the data, one finds that women receive less pay than men, on average. Concluding that this is wage discrimination may be erroneous because it may be the case that the majority of women choose to work part-time or in occupations that have a lower salary, both of which result in them having lower pay out of choice and not due to discrimination. In this example, the control variables you will need are whether individuals are working full- or part-time, and the type of occupation they are in.

 Things to avoid

 Do not ‘stitch’ your paper together. That is, do not get four people to do four different parts of the paper and then simply staple them together. You have to spend time ensuring each paragraph flows from one to the other and that the overall paper is coherent. This can only be achieved if each member understands what everyone else  is writing.

  Do not just ‘present’ information. For example, “this graph shows that unemployment has been decreasing over time”. Instead, tell me how decreasing unemployment is related to your hypothesis. Also in your literature review do not just say: “this paper talked about how unemployment increased over time”. Instead, tell  me how the paper’s findings relate to your hypothesis and how your own paper is different. That is, as much as you can, relate all your statements to your hypothesis.

Avoid broad hypotheses. For example, ‘is an appreciation of the exchange rate good for Australia?’. Because many factors affect the exchange rate, you have to take a moment to think about what you are actually interested in. For example, the exchange rate appreciates because of the demand for Australian resources. This may affect the economy differently from an exchange rate appreciation caused by high interest rates in Australia. In addition ‘good for Australia’ is another vague term. Does this mean Australia’s GDP? Or the average level of happiness in the country?

The broad hypothesis could be made more specific as follows: ‘Is an appreciation of the exchange rate caused by the demand for natural resources good in Australia?’.  This is an "explanation-type" hypothesis. Again, the more specific you are, the better. Be wary of specifying topics related to broad macroeconomic indicators like GDP. Numerous things affect GDP. Unless there is a specific strong effect you are looking for (e.g. a recession) it is better to think about a component of GDP like the manufacturing sector, education sector etc.

Sample topics

(you still have to define your own specific hypothesis for each; the less that your hypothesis looks like what I've written, the better!)

1. Do higher education subsidies result in overeducated graduates? You will need data  that can capture/represent overeducation.
2.  Is income inequality higher in richer or poorer countries? Form a hypothesis based  on possible mechanisms, including wages to the high versus low skilled.
3. How does trade policy affect different sectors of the economy? Form a hypothesis based on arguments of comparative advantage, protectionist policies, import substitution, and/or economic stability.
4.  Are developed economies more likely to be characterised by monopolies or small homogenous firms? You will need data on the size/revenue of firms, and be able to compare with less developed countries, or a developed country over time.
5. Whom should childcare benefits be targeted at (e.g. the poor or the middle-class, or the rich?). You will need data on time when childcare benefits have changed, and the impact this has had on the targeted group.
6. How important is the free-market versus socialism in explaining economic growth? Consider the welfare state Nordic countries, contrasted with the market-oriented economies  of the U.S.A. and Hong Kong.
7. Developed countries tend to import less capital-intensive goods. Consider the Leontiff Paradox. Consider comparing a country that has moved from ‘developing’ to ‘developed’ status.
8.  If countries are growing (in an economic sense), why is youth employment and their standards of living decreasing? You will need data of employability and living standards differentiated by age.