Market Research Glosary

A Posteriori

is an approach where a theoretical framework is developed from the research (after it has been conducted).

A Priori

is an approach where a theoretical framework is developed before the research is conducted.

Accompanied Shopping

is a form of observation study where an interviewer accompanies a respondent (with his or her agreement) as they go shopping.

Achieved Communality

a term used in factor analysis that represents the proportion of variance in an original variable accounted for by all the extracted factors. Each original variable will have an achieved communality value in the factor analysis output.

Acquiescence Bias (aka “yea saying” or “friendliness effect”)

A systematic bias caused by some respondents tending to agree with whatever is presented to them. Such a bias may be caused by either respondents or interviewers being overly friendly during interviews.

Ad Hoc Research

is research that is specifically designed to address a particular problem or issue. Ad hoc research is conducted when there is insufficient existing research within the marketing information system to answer all the questions.

Affective Component

is one of the three components of attitude that is concerned with individuals’ emotions or feelings towards an object or idea.

Aggregate

is a summary measure made by compounding two or more separate measures, eg national income and price index numbers.

Aided Awareness

see prompted awareness.

Alternative Hypothesis

is a statement that suggests a difference or an effect is present (ie there is an alternative). The alternative hypothesis is adopted when the null hypothesis has been disproved.

Analysis of Covariance (ANCOVA)

An analysis of variance procedure in which the effects of one or more metric-scaled extraneous variables (covariates) are removed from the dependent variable data before one conducts ANOVA.

Analysis of Variance (ANOVA)

is a statistical technique for examining the differences among means for two or more populations.

Anchor Label

is label used to define anpppp of a measurement scale.

Anonymity

involves concealing respondents’ identities from interviewers and/or researchers.

Area Sampling

is a type of cluster sampling where geographical areas are the clusters.

Arithmetic Mean

see mean.

Association Matrix

see contingency table.

Association Technique

is a form of projective technique where participants are presented with some stimulus material and they are then asked to respond with the first thing that comes to their minds.

Atomistic Test

is a test that aims to assess participants’ reactions to individual elements of a product or concept (in contrast to a holistic test that looks at a product or concept as a whole).

Attitude

is an individual’s learned predisposition to behave in a consistent manner towards an object or idea. There are three components of attitude: (i) a cognitive component - knowledge and beliefs (ii) an affective component - feelings and emotions (iii) a conative component - behaviour (usually measured in terms of likelihood to buy).

Attitude Scaling

is the development of measurement criteria used to measure individuals’ attitudes.

Attribute

is a qualitative characteristic of an idea or object being evaluated, eg gender is a attribute but age is a variable.

Attribute Analysis

is a technique that is designed to develop lists of characteristics, uses or benefits relevant to a particular product category.

Audimeter

see people meter.

Audit

is a data collection process based on counting physical objects, such as records or inventory.

Average

is a general term that is used to represent or summarise the relevant features of a set of values. The arithmetic mean is often used as a measure of average, but the median and the mode can also be used to summarise a set of values.

Average Opportunities to See

see frequency.

Awareness

is a measure of respondents’ knowledge of an object or an idea. There are two main measures of awareness: spontaneous (or unaided) and prompted (or aided) awareness.

Baby Boomers

are those people who were born between 1946 and 1964.

Back Checking

see validation.

Back Translation

is a validation process where a survey is first translated into another language and then translated back into the original language by a different person. The objective is to ensure that the original translation is accurate.

Balanced Scale

is a scale with an equal number of favourable and unfavourable categories.

Banner Format

is a style of data tabulation where the responses from each sub-group are listed for each question in the survey.

Banner Point

is a column heading in banner format tabulations and corresponds to a particular sub-group of the sample. It is usual practice to have the independent variables as banner points.

Bar Chart

is a type of graph where the data is displayed in the form of bars that can be arranged vertically or horizontally.

Base

is the required number of interviews to be completed.

Base Line (aka Bench Mark)

is a measurement against which another measurement will be compared.

Baysian Analysis

is a mathematical procedure based on decision theory that aims to estimate the value of a market research project.

Bench Mark

see base line.

