A - B

Click a term to expand its definition, or expand all definitions collapse all definitions

• aggregate statistics

  Information, written as numbers, about groups, not individuals; the term can also mean the study of this information.

• analysis of variance

  A statistical test showing the effects of an independent variable on a dependent variable. It is a technique used to determine whether there are statistically significant differences of means between two or more groups.
  

• anecdotal evidence

  What people say about something; not proven by "hard" (experimental) research.

   

• anonymity

  The state of having your individual identity protected, especially from researchers. For example, someone would be anonymous who answered a mail survey that didn't ask for the person's name or any other ID.

• applied research

  Applied research is research done for an express purpose to solve an identified problem.

• assessment

  A test or other way of measuring something, such as a person’s mental health or goals or needs; often the first test in a series of tests, or a test given before treatment starts.
  

• association

  A measure of whether and how closely certain values (numbers, amounts) in a study go up or down at the same time.

• attrition

  The drop-out rate among people who are being studied. People may quit because they want to, or may not be able to stay in the study group (because of illness, lack of time, moving to another city, etc.), or they may not fit into the study anymore (if they get a job or marry, for example, in a study about single people who are not working).

   

• bar graph

  A drawing that uses bars for various groupings. The height of the bar shows how many things or people are in that grouping.

• baseline analysis

  An analysis of data collected before a treatment or intervention is applied.

• benchmark

  A standard, test, or point of reference (often a number).

• bias

  Bias is a loss of balance and accuracy in the use of research methods. It can creep into research via the sampling frame, random sampling, or non-response. It can also occur at other stages in research, such as while interviewing, in the design of questions, or in the way data are analysed and presented. Bias means that the research findings will not be representative of, or generalizable to, a wider population.
  

• bimodal distribution

  A range of scores that has two most frequent scores instead of one.

• biographical research

  Biographical research (or narrative approach to research) is primarily qualitative, and includes gathering/ using data in the form of diaries, stories and life histories.
  

• bivariate analysis

  The study of two things (amounts, values, variables) and how they are connected.

• bottoms-up pooling

  A statistical method that combines data from different groups by first comparing data across individual groups and then across the whole set.