32931 Technology Research Methods
Research Proposal Workshop
Prof David McGloin
[email protected]
Natural Language Processing
Acknowledgements + Further Reading
Much of this workshop is based on material developed by the previous TRM coordinator, Barry Jay.
Additionally, some is based on “The Craft of Research” 4th edition, University of Chicago Press, Booth, Colomb, Williams, Bizup and Fitzgerald
Workshop Introduction
In 32144 Technology Research Preparation (TRP) you worked on developing a literature review on your topic of study.
If you are an honours student you may not have completed this task as yet.
In TRM we are interested in developing your understanding of the literature into a research proposal.
Traditionally the proposal forms the first chapter of your thesis, while the literature review forms the second chapter.
This workshop will allow you to start to develop your research proposal and to understand the importance of this type of document in your future research.
Activity 1
In your groups you should introduce yourselves and give each other a quick overview of your research project.
You will be doing some work throughout the workshop in your groups as you start to develop your proposal outline.
Reading
In TRP you will have developed a literature review in the area of your proposed research topic.
A critical part of of any research degree is developing:
A deep understanding of the research literature
An ongoing evaluation of the literature
A paper a day?
Summary of your reading every day
A broad perspective of your general area and related areas
How to do research
The ‘Feynman Algorithm’
Write down the problem.
Think very hard.
Write down the solution.
How to do research
The Polya Principles:
Make sure you understand the problem.
Come up with a plan for solving the problem.
Check that the plan works.
Stupidity in Research
“How could I possibly frame the questions that would lead to significant discoveries; design and interpret an experiment so that the conclusions were absolutely convincing; foresee difficulties and see ways around them, or, failing that, solve them when they occurred?…I remember the day when Henry Taube (who won the Nobel Prize two years later) told me he didn’t know how to solve the problem I was having in his area. I was a third-year graduate student and I figured that Taube knew about 1000 times more than I did (conservative estimate). If he didn’t have the answer, nobody did.
That’s when it hit me: nobody did. That’s why it was a research problem…The crucial lesson was that the scope of things I didn’t know wasn’t merely vast; it was, for all practical purposes, infinite. That realization, instead of being discouraging, was liberating. If our ignorance is infinite, the only possible course of action is to muddle through as best we can.”
Forms of Proposal
You may come across different types or forms of research proposal.
Thesis proposal – a outline of your topic, the problem you wish to address and the importance of doing so.
Typically written in the first phase of your research, for initial candidature confirmation.
Thesis proposal [post hoc] – a version of your proposal written for publication as part of your final thesis submission.
Typically written with the experience of having done the work proposed in the outline proposal
Grant Research Proposal – a proposal aimed at securing research funding.
Introductory Chapter
The typical format of the introductory chapter of your thesis will be something like:
Thesis Statement.
Background (TRP lite)
Aims, objectives, significance
Research Methodology
Outline of results
Outline of dissertation.
Note that each thesis is different and this structure is not always suitable.
Thesis Approach 1:Focussing on your research question
Thesis Statement
A Thesis Statement is
“A statement or theory that is put forward as a premise to be maintained or proved.”
This is a useful starting point for your introductory chapter, and will often come at the end of your introductory paragraph.
Proposal Topic
The topic is the area your work is about.
A broad topic can be described in perhaps three of four words:
“The history of commercial aviation”
“Electromagnetically Induced Transparency”Focussed Topic
Our broad topics are much too broad:
There will be too many sources for us to consider
The information you want is buried within a mountain of information you are not interested in
Difficult to draw anything other than general statementsFocussed Topic
We need to narrow our topic focus
Be careful not to narrow too much or there will be no information to find
Question: Could this be a good thing?
Activity 2
Make a new file on your laptop, called “TRM Research Proposal” (or something equally informative.
Write down your broad research topic area, followed by your focused research area.
Get feedback from your groups as to whether your focusing seems sensible and makes sense.
