The Do's and Don'ts of Social Research
The final week's whole substantive content is a checklist the subject did not write. It is a good checklist. It is also shorter than it looks, silent exactly where you most want a number, and pointed at a book you do not have.
What the checklist is and where it comes from
The last topic of the subject supplies one piece of substantive content, and it is not the subject's own. Seven consecutive slides carry images of a checklist headed "'Dos and Don'ts' of social research — How (not) to do a social research project", reproduced from a student-researcher toolkit that accompanies a methods textbook and was published on the publisher's website in 2012.
The counts of twenty-three and twenty-one are this library's; the source states no total and prints the items unnumbered. The subject's own contribution in the final week is the decision to include the checklist and the sentence explaining why.
The list, regrouped by where it bites
The checklist is printed as two flat lists: everything to do, then everything not to do. That ordering separates instructions that belong together — the two supervisor items sit more than twenty apart, and the two data-storage items are exact inversions printed in different halves of the page.
THE CHECKLIST REGROUPED — PARAPHRASED, NOT REPRODUCED
| Where it applies | What the "do" side asks | What the "don't" side warns against |
|---|---|---|
| Starting and framing | Begin thinking about the research early; settle specific research questions; choose a strategy, quantitative or qualitative, and then a design | Leaving everything to the last minute; working from vague, unfocused questions |
| Time, planning and realism | Build a timetable so the work does not end in a panic | Overestimating the time and money you have; forgetting deadlines; expecting everything to go to plan; imagining a good dissertation can be written the night before |
| Records | Keep a research diary; note bibliographic details as you go rather than reconstructing them later | Failing to record what you did and when |
| The literature | Search thoroughly, and keep the search updated as the project runs | Assuming nothing has been written about your topic |
| Access, approval and ethical conduct | Negotiate access, run a pilot study and obtain ethical approval before you start; make sure participants know their rights and consent; be reflexive about the ethics and politics of the work | Starting before access and approval are in place; disregarding your discipline's code of ethics; risking your own safety in the field; using sexist, racist or disablist language in what you write |
| Sampling and fieldwork | Choose a sampling technique that suits the design; stay alert to the interaction dynamics of face-to-face work and what weak rapport does to the validity of the data; keep the research questions in view while collecting | Expecting a high response rate from a survey or questionnaire; being pulled off course by peripheral questions, unless the design is deliberately open-ended |
| Data handling and tools | Allow generous time for transcription; store data securely and keep backups, in a place only you can reach; learn your equipment and software before you need them | Underestimating how long transcription takes; storing data anywhere other people can get to it |
| Analysis | Start analysing as soon as data begin to arrive | Assuming collection has to finish before analysis starts; believing a software package is required to analyse data |
| Supervision | Meet your supervisor regularly and act on the advice; show drafts early | Skipping meetings or disregarding the advice; adopting an unconventional style of analysis or writing without discussing it first |
| Writing and submission | Start drafting early; include every section the report requires; be realistic and honest about the limitations of your methodology | Ignoring your institution's requirements on length, format and presentation |
| After the work is finished | Write to thank your participants and send them a summary of the findings, where that is appropriate | Forgetting to acknowledge participants, your supervisor, any funding body and anyone else who helped |
Every cell is a paraphrase. The original wording and order belong to the 2012 toolkit and are not reproduced here.
It is not forty-four independent items
Four pairs state the same instruction twice, once as something to do and once as something not to do. Keeping a research diary and not failing to record what you did and when are one instruction. So are the two supervisor items. The two storage items are exact inversions of each other: store your data where only you can reach it, and do not store it where others can. Building a timetable and not writing the thing the night before are the same discipline stated from two ends.
Said twice
- Keeping a record of the work as it happens
- Using your supervisor and acting on what they say
- Securing your data against other people's access
- Working to a timetable instead of a deadline
Said once, and easy to skip
- Your own physical safety while doing fieldwork
- The language you use in the written work
- What weak rapport in a face-to-face interview does to data validity
- Keeping the literature search running after the initial search is done
- Being honest about the limitations of your own methodology
Collapsing the four pairs leaves at most forty distinct instructions rather than forty-four. That arithmetic is this library's; the source states no count at all. The more useful observation is where the repetition falls. A checklist that says a thing twice is usually emphasising rather than padding, and all four repetitions are about process discipline — recording, supervision, storage, timing — rather than about method.
