Market Intelligence Group - know thy customer


A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z







A classification of residential neighborhoods; a marketing segmentation system that enables consumers to be classified according to the type of area they live - usually just by using the consumer's postcode.

Ad Hoc


In basic terms this means 'as and when' required. Generally applied to single surveys which are designed as a 'one-off' rather than continuous on-going research.

Ad Testing

Closely, related to promotion testing, ad testing (a.k.a. advertisement testing) refers to various methodologies that focus exclusively on gauging the, perception, effectiveness, or targeting of advertisements in a market be they single adverts or a series. The major difference between ad and promo testing is that promo testing is usually carried out on a larger scale to measure the reaction to a campaign.

Ad testing can deal with adverts of all types across the spectrum and can be employed at any stage throughout the advertisement development process.
The most frequent use for ad testing is to identify the most effective advert(s) at the prototype stage which helps in turn to eliminate the likelihood of expensive and ineffective campaigns. Furthermore, it tests the appropriateness of an advert to its audience.

Some use the method to measure an adverts competitiveness against a major competitor’s. In this case researchers ask respondents to read magazines or papers or ask them to watch a television program with an advertisement break and then ask the respondents to recall the adverts that caught their attention or ask them for key information that the adverts contained. A deeper way of using this method is by conducting face-to-face interviews or using focus groups to gain consumer feedback to potential adverts or advertising campaigns.
Researchers aim to analyze the ability of the respondents to recall information, the affect that the advert has on a respondent, its persuasive power, whether respondents identify with the given setting in the advert, understanding of the appeal, and the respondents’ perception of the brand in question.


An agency can be any individual, organization, department or division (which includes those belonging to the same organization as the client) responsible for or acting on all or part of a market research project.


This is the processing of data collected (which can be qualitative or quantitative data) within a market research project, which allows us to draw summary and conclusions in relation to the project



Balanced Scale

Used mainly in questionnaire formats, a balanced scale is one on which the number of positive and negative categories are equal, leaving the respondent with a fair decision. 

Bar Chart

A chart which displays data in the form of bars either vertically or horizontally with the length of the bars relating directly to their value.

Bivariate Analysis

This is the term for when two variables are being measured at one time.


A brand is a specific name, symbol or design which is used to distinguish a particular seller's product from any other.

Business to Business Research (B2B)

B2B research is research that relates to businesses rather than consumers. Whereas in the case of consumer research, respondents could be asked of their opinions upon a certain product or service, B2B is concerned solely with business related issues.




CAPI stands for computer assisted personal interviewing. It is performed with the respondent face-to-face with the interviewer, using lap-top computers or other such devices (e.g. IPAQ's). The interviewer uses the computer to prompt the questions and the interviewee's responses are immediately inputted back into the computer which enables the program to select the next appropriate question. This enables the interview to pursue many different, yet relevant, routes depending upon the responses given by the interviewee. This form of surveying is particularly effective because due to the fact that the data is already on the computer, the analysis becomes quicker than other methods.


CATI stands for computer assisted telephone interviewing. It is similar to CAPI but conducted over the phone, rather than face to face.




CAWI stands for computer assisted web interviewing. This is based on the same principle as the two above, apart from that it is conducted over the internet.


This is a study which incorporates every available element within a defined population.


Used in the analysis of data, CHAID stands for chi-square automatic interaction detection. It identifies market segments that differ in response, screens out extraneous predictor variables and collapses the number of categories of explanatory variables.

Chi-Square Test

Used for making a comparison in nominal variants between hypothsised samples and a population samples. The method is used in both univariate or bivariate analysis. 


A client is anybody, be it an individual, organization, department or division, who is responsible for the commissioning of, or who agree to subscribe to a market research project.




Observations or measurements made that apply to an area of the the marketing system.

Data Editing

During the process of data editing, the data collection instruments are tested to ensure that the  maximum accuracy in terms of results has been attained. Checks are made on legibility, consistency and completeness and above all the process aims to avoid ambiguity. 

Data Processing

Data processing is the inputting of data into a database format. Once in this format, data can be tabulated and conclusions drawn as to the results of the market research findings.

Data Sources

There are four main sources of marketing data. These are respondents,  analogous situations, experimentation and secondary data. For more information on any of these sources, see glossary definitions.

Depth Research

A term which is used to describe a multitude of data-collection techniques and procedures although it is usually used in qualitative research with individuals.

Desk Research


This is a research technique which involves the market researcher collating and drawing together secondary sources of information. These secondary sources can serve as complimentary to primary sources or can be collated to form part of a larger unrelated project.

Diary Methodology

Diary methodology is a procedure which is used to compile consumer purchase data or media habits. During this procedure, respondents will be required to complete a written report which details their behavior over a certain period.





Editing is the procedure that takes place in order to ensure that a questionnaire has been correctly completed.


