Research Methods: the Philosophies, Designs and Sources Explained

Research philosophy, design choices and the layered models used to justify a method.

What this section covers

Research methods is the part of a business course that decides whether the rest of a dissertation is worth anything, and it is taught as a sequence of choices that each constrain the next.

The entries follow that sequence. Philosophy — what you believe counts as knowledge. Approach — whether you are testing a theory or building one. Strategy and technique — what you will actually do. Sampling and analysis — who you will ask and what you will do with the answers.

How the entries relate

Saunders' research onion is the frame the others sit inside. It is drawn as layers for a reason: each outer choice narrows the ones underneath, so a project that picks a technique first and reasons backwards usually ends up with a method that does not answer its own question.

The interpretivist philosophy entry is the inner layer made explicit. It matters because the two dominant positions produce different questions, different data and different claims from the same starting topic, and a dissertation that never states its position tends to mix them.

The consumer research process entry is the same discipline applied to a commercial project instead of an academic one, with the stages named in the order they constrain each other and the errors attached to the stage that produces them.

Where to start

Read the research onion first, then the interpretivist entry, then the process entry. That order is the argument: the frame, the position, the practice.

If you are writing a methodology chapter, the useful discipline is to state each layer and give a reason drawn from your own question. The marks are in the justification; a chapter that names a philosophy without saying what it rules out has described a label.

What separates a method from a procedure

Two criteria run through every entry here, and they are constantly treated as one.

Reliability asks whether a measurement is consistent — would the same instrument give the same answer again. Validity asks whether it measures what it claims to. The relationship is asymmetric: an instrument can be reliable and measure the wrong thing, but it cannot be valid without being reliable.

Almost every weak methodology chapter reports the first and asserts the second.

What a methodology chapter has to answer

The entries are written against the questions a supervisor actually asks, and there are five.

What are you trying to find out? A research question, not a topic. "Employee motivation" is a topic; "whether a shift from individual to team bonuses changes retention in a particular firm" is a question, and only the second constrains a method.

Why does that question need answering? A gap in what is known, or a decision somebody has to take. This is where a literature review earns its place, and where a chapter that summarises reading without saying what it leaves open loses marks.

What counts as an answer? The philosophy layer. Whether you are looking for a measurable regularity or an account of how people understand something decides everything below it.

What will you actually do? Design, method, sampling, instrument — the parts most students start with and the parts that depend on all three answers above.

What will you not be able to claim? The limitations, written as a consequence of the design, not as an apology at the end. A sample of one organisation cannot support a claim about a sector, and saying so is a strength.

The commonest structural error

Choosing the technique first. A questionnaire is easy to design, easy to distribute and easy to describe, so a great many projects begin there and reason backwards to a philosophy that would justify it. The result is a chapter in which every layer is present and none of them constrains the next, which a reader notices immediately.