Order 1369262: Planning

This is detailed practical guidance with examples. Take time to look at it before you write your first assignment. Do not worry if the technical risk management material in the examples is too difficult to understand. Just look at how the technical material is introduced and linked.

Structure

The instructions for some assignments give you a particular structure to use. Dissertations are a bit like this although you have choices and some flexibility.

But for most assignments you have freedom over the structure. We usually recommend you start with an introduction and end with a conclusion, but what goes in between is up to you and should match the process of analysis and argument that you use in your piece.

It is usually a good idea to use sub-headings to help show your structure clearly and keep you focused.

For MANG6182 both the ‘essay’ and the ‘report’ can follow this simple structure. There is nothing special about the ‘report’.

Introductions

Alasdair and I would like your introductions to be succinct and explain the analytical approach taken in the essay/report as a whole. The following example of an introduction is NOT based on this year’s assignments but still shows the difference between a poor introduction and a good one. Here’s the poor version.

“Risk is a key concept in management today, with more and more organizations adopting risk management methods and following risk management regulations. Risk has become a fundamental concern (Power, 2004) of government and wider society. Risk, and the communication of risk, have also become a major concern for clinicians, who are increasingly expected to explain risks to patients as part of obtaining their consent to medical treatments such as surgery. In this context, it is important to understand the effect that talking about risk has on the way people think and, in particular, on the way they make decisions and the quality of those decisions. Motivated by this, Dowie wrote “Against risk” to argue that the use of the word “risk” was unhelpful in this medical context and more broadly. The article discusses the issues in principle through an imaginary conversation involving Humpty Dumpty, but reaches more practical and succinct conclusions and proposals later. One of those proposals is the replacement of various risk phrases with risk free alternatives.

This article will discuss the risk and risk free phrases proposed by Dowie before critically analysing them. The article ends with a conclusion.”

What is wrong with this introduction? Why is this below average despite good English and a citation? The first paragraph is too long and begins with very general waffle. The second paragraph is far too generic and gives us no idea of the strategy used to analyse the issue. We are simply promised some “critical analysis.” The one positive point is that this second paragraph at least makes clear that it is the risk phrases that are the topic of the report, not the whole article by Dowie.

Here’s a better introduction to the same essay:

“Jack Dowie’s article, ‘Against Risk’ argues that using the word ‘risk’ liberally in health decisions with patients is unhelpful and confusing. In addition to discussing the idea in principle, Dowie offers a table with risk phrases and risk free alternatives. This report focuses on his table of alternative terminology. It begins by explaining the alternative phrases and Dowie’s reasons for suggesting them, then goes on to review the disappointingly limited empirical research on the effect of using risk terminology on decision quality, before concluding with some suggestions on how empirical research could be done to test Dowie’s proposals in a practical way.”

Why is this better and far above average? It’s better because it doesn’t waste words on general waffle but gets on with it. It’s also much, much better because it explains what is coming in a way that shows the analytical approach taken. It’s clear from this succinct overview that, despite Dowie’s reasons for suggesting better terminology, almost nobody has tested if the phrases work better or if risk phrases have unwanted effects. Consequently, the report will end with some suggestions on how it could be done. This is much more interesting and promising than just some “critical analysis.”

Sub-headings

Your sub-headings should help to show you analytical approach, or at least the topics you are covering. They should be consistent with the analytical approach you have chosen and that you explain in your introduction.

For example, imagine the task is to write about an empirical study of something. A poor set of sub-headings would be:

Introduction

The study performed

Critical analysis

Conclusion

There are not enough sub-headings and they are rather general. A better approach would be to think about the study and decide what you want to say about it. Suppose you look at the study and you decide that it has so many problems that it is impossible to reach any useful conclusions, but you have some ideas for a better study that could be done in future. In your introduction you would explain that you will explain the study and highlight methodological flaws, then suggest a revised design of study that could be carried out. Your sub-headings could then be:

Introduction

The study performed

Alternative explanations of the results and methodological weaknesses

A proposed study design

Conclusion

These are much better sub-headings. There are more of them and they are more specific. They explain the analytical approach taken.

Alternatively, suppose you were writing about the same topic but you thought that the study design was very good and that the conclusions are correct and should be applied in a practical way in companies. You could say in your introduction that you will explain the study performed and its conclusions, then discuss their implications for company practice of risk management. Your sub-headings could then be something like:

Introduction

The study performed

Key conclusions from the study

Implications for risk management practice

Conclusion

Again, the argument is signposted nicely by the sub-headings.