Benefit Segmentation

is the dividing of potential consumers into sub-groups according the benefits sought.

Best Light Phenomenon

is when respondents bias their answers in a market research project so that they can then appear in the best possible way to those who are looking at the responses.

Biased Question

is a question that is phrased in such a way it influences the respondent’s opinion and the resulting data.

Biased Sample

is a sample that does not contain units in the same proportion as the population of interest.

Bimodal Distribution

is a frequency distribution with two modes.

Binomial Test

is a statistical test of dichotomous data (where there are two possible outcomes) to check whether the research data is significantly different to what would be expected.

Bipolar Adjectives

are two opposing adjectives that define the opposite ends of a scale (such as a semantic differential scale).

Bivariate Analysis

is the analysis of a relationship between two variables.

Bivariate Correlation

see product moment correlation coefficient.

Bivariate Regression

is a procedure for deriving the equation that relates a single metric dependent variable and a single metric independent variable.

Blind Testing

is the testing of products with potential consumers where brand names, packaging and other identifying items have been removed.

Blocking Factor

is the relevant external variable that is used to group (or block) experimental units into groups so that the experimental group and the control group are matched.

Booster

refers to additional interviews involving a particular sub-group of a sample to ensure there are sufficient members of the sub-group in the resulting sample.

Branching Question

see skip pattern.

Brief

is a statement (usually in writing) of a business problem that could be alleviated by conducting some marketing research. Briefs are normally written by a client company for a research supplier and they usually have a background and an objectives section. A suggested methodology section can also be included.

Bubble Test

see thematic apperception test.

CAGR - Compound Annual Growth Rate

is the average annual percentage growth rate for a series of n observations. The formula for determining the CAGR % is as follows: ((last value/first value)^(n-1))-1

Call Disposition

is a tabulation of the outcome of calls made during a computer-aided telephone interview (CATI) survey.

Call-back

is a repeat telephone call to a potential respondent to see if they can participate in a survey.

CAPI

is computer-aided personal interviewing, where the responses in a personal interview are keyed directly into a computer and the administration of the interview is managed by a specifically designed programme. The programme checks for invalid responses and will not accept responses outside prescribed limits, hence subsequent editing and keying in of data is avoided.

Cartoon Test

see thematic apperception test.

Casewise Deletion

is when an entire questionnaire from a respondent is removed from the analysis because some of the questions have not been completed.

CASRO

is the Council of American Survey Research Organisations and it is a trade organisation for those who are actively involved or concerned with marketing and opinion research.

Categorical Scale

is a scale that asks respondents to choose from a limited number of alternatives. There are three main types of categorical scale; semantic differential, stapel and Likert.

Categorical Variable

is a variable that is based on non-metric data (ie data that cannot be analysed statistically such as that from a nominal or ordinal scale).

CATI

is computer-aided telephone interviewing where the responses are keyed directly into a computer and administration of the interview is managed by a specifically designed programme. The programme checks for invalid responses and will not accept responses outside prescribed limits, hence subsequent editing and keying in of data is avoided.

Causal Relationship

is where one can infer that two variables are related in some way. Three conditions have to be met before a causal relationship can be inferred; there has to be evidence of association (concomitant variation), the dependent variable has to change after the independent variable has changed (temporal ordering) and all other possible causes have to be eliminated.

Causal Research

is a type of conclusive research that aims to collect data on causal relationships.

Causal Variable

see independent variable.

Cell

refers to a sub-sample of respondents in a research study.

Census

is the collection of data from all available units in a population of interest.

Central Limit Theorem

is a theorem that states that the sampling distribution curve (for sample sizes of 30 and over) will be centred on the population parameter value and it will have all the properties of a normal distribution.

Central Location Interviewing

is when face-to-face interviews are conducted at one or more specified locations (eg mall intercept interviewing).

Centralised Research Function

is where the marketing researchers in an organisation are located together in one department.

Chat Room

is where respondents in separate locations all participate in a virtual group discussion by using their PCs. The discussion has a moderator and comments are communicated in writing on members’ screens.

Chi Square Distribution

is a skewed distribution whose shape depends on the number of degrees of freedom. As the number of degrees of freedom increases, the distribution becomes more symmetrical.

