Effective research begins with converting general interests into specific, answerable questions that specify the variables, populations, and context of investigation. A well-formulated research question is narrow enough to investigate thoroughly yet broad enough to contribute meaningful knowledge. The specificity of your question directly determines what design, measures, and analyses are appropriate.
Start with a broad topic of interest and practice narrowing it through iterative questions: 'What specifically?', 'In whom?', 'Under what conditions?' Compare overly broad questions ('Does stress affect cognition?') with specific ones ('Does acute time pressure impair decision accuracy in high-stakes financial tasks?').
From the scientific method, you know that empirical research is a cycle: observe, hypothesize, test, revise. Research question formulation is where that cycle begins, and the quality of the question determines the quality of everything that follows. A vague starting question does not produce a vague study — it produces an incoherent study, because every downstream decision (what to measure, who to study, what design to use, how to analyze results) requires the question to be specific enough to constrain the choices.
The transformation from interest to question follows a predictable narrowing process. Interests are broad: "I want to study stress and memory." That is a topic, not a question. The first narrowing move adds specificity about the outcome: "Does stress impair memory encoding or retrieval?" Still broad — what kind of stress? What kind of memory? Who are the participants? Under what conditions? A well-formed research question specifies: the independent variable (the thing that varies), the dependent variable (what is measured), the population (whose behavior you are studying), and the context (under what conditions). Compare "Does stress affect cognition?" with "Does acute cortisol elevation — induced by a standardized social evaluative threat — impair free recall of neutral word lists in healthy adults?" The second question directly implies a design: an experimental cortisol manipulation, a free recall task with neutral stimuli, and a sample of healthy adults. The design is entailed by the question.
Specificity also determines what counts as an answer. A vague question can always be answered with "it depends" — which provides no traction for building knowledge. A specific question can be answered "yes, under these conditions" or "no, this manipulation did not produce that effect" — both of which advance the field. One of the most valuable discipline-specific skills in psychology is learning to ask questions that are falsifiable, not just interesting. If your question could not, in principle, come back false, it is not a research question — it is an assertion dressed up as inquiry.
A final point about iteration: research questions are not set in stone at the start. The literature review frequently reveals that your initial question has already been answered, answered with important limitations, or involves a conceptual confusion that prior work has clarified. This is expected. The question is refined, not abandoned, when you discover prior work. A question that builds precisely on a gap in the literature — "Prior studies examined acute stress; none have examined chronic stress" — is stronger than one generated from curiosity alone, because it is positioned relative to what is already known. The specificity you build in question formulation is what allows you to locate that gap.