Conclusion

Writing a conclusion for most assignments is quite simple. Your conclusion should restate your argument but with the examples, evidence, definitions, and other supporting material taken out. Now that the reader has had all that detail, you can restate your basic argument in just a paragraph or two.

Do not introduce new ideas or arguments. Keep it fairly short or you will be repeating yourself too much.

(Conclusion chapters in dissertations are different.)

Clear citations

It is vital at all times to make clear whose ideas or work you are stating. The convention is that if you don’t say whose ideas you are giving then you are claiming them as your own. Here’s an example that fails to make clear whose material is being given. It could be something in a discussion of the value of calculation and modelling, for example.

“The Unconscious Thought Effect has been demonstrated in a variety of preference choice tasks (e.g. choosing which car to buy) where there is no one decisive factor. People who think consciously about the decision for a while then decide will usually make worse choices than people who think consciously for the same amount of time but then have a few minutes of doing something else before making a snap decision. This is due to unconscious thought occurring between the initial deliberation and the final choice. This suggests that, though people engage in a lot of conscious thought when they tackle complex problems, they should do more unconscious thought (Dijksterhuis and Nordgren 2006).

However, another explanation is that the conscious thought usually involves wrestling with various aspects of the choice in sequence and that taking a decision immediately after that activity tends to leave the most recently considered factors over-weighted in the decision. Leaving the problem for a few minutes reduces this over-weighting of recently considered factors. In other words the so-called Unconscious Thought is really just a process of forgetting and memory consolidation. In some studies the quality of decisions has been compared with the strategy of rating each factor in the decision and then using a mathematical calculation to combine the information. This very conscious strategy performs best of all and suggests that the right kind of deliberation is the best strategy.”

Although these two paragraphs are clear and there is even some excellent critical thinking in the second paragraph, there is a problem. It is not clear where the ideas have come from. There is one citation but it appears only at the end of the first paragraph. It could refer to the sentence in which it appears, to the whole paragraph, or to just some sentences within that paragraph. And how do we know if it also is the source of the critical analysis in the second paragraph? You can’t get credit for critical analysis if it’s not clear that it is your work.

Here’s the same material properly attributed.

“What Dijksterhuis and Nordgren (2006) mis-name The Unconscious Thought Effect has been demonstrated in a variety of preference choice tasks (e.g. choosing which car to buy) where there is no one decisive factor. Dijksterhuis and Nordgren (2006) describe a number of studies where people who thought consciously about a decision for a while then decided made worse choices than people who thought consciously for the same amount of time but then had a few minutes of doing something else before making a snap decision. They theorise that this is due to unconscious thought occurring between the initial deliberation and the final choice. Their conclusion is that, though people engage in a lot of conscious thought when they tackle complex problems, they should do more unconscious thought.

However, another explanation is that the conscious thought usually involves wrestling with various aspects of the choice in sequence and that taking a decision immediately after that activity tends to leave the most recently considered factors over-weighted in the decision. Leaving the problem for a few minutes reduces this over-weighting of recently considered factors. In other words the so-called Unconscious Thought may be just a process of forgetting and memory consolidation. In some studies by Dijksterhuis and Nordgren the quality of decisions has been compared with the strategy of rating each factor in the decision and then using a mathematical calculation to combine the information. This very conscious strategy performs best of all and suggests that the right kind of deliberation is the best strategy.”

Not only are the names of the theorists mentioned more times, but they are also identified by the pronoun “they”, so that it is clear at all times whose ideas and work are being talked about. (As an added refinement their sneaky trick of naming an effect using the theory they believe explains it is flagged. A better name would be something like “Delayed Choice Improvement Effect”, because it does not presume the explanation.)

Citations for different types of support

If you cite someone because they suggested an idea or made a claim then the support they provide is much less than if they had actually done some research and got some facts. Always make it clear what kind of citation you are giving by the way to introduce the citation.

Here’s a short sentence where it is not clear what kind of citation it is:

“Intrinsic Risk Attitude indicates the relationship between risk attitude and strength of preference (Smidts, 1997).”

Here’s that same citation, but this time factually more accurate and making clear that it was Smidts’ definition that Smidts was being cited for:

“Smidts (1997) defined Intrinsic Risk Attitude as the functional relationship between risk attitude, as measured by the utility function u(x), and strength of preference, as represented by the value function v(x).”

This is important because this is only one way to define Intrinsic Risk Attitude and far from the most popular.