Elasticity  is used to measure the exact amount of volume shift responding to the variable under scrutiny.

Element Sampling

During this procedure, every unit in a specific population has an equal probability of being chosen for research.

Enabling Technique

This is a technique that is used within a qualitative interview that allows the respondent to articulate their views.


An estimate is an assumption based upon a smaller sample that is then applied as a hypothesis for a whole population.

Experimental design

This is a test whereby the researcher performing it has control over one, or a few independent variables and manipulates them to retrieve the most relevant data.

Exploratory Research

Used in the early stages of the decision-making process, exploratory research is used to assess the situation in hand with the minimum cost and time possible. The process must be flexible due to the unknown quantities that may be encountered. Versatility and a wide-ranged approach to the preliminary investigation are key. The exploratory research can draw on interviews, observations, group interviews, secondary data sources and case histories.  



F Test

An F Test explores the probability that a specific result may have been down to chance and therefore inconclusive.

Face Validity

This is a simple test which ensures that a measurement or research actually targets the areas that it should.

Factor Analysis

Factor analysis of data identifies the underlying factors that explain the correlation among a set of variables. It is used to identify a new, smaller use of uncorrelated variables to replace the original set of correlated variables. It also identifies a smaller set of salient variables for use in subsequent research. It is used in various types of research including product research, customer satisfaction, market segmentation and consumer profiling.

False Accuracy

False accuracy occurs when a collection of data gives the impression of being accurate when in reality, only a low degree of accuracy exists. This problem can arise when results are being rounded to fit a pattern. e.g. rounded to decimal places etc.


Fieldwork can be conducted by observation, surveys (such as face to face interviews, telephone interviews and web interviews) or experiments. It is the basic term for the live collection of primary data from external sources. Fieldwork is either co-ordinated by an in-house fieldwork department within a market research agency or an external fieldwork company. Once fieldwork has been conducted data processing is usually the next step.

Focus Groups (aka Group Discussions)

In a focus group, respondents (normally between 8-10 people) are gathered together in order to gauge their responses to specific stimuli. Groups are guided by a research moderator who often uses a topic guide to control the discussion to ensure it meets the initial research objectives. The data generated is probably most applicable to exploratory work. The technique falls under the broad category of qualitative research.

Freelance Market Researcher

These self employed market researchers usually work alone. They would typically have many years experience in the field of work and are contracted by market research agencies and consultancies in busy periods, effectively as an extra resource. In many cases they usually help out with qualitative research techniques (for example conducting depth interviews or group discussions) but they can also be utilized for quantitative research projects.


Frequency measures the number of times that an incident or event occurs. Used in data-analysis it details the number of results that fall in the various categories.




Geographics refers to any of the various methods used to sub-divide a list based upon geographic or political boundaries.

Grid Test

This is a method where in which to variables can be tested for data at any one time.

Group Discussions (aka Focus Groups)

In a focus group, respondents (normally between 8-10 people) are gathered together in order to gauge their responses to specific stimuli. Groups are guided by a research moderator who often uses a topic guide to control the discussion to ensure it meets the initial research objectives. The data generated is probably most applicable to exploratory work. The technique falls under the broad category of qualitative research.

Group Dynamics

Group dynamics refer to the ways in which people interact and relate to one-another in a group situation. A successful moderator will be able to manage these dynamics to produce a relevant and informative discussion by employing a variety of different techniques which obviously involves leading the conversation away from trivial subjects when needed.

Growth Rate

Usually measured as a percentage, growth rate records the changing size of a population by plotting the change from a specific point in time. It is calculated by dividing the total increase/decrease in population during a set period by the average population during that period.



Hall Test

During this procedure, a number of respondents are invited to attend a fixed location (traditionally a hall from which the term originates) and then are asked to respond to stimuli, usually individually.

Hedonic Scale

This scale is used to measure the generalized overall views and opinions of a product or service.


Heteroscendasticity is the term used to express a non-consistent variance during regression analysis.

Homogenous Groups

Groups in which the people or objects are extremely similar if not identical in their characteristics or behavior.


The opposite of heteroscendasticity. It is used to describe a state of constant variance in regression analysis.


The generic term used to describe every person living in one unit of housing.



In-Depth Interview

The in-depth interview is a method used primarily in qualitative research and is carried out upon individuals (one-on-one), rather than in a group discussion scenario. The interview takes part with a researcher and a respondent (the interviewee). The researcher normally uses a topic guide to guide the discussion and it usually lasts between 30 minutes to an hour. The interview helps the researcher to gain detailed and in depth information.


This is the payment made to a respondent in return for their participation in a research project and depending on the complexity and duration of the research, the incentive will vary.

