Chapter 3 Developing a research plan

3.1 Research proposals

 Like most engineering work, a research project should be structured and costed before
it is commenced.
 This should ensure that the project plan is coherent and viable and acceptable to the
project team and that all resources are available.
 The project description is usually planned with one or more prospective supervisors,
includes the background to the research, the contribution it will make, the resources
required, the tools to be used and the likely outcomes.

2/4/2024 1
3.2 The elements of a research proposal
 Generally writing a research project proposal is like starting to write a final research report.
 Figure 3.1 shows the basic structure of a research proposal.
 A research proposal includes:
 a title,
 list of team members,
 a project summary
 a project description, and
 budget.

2/4/2024 2
2/4/2024 3
Figure 3.1 Generalized form of a research proposal.
 The elements of a research proposal are discussed in more detail below:
3.2.1 Project title
 As is the case for the titles of scientific papers, the title of a research project should be clear
and specific.
 More than ten words are recommended for the title.
 Acronyms should not be used as the assessment panel might not be familiar with their use.
 The title should indirectly indicate the engineering discipline of the research so that the
readers can direct the project proposal to those who have relevant experience in the discipline.

2/4/2024 4
3.2.2 Research team
 It is rare for a project to involve only one researcher.
 The evidence for this is the very low number of single author research articles that are
published.
 Thus the research team for one student will include at least one academic advisor.
 In addition the project is likely to need the support of other experts in specialised fields.
3.2.3 Project summary
 The project team needs to provide a short, clear statement about how the project will benefit
their organization and how it fits into the goals and objectives of the research field.
 The project summary should also contain keywords so that the literature in the field can be
thoroughly surveyed for competing technologies and research outcomes.

2/4/2024 5
3.2.4 Project outline
 As is the case for all research reports, the project outline should include a number of
sections.
 The research question which covers the aims and relevance of the project.
 One might then add a number of specific aims which clarify the methods of approach to be
used.
 The background section should contain a review of the literature and an argument that the
proposed research will create new important, relevant knowledge.
 This analysis might include some of the initial work published by the research team. This can
induce confidence in the assessment panel reviewing the research application.
 The research methods must suit the research question and the research aims in a clear and
logical way.

2/4/2024 6
 The data analysis section should outline the data processing required to remove interfering
effects including calibration methods, and the type of statistical analysis that will be used to
prove that the outcomes are valid ‘beyond reasonable doubt’.
 The project plan must clearly state a list of deliverables which will arise from the project.
 Deliverables are things which can be accessed by others such as reports, software code,
presentations, images, etc.
 It is important that novice researchers understand that their own understanding of existing
information is not a project deliverable.
 However, reports, presentations, patents, designs, objects, etc can be delivered to the granting
authority as an indication of progress in the research project.
 New knowledge is only created when it is publicly available. In some cases, the research team
might specify the conferences and journals where their results will be disseminated.

2/4/2024 7
 The timelines are particularly important in any planning process. This ensures that the
granting organization can be kept up to date with research progress in a formal manner.
 For commercially funded research projects, there might be a requirement for monthly
meetings to report progress.
 For government research projects the reporting might be required annually.
 Regardless of the formal reporting schedule, the research team should meet regularly to
review progress against the projected timelines contained in the initial project proposal.
 Simple representations of the time lines using a spreadsheet or a simple text document can be
used to indicate the work breakdown and the timing of deliverables (see Table 3.1).

2/4/2024 8
Table 3.1 Simple representation of timelines for a project. The crosses represent major time
requirements. Q refers to one quarter of one year. Only the major deliverables ◊ should be
included. The work titles must be specific (not generic as presented in this illustration).
Work Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12
Literature review X X X X X X X
Equipment arrival X X ◊
Equipment calibration X X
Sample preparation X X X X X
Data acquisition X X X X X ◊
Statistical analysis X X
Numerical modeling X X X X X X ◊
Report 1 submission X ◊
Report 2 submission X ◊
Theoretical development X X X X X ◊
Model optimization
Final report submission X X X X X ◊
9
 Concurrent engineering principles should be used throughout the project and clearly indicated
in the timelines given in the research proposal.
 Thus the timelines should indicate that a number of activities will be driven simultaneously.
Table 3.1 gives an example of these concurrent timelines.
 At the start of the project, the literature review (an update on the literature review presented in
the project outline) might run simultaneously with the ordering and installation of the test
equipment, the acquisition of the samples to be tested and the organization of the data entry
systems.
 In this way, the granting authority should be confident that even when there is a delay in one
part of the project, other aspects can and will proceed.

2/4/2024 10
3.2.5 The budget
 With the timelines defined in Section 3.2.4, it is possible to estimate the budget required to
under take the project.
 The budget will require cash and ‘in kind’ support. The ‘in kind’ support is a list of resources
available to the project which do not require direct financial support.
 For example, the host institution might continue to pay the salaries of existing staff who work
on the project, they might provide facilities (space, electricity, internet access, computing
support, etc) without the need for additional funding.
 In some cases the host institution will seek a contribution to these costs calculated as a
percentage of the total cash cost.
 The research team must engage in detailed planning to ensure all resources, including human
resources, are available.

