Chapter 18 — Research methods I

Lesson map

Lesson 18 published. Map, example, EX052.

flowchart TD
  N0["Learning outcomes"]
  N1["18.1 Questions that can be answered"]
  N2["18.2 Choosing methods"]
  N3["18.3 Sampling"]
  N4["18.4 Ethics in research"]
  N5["Synthesis"]
  N0 --> N1 --> N2 --> N3 --> N4 --> N5
  N5 --> EX["Worked example"]
  EX --> Q["Exercise EX052"]

Logical drawing

flowchart LR
  N0["Learning outcomes"]
  N1["18.1 Questions that can be answered"]
  N2["18.2 Choosing methods"]
  N3["18.3 Sampling"]
  N4["18.4 Ethics in research"]
  N5["Synthesis"]
  N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> Q["EX052"]

[Learning outcomes]
    |
    v
[18.1 Questions that can be answered]
    |
    v
[18.2 Choosing methods]
    |
    v
[18.3 Sampling]
    |
    v
[18.4 Ethics in research]
    |
    v
[Synthesis]
    |
    v
[Worked example] --> [Exercise EX052]

Learning outcomes

You will (1) form answerable questions; (2) choose methods; (3) plan data collection ethically.

18.1 Questions that can be answered

Research starts with a question precise enough to answer. "Is the clinic network secure?" cannot be answered; "what share of clinic devices sit outside a planned network group?" can. Priya practises narrowing questions by naming the population, the measure and the time frame. A good question also matters to someone: the clinic team would act on the answer. Her tutor warns that the most common weakness in student projects is a question too broad to answer with the time and data available.

18.2 Choosing methods

Different questions need different methods. Counting misplaced devices needs a survey of inventory and network data. Understanding why staff share the wireless passphrase needs interviews. Testing whether a firewall rule blocks traffic needs an experiment in a lab. Priya maps each sub-question to a method and lists what data each needs. Mixing methods is common in applied security research, because technical measurements explain what is happening while interviews explain why people behave as they do. She notes the cost and time of each method so the plan stays realistic for one term.

18.3 Sampling

Studying everything is often impossible, so researchers sample. If Priya can review only one week of logs, she should choose a typical week, not one with a holiday or an outage, and state why. If interviewing staff, she should include different roles rather than only those who volunteer first. Biased samples lead to conclusions that do not hold. Stating the sample and how it was chosen lets readers judge how far findings apply to the whole clinic network. She records the sampling rule before collecting data.

18.4 Ethics in research

Research involving people or their data needs ethical review. Priya's university requires her to describe what data she will use, how she will protect it, how participants consent to interviews and how findings will be reported without identifying individuals. Using sanitised logs and anonymous interview notes reduces risk. Ethical review sometimes feels like paperwork, but it protects participants and the researcher, and it often improves the study by forcing clearer thinking about purpose and data. Approval must be granted before any data are collected.

Worked example

Priya narrows her question to "what share of devices seen on clinic networks in one typical week sit outside planned groups?"

Synthetic case. The question is synthetic.

Errors and limits

Choosing a convenient but unusual sample period can make results unrepresentative.

Exercises

• EX052 — Narrowing a question. Rewrite "is the network secure?" as an answerable question.
  Show marking criteria

  Rubric: 3 = names population, measure and time frame;; 2 = narrower;; 1 = still vague;; 0 = none.
• EX053 — Interview need. Why use interviews alongside log analysis?
  Show marking criteria

  Rubric: 3 = explain why people act, not just what;; 2 = reasons;; 1 = unnecessary;; 0 = none.
• EX054 — Sample bias example. Give one way a log sample could be biased.
  Show marking criteria

  Rubric: 3 = unusual week such as holiday or outage;; 2 = bad week;; 1 = vague;; 0 = none.

Synthesis

Clear questions, matched methods, fair samples and ethical review make research credible.

Continue studying — chapter 18 of 40

← Chapter 17 (draft) · Course chapter index · Chapter 19 (draft) →