Internet Research

The use of internet research has grown massively over the last few years due to the increasing popularity of the internet for both business and leisure purposes. The research itself can be qualitative and quantitative. Quantitative internet research can be performed as a web-survey where the respondents reply to questionnaire-based emails. This means that with the click of a mouse, the results are back in the inbox of the researcher ready for analysis. Qualitative research can be carried out for example by setting up group forums on the internet, or by setting up group discussions using web-cams, thus mimicking the conventional group-discussion format.


An interview is the general term for the method used to draw information from a respondent. It can take several guises e.g. Face-to-face, in depth, telephone interview, group discussion etc.



Judgement Sample

The term used for a a sample which consists of respondents who have been selected with the understanding that their opinions, behavior, characteristics and so on, will be representative of the population as a whole.



Key Verifying

This is the process by which two people input the same data to avoid the distortion of data at the data input stage.



Laboratory Experiments

This term relates specifically to experiments that are carried out in a controlled environment.

Likert scale

The Likert Scale allows the respondent to specify an opinion on statements which relate to the subject being researched.

Longitudinal Research

Longitudinal research is used in most instances where continuous performance monitoring is taking place which consists of repeatedly analyzing a fixed sample of population elements.  



Market Research Brief

Also known as a research brief, this is a basic plan which guides the data collection and analysis phases of the research project. It acts as a framework which details the type of information to be collected, the data sources and the data collection procedure.


A simple average value of a set of data.


Another form of averaging data, the median value is the middle value of a set of data when the data is arranged in terms of magnitude.


(e.g. 1234567 - Four would be the median.)


An average which is based on identifying the single most frequent value in a collection of data.


(e.g. 2 ,3 ,2 ,1 ,6 ,5 , 2 , 3 - Two would be the mode)

Modelling / Simulation

The application of specific assumptions to a number of variable factors and the relationships that exist between those factors. Usually used in hypothetical situations, the models can be verbal, mathematical or graphical.

Monadic Rating

This is where a respondent is asked to rate an item on some form of scale without comparison to another product or service.

Multiple-Choice Question

A question which the respondent may only answer within a predetermined selection of responses.



Nominal Scale

A scale of measurement in which numbers don't represent value, instead their use is to label or categorize an object or an event.

Non-Balanced Scale

A non-balanced scale is one which is unfairly weighted towards one end of the response spectrum and can often cause response bias.

Non-Probability Sample

This is a sample which has been chosen with disregard in terms of representing the wider population.

Non-Sampling Error

A non-sampling error refers to any cause of bias or inaccuracy with the results that has been caused by anything other than the sampling.




Observation is the collection of data through non-verbal means which can stand alone as research on its own or compliment other relevant research. It is the process of recognizing and recording relevant objects and happenings.

Observational Methods

These allow the recording of behavior when it takes place  therefore allowing for the elimination of such errors associated with the recall of behavior. It allows for the data to be more precise with less margin for error.

Omnibus Surveying

An omnibus survey covers a variety of topics, usually for different clients. The majority of the samples used are generally nationally representative and are made up of people who are in general demand. The charges for this element of research are based upon the size of the questionnaire produced in either space or questions required. It is a form of cost effective quantitative research - clients can share much of the cost of the survey. Questions normally cost around £1,000 each.

Open-Ended Question

A question to which the answer has no definite response.




A panel is a long-standing sample that is retained by a market research agency from which data can be attained. It is most useful for continuous research whereby the same set of respondents are used on a continuous basis over time.

Performance-Monitoring Research

Once a marketing plan has been implemented, performance monitoring research simply monitors the progress of the plan. It ensures that the progress of the program is as planned as a deviation from the initial projected route can lead to the plan being carried out improperly or the results pattern inexplicably changing.

Pie Chart

A circle which is divided into segments which vary in size relating to the value which they represent.

Postal Research

Postal research is carried out by sending questionnaires or journals, which are to be completed and returned, through the post. It is perhaps the most traditional form of data collection. The main drawback with postal research is response rates - it is difficult to achieve a high response rate.

Projective Technique

These are used by moderators or research interviewers. They are used in interview situations when the interviewer will use the projective technique to enable the respondent to distance themselves from their personal views. It enables the researcher to obtain 'the real views' that perhaps wouldn't be revealed without using the interview technique.



Qualitative Research

Qualitative research is in-depth research performed on a small scale to provide detailed in depth results and data. Qualitative research can be performed over the phone, via group discussions or through one-to-one interviews. Discussions are normally aided by using a 'topic guide' - this outlines the basic structure of the interview/group and indicates the general direction in which the interview/group should be led. Questions are of open nature as opposed to closed questions which are used in quantitative research.

Quantitative Research

Quantitative research is performed on a far larger scale compared with qualitative research (in terms of the sample size) and helps to provide accurate statistical data from which conclusions can be drawn. Questions tend to be closed as opposed to open.