Here’s a citation where, again, it is not quite clear what the citation is for:

“According to Zhang et al (2008), the four key road safety problems in China are: pedestrians, night time driving, vehicle passengers, and motorcycles.”

How did they come up with these? What was it based on? Did they just give their opinions or was there more behind this? Here’s the same citation but given in a way that more clearly shows the weight that should be given to it:

“On the basis of a wide ranging analysis of official Chinese accident statistics, Zhang et al (2008) identified four areas where there was high potential for reducing fatalities: pedestrians, night time driving, vehicle passengers, and motorcycles.”

This version is also more accurate in stating what Zhang et al actually wrote.

Critical analysis of a research study

A lot of marks are given for critically reviewing the literature rather than just repeating claims made by other researchers. Here’s an example to show exactly what I mean. The first version shows how not to do it.

“Golomb et al (2012) studied the link between consumption of chocolate and Body Mass Index (BMI) and found that people who ate chocolate more also had lower BMI scores. Golomb et al discuss the chemicals in chocolate that may lead to this beneficial effect.”

Here’s a much better alternative:

“Golomb et al (2012) found a negative correlation between the objectively measured Body Mass Index (BMI) of 1,018 people and their self-reported frequency of eating chocolate. Although statistically significant, with the large sample, the size of the effect was small. Golomb at al discuss chemicals in chocolate that may assist weight loss, but it is also possible that many overweight people avoid chocolate, or at least report that they do. No specific causal relation was established by the correlation and no distinction was drawn between high cocoa dark chocolate and high sugar milk chocolate.”

The second version is dramatically better because of the following improvements:

1. Negative correlation is more specific than a vague link of unspecified statistical form.

2. The specific variable that linked to BMI is given: “frequency of eating chocolate”. In the study, quantity of chocolate eaten did not correlate with BMI.

3. The size of the effect is mentioned, carefully distinguishing it from the statistical significance, which is a very different thing.

4. More is said about how the study was done. The sample size is given but, more importantly, the way the variables were measured is given, highlighting the fact that the chocolate consumption was done by the unreliably method of just asking people to say what they ate and how often.

5. The difference between correlation and causation is pointed out, with a specific alternative explanation of the correlation found.

6. A rather obvious distinction about the type of chocolate eaten is pointed out that was ignored in the study.

To improve this further you can add some ideas for a better test – which might be the study you perform or something like it but does not have to be. For example, you could add something like this:

“Demonstrating that eating chocolate promotes weight loss would require a double-blind experimental study with randomly selected subjects and a control group. However, it is possible that some people who are overweight are chocolate lovers, and that is part of their problem, while others manage to avoid chocolate or convince themselves that they do, while maintaining an unhealthy weight by other means. To understand if these behaviours happen at all, a first step would be to interview a small selection of people in their homes and ask them about their chocolate eating habits and perceptions of them.”

Critical analysis of a definition

Definitions very often deserve some critical analysis. Here’s a point with no critical analysis.

“Risk and uncertainty are not the same idea. According to Knight, risk involves known probabilities whereas uncertainty does not.”

Now here’s that point with some critical analysis added:

“Risk and uncertainty are not the same idea. A number of suggestions as to the difference have been made. According to Knight, risk involves known probabilities whereas uncertainty does not. This idea is based on a Frequentist interpretation of probability so for a Bayesian this is not a distinction at all. One could easily supply a probability subjectively or by using an uninformative prior distribution so, what at first seems to be Knightian uncertainty becomes risk. Amended to refer to frequencies, this might be a useful distinction between two concepts but it probably does not agree with general usage of the words by most English speakers, and is not supported by ordinary dictionary definitions.”

Critical analysis of a theory/model

Theories also need to be critically analysed.

“According to risk homeostasis theory (Wilde 1998) humans have a preferred level of risk and adjust their behaviour to maintain it, responding to varying levels of incidents.”

Don’t just state that and try to apply it. This is a ridiculous idea so explain why.

“According to risk homeostasis theory (Wilde 1998) humans have a preferred level of risk and adjust their behaviour to maintain it, responding to varying levels of incidents. As a general theory this seems very unlikely to be true. Consider a person sitting at home watching TV and exposed to virtually no risk at all. Should we predict that the person will need to do something dangerous to compensate, such as juggling with knives? ”

Making suggestions

Here are some things you might want to suggest and examples of well-made suggestions. (Each of these might be followed or preceded by considerably more argument and evidence. This just shows one form of words to introduce the ideas.)