ABBA design: a research design in which one set of participants is given the first condition (A) and then the second (B), while the other set is given condition B first, followed by A.
action research: a method of undertaking social research which acknowledges that the researcher's presence is likely to influence people's behaviour, and so incorporates the researcher's involvement as a direct and deliberate part of the research, with the researcher consciously acting as change agent.
action research cycle: the circular process of action research, which proceeds from a diagnostic stage to the development of a change strategy, to the action and implementation stage, then to the evaluative stage, and back through the cycle again.
ambiguous questions: questions which have more than one possible meaning.
analysis of variance: a statistical technique which allows researchers to locate significant differences between several different samples.
anecdotal evidence: evidence which is purely based on hearsay and personal experience rather than systematic investigation.
ANOVA: the usual abbreviation for analysis of variance.
ANOVA interaction diagram: a type of diagram which illustrates the relationships between variables identified by an ANOVA.
anti-positivism: an approach to research which emphasises the human interpretation of meanings and implications; introduced as a challenge to hard-line positivism.
artefacts: distortions in data or research findings brought about by artificial situations or uncontrolled variables.
attribution: the process of giving reasons for why things happen.
average: a typical score from a set of data. In everyday use, this term generally refers to the mean.
axis: a vertical or horizontal line used to show distances or amounts on a chart or graph.
bar chart: a graphical representation of a set of data which uses columns or bars to indicate different amounts of each category.
behavioural coding: a method of recording observations which involves using different categories. Each unit of behaviour is observed by checking one of those categories.