4.2.3.1 - Hypotheses

What A Hypothesis Does

A hypothesis is the researcher's prediction. It states the pattern the researcher expects to find and should be set before examining the results being used to test it.

The key word is testable. A hypothesis must be specific enough for someone to collect data and decide whether the results support it. "Caffeine affects performance" is too vague because it does not say what caffeine dose is used or how performance is measured. “Adults given a drink containing 100 mg of caffeine will correctly recall more words from a 20-word list after a two-minute delay than adults given an otherwise matched caffeine-free drink” is more precise. It specifies the manipulated amount and the scored outcome. Coffee volume alone would not specify caffeine dose, because different coffees contain different amounts.

AO1: the core distinction. AQA requires you to know two types of research hypothesis:

This lesson focuses on research, or alternative, hypotheses. A null hypothesis says there will be no difference or relationship, or that any observed effect is due to chance. In some broad "write a suitable hypothesis" questions, AQA mark schemes may allow a null form. However, if the question asks for a directional or non-directional hypothesis, a null hypothesis does not answer that command.

Directional Hypotheses

A directional hypothesis predicts which way the result will go.

A directional hypothesis tells the examiner which condition is expected to score higher, lower, faster, slower, more or less. In a correlational study, it predicts whether the relationship will be positive or negative.

AO1: experimental example.

Participants who revise with spaced practice will recall more psychology key terms from a 30-item test than participants who revise in one long session.

This is directional because it predicts which group will recall more key terms.

AO1: correlational example.

There will be a positive correlation between hours slept the night before a test and score on a concentration task.

This is directional because it predicts the type of relationship: as sleep increases, concentration score is expected to increase.

AO2: spotting direction in a scenario. If a researcher predicts that students using flashcards will score higher than students using rereading, that is directional. If they predict that social media use will be negatively correlated with sleep duration, that is also directional. The direction can be an increase, a decrease, a positive correlation or a negative correlation.

AO3: when it is justified. A directional hypothesis is most appropriate when previous research or a clear theory gives the researcher a good reason to predict one outcome. The reason must exist before inspecting the new results. Choosing the direction after seeing the scores would turn an observation into a prediction after the event.

Non-Directional Hypotheses

A non-directional hypothesis predicts a difference or relationship without choosing its direction.

A non-directional hypothesis is still a proper prediction. It is not vague. It must still name the variables clearly, but it leaves open which condition will be higher, lower, faster or slower.

AO1: experimental example.

There will be a difference in the number of psychology key terms recalled from a 30-item test between participants who revise with spaced practice and participants who revise in one long session.

This is non-directional because it predicts a difference but does not say which revision method will produce better recall.

AO1: correlational example.

There will be a correlation between hours slept the night before a test and score on a concentration task.

This is non-directional because it predicts a relationship but does not say whether the relationship will be positive or negative.

AO2: spotting non-direction in a scenario. If a researcher says "there will be a difference in obedience scores between participants tested by an authority figure and participants tested by a peer", the prediction is non-directional. It does not say which group will obey more.

AO3: when it is justified. A non-directional hypothesis is appropriate when the research area is new, when previous findings are contradictory, or when a theory does not clearly predict the direction. It leaves the direction open, but the variables should be just as precisely operationalised as in a directional hypothesis.