The nearest thing to a benchmark, and why it gives you no number
One item comes closer than any other to a performance expectation: do not expect a high response rate from a survey or questionnaire. It is the only item on the list about how much of something you should anticipate, and it carries no figure — no threshold, no range, and no definition of what would count as high.
The subject supplies two figures about response elsewhere, and neither can be read against this item or against the other.
THREE STATEMENTS ABOUT RESPONSE, NO TWO COMPARABLE
| Statement | What it quantifies | Why it cannot be compared with the others |
|---|---|---|
| The checklist's warning not to expect a high response rate | Nothing | No figure of any kind, and "high" is undefined. It sets an expectation without a scale |
| An earlier week's slide: a return of 15% on postal surveys may be a good response, with internet methods said usually to improve the percentage | A return rate on posted questionnaires | Unattributed, hedged twice, with no year, country or survey type attached, and no figure at all given for the internet improvement |
| A set reading from another discipline: journal editors are usually happy with compliance rates of at least 70% | The proportion of people contacted who end up as subjects | A recruitment rate, not a questionnaire return rate, and drawn from a field with different conventions and different study designs |
The 15% and 70% figures are each that source's own, marked as illustrative in the material they come from. Neither is a standard and neither should be quoted as a target for a project management study.
Two items correct a belief rather than a behaviour
Forty-two of the items tell you to do something or to stop doing something. Two do neither. They tell you that something you may believe is not true: that you have to finish collecting before you can start analysing, and that you have to use a software package to analyse data. Both concern analysis, and they are the only items on the list addressed to an assumption.
The distinction matters because a false belief is not fixed by a reminder. You can be told to keep a diary and simply do it; being told that analysis need not wait for collection changes how you schedule the project months earlier.
What follows from the two belief items
The only obligations that fall after the work is finished
Two items sit outside the project entirely. One asks you to write and thank your research participants and give them a summary of the findings where that is appropriate. The other warns against forgetting to acknowledge participants, your supervisor, any funding body and anyone else who supported the work. Everything else on the list is about producing the research; these two are about what you owe once it is produced.
They are worth separating from the rest because they are the items most likely to be skipped, and because one of them can become an obligation you incurred much earlier. A promise of feedback made during recruitment — often the thing that secured access in the first place — outlives the submission date. See voluntary participation and informed consent.
The closing pair, made concrete
- Every person or organisation who gave access, time or data is on a list you kept as you went, not one you are reconstructing now.
- Anything you promised participants during recruitment — a summary, a copy, a briefing — is scheduled as a task with a date, not left as an intention.
- The summary you send participants is written for them, and breaches none of the confidentiality or anonymity you undertook to maintain.
- The acknowledgement covers participants, supervision, funding and any other support, and names nobody you did not clear.
Where the checklist points, and where you cannot follow
Twelve of the forty-four items carry a cross-reference — to a numbered chapter or a numbered FAQ of the methods text the toolkit accompanies, and in one case to a professional association's equality guidelines. None of that material is in the supplied dataset.
One item names running a pilot study among the preparations to make before you start. The pilot stage is named repeatedly across the subject — in the proposal's research design list, in an earlier week's reading, and here — and is explained nowhere. This is its fourth appearance as a named requirement without a definition, a procedure or a worked example. It is recorded and not filled.
Using it as a checklist rather than reading it as one
Read straight through, a forty-four item list produces agreement and no change in behaviour. The items are gates, and gates only work at boundaries. The sequence below is this library's; the source supplies no timing, no order and no instruction about when to use the list.
