Agency

Any individual, organisation, department or division, including those belonging to the same organisation as the client, responsible for or acting as a supplier on all or part of a research project.

CAPI
Computer Assisted Personal Interviewing is conducted face-to-face, usually employing laptop computers. The interviewer is prompted with the question by the computer and the appropriate response codes are keyed in directly according to the respondent's answers. Routing procedures use these codes to determine which question appears next. Since the data is entered directly into the computer, analyses can be produced quickly.

CATI
Computer Assisted Telephone Interviewing, unlike CAPI, is conducted over the telephone rather than face-to-face.

CAWI
Computer Assisted Web Interviewing, unlike CAPI, is conducted over the Internet rather than face-to-face.

Client
Any individual, organisation, department or division, including those belonging to the same organisation as the research agency, responsible for commissioning or who agree to subscribe to a market research project.

Coding
The process of allocating codes to responses collected during fieldwork facilitating analysis of data.

Continuous Research
A survey conducted on a regular and frequent basis among parallel samples within the same population or a survey in which the interviews are spread over a long period of time.

Depth Interviews
This phrase is used to describe a variety of data collection techniques, but mainly for qualitative research (qv) undertaken with individual respondents rather than groups.

Desk Research
The collation of existing research results and data from published secondary sources for a specific, often unrelated, project.

Fieldwork
The live collection of primary data from external sources by means of surveys, observation and experiment.

Group Discussions/Focus groups
A number of respondents gathered together to generate ideas through the discussion of, and reaction to, specific stimuli. Under the steerage of a moderator, focus groups are often used in exploratory work or when the subject matter involves social activities, habits and status.

Hall Tests
A group of respondents are recruited to attend a fixed location, often a large room or hall, where they respond - usually as individuals - to a set of stimuli.

Interview
Contact with a respondent, or group of respondents, in order to obtain information for a research project.

Modelling/Simulation
The application of specific assumptions to a set of variable factors and the relationships which exist between them. Used to experiment with "what if" scenarios, models may be mathematical, graphical or purely verbal.

Multivariate Analysis
A range of analysis techniques which can examine quantitative data in more depth than can usually be obtained from a basic cross-analysis of the data by, for example, age, sex and social grade. The essence of this range of approaches is that the information is analysed in a way that permits patterns to emerge from within the data itself - ie based on the responses of the informants - rather than being imposed in advance, perhaps incorrectly or simplistically, by the researcher.

Mystery Shopping
The collection of information from retail outlets, showrooms etc, by people posing as ordinary members of the public.

Observation
A non-verbal means of obtaining primary data as an alternative or complement to questioning.

Omnibus Surveys
A survey covering a number of topics, usually for different clients. The samples tend to be nationally representative and composed of types of people for which there is a general demand. Clients are charged by the market research agency on the basis of the questionnaire space or the number of questions required.

Panels
A permanent representative sample maintained by a market research agency from which information is obtained on more than one occasion either for continuous research or for ad hoc projects.

Postal Research
The collection of primary data using a self-completion questionnaire or diary distributed or returned by post.

Qualitative Research
A body of research techniques which seeks insights through loosely structured, mainly verbal data rather than measurements. Analysis is interpretative, subjective, impressionistic and diagnostic.

Quantitative Research
Research which seeks to make measurements as distinct from qualitative research.

Respondent
An individual or organisation from whom information is sought, directly or indirectly, which could, in whole or in part, form the results of a research project.

Sample
A part or subset of a population taken to be representative of the population as a whole for the investigative purposes of research.

Semiotics
A form of social description and analysis which, used in research, puts particular emphasis on an understanding and exploration of the cultural context in which the work is taking place. Advertising and other images (including overt and implied symbolism), language, societal assumptions, media content and style, packaging design, etc, are evaluated since they provide the cultural framework within which, for example, purchasing patterns develop and can be influenced.

Social Grade
The socio-economic classification system used by the National Readership Survey (NRS) and generally for market research in the UK consists of six social grades: A, B, C1, C2, D and E. The publication Occupation Groupings: A Job Dictionary can be used to determine the social grade of respondents.

Survey
The systematic collection, analysis and interpretation of information about some aspect of study. In market research the term is applied particularly to the collection of information by means of sampling and interviews with the selected individuals.

Tabulation
Putting data collected during research into tables. Cross-tabulation involves a two dimensional table, based on answers to two of the questions included in a survey.

Viewing Facility
Premises used for conducting market research, particularly group discussions. Rooms have observation suites where the client can observe the proceedings undetected by the respondents either by video link or through a one-way mirror.