2/4/2024 11
 Conflict of interest: When a research team receives funding from a commercial partner,
some ethical issues might arise. In particular, researchers must ensure that the results are
presented in an unbiased manner and that all conflicts of interest are clearly stated in any
publications, presentations and reports. This can be done in the acknowledgements section of
the paper.
Example 3.1 Conflict of interest
 An investigation into a bridge collapse is seen to be biased if conducted bythe company that
constructed the bridge.
 An assessment of the health effects of medical implants is seen to be biased if funded by the
manufacturer of the implants.

2/4/2024 12
3.4 Design for outcomes
 The research plan must be designed to give definitive answers to the research question.
 This must involve statistical support for the results.
 Multi-parameter investigations involve a number of variables which might or might not be
controlled. This depends on the nature of the investigation.
 Thus if the researchers measure one parameter (the dependent variable) as a response to
changes in another parameter (the independent variable) the consistency of the data is a
measure of the strength of the data.
 This can be simply observed by looking at the scatter in the data points about the trend line,
but a measure of this scatter is very important in establishing the accuracy of the trend line.

2/4/2024 13
 Almost all scientific and engineering papers use figures to represent the experimental
method as a block diagram and to summarise the results using graphs.
 These graphs must be self-sufficient; that is, the reader should be able to gain a level of
understanding from the graphs and the associated captions without referring to more detailed
explanations given in the text.

3.4.1 One-dimensional data

 Some research projectsseek to gain one piece of information only– a single number. We will
refer to this as one-dimensional data.
Example 3.2 One-dimensional research questions
 What is the mean voltage measurement?
 What is the mean size of the nanoparticles?
 What is the optical reflection coefficient of soil?

2/4/2024 14
 The presentation of one-dimensional scalar data is relatively simple.
 A single number with an error marking the upper and lower bounds constitutes a one-
dimensional measurement.
 The same measurement of a number of samples represents an array of one-dimensional
numbers.
 These data are commonly represented by a histogram, where the range of the measured
values is displayed (see Figure 3.2).
 The mean value of the dataset can be presented as a one-dimensional result of the
investigation.
 The standard deviation from the mean value is a measure of the variations in the
measurements.

2/4/2024 15
Figure 3.2 Two hundred voltage measurements were made from a strain gauge attached to a
static beam. The measurements are plotted as a 20 bin histogram to illustrate the variation in the
measurements. The bins or intervals are ranges of the measured voltage.

2/4/2024 16
3.4.2 Two-dimensional data
 If a variable is measured as a function of time (for example), then a two-dimensional dataset
is generated by measurements.
 Commonly this will be represented by a line graph or scatter plot (see Figure 3.3).
 Note that if the data represent a continuous function a line should be used, but if the points are
experimental measurements the individual pointsshould be plotted without connecting lines
 A theoretical line fit can be used to illustrate the theoretical relationship.

2/4/2024 17
Figure 3.3 Scatter plot of the measured voltage as a function of time. The linear fit (continuous
straight line) has a slightly positive slope, which indicates that the voltage might depend on the
time
2/4/2024
elapsed. 18
 Now that the one-dimensional problem has developed into a two-dimensional problem, so
also the possibility of a third parameter might influence the dependent variable.
 Consider the possibility that temperature might affect the results.
 In making the measurements, assume that the temperature and the time of measurement were
recorded.
 The research team is now in a position to evaluate the impact of temperature on the
measurement.
 This leads us to a three-dimensional analysis.

2/4/2024 19
3.4.3 Three-dimensional data
 If a third parameter is recorded during the measurements of the original parameter, scatter
plots on the same axes can still be used.
 It is quite simple to distinguish some points on the basis of this additional parameter.
 The data are now three-dimensional, but a two-dimensional representation can still be used.
 Assuming the additional parameter measured was temperature,
 Figure 3.4 includes circled points where the temperature was relatively high when compared
to the total data set.
 It is possible to treat the high temperature values independently of the total population to
decide statistically if there is a significant difference.

2/4/2024 20
Figure 3.4 Scatter plot of the measured voltage as a function of time. Those values with a
temperature in the highest 10 percentile have been circled. By applying an appropriate statistical
technique it is possible to determine if the temperature effect is significant.
2/4/2024 21
3.5 The research tools
 In every research project, the challenge is to validate the results using independent means.
 The research plan must clearly outline the techniques which will be used to undertake this
validation.
 The strongest case can be made if the experimental or numerical results can be
validated against a theoretical model.
 If two very different and independent methods yield results which are statistically identical,
then it is highly likely that both tools used in the research are valid.
 There still remains a small probability that both techniques yield results with the same bias;
that is, both the results gained from the research project and the published results are
incorrect.
 This might occur if the same error is common to both methods of analysis.

2/4/2024 22