Planning, Procedures and Ethics

Designing Research That Works

Collecting data is only valuable if the data accurately reflects what you intended to measure. Two fundamental concepts underpin the quality of any psychological study: reliability and validity. Beyond quality, research involving human participants also carries ethical responsibilities — and in the UK, these are set out by the British Psychological Society (BPS).

Standardised Procedures

A key way to improve reliability is to use standardised procedures. This means every participant experiences exactly the same conditions: the same instructions, the same environment, the same timing, the same stimuli. Without standardisation, differences between participants’ performance could reflect differences in how the study was conducted rather than genuine variation in the variable being measured.

Standardised instructions to participants are a core part of this. Written instructions ensure that every participant receives identical information before the study begins. The researcher delivers the same briefing — or participants read the same text — regardless of who is conducting the session. This also means that if another researcher wants to replicate the study, they can follow the same procedure and expect comparable results.

Key Fact

Reliability = consistency (getting the same result each time the study is run under the same conditions). Validity = accuracy (measuring what you actually intend to measure, in a way that reflects real behaviour). A study can be reliable without being valid: for example, a memory test that consistently scores the same people highly may be measuring intelligence rather than memory.

Randomisation and Allocation to Conditions

Randomisation refers to using chance to allocate participants to conditions. In an independent groups design, random allocation means each participant has an equal chance of being placed in condition A or condition B. This controls for individual differences between conditions — it reduces the risk that one group is systematically different from the other in a way that could affect results.

Randomisation can also be used to determine the order in which stimuli are presented (for example, if a researcher is testing memory for a list of words, the order in which the words appear should be randomised to prevent order from affecting which words are recalled).

Counterbalancing is used in repeated measures designs to reduce order effects. When the same participants take part in multiple conditions, there is a risk that their performance in later conditions is affected by having already done earlier ones — either improved by practice or reduced by fatigue. Counterbalancing divides participants so that half do condition A then condition B (AB order), and the other half do condition B then condition A (BA order). If any order effect exists, it is spread equally across both conditions and cancels out.

Counterbalancing is distinct from randomisation of stimuli. It specifically addresses the sequence of experimental conditions when participants take part in all of them.

Extraneous Variables and Controlling Them

Any variable that could affect the dependent variable other than the independent variable is an extraneous variable. If not controlled, extraneous variables become confounding variables — they provide an alternative explanation for any difference found between conditions. Researchers control extraneous variables through standardised procedures, random allocation, controlled settings, and matched pairs or repeated measures designs.

Reliability and Validity in Research Planning

When planning research, psychologists consider reliability and validity at every stage. Sampling methods affect external validity (how well results generalise to the target population) — a biased sample reduces it. Experimental designs affect internal validity — poor control of extraneous variables means the IV may not be the real cause of any difference found. Quantitative and qualitative methods differ in their reliability and validity trade-offs: quantitative data is more reliable and easier to compare, but may miss the depth and meaning of experience; qualitative data captures richer detail but is harder to replicate consistently.

Psychological research involves human participants, which creates ethical responsibilities. The BPS sets out guidelines that researchers must follow. Key ethical issues include informed consent — participants must be told what the study involves and agree to take part freely, without pressure. They have the right to withdraw at any point without penalty.

Confidentiality means that participants’ personal data must be kept private and not shared without consent. Protection from harm requires researchers to ensure no lasting physical or psychological damage results from participation. After the study, a debrief must be provided: participants are told the full purpose of the study and are given the chance to ask questions or withdraw their data.

Where deception has been used (withholding the true purpose to prevent demand characteristics), the debrief is especially important. Deception should only be used if it is unavoidable and it must not cause participants lasting distress. The right to withdraw data after the study is part of the debrief process.