Focussed Topic
If we recast our topics as sentences we see that the broad topics don’t change very much:
This has the effect of turning our static topics into claims, but has little other value, as they don’t direct us down any useful research avenues.
Focussed Topic
If we recast our focussed topics as sentences we see that they might form interesting claims:
The claims now have a little more substance, but still fall someway short of a convincing proposal. We have no question we can address.
Activity 3
Turn your focussed research topic into a claim.
Developing a Question
Some things to think about:
The history of your topic
The structure and composition of your topic
The categorisation of your topic
Develop negative questions
Why has EIT not been carried out with RF fields?
Develop speculative questions.
Extend questions posed in the literature
So what?
Probably the most important question of all, and one we will return to is “so what?”
Why is studying the thing you wish to study of any importance?
So what if no one can control EIT using RF fields? What do we lose?The answer could be nothing at all, and to some extent that is OK, but ultimately your readers, or funders, will wonder the same thing.
A key component to developing a sensible proposal is offering an appropriate motivation.
Structuring a Research Question
- Topic: I am trying to learn about/working on/studying _
- Question: because I want to find out who/what/when/where/whether/why/how _
- Significance: in order to help my ‘reader’ understand _
- Question: because I want to find out who/what/when/where/whether/why/how _
I am trying to study RF control of EIT
because I want to find out whether low-power compact experiments can be built
in order to help my reader understand their utility as components of future quantum control experiments.
Activity 4
Develop your key research question.
Topic
Question
Significance
Some thoughts on research problems
The ‘pleasure of finding things out’ is a common argument for undertaking academic research.
This is a good starting point for most work.
It seems reasonable that research should aim to have an impact, and to matter to others beyond yourself, however.
To have impact you need to address a problem that others in your community (your ‘readers’) see value in, or need to have a solution to.
Practical and Conceptual Problems
One version of the research cycle can be stated:
Practical problems lead to research problems, and the answers are intended to help solve the practical problems.
Practical problems:
- Originate in the world
- Are based on some cost to society
- Are solved by taking action in the real world
Practical and Conceptual Problems
One version of the research cycle can be stated:
Practical problems lead to research problems, and the answers are intended to help solve the practical problems.
Research or conceptual problems:
- Originate in your mind
- Are based on incomplete knowledge or flawed understanding of how something in the world works
- Are solved by gathering useful information
Problems
Research is about the solution to problems.
Typically problems have a condition and consequences. Consequences typically impose some unwanted cost that you (or your ‘readers’) are unwilling to pay.
In a practical problem we may have:
Condition: I missed the bus.
Consequence: I’ll be late for work and may lose my job.
Condition: The ozone layer is thinning
Consequence: The incidence of skin cancer will increase.Problems
Conceptual Problems can also be broken down in this way. The condition in this case can be any issue for which there is some tangible cost.
The consequences are, however, always some aspect of not knowing or understanding.
In developing a key research question, we have framed a problem:
Condition: I am trying to study RF control of EIT because I want to find out whether low-power compact experiments can be built.
Consequence: in order to help my reader understand [the more important question] of their utility as components of future quantum control experiments.
Thesis Statement 1
If we have developed a sufficiently robust research question which addresses a sufficiently interesting research problem, we can then develop our overarching thesis statement:
I am trying to study RF control of EIT because I want to find out whether low-power compact experiments can be built in order to help my reader understand their utility as components of future quantum control experiments.
That RF control of EIT will allow the development of compact and low-power systems that will form components of future quantum control experiments.
Activity 5
Write down your thesis statement based on your key research question.
Thesis Approach 2:Forming a thesis from a topic and keywords
Alternative Approaches
There are, of course, different approaches to develop a research proposal.
An alternative is to start with your proposal topic and some keywords.
Using my topic:
Electromagnetically Induced Transparency
Keywords:
Mismatched wavelengths; Gaseous systems; Compact; Low power
Approaches
Now you need to consider the systems or approaches that you wish to develop or improve upon.
