Hero
Most research problems are design problems wearing a statistics costume.
By the time an analysis will not behave, the decision that caused it was usually made months earlier — in how the question was framed, who was sampled, or what was measured. This is guidance on getting those decisions right while they are still cheap to change.
Overview
"I want to study social media and adolescents" is a topic. A research question names a population, a relationship and something measurable. Most of the difficulty in a project is decided here, before any reading.
Choosing a design that fits the question
Cross-sectional, longitudinal, experimental, quasi-experimental, qualitative, mixed. The design should follow from the question — not from what your department did last year, and not from what is easiest to get past a committee.
Sampling, honestly
Who you can actually reach, what that does to your claims, and how to say so in your limitations instead of hoping nobody notices. Convenience sampling is not a sin; pretending it was random is.
Measurement
Choosing instruments that have been validated on a population resembling yours, understanding reliability and validity as properties of a use rather than a test, and knowing when translation or adaptation makes an instrument a different instrument.
Choosing the analysis before you collect
The analysis follows the design and the level of measurement. Deciding it in advance is what stops the results deciding the method.
Qualitative work as method, not as fallback
Thematic analysis, grounded theory, phenomenology and narrative approaches have their own standards of rigour. Choosing qualitative because "the numbers were hard" is visible in the write-up.
Ethics that are real rather than a form
Consent, assent, confidentiality, what to do with a disclosure, and the particular care that psychology research with students, children and clinical populations requires.
Who it is for
- MA/MSc students designing a dissertation
- MPhil and PhD scholars at proposal stage
- Anyone whose committee has sent a proposal back
- Anyone about to collect data who wants a second opinion before rather than after
What we do not offer
Stated plainly, because vagueness here helps nobody:
- We do not write your thesis, dissertation, chapter or paper — not in draft, not as "a starting point", not as "polishing".
- We do not run your analysis and hand back results for you to present as your own work.
- We do not guarantee publication, acceptance or a grade. Nobody honest does.
- We do not arrange, broker or recommend paid placement in any journal.
- We do not sell authorship, and we do not accept it on work we did not do.
- We do not provide therapy or counselling. Research is stressful; that is a real thing and this is not the place for it. Tele-MANAS is free on 14416, any hour.
Process
- You describe where you are — the question, the stage, the constraints, the deadline.
- We tell you honestly whether this is something we can help with, and whether it is something your supervisor should be answering instead.
- We work through the decisions with you — you making them, with the reasoning made explicit.
- You write it. Always.
FAQs
No. We can help you prepare better questions for them, and think between meetings. The supervisory relationship is academic and cannot be outsourced.
Can you tell me which test to run?
We can help you work out which test your design and measurement imply — which is more useful, because you will have to justify it.
I have already collected the data and it will not work.
Common, and worth talking about. Sometimes there is a defensible analysis; sometimes the honest answer is a limitation stated clearly. We will not manufacture the first to avoid the second.
Is my topic any good?
A fair question and one we can help you interrogate — feasibility, scope, whether it is answerable in the time you have.
Integrity disclaimer
You write your own work. We help you think about it.
That line is the whole of it, and it is not negotiable at any price, for any deadline, for anybody.
Guidance means you understand your design, your analysis and your argument well enough to defend them — in a viva, to a reviewer, to yourself in five years. Work produced by someone else and submitted as your own is misconduct, and the person who loses is you: a degree can be withdrawn years later.
We will say no to a request that crosses this line, and we would rather say no early than take it on and disappoint you at the point it matters.
Booking
If you are stuck on a design, a sample or an analysis decision, ask. Tell us where you are, what you are working on, and what has you stuck. If it is something we can help with, we will say so. If it is not — or if it is something you should be taking to your supervisor — we will say that instead.