- Before you commit to a topicRun the framing group: specific questions, a strategy, a design. Vague questions are the item that costs the most later
- Before you approach anyoneRun access, approval and ethical conduct in full, including the pilot and your own safety. This is the group where proceeding out of order cannot be repaired
- Before the first data arriveRun data handling and tools: storage, backups, transcription time, and familiarity with your equipment
- As data arriveRun the analysis group. This is where the two belief items apply and where the schedule usually needs adjusting
- ThroughoutRecords and supervision — the two groups the checklist states twice, and the two most often deferred
- Before submissionRun writing and submission: your institution's requirements on length, format and presentation, and an honest statement of methodological limitations
- After submissionThe acknowledgement pair, which nothing else in the project will remind you about
What to carry forward
- The list is good practice, honestly attributed, and it is the whole substantive content of the subject's final week.
- It is not forty-four independent instructions. Four pairs restate one another, and all four concern process discipline rather than method.
- Its one near-benchmark is unquantified, and the subject's two other response-rate figures measure different things in different fields. No target can be built from them.
- Two items correct a belief rather than a behaviour, both about analysis, and both change how you schedule rather than what you do on the day.
- The acknowledgement pair is the only obligation the list places after the work is finished, and it can be an obligation you incurred at recruitment.
- Twelve items point into a text that is not supplied, so treat the checklist as an index of what to go and read rather than as self-contained guidance.
Frequently asked questions
Is this checklist the subject's own material?
No. It is reproduced in the final week's slides from a student-researcher toolkit accompanying a methods textbook, published on the publisher's website in 2012. The subject's contribution is the decision to include it and a sentence explaining that the link it would otherwise point to is unreliable.
Why does this page not reproduce the list as printed?
The checklist carries a commercial publisher's copyright line. Every item here is restated in this library's words, regrouped by where it applies rather than printed in the published order. If you need the original wording, it belongs to the toolkit rather than to the subject that reproduced it.
What response rate should I expect from a survey?
Nothing in the supplied material supports a figure. The checklist warns against expecting a high response rate and gives no number; an earlier week quotes an unattributed 15% return on postal surveys as something that may be a good response; and a set reading from another discipline mentions compliance rates of at least 70%, which measures the proportion of people contacted who become subjects. The three are not comparable.
What is a research diary and how do I keep one?
The checklist tells you to keep one and the subject never explains it. No week defines what goes in it, how often it is written or how it is used when you come to write up, and this library does not invent a procedure. The related instruction that is stated — record what you did and when — is the part you can act on immediately.
Does the checklist really say I can start analysing before collection finishes?
Yes, twice over. One item tells you to start analysing as soon as you collect, and a second tells you not to assume you have to collect everything first. That is one of only two items on the list that correct a belief rather than a behaviour, and it has direct consequences for how you schedule a project.
How should I use a forty-four item list in practice?
As gates at project boundaries rather than as something to read once. Group the items by the point at which they apply — before you commit to a topic, before you approach participants, before data arrive, as data arrive, before submission, after submission — and run only the relevant group at each boundary. That grouping is this library's; the source supplies no timing.
References and source attribution
- The supplied teaching source: the slide deck for the final topic of a master's-level research methods subject in project management. Author, institution and year are not stated in the supplied files. Seven of its slides are images and carry no extractable text.
- A student researcher's toolkit accompanying a methods textbook, published on the publisher's website with a copyright line dated 2012 — the origin of the checklist. Neither the text nor the publisher is identified in the supplied material, and the toolkit itself is not supplied.
- The chapter and FAQ cross-references carried by twelve of the checklist items refer to that same unsupplied text and cannot be followed from the supplied dataset.
- Bryman, A. 2016, Social Research Methods, 5th ed., Oxford University Press, Oxford — among the social research methods works cited in the supplied source's weekly reference lists. It is not identified in the material as the source of this checklist.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London — among the general research methods works cited in the supplied source's weekly reference lists.
Suggested questions for Ask KEVOS
- Which items in the do's and don'ts checklist say the same thing twice?
- What does the supplied material actually say about survey response rates, and why can the figures not be compared?
- Which parts of the checklist apply before I approach participants?
- What does the checklist say about analysing data before collection is finished?
- Which checklist items point at material that is not supplied with the subject?