In my example, this will be:
Optical–optical control; Microwave-optical control; RF-Microwave control; RF-optical control
The we make a table: columns are keywords, rows are approaches.
Hopefully we end up we the ‘best score’ for one row (and ideally the one you are most interested in).
Analysis Matrix
Modify as needed…
It may be that your table/matrix does not give you a clear answer, or maybe forces you to re-think the keywords and approaches.
This is fine – you can iterate until your scoring matrix gives you a useful target.
Thesis statement
We can construct a simple thesis statement combining the keywords and systems.
That the use of RF-optical field coupling will enable the first demonstration of mismatched wavelength EIT in a gaseous system in a compact, low-power configuration.Try and use all your keywords.
Activity 6: Keyword Approach
Build your own keyword/systems matrix.
Do your keywords make sense? Discuss the definition of at least one with your group.
Then construct your own thesis statement.
Dissertation Title
You should be able to derive a sensible dissertation title from your topic and thesis statement:
Electromagnetically induced transparency: RF-optical field coupling will enable the first demonstration of mismatched wavelength EIT in a gaseous system in a compact, low-power configuration.
A bit unwieldy, so extracting key points:
RF-optical field coupling in Electromagnetically Induced Transparency
or RF-optical field coupling for low-power electromagnetically induced transparency
Is the loss of the other parts important?
Dissertation Title
RF-optical field coupling in Electromagnetically Induced Transparency
My thesis is in fact called:
Radio and optical field effects on electromagnetically induced transparencyThis is surprisingly close the title we have just generated. The real title is different, in part, as the work I did is broader than the new title suggest.
Activity 7
Write down your thesis title using the approaches above.
Background
Background
Your introductory chapter needs to set the scene.
Your second chapter is typically your in-depth literature review.
You wrote this in TRP (most likely)…is it up to date?
The introduction needs to offer a summary of the literature review, or broader research area, putting your work in context and demonstrating that there is novelty in your approach or study.
Objectives and Readers
Summary of Chapter 1
Thesis Statement
Background and Novelty
Readers/Stakeholders, your objectives, deliverables, significance
Research Methods and Approaches
Summary of what you are about to read
Deliverables
In an abstract sense your dissertation is your defence of your thesis statement.
Of course as your thesis is written with hindsight, it is likely your thesis and defence may morph to fit the material you have at the end.
In the humanities the defence may simply be words, words, words…
In STEM, we tend to deliver artifacts of some form
An artifact may be a piece of software, an algorithm, a performance, experimental data,a case study, a digital object, a physical object.
Deliverables
Example:
I will deliver a theoretical model of RF-optical EIT
I will deliver a proof of concept experimental verification of RF-optical EIT
I will deliver a detailed examination of optical polarization effects in RF-optical EIT
Crudely we might expect one deliverable per year of research (could be tied to a research paper, for example)
Activity 8
Identify the 2-3 key deliverables for your project.
Objectives: Success
You should always understand what successful completion of your objectives entails. What does success look like? E.g.
Software should meet the original specifications
An experiment should yield meaningful results (which could be positive or negative)
A performance should cater to the desired audience
A procedure or algorithm should have been applied in a meaningful context
A prototype must demonstrate something is feasible
Objectives: Success
In my case, success might mean:
- I will deliver a theoretical model of RF-optical EIT
Success means a model that accounts for a realistic atomic/molecular system, including field polarisations and enables verifiable absorption plots to be created. - I will deliver a proof of concept experimental verification of RF-optical EIT
Success means a demonstration of the effect in rubidium. - I will deliver a detailed examination of optical polarization effects in RF-optical EIT
Success means experimental data illustrating predicted differences in absorption profiles when the optical field polarization is altered.
Activity 9
Take each of your deliverables and indicate what success means in each case.