Chief Income Earner

is the person in the household with the largest income, whether from employment, pensions, state benefits, investments or any other source. Where there are two people with the same income, the researcher should specify who to include in the study.

Chief Shopper

is the member of a household who is responsible for the majority of household purchasing decisions.

Chi-square Test

is a non-parametric statistical test that compares research data with the expected results from a hypothesis.

Chronological Age

is individuals’ ages in years and months (which may be different to their cognitive age).

Clarifying

is the repeating or re-phrasing of an existing question to get a further explanation of an answer provided by a respondent.

Classification Information

involves socio-economic and/or demographic information on participants in a market research study.

Classified Scale

see nominal scale.

Client Facing Staff

are marketing researchers who are the link between research clients and the marketing research department or supplier.

Clients

are the ultimate consumers of marketing research, they create the demand for research services and pay for the research that is conducted. Clients can be internal or external. Internal clients are other users of research in the same organisation. External clients are based in other organisations.

Closed-ended Questions

provide respondents with a pre-determined list of possible answers.

Cluster Analysis

is an analytical technique that arranges research data into mutually exclusive and collectively exhaustive groups (or clusters) where the contents of each cluster are similar to each other, but different to the other clusters in the analysis.

Cluster Sampling

is a type of probability sampling where a population of interest is divided into mutually exclusive and collectively exhaustive sub-groups (or clusters) and a sample of clusters is selected. From the selected clusters, a sample of units is drawn.

Clutter Reels

are video tapes with a number of TV commercials including the one(s) being tested and they are used to assess a commercial’s ability to stand out from the rest.

Code and Tab Plan (aka Tab Plan)

is a plan that details the column and row headings in the cross-tabulations that are produced from the data.

Code Book (aka Coding Frame)

is a set of instructions regarding the allocation of codes to research data. It helps researchers identify and locate the variables to be used in data analysis.

Code of Conduct (or Ethics)

all professional marketing research societies have a code of conduct that details the rights and responsibilities of those involved with marketing and opinion research.

Coding

is the organising of responses into categories and the assignment of a unique code to each response prior to data entry.

Coding Frame

see code book.

Coefficient of Determination (R Squared)

is the exact percentage of variation shared by two variables, obtained by squaring the product moment correlation coefficient.

Coefficient of Variation

is a measure of variability (or dispersion) of a distribution and it is equal to the standard deviation expressed as a percentage of the mean.

Cognitive Age

is an individual’s perceived age, which may be different to their chronological age. For example pre-teens aspire to be older and 40 year olds may well aspire to be younger.

Cognitive Component

is one of the three components of attitude that is concerned with individuals’ knowledge and beliefs.

Cohort Analysis

is a type of multiple cross-sectional design where the population of interest is a cohort whose members have all experienced the same event in the same time period (eg birth). The samples are usually drawn at regular time intervals.

Collectively Exhaustive

A set of alternatives is collectively exhaustive when they include all possibilities.

Communality

in factor analysis is the amount of variance a variable shares with all the other variables being considered. It is also the proportion of variance explained by the common factors.

Comparability

is the extent to which research results can be meaningfully compared.

Comparative Scales

are a type of scale where one object is compared with another and a relative measure of preference is obtained, eg do respondents prefer one or another soft drink? The main types of comparative scales are: paired comparison, rank order, constant sum and Q sort.

Completion Rate

is the proportion of qualified respondents who complete the interview.

Completion Technique

is a form of projective technique where participants are asked to complete an incomplete situation.

Complex Questions

are questions containing words that are unfamiliar to respondents.

Composite Variable Index

is an index that combines a number of separate variables, eg education, income and occupation being used to form one overall measure of social class.

Conative Component

is one of the three components of attitude that is concerned with individuals’ tendency to behave in a particular way towards an object or idea. It is usually measured in terms of likelihood to buy.

Concept

is a description of a proposed product or service consisting of attributes and benefits.

Concept Testing

is the investigation of potential consumers’ reactions to a proposed product or service.

Conclusions

are a summary of the research findings.

