okay, here goes: A thread on tips for writing a postgrad research proposal in the Hum/Soc Sciences. There's a lot going on here, so bear with me.
all of this stuff comes from a postgrad research course i convened + my own reflections. https://twitter.com/AviceBennerCho/status/1277613052284854273

1. Three qs to ask yourself:
- What do I want to know?
- What is known?
- How will i find out what I want to know?
these correspond with:
- context + background of study
- lit review + conceptual/theoretical frameworks
- methodology ( http://all.of.it .)

1.1 the next important thing to do is to break up your word/page count according to section. Depending on discipline and topic, lit review is 25-30%, methodology is 40-45%. You need no more than 25% for intro + context, aims, questions and problem statement.

note: some proposals require abstracts and rationales - check with your dept guidelines. An abstract is 200-250 words and a rationale usually no more than a page and a half. factor these elements into your remaining 25%.

1.2 sections:
title
(rationale)
context/background
problem statement
research questions
aims
literature review
methodology:
- research philosophy
- research design
- empirical context
- sample
- data collection tools
- data analysis
- ethics, limitations and trustworthiness

2. Context of study/ background sections require you to give the narrative of the research problem up til now. depending on scope you could go back and discuss from 10, 20, even 100 years ago to now.
Cover 'who, what, where, when, how' to make clear the problem - the 'why'

3. define a research problem: what is the problem you want to address? go SUPER specific. consider:
'i want to study the effects of gaming on learning'
vs
'i want to study the effects of role-player gaming on logic and problem solving skills among Grade 7 maths learners'

now that you have a problem, your problem statement needs to describe:
- what the problem is, and why it's a problem
- how your research seeks to address +understand it
- the end goal of the study - what insight will the findings offer (e.g. new approaches, understanding, ideas)

4. Res. Questns - work from abstract to concrete, i.e. let your main RQs frame, and your sub-RQs operationalise. See tweet after next for e.g.
don't ask 'why' questions unless the premise can be proven. i.e. 'why do teachers show up late' needs proof that the maj. do.

a better approach is to break the 'why' down:
'why do teachers show up late'
vs
'what are the factors influencing teacher lateness and absenteeism?'

RQ example:
1. How do Gr 11 English teachers address issues of linguistic identity?
1.1 What skills and values are promoted by Grade 11 teachers in their teaching?
1.2 How do learners articulate their linguistic identities based on what they learn in the English class?

RQs are also important for your methodology - each sub-question should correspond with a data collection tool. so, in the example of the question above,
1 = teacher interview
1.1 = observation
1.2 = learner focus group
observation is a tool on its own + used to triangulate finds

5. Aims. what do you want to get out of the research? What do you want to achieve, demonstrate etc? these =/= answering your RQs. they're the 'why'. e.g.
- understd how gaming can be a tool for effective learning
- determine which elements of RPGs encourage problem solving
etc

your big study might want to understand X by studying A, but along the way to X you may also need to prove that a particular method works, or identify gaps in a policy, or understand how propaganda makes research problem A possible. So think of aims like a roadmap.

6. Lit review (this is my fave shit, buckle up kids)
LR is where you really get to flex your knowledge. think of it as an essay where you present a shit ton of evidence. it needs a beginning/middle/end, and a throughline of argument that follows to a logical conclusion.

LR is where you can make a strong case for your study. Try one of 3 broad styles:
methodological: how have ppl approached studying this issue?
thematic: organised acc. to main ideas/themes of the problem
historical: develpt of the literature on the problem over time

style is determined by problem and your proposed methodology. i used a dual structure:
Pt 1: history of lang-in-education in SA from colonialism to now, policy shifts + contemp context and inequalities
Pt 2: Sociolinguistics, critical pedagogy and symbolic power

your lit review also shld cover:
who -has been writing on this
what -are the big conceptual + ethical debates, methodolgicl issues etc
where -does the chat take place? is there a geog bias in the field?
when -are the big moments in the field?
why -is this selectn of lit imprtnt?

a deft LR will cover the above features while following a clear structure and a line of argument that makes the importance of the study very clear to the reader.
think of it as an inverted triangle - work from broad to specific, so reader gets to the methodology fully clued up.

6.1 frameworkssssss
theoretical f/w: describe the existing theory your study is rooted in (such as intersectional feminist or Marxist theory)
conceptual f/w: can be developed from your theoretical framework to more closely align to how you want to conduct your study

theoretical f/ws and conceptual f/ws are often confused. think of it this way:
TFWs discuss -who- you are in conversation with, and why its relevant to your study
CFWs present and structure the ideas you're introducing to the conversation.

so a CFW may take from your theoretical interest in Foucault that you want to look at governmentality.
your topic is SA youth care centres.
your CFW could be a model showing the interplay between institutions, governmentality, social policy and inequality

frameworks are tools that structure what you know. they can help you figure out your RQs and even analyse your data. they emerge from your LR, which is why it's so important to have quality + diversity in what you read.

Also soz but they change. it's the nature of the beast.

methodology coming up next. i need a cup of tea. brb.

7. Methodology. apt that 7 is methodology bc that's the number of truth and knowledge yoooo.
woowoo shit aside, this is the single most important section of a proposal. a reviewer needs to know your work will be feasible and above board.