Objectives
Your objectives will form the contract between yourself and those supporting your research (either financially, materially or via supervision)
Objective = Deliverable + Criteria for success
Ideally these should contain your keywords as well. Note mine have drifted somewhat from my keywords, and perhaps have become higher level and maybe less ambitious in the process.
Have your objectives preserved your keywords? Does this pose a problem?
Stakeholders
In the context of a dissertation, we need to persuade, as a minimum, our readers that the work has some value.
The reader forms our initial stakeholder.
More broadly we can ask the questions
Who is this work for? Who benefits?
Stakeholders
A lot of academic work is compartmentalized within the Academy.
This is fine as far as it goes, but overlap with other communities is both to be encouraged and deemed to be increasingly important for academic research.
While figuring things out for their own sake has an intrinsic value, writing for a small niche audience puts your work at high risk of irrelevance.
In putting your work in its proper context, it is important to consider potential stakeholders for your work.
These could be other academics, different aspects of our community, industrial stakeholders, governmental stakeholders.
Stakeholders
We also need to be careful about identifying stakeholders for the research project you are engaged in and for your dissertation, as these might be very different.
Your work could be in collaboration with a defence organization working on new power sources for autonomous drones.
Your project deliverables might align with with project funders requirements.
However, they may never read, or examine your dissertation.
Activity 10
Discuss who potential stakeholders might be for your different projects.
Who is the likely key stakeholder/reader of your thesis?
If you were using your thesis abstract as part of a job application, who might this appeal to sufficiently to hire you?
Aims and Significance
Aims
The aims of your project are what the stakeholders want (and hopefully what you want too).
They are typically more concrete that your thesis statement, and a little more abstract than the deliverables.
Aim 1: To understand the role that a significant wavelength mismatch between the coupling and probe fields plays in EIT.
Aim 2: To build a compact EIT system in which the coupling field is a non-optical field.
Aim 3: To understand if absorption in a gaseous EIT system can be usefully controlled by changing the field polarisation
Activity 11
Write down the aims of your stakeholders.
Do the aims and objectives form coherent pairs?
If there is not a one-to-one mapping, do you need to adapt either aims or objectives?
Do your key-words appear in your aims and objectives?
Quick Summary
Topic: Electromagnetically Induced Transparency
Keywords: Mismatched wavelengths; Gaseous systems; Compact; Low power
Thesis: That RF control of EIT will allow the development of compact and low-power systems that will form components of future quantum control experiments.
Objective 1: I will deliver a theoretical model of RF-optical EIT
Success means a model that accounts for a realistic atomic/molecular system, including field polarisations and enables verifiable absorption plots to be created.
Objective 2: I will deliver a proof of concept experimental verification of RF-optical EIT
Success means a demonstration of the effect in rubidium.
Objective 3: I will deliver a detailed examination of optical polarization effects in RF-optical EIT
Success means experimental data illustrating predicted differences in absorption profiles when the optical field polarization is altered.
Aim 1: To understand the role that a significant wavelength mismatch between the coupling and probe fields plays in EIT.
Aim 2: To build a compact EIT system in which the coupling field is a non-optical field.
Aim 3: To understand if absorption in a gaseous EIT system can be usefully controlled by changing the field polarisation
Quick Summary
Some of my keywords do not appear in my aims or objectives.
This suggests that either I am not concentrating on the correct things (compact, low power) or that maybe I have overlooked an important area (polarization).
Should probably go back and redraft.
Significance again
Each objective is expected to make a significant contribution to the aim or aims.
While it should straightforward to tie the aims of your project to the central thesis statement, it can be trickier to tie the objectives to the aims.
This is what your significance section needs to set out.
The thesis is important to the reader (the stakeholder)
Achieving the aims supports the thesis
Achieving the objectives supports the aims
Example
To date there has been little consideration of EIT in regimes with widely mismatched probe and coupling fields (Aim 1); exploring a solution, via a theoretical model, to the understanding of whether non-optical fields can control absorption (Objective 1) will open up feasibility studies of future low-power, compact experimental system useful for future quantum control systems.