This contains a group of questions and is used as an investigative market research tool in order to gain information from a respondent. It is used mainly in quantitative research. Questionnaires normally contain closed questions but open questions can also be included.

Quota Sample

A quota sample is one which has been selected to represent the population, sometimes incorporating certain control characteristics.



Research Design

Also known as a market research briefing, this is a basic plan which guides the data collection and analysis phases of the research project. It acts as a framework which details the type of information to be collected, the data sources and the data collection procedure.

Research Report

The research report is the compilation of findings from a piece of research. These findings are normally presented in the form of a report or a PowerPoint document.


A respondent is the person whose views and opinions are required by the researcher.

Response Bias

Response bias is the term for inaccurate response data which could have been caused by a number of factors such as monotony, tiredness, aiming to please etc.




In research terms, a sample refers to a group of interviewees or respondents who are chosen to represent the population as a whole. The sample provides the data within the market research project.

Secondary Research

Secondary research utilizes data sources that already exist. The technique used to collect the information is called desk research (for a definition please see desk research). Using the web or visiting libraries would be two forms of desk research.   


Used in research, semiotics is a type of social description and analysis which places specific emphasis upon an understanding and exploration of the cultural context within which the particular work is taking place.

Social Grade/Socio-Economic Grouping

Socio-economic groupings are a way of separating various sections of society. There are six social grades in the UK socio-economic grouping system. These are:

  • A: Upper middle class. Higher level managers, administrators and  professionals.

  • B: Middle class. Intermediate manager's administrators and professionals.

  • C1: Lower middle class. Supervisory or clerical workers and junior manager's administrators and professionals.

  • C2: Skilled working class. Skilled manual workers.

  • D: Working class. Semi and unskilled manual workers.

  • E: Those at the lowest level of subsistence. State pensioners or widows (no other earnings), casual or lowest grade workers.

Statified Sampling

A sampling procedure which is carried out in two stages. The population is divided into two separate groups, mutually exclusive and collectively exhaustive, and a random sample is taken from each of the two.


The systematic gathering, analysis and elucidation of information about any aspect of study. In reference to market research it specifically applies to the gathering of information through processes such as sampling or interviews with selected candidates, respondents or interviewees.




When a sample is too small to use a Z -test to analyze the results, a T-test is used to analyze a single mean value.


Tabulation is simply placing the results and data obtained from research into data tables. (See cross tabulation)

Target Population

The basic term for the population that is being studied.

Test Market

A test market is used as a trial market for a new product or service.

Time series analysis

During this form of analysis, data is recorded at set intervals in time.

Topic Guide

A topic guide is a brief guide which is used by a researcher in qualitative research interviews or groups discussions. It enables the interview to retain relevance and ensures that the project meets the initial research objectives.


Tracking refers to studies that monitor consumer behavior towards a product, brand or service over a period of time and is carried out over a longer timescale than ad-hoc research.

Trimmed Mean

The trimmed mean is obtained by taking a percentage of the high and the low levels of data and then finding the mean.

Two Way Focus Groups

This is where one focus group observes another and then discusses what they have taken from the observation.



Unbiased Estimator

An unbiased estimator is when the mean of the sample distribution and the estimated distribution is equal.

Unbiased Samples

These are samples that derive from an unbiased source thus ensuring that any sampling error is a result of the randomness and nothing else.

Unidimensional scaling

These are procedures that  are designed purely to measure only one single attribute of an object or respondent.


This is an ordinal scale which has a positive end and a negative end.

Unstructured observation

Unstructured observation is where a study takes place with an observer or moderator simply watching and taking notes on the behavior on display.

Unstructured Segmentation

This is the process of market segmentation which uses data and analysis when no previous ideas or views are held regarding the segmentation of that particular market in any way.




A post-research essential; Validation in the process by which a respondent verifies that the interview was conducted correctly.


Validity in reference to market research is concerned with whether the purpose of the research was fulfilled accurately.



A verbatim is a record of actual spoken comments by a respondent. These are used to back up any research findings in the final report.

Video Focus Groups

These are focus groups that instead of meeting face-to-face, carry out the group via a video conferencing link.

Viewing Facility

Viewing facilities are the premises used for conducting group discussions or depths interviews in qualitative research. These premises can contain rooms where when in, the respondent can be observed by the client using one-way mirrors or via a video link. It is thought that because the client is not physically present in the discussion, the respondents may be less inhibited and more honest and open about their opinions and feelings.



Word Association



A technique in which the respondent is confronted with a word and is asked to respond with the word that immediately comes to mind.




A test which draws comparisons between a sample test and a hypothesized test. It is used when working with interval data and any data on a large scale.

Free resources

Article- What is the importance of Data Analysis?

Article- Why Employee Satisfaction Survey?

Download e-books