Conclusive Research

involves the use of highly structured techniques (such as questionnaires with closed questions) with statistically representative samples in order to prove or disprove hypotheses.

Concomitant Variation

is when two variables occur or vary together. It is one of the conditions that has to be met in order to infer a causal relationship.

Concurrent Validity

is the degree to which two different measuring systems produce correlating results. It is often used to determine the validity of new measuring techniques, by comparing them with established techniques.

Conditional Probability

is the probability of a research outcome occurring if a state or condition was to exist.

Confidence Coefficient

see confidence level.

Confidence Interval

is a range of values centred on the sample estimate that is known to contain the true value with a given degree of confidence.

Confidence Level (aka Confidence Coefficient)

is a percentage (usually 95%) that reflects the degree of certainty that the true value lies within the confidence interval. It is the minimum probability of not rejecting a true null hypothesis (committing a Type I error) and is equal to one minus the significance level.

Confidence Limits

are the two values from each end of a confidence interval.

Confounding Variables

see external variables.

Conjoint Analysis (aka Trade Off analysis)

is a statistical technique that measures the relative importance individuals attach to particular product attributes.

Consistency Check

involves identifying completed questionnaires with data that are out of the permissible range, logically inconsistent or have extreme values. Data that is out of the range of the coding scheme is inadmissible.

Constant Sum Scaling

is a type of comparative scale where respondents are asked to allocate a fixed amount (or constant sum) of points, dollars or anything among a set of objects according to a criterion.

Consumer Confusion Study

is a study that aims to measure the tendency of consumers to confuse the company that makes a particular brand with another company, or to confuse one brand with another.

Consumer Panel (aka Panel)

is a group of selected research participants who have agreed to provide information at regular specified intervals over an extended period of time.

Contingency

is the difference between an actual frequency and an expected frequency in a table.

Contingency Table (aka Association Matrix)

is a cross-tabulation table that contains a cell for every combination of categories of the two variables.

Continuous Data

is that from a measurement scale where it is permissible to calculate intermediate values.

Continuous Rating Scale (aka graphic rating scale)

is a type of non-comparative scale that offers respondents a form of continuum (such as a line) on which to provide a rating of an object according to a criterion.

Continuous Research

is any research that involves the regular, on-going collection of data, eg consumer panels and epos data.

Contrived Observation

is the observation of behaviour in an artificial setting.

Control

is an experimental sub-group where the independent variable is not manipulated. It is used as a benchmark for comparison with the experimental group.

Convenience Sample

is a type of non-probability sample where the units have been selected because they are convenient for the researcher (which may not be the optimum sample for the research project).

Cookie File

is a computer file that is secretly added to the hard disk of someone who visits a web site that sends them. Their purpose is to track web site visitors, however the use of cookies in marketing research is not approved by ESOMAR.

Correlation

is the existence of a relationship between two variables (which may or may not be a causal relationship - correlation on its own does not infer causality).

Correlation Coefficient

see product moment correlation coefficient.

Correspondence Analysis

is a perceptual mapping technique that is based on data where respondents are asked to identify only the attributes that relate to (or correspond with) the subject of the study.

Counter-biasing

is a technique that is used to increase respondents’ willingness to answer questions honestly. Questions are prefaced with statements that attempt to justify an answer that may go against social group norms, eg “recent surveys have indicated that the majority of people have difficulty saving every month”.

Covariance

is the extent of a relationship between two variables, whereby a change in one variable implies a change in the other.

Covariate

is a metric independent variable (ie based on data that can be analysed such as that from an interval or a ratio scale).

Coverage

is the proportion (usually expressed as a percentage) of a population of interest that has been exposed to a particular advertisement.

Creative Development Research

is a type of qualitative research that is used to formulate advertising. It can work at three stages in the advertising development process (i) Defining the strategy - ie what should the advertising be saying? (ii) Defining the execution - ie how should it be said? (iii) Testing a chosen execution.

Cross-cultural Analysis

is the collection and analysis of data from different countries (or cultural units) that compares the findings from different countries.

Cross-sectional Design

is a research design that involves the collection of data from a sample only once.

Cross-tabulation

is the simultaneous tabulation of two or more variables.

