Questions and Hyptheses.

Formulating a Question.

The research process is led by a research question: a precisely framed question that your research is designed to answer. Posing your research question appropriately is not a trivial matter, and it will only be possible to pose a sensible research question when you have a considerable familiarity with your research area and the kind of work that you can hope to do.

Personally, I prefer research questions to have several clauses - a general question (which you can't hope to tackle head-on), followed by a sequence of more narrow questions, such as:

"Can consumer decision models be used as the basis for design guidelines for e-commerce sites? Can the cognitive consumer decision-making model of Howard and Sheth be operationalised in an e-commerce design methodology? How might such a methodology be applied to the on-line selling of a high-end of the market lifestyle product? Is the resulting site effective?

A sequence such as this shows your thinking proceeding from a broad category of enquiry down to a fine point of practical application. Chisel this point as finely as you can, because you have to have a clear answer to your most precise question, and this will propagate tentative conclusions back to the more general questions. Suppose that the e-commerce site designed to answer the most precise question posed above turned out to be effective - you might then conclude with a fair degree of confidence that you had successfully operationalised a marketing theory (and others could use your methodology with similar success), and tentatively suggest that other consumer models might also be worth operationalising in a similar way.

Proposing a Hypothesis.

'Hypothesis' might be said to be a polite term for an informed guess. When you have formulated your research question, then you should state (in reasoned terms) what you think the answer will be. The aim of the remainder of the research process is to try and find reasons for either increasing the confidence you have in your hypothesis, modifying it to a form in which you have more confidence, or rejecting it outright. Each of these outcomes might be deemed to be a 'successful research outcome' - depending on the precise nature and circumstances of your research.

In the case of the sample question above the hypothesis might be framed as:

"Well-grounded marketing theory can provide a sound basis for e-commerce site design. Cognitive marketing models are well-suited to being transformed into structured design procedures; and a structured design procedure derived from the well-respected Howard-Sheth model should, all other things being equal, be a sound basis for the design of e-commerce sites which involve the consumer in a complex purchasing decision."

When you have such a hypothesis, then you proceed as if it is true, and effectively challenge real experience to knock it down. If experience fails to knock it down, then your confidence in its truth rises.