Questions: Formulating Empirical Questions and Hypotheses

Question 1

A researcher states: 'I hypothesize that stress affects sleep quality.' Why does this fail as a scientific hypothesis?

• A. It uses the word 'hypothesize' incorrectly, which should only appear in formal academic writing
• B. It does not specify the direction of the expected relationship, the measurable variables, or the mechanism linking stress to sleep
• C. It is too broad to be tested in a single study and would need to be split into sub-hypotheses
• D. Stress and sleep are correlated variables and cannot be tested experimentally for causal relationships

Answer: (B)

A well-formed hypothesis does three things this statement does not: (1) specifies direction (does more stress lead to worse sleep?), (2) names variables precisely enough to be measured (what is 'stress'? what is 'sleep quality'?), and (3) grounds the prediction in a mechanism (why would stress affect sleep?). 'Stress affects sleep quality' is a research question — it states what the researcher wants to know. A hypothesis commits to a specific, falsifiable prediction.

Question 2

Which of the following is NOT falsifiable, and therefore fails as a scientific hypothesis?

• A. Sleep-deprived participants will recall fewer words on a 20-item list than well-rested controls
• B. Students who study in a quiet environment will score higher on exams than those in noisy environments
• C. Human behavior is sometimes influenced by unconscious mental processes
• D. Participants who receive CBT treatment will show a 10% greater reduction in anxiety scores than control participants

Answer: (C)

Option C is not falsifiable because 'sometimes' provides an escape from any disconfirming evidence. Any single observed instance of unconscious influence confirms it; no amount of evidence showing deliberate behavior could refute it, because the claim only requires the phenomenon to occur 'sometimes.' A falsifiable hypothesis must specify conditions under which data would contradict it. Options A, B, and D all predict specific, measurable effects — data showing the opposite would directly refute each of them.

Question 3

A directional (one-tailed) hypothesis is generally preferable to a nondirectional one because it provides more statistical power.

• T. True
• F. False

Answer: False

Directional hypotheses are only appropriate when theory and prior evidence strongly support a specific direction. While it's true that a one-tailed test concentrates statistical power in one direction, using a directional hypothesis when the evidence doesn't warrant it introduces bias: you are predicting a direction for convenience or publication reasons rather than because theory demands it. The choice should be driven by the state of evidence, not statistical advantage. Using directional hypotheses simply for more power is a form of researcher bias baked in before data collection.

Question 4

A hypothesis is only as falsifiable as its variables are precisely defined and measurable.

• T. True
• F. False

Answer: True

Falsifiability requires specifying what would count as evidence against the hypothesis — and you cannot do that without knowing what you're measuring. If 'stress' is undefined, any measurement could be dismissed as not capturing the 'real' stress the hypothesis was about. Precise operational definitions convert a vague claim into a specific prediction that can be contradicted by data. This is why operational definitions and hypothesis quality are inseparable: vague constructs permit endless reinterpretation that evades falsification.

Question 5

What distinguishes a research question from a well-formed hypothesis, and why does the distinction matter for empirical research?

Model answer: A research question asks what you want to know ('Does social media use affect adolescent mood?'). A well-formed hypothesis is a specific, falsifiable prediction about what you expect to find and why — specifying direction, measurable variables, and a grounding mechanism. The distinction matters because only a hypothesis is testable in the scientific sense: it specifies what data would confirm or contradict it. A research question merely frames a curiosity; a hypothesis converts that curiosity into a commitment — and therefore into something that can be wrong. Being wrong is exactly what science requires, because that is how theories are tested and refined.

The key criterion is falsifiability: a hypothesis must stick its neck out far enough that data could come back and cut it off. Research questions don't do this. A question can be answered by any finding; a hypothesis is confirmed or refuted by specific outcomes. This is why hypothesis formulation is the gateway to the rest of the empirical research process.