Cross-validation

is a test of validity for a regression model that involves using comparable data to check the validity of an original estimation.

Culture

is the total sum of learned beliefs, values and customs that serve to regulate the consumer behaviour of members of a particular society.

 Data

are research facts that are based on respondents’ answers to questions.

Data Cleaning

see editing.

Data Collection Instrument

is any device that is used to gather information from respondents, eg questionnaires, video recorders, tape recorders.

Data Deck

is a complete set of responses in a given study.

Data Entry (aka Transcribing)

is the act of inputting data into a database.

Data Processing

is the transformation of data into information.

Data Reduction Plan (aka Edit Instructions)

is a set of instructions for editing and coding the questionnaire and specifies how the data should be evaluated for inconsistencies, skipped questions etc and then verified.

Database

is a centrally held collection of data that allows access and manipulation by one or more users.

Day After Recall

is a standard measure of the impact of a TV commercial where respondents’ recollection of a commercial is tested the day after they are exposed to it.

Debranding

is the removal of brand names, packaging or other material that would enable research participants to identify a brand.

Debrief

is the presentation of research findings to clients.

Decentralised Research Function

is where the marketing researchers are spread throughout an organisation and are not located together.

Degrees of Freedom

is the number of ways in which values could be assigned within a system. For example a table of observations with p rows and q columns has (p-1)(q-1) degrees of freedom. (In a sample of size n grouped into k intervals, there are always (k-1) degrees of freedom, because if (k-1) frequencies are specified, the other is determined by the total size n.)

Delphi Approach

is an approach to forecasting where a group of people produce separate forecasts and then each member is given information about the other forecasts and asked to revise their original estimate. The objective of the approach is to converge forecast estimates.

Demand Artefacts

are the responses given when respondents attempt to guess the purpose of the questions being asked. Demand artefacts can be the result of experimental conditions.

Demographic Information

is based on the age, gender, life-cycle stage and occupation of consumers.

Dendogram

is diagram that shows a hierarchy and the relation of subsets in a structure. It branches like a tree and is usually read downwards from the main trunk.

Deontology

is a non-consequential approach to evaluating ethics, whereby the degree of ethicalness depends on the intentions behind the decisions rather than the outcomes or actions that result.

Dependence Techniques

are types of multivariate analysis techniques that are used when one or more of the variables can be identified as dependent variables and the remaining variables can be identified as independent.

Dependency

is when one variable is influenced to an extent by another variable.

Dependent Variable (aka Effect Variable)

usually denoted as y, is a variable that is influenced to some extent by one or more other (independent) variables.

Depth Interview (aka One on Ones)

is an unstructured personal interview with a single respondent that is conducted by a highly skilled interviewer. The purpose of depth interviews is to understand the underlying motivations, beliefs, attitudes and feelings on a particular subject.

Descriptive Research

is a form of conclusive research that aims to describe a product or market or identify associations among variables.

Desk Research

is the systematic examination of all available secondary data in the context of a particular marketing research problem.

Deviation

is the difference between the mean and an observed value.

Diary Panel

is a type of consumer panel where participants record activities or events in a diary.

Dichotomous Questions

are questions with only two alternatives, eg agree/disagree or yes/no.

Direct Observation

is when behaviour or events are observed while something is happening.

Direct Questioning Techniques

are ways of asking people directly for information, such as personal or telephone interviews and mail surveys.

Discrete Data

is that from a measurement scale consisting of a number of separate values where intermediate values are not permissible, eg the number of cars per household.

Discriminant Analysis

is an analysis technique where the dependent variable is non-metric (ie nominal or ordinal in nature) and the independent variables are metric (ie interval or ratio in nature).

Discussion Guide

is an outline of the subjects to be discussed during group discussions and/or in-depth interviews.

Disguised Observation

is the observation of behaviour without participants’ knowledge.

Disguised Questioning

is any form of questioning where respondents are unaware of the true purpose of the questions.

Disproportionate Stratified Sample

is a type of probability sample where the probability of a unit being selected from a stratum is not proportional to the number of units in the strata. This sampling approach is used when there are strata in the population of interest that are quite small but very important and they may not be adequately represented in a survey if other sampling approaches are used.

