Aims in psychological research

Research begins with a clear purpose, which guides the entire study. An aim is a statement that describes what the investigation intends to achieve.

For example, in a study on animal learning, researchers might aim to investigate whether secondary positive reinforcement could be used to train elephants to complete a trunk wash.

Hypotheses and their role

Once the aim is established, researchers create testable predictions. A hypothesis is a precise, testable statement that predicts the expected outcome of an investigation.

Key components in hypotheses

Hypotheses often involve variables, which are factors that can change or be measured in a study:

Types of variables:

• Independent variable (IV) - The factor manipulated by the researcher to observe its effect.
• Dependent variable (DV) - The outcome measured to see if it changes due to the IV.
• Co-variables - In correlational studies, these are two variables measured to assess their relationship without manipulation.

For instance, in a study on personal space, researchers might predict that preferred interpersonal distance following oxytocin administration would differ depending on whether the person was high or low in empathy.

Operationalisation of variables

To make hypotheses testable, variables must be clearly defined in measurable terms. Operationalisation involves clearly defining the independent variable (IV) and dependent variable (DV) in an experiment or the co-variables in a correlation.

Operational definitions are needed to describe how the variables will be observed or measured.

Steps in operationalisation

A researcher interested in the effects of sleep deprivation (IV) on level of stress (DV) must consider:

Key decisions in operationalisation:

• What is sleep deprivation? Is it missing one night's sleep or having less than 10 hours' sleep over a 72-hour period?
• How will the level of stress be measured – a score from a questionnaire or a physiological measure (e.g. heart rate monitor, beats per minute)?

Making these choices clear in the hypothesis is operationally defining the variables.

Types of hypotheses

Hypotheses come in different forms depending on the research design and the level of prediction.

Main types of hypotheses

• Experimental hypothesis - Predicts an outcome for experiments (IV and DV present). Example: 'Participants who experience 24 hours' sleep deprivation will have significantly higher levels of stress as measured by a heart rate monitor compared to participants who experience no sleep deprivation.'
• Alternative hypothesis - The alternative to the null hypothesis and used for all types of research (including correlations). Example: 'There will be a significant relationship between the reaction time in seconds and amount of sugar consumed (in grams) in the previous two hours.'
• Directional hypothesis - States the kind of difference or correlation, such as higher/lower, better/worse, or positive/negative.
• Non-directional hypothesis - States there will be a difference or correlation but does not state the direction of the difference, effect, or relationship.
• Null hypothesis - States that results are due to chance (no significant relationship or difference) and lack statistical significance.

Recognising hypothesis types

• Look for an IV and DV to identify experimental hypotheses.
• Check if the prediction specifies direction (e.g., 'higher' or 'positive') for directional types.
• Null hypotheses typically state 'no significant difference' or 'results due to chance'.

Data analysis and hypothesis testing

After collecting data, researchers use statistical methods to determine if the results support the hypothesis. Data analysis involves inferential statistics, which show the probability (p value) that the results are due to chance. Data analysis shows which hypothesis should be accepted (alternative or null).

Process of hypothesis testing

1. Collect and analyse data using inferential tests.
2. When analysis is complete, the researchers will reject the null hypothesis if the p value is less than 0.05 (1 in 20).
3. They will accept the null hypothesis if it is more than 0.05 (1 in 20).