Research Methods and Practical Requirements

What you'll learn

• How psychologists plan investigations using hypotheses, variables, designs, sampling and ethics.
• How to compare experiments, interviews, questionnaires, observations, case studies and correlations.
• How to analyse data using GCSE descriptive statistics, tables, charts and graphs.
• How to judge reliability, validity and bias in novel sources and core studies.

The big picture: research methods across GCSE Psychology

Research methods are the tools psychologists use to collect and analyse evidence about behaviour and mental processes. In OCR J203, research methods can appear in both components: Component 01 often focuses on designing an investigation, while Component 02 often asks you to apply methods to a novel source. They can also be assessed inside topic questions.

You are also encouraged to carry out small, ethical practical activities. This matters because research methods are much easier when you have actually tried planning, collecting and analysing data yourself.

Research methods also link to the named core studies across the course. For example, Bickman (1974) can be discussed as a field experiment, Tandoc et al. (2015) used questionnaire-style self-report and correlational data, and Wilson, Kopelman & Kapur (2008) is useful when thinking about case-study evidence. The dual-purpose brain studies are Wilson, Kopelman & Kapur (2008) and Daniel, Weinberger, Jones et al. (1991).

Planning Research

Aims, hypotheses and variables

An aim is the general purpose of a study. A hypothesis is a testable prediction about what the researcher expects to find.

An experimental hypothesis predicts a difference or relationship. A null hypothesis predicts no difference or relationship, except any difference due to chance. A directional hypothesis predicts the direction of the result, while a non-directional hypothesis predicts that there will be a difference or relationship but does not say which way.

A variable is anything that can change. In an experiment, the independent variable is what the researcher changes, and the dependent variable is what the researcher measures. An extraneous variable is an unwanted factor that could affect results, while a confounding variable changes systematically with the independent variable and makes the result hard to interpret.

In a correlation, there is no independent or dependent variable because nothing is manipulated. Instead, psychologists measure two co-variables.

Experimental designs

An experimental design is how participants are arranged across conditions.

• Independent measures design: different participants take part in each condition. This avoids order effects, but participant differences may affect results.
• Repeated measures design: the same participants take part in every condition. This controls participant differences, but order effects such as practice or tiredness may occur.
• Matched pairs design: participants are paired on relevant characteristics, then each person in the pair does a different condition. This reduces participant differences, but matching people is time-consuming.

Counterbalancing means changing the order of conditions for different participants to reduce order effects.

Populations and sampling

A target population is the group the researcher wants to make conclusions about. A sample is the smaller group of people who actually take part.

Common sampling methods include:

• Random sampling: everyone in the target population has an equal chance of being chosen. It can reduce researcher bias but needs a full list of the population.
• Opportunity sampling: using people who are available. It is quick but may be unrepresentative.
• Volunteer sampling: participants choose to take part, often after an advert. It is practical but may attract a biased sample.
• Stratified sampling: the sample contains the same proportions of key groups as the target population. It can be representative but takes planning.

Ethical guidelines

Psychologists should follow the British Psychological Society Code of Ethics and Conduct. Key principles include informed consent, avoiding unnecessary deception, protection from harm including psychological harm, the right to withdraw, confidentiality and debriefing after the study.

Doing Research

Experiments

An experiment tests cause and effect by manipulating an independent variable and measuring a dependent variable while controlling other variables.

• Laboratory experiments have high control, so they can be reliable, but may lack ecological validity.
• Field experiments happen in a real-world setting, so behaviour may be more natural, but control is lower.
• Natural experiments use naturally occurring differences, so they can study real situations, but the researcher cannot fully control the independent variable.
• Quasi-experiments compare existing groups, such as age or biological sex, so random allocation is not possible.

Interviews and questionnaires

An interview asks participants questions face to face, by phone or online. Structured interviews use fixed questions; unstructured interviews are more flexible. Interviews can produce rich detail, but interviewer bias and social desirability bias may affect answers.

A questionnaire is a written set of questions. Closed questions give fixed options and produce quantitative data. Open questions allow fuller written answers and produce qualitative data. Questionnaires are easy to standardise, but people may misunderstand questions or give socially desirable answers.

Observations

An observation involves watching and recording behaviour. Observations may be naturalistic or controlled, overt or covert, participant or non-participant.

Researchers often use behavioural categories, which are clear labels for behaviours being counted. They may use event sampling, where every example of a behaviour is recorded, or time sampling, where behaviour is recorded at set intervals.

Case studies and correlations

A case study is an in-depth investigation of one person, group or event. It can provide rich, detailed evidence, especially for rare cases, but findings may not generalise to everyone.

A correlation measures the relationship between two co-variables. A positive correlation means both variables increase together. A negative correlation means one increases as the other decreases. A zero correlation means there is no clear relationship.

Analysing Research

Types of data

Quantitative data are easier to summarise in graphs and descriptive statistics, but may lack depth. Qualitative data can give insight into thoughts and experiences, but can be harder to analyse objectively. Primary data are tailored to the study, while secondary data are quicker to access but may not perfectly fit the research aim.

Descriptive data and GCSE maths

Descriptive statistics summarise a set of results. GCSE Psychology uses descriptive maths only; you do not need inferential tests, p-values or critical-value tables.

The main averages are:

• Mean: add all scores and divide by the number of scores. It uses all values but is affected by extreme scores.
• Median: the middle score when data are ordered. It is useful when there are extreme scores.
• Mode: the most common score. In grouped data, the modal class is the group interval with the highest frequency.
• Range: highest score minus lowest score. It shows spread, but is affected by extremes.

You should also be comfortable with fractions, decimals, percentages, ratios, estimation, standard form, significant figures and decimal places. Standard form writes very large or small numbers as $a\times 10^n$, where $1\le a<10$.

A normal distribution is a symmetrical bell-shaped pattern. Most scores cluster around the average, with fewer very low and very high scores. In a normal distribution, the mean, median and mode are in the same central place.

Tables, charts and graphs

A frequency table shows how often each score or category occurs. A tally table uses tally marks to count frequencies before totals are written.

Choose the graph that matches the data:

• Bar chart: separate categories.
• Histogram: continuous grouped data, with bars touching.
• Pie chart: proportions of a whole.
• Line graph: change over time.
• Scatter diagram: relationship between two co-variables.

Reliability, validity and bias

Reliability means consistency. A reliable method should give similar results if repeated in the same conditions. Observations can improve reliability by using clear behavioural categories and checking inter-observer reliability, where two observers compare records.

Validity means accuracy: the study measures what it claims to measure. A highly controlled lab study may have strong internal validity, but lower ecological validity if the task feels artificial. A representative sample improves population validity.

Common sources of bias include demand characteristics, investigator effects, observer bias, social desirability bias, sampling bias, volunteer bias, response bias and order effects. You can reduce bias using standardised instructions, random allocation, counterbalancing, anonymity, clear behavioural categories, pilot studies and independent observers.