Disqualifier

is an answer to a question that makes the respondent ineligible to participate in the research project.

Double Blind Test

is a product test where both the researcher administering the test and the participants are unaware of the complete identity of the products being tested.

Double-barrelled Questions

are those that ask two questions at the same time. They can confuse respondents and answers to such questions are uninterpretable, because it is not possible to determine to which question the answer refers.

Dual Moderator Group

is a group discussion with two moderators, each moderator having a different role, eg one to ensure the smooth running of the group and the other to discuss or explain key issues.

Dummy Variables

are a way of respecifying categorical variables in data analysis by giving them either of two values (eg 0 or 1).

Edit Instructions

see data reduction plan.

Editing (aka Data Cleaning)

involves examining each completed questionnaire to ensure that the proper sequence of questions has been asked, the answers are clear and consistent and they have been correctly marked.

Effect Variable

see dependent variable.

Eligibility Criteria

are specified characteristics that potential participants must possess in order to be involved in a particular research project.

Empty Nesters

are those people whose children have left or are about to leave the family home.

Enabling Techniques

see projective techniques.

End Piling

is a phenomenon where many survey responses fall into just a few categories at the end of a measurement scale.

EPOS Data

is electronic point of sale data and it measures sales of products as they are being paid for by consumers.

Error

see total error.

ESOMAR

ESOMAR is the world association of research professionals. Founded in 1948 as the European Society for Opinion and Marketing Research - ESOMAR unites 4,000 members in 100 countries, both users and providers of opinion and marketing research. ESOMAR's mission is to promote the use of opinion and market research for improving decision-making in business and society world-wide. The address is Vondelstraat 172, NL-1054 GV Amsterdam, (+31-20-6642141).

Executive Summary

is a document that summarises all the sections of a market research report.

Experience Survey

is a survey involving participants who have knowledge of a particular situation.

Experiment

is the process of manipulating one or more independent variables and measuring their effect on one or more dependent variables, while controlling for external variables.

Experimental Design

is a set of experimental procedures specifying: the test units, sampling procedures, independent variables, dependent variables and how external variables are to be controlled.

Experimental Error

is the error caused by the conditions of the experiment itself and it creates uncertainty that the observed effects may not be due to the manipulation of the independent variable.

Experimental Treatments

are the different conditions created by manipulating the independent variable.

Experimental Units

are the objects or people who are part of the experiment.

Expert Opinion Survey (aka Key informant Technique or Specialist Research)

is a form of exploratory research that involves discussing a research problem with someone (or a group of people) with experience on a particular subject.

Exploratory Research

refers to the initial investigation of a problem that uses unstructured techniques (such as group discussions or in-depth interviews) in order to develop hypotheses and/or understand a problem further.

Exponential Smoothing

is a statistical technique that calculates a moving average where the most recent data are given a different weight to earlier data.

Extended Group Discussion

is a group discussion that has been designed to last around 3-4 hours (compared with the usual time of around an hour and a half).

External Secondary Data

are existing data that have already been collected by other organisations.

External Validity

is the extent to which experimental results can be projected to a population of interest.

External Variables (aka Extraneous Variables or Confounding Variables)

are factors that are not manipulated as part of an experiment, but they may exert some influence on the dependent variable under study.

Extraneous Variables

see external variables.

Eye Tracking Research

involves the use of various mechanical devices to record participants’ eye movements when they are looking at some form of stimulus (such as a press advert in a newspaper).

F Test

is a statistical test of the equality of the variances of two populations.

Fact

is a statement that is objectively true and can be verified.

Factor Analysis

is a form of multivariate analysis that takes a large number of variables or objects and aims to identify a small number of factors that explain the interrelations among the variables or objects.

Factorial Design

is a type of statistical experimental design where units are assigned to groups that represent all possible combinations of the independent variables of interest.

Factory Staff

are those who are mainly involved with conducting marketing research and have limited contact with research clients.

Field Experiment

is one that is conducted in a natural setting (where the external validity is usually higher than the internal validity).

Field Force

refers to the interviewers and supervisors who are involved in data collection.