7.1 research philosophy: this is your theory of knowledge and of being. popular paradigms incl. constructivist, positivist, critical realist. Each philosophy has a particular approach to epistemology + ontology.

a positivist might say reality is only about what we can see and observe, while constructivists and realists are concerned with the underlying and unseen factors influencing the reality that we see. You see why this directly impacts your choice of methods?

7.2 research design: is your research quantitative, qualitative, or mixed methods? Are you doing a case study? Secondary data analysis? Life history? Experiments? Defend your choice of whether qual/quant/mm, and also your choice of overarching method.

certain methods don't play well with certain paradigms - think about how positivism would interact with narrative analysis! there needs to be cohesion between R Philosophy, R Design, and data collection tools.

7.3 empirical context: this is the context of your actual research sites. provide a tight, concise description specifically in terms of the research you are going to conduct. You can for e.g. give more details about local dynamics, specific actors and so on.

e.g. in my study i gave 'general' context in intro + lit review. in the methodology i introduced the participant schools and participant teachers, describing their institutional features + educational philosophies, + the teachers' professional profiles.

7.4 sample - pls do not sleep on sampling. how are you going to choose participants? does this strategy fit with your philosophy, design, and research problem?
e.g. what would happen if you used a representative sample to ans the q 'what are factors influencing teacher lateness'

vs what would happen if you used a snowball sample? Your choice of sampling strategy can skew your data, whether you're doing a quant/qual study. explain why you've chosen it and how this will ensure quality, trustworthy results. you can use more than one sampling strategy, but

it must be appropriate to the data collection tool. this is also true if you're doing document/content/literary analysis. your criteria for selection must appropriately define how you'll sample texts/media and how you intend to control for biases.

this is from one of my ppts about sampling - it's not just about who, but about how many, as well as:

7.5 data collection tools - how will you gather data? examples of tools incl. interviews, observations, journals, photo elicitation. you will need research instruments, such as interview guides or surveys. and you need to discuss and justify your choice of tools, with literature.

*sometimes when you collect the data matters as much as how. you will need to provide a logical sequence of data collection. for e.g., it may be better to schedule interviews/focus groups after you've finished observations - you could skew your findings by introducing new ideas.

then again, if you're doing action research or similar, you may want to see what happens if you introduce a new idea/practice to a setting once you've established a baseline. this is why your choice of research design is so important to everything else in the methodology.

if you take anything away from this thread, it's The Table. In one little piece of work you end up distilling half your methodology - it's clear, it's simple, it's effective. your reviewer will love you for it.

the table also serves as a checklist for you - it makes sure you know exactly what you're doing, with who, and how many of them (laugh if you have filth for brains). it also tells you what instruments you need to develop.

7.5 data analysis - how do you plan to analyse your data once you have it? there are a range of methods for both quant and qual analysis.
main methods in qual: discourse, content +narrative analysis; grounded theory
quant: descriptive + inferential stats (which give rise to oths)

before you analyse you need to sort - and scrub your data for inaccuracies, duplications, and fact check certain claims (esp. historical stuff). then you code your data, which helps with picking up general themes later on.

you could use a software package to do your analysis, in which case discuss how you intend to do so. if you're doing manual analysis, discuss the steps you're going to take to process, organise and analyse the data, including if you're using particular frameworks.

i developed an 'analytical toolbox' from my original conceptual framework. i created sets of indicators for each analytical category i was studying, and interpreted my data against these indicators and categories.

8. Ethics. I separate ethics from methodology bc ppl like to treat ethics as an afterthought. it's not. esp if you're working with vulnerable people, kids, marginalised or underrepresented groups - a rigorous ethics section is vital.

8.1 Ethics can be divided into 2 dimensions. procedural ethics is the practical stuff - getting permission from gatekeepers like the govt, the organisation, participants. here you need to clearly state how you'll provide information on the study, the risks/benefits + ways you

will protect participants' information, their identities, etc. consider also the language you'll be conducting the study in, barriers to participation like literacy or disability. procedural ethics is about establishing a relationship of accountability. this is important.

but the deeper stuff is about honouring your participants' trust. think about ethical issues that might emerge specific to your study. then think about how you might resolve those issues, and offer some solutions. one example i can think of is when we have focus groups with kids

and they become overwhelmed and start to cry. i switch off the recording device and lead the group in a soothing ritual to calm the child down, and we have a few minutes to laugh and relax again before the child gives me permission to record again.

important to this is that EVEN IF that kid said the most amazing thing when the recorder was off - you do not repeat it. you violate your ethics when you do that. step outside your study for a second and think about how your participants might experience it.

that's the key to good ethics. so when writing up your ethics, consider how your study might bring harm to someone and what you're going to do to protect them - whether that means paying for translators, getting a trained assistant, or simply establishing ground rules.

8.2 Limitations: your study can't do everything, so what's the important stuff it's missing? maybe your methods don't allow you to provide a representative case study. maybe your post-colonial analysis is confined to the SADC region. maybe you can't get access to particular

participants, which affects the breadth of your analysis. be thoughtful when describing your study limitations, without self-flagellating. your reviewer wants to see that you know your study is valuable and feasible in spite of the things it can't deliver on.

my eyes are fried right now. i'll return to this thread tomorrow to discuss validity, triangulation and reliability. :)