The development of such a system (Aim 2) will be underpinned by the first verification that EIT is possible in a widely mismatched system in rubidium vapour (Objective 2).
To fully verify the model developed in Objective 1, polarization effects will be explored (Objective 3) with the goal of delivering finer control of the absorption of the probe beam via field control that is not dependent on changing the wavelength (Aim 3).
Summary
Identify stakeholders.
Identify the significance of the topic to those stakeholders (your project is not just for your own personal fulfillment)
Develop aims and objectives
Link these to your overall thesis statement.
Aims: meaningful to stakeholders
Objectives: more technical, measurable (KPIs?)
Significance: meaningful to stakeholders
Real World Implementation
ARC DP Example
Writing
Technical writing for others
Having something to say (hopefully something interesting)
Know your reader
Use a standard organization
Keep it as simple as possible.
Have something to say
Technical writers should have writer’s block.
No thesis statement? No paper.
Know your reader
What is their motivation? Why are they reading?
What do they already know?
What do they want to get from reading your paper?
You can think of the ’reader’ in an abstract sense, or target your writing to specific person ( a reader).
What is it you need to tell your reader?
Reader Motivation
You can make certain assumptions about the reader’s motivation – that they have some familiarity or interest with the general topic area.
You should then try and make them interested in the thing you are most interested in by writing your paper.
One technique is the three part hook
You want to delete spam (second person)
Spammers get smarter everyday (third person)
Together, we can customize your filters (first person plural)
Activity 12
Develop a three part hook for your research project (or a paper you are working on), addressed to your stakeholders.
Expanding your audience
A technique to develop your writing towards a wider audience is to increase the readership with each draft:
a diary/lab book (personal record; will you understand it in a month)
a letter/report (to your supervisor or group member – they know what you know)
a paper (to an audience – they know some of what you know)
ME – YOU - THEMUse a standard structure
This makes it easy for the reader.
Typically a structure is imposed for most STEM based technical writing: mainly paper templates.
Theses have a similar structure – but the important thing is that there is a clear flow and minimize cross-referencing.
Types of reports
Summarising data (IPCC on Climate Change)
Posing a question (Why so many programming styles?)
Answering a question (Smoking is bad for you.)
Presenting options (Renewables or more coal)
Making recommendations (Knock down Building 2 and replace with UTS Central)
Present an argument (An analytical approach is better than a machine learning solution)
Report purposes
Report Structure
Organise for reader not writer
Lots of approaches
known ⇢ unknown
old ⇢ new
simple ⇢ complex
general ⇢ specific
specific ⇢ general
problem ⇢ solution
Activity 13
Your literature review should frame your research problem in terms of a shift from old to new, concluding with where your contribution will lie.
Write a paragraph about your topic in this style.
Draft a second paragraph outlining how your thesis will solve a specific problem.
Why bother with a thesis?
You should be able to succinctly summarise the work you have carried out.
Your final thesis submission will have a very short summary.
Distilling your work into a simple digestible package is worthwhile as it lets you (and others) understand the core purpose of your work.
100 words in 4 sentences:
Problem
Idea
Approach
Outcome
100 words
Title: RF-optical field coupling in Electromagnetically Induced Transparency
Electromagnetically induced transparency (EIT) is typically limited to use of matched wavelengths for the probe and coupling fields. Control of probe absorption using widely mismatched wavelengths would open up opportunities for sophisticated control of atomic and molecular systems. This project will explore the use of RF control of optical probe fields in atomic rubidium. In doing so we will demonstrate the viability of this new approach and lead to compact and low power quantum systems, such as inversionless lasers.
Activity 14
Write a 100 word summary of your project using the problem, idea, approach, outcome structure.
Writing: Editing
Editing
Editing is a methodical way of reading.
No sentence should require a second reading.
If you can work this trick, tell me how….
Homework
As a simple exercise trying taking any page or two of text you have written and editing it for clarity, as well as seeking to remove superfluous words.
As a more advanced exercise, see if you can take a page of a published paper and do the same trick – how much of the writing is really needed.
(Some grant application place very stringent page/word limits (some have character limits), so this type of editing is a good skill to have).
Finally for those who are not too shy, try swapping some of your written work (a page or two) with one of your TRM colleagues and edit each other work. Does this make it more readable? More understandable?
Ongoing Homework…
When writing technical documents, always seek feedback – from your supervisor, from your peers, sometimes from lay readers.
One way to formalise this is through a form of journal club, where you distribute preprints of each others work and then discuss and criticise.
Do any of you belong to journal clubs?
Ways of reading
Read aloud
Read for structure (flow, clarity)
Read for logic
Reading by deletion
Read for presentation (headers and footers, page numbers etc)
Headings as vertical structure
Keywords as horizontal structure
Reading Aloud
This is an excellent way to monitor grammar and flow.
Flow: Does your writing tell a story?
It also lets you root out sloppy mistakes.
Structure
Does the reader know in advance what the structure is?
Is any agreed structure followed?
Are structural promises or signposts within the document adhered to?
Presentation
Often the presentational aspects (figure captions, headers, footers etc) are overlooked, or not proof read as closely as the main text.
Do a visual check. Are all figures and captions where they need to be? Are figures of appropriate size and quality for printing?
Also, might it be better to replace some of your text with a figure (picture tells a 1000 words…)
Logic
Are all terms and acronyms introduced?
Are their unflagged forward references?
Are arguments clear (‘because’, ‘hence’, ‘since’)?
[where appropriate]
: focus on facts not value judgements (or vice versa)
Criticise documents, not people.
Avoid negatives – use positive statements
Think hard about the passive voice (‘I carried out a study’ versus ‘A study was carried out’)
Keep track of quantifiers (‘all’, ’a large number of’ etc)
Never say ‘never’ (or be very careful if you do).
Dot points should have consistent grammar. Do mine?
First person?
Deletion
What happens if quantifiers are removed?
Can you get rid of sub-ordinate phrases and clauses?
Headings
Do your headings tell the story you want to tell?
Are the sections appropriately self-contained?
Keywords
These can be used to cross-link sections:
My intro has:
“There are different issues for researchers when choosing a journal: readership, scope and importance.”
Use these keywords throughout. Don’t change them to, for example, “audience”, “breadth” and “prestige”.
Brevity
Brevit in technical writing, usually (but not always) offers clarity.
This can be true in other forms of writing too:
Hemingway, (so the legend goes), made a bet with some friends – that he could write a complete story, in six words.
Money changed hands.
Hemingway won:
“For sale: baby shoes, never worn.”
Methods
Your Methods
What research methods will you use in your project?
Have you discussed these with your supervisor?
Do you know how to use your chosen methodology?
Does this impact your planning?
Methods linked to objectives
For each objective that you have, you need to think about the appropriate methodology that is required to achieve it.
Who introduced the method (or is it new)?
Is your use of the method standard?
Is your use of the method unusual?
Are your methods mixed?
Methods may involve: experiments, writing software, integrating software, proofs, case studies, observations/surveys, statistical analysis
Theory and Practice
The best laid plans.
Having gone through this process, it is fair to ask, how did it work out?
Well, far from enabling low power control of EIT, I demonstrated that it would, in fact, need lots of either microwave or RF power to see the effects we wanted to.
This rather killed the major objectives of the project.
So I had to revise. We were able to show polarization effects did play a role in the absorption profiles.
We were also able to control optical-optical EIT using an RF field (so three fields) and then to generalize such effects to multi-field control.
Questions?