Field Supervisor

is the person who is responsible for selecting, hiring and training interviewers. He or she is also responsible for the data collection phase of the survey and following the agreed instructions.

Fieldwork

is a general term that refers to any data gathering process.

Filter question

is a question in a questionnaire to ensure that respondents meet the required criteria for a subsequent question (or questions) in a survey.

Findings

are information that answer a research question.

Fixed Field Code

is a code in which the number of records for each respondent is the same and the same data appear in all the same columns for all respondents.

Fixed Sample

is the repeated observation of the same sample of respondents over a period of time.

Focus Group (aka Focus Group Interview or Group Discussion)

is an informal discussion with a small number of selected participants (usually 8-12) in the company of a skilled moderator who guides the discussion without influencing its outcome. The ideal number of participants depends on the subject matter being discussed, eg complex subjects may be better discussed with fewer participants - possibly 4-6.

Folder Test

is a qualitative press advertising test where the advertising is placed in a folder with competitive advertising (and editorial extracts) and respondents are asked to page through it.

Forced Rating Scale

is a scale that does not allow a neutral or no opinion choice.

Frequency (aka Average Opportunities to See)

is the average number of times an advert has been exposed to a specified television audience or universe .

Frequency Distribution

is a representation of the number of counts of objects or responses, usually in the form of a table or graph.

Friendliness Effect

see acquiescence bias.

Friendship Pair Interview (aka Paired Depth or Triad)

is a type of depth interview where there are two respondents (instead of the usual one). Friendship pair interviewing is often used for research with children and young people (where they know each other) in order to avoid respondents being intimidated and biasing their responses accordingly.

Frugging

is fund raising under the guise of research and it is one of the reasons why potential participants in market research projects are reluctant to take part.

Full Service Suppliers

are research suppliers who are capable of completing an entire research project.

Funnel Approach

is a way of ordering questions in a questionnaire so that general questions are asked before specific questions. This ordering avoids the responses to specific questions biasing the answers to general questions.  

Hall Test

is where research participants are invited to a central location to participate in some (usually quantitative) market research.

Head of Household

is that member of the household who is responsible for the household having that accommodation, either by owning, renting or having it rent-free. Where two or more people share this responsibility, the researcher should specify who to include in the study.

Hidden Issue Questioning

is a technique used during depth interviews that aims to identify significant personal views that would otherwise not be revealed by respondents using a direct approach.

Hierarchy-of-effects Model

is a marketing behavioural response model consisting of stages through which a buyer is presumed to go, including: awareness, knowledge, liking, preference, intention to buy and purchase.

Histogram

is a vertical bar chart where the height of the bars represents the data.

History Error

occurs in experiments when an unexpected (but significant) effect occurs that has an impact on the dependent variable being measured.

Holistic Test

is a test that aims to assess participants’ reaction to a product or concept as a whole (in contrast to an atomistic test that examines reactions to the individual elements).

Home Audit (aka Pantry Check)

is a survey where an interviewer enters a participant’s home to take an inventory of specific products present at that time.

Home Use Test (aka Product Placement Test)

involves participants evaluating products in their own homes, or more generally, in a natural usage environment.

Hypothesis

is an unproven statement that aims to describe a relationship or phenomenon that is relevant to a business.

Hypothesis testing

is a statistical procedure used to compare a sample mean to a specified value or to compare a pair of sample means.  

Ideal Population

see population of interest.

Identification information

lists details such as the name and address of participants in a market research project whereby they can be identified.

Implicit Alternatives

are alternatives in a question that are not expressed openly. Such alternatives may have lower chance of being selected and therefore the data may be biased.

Implicit Assumption

is when all respondents are assumed to have the same level of knowledge on a subject. Questions with implicit assumptions can lead to instrument error and/or respondent error.

Implied Population

is the population as suggested by the sample, ie it is that part of the population of interest that was available for the research. When a convenience sample is used or where there is sampling frame error, the implied population could be significantly different to the population of interest (or the ideal population).

Imputed Response

is a substitute for a missing response that is based on the pattern of other responses from a survey respondent.

Inability Error

is where respondents are unable to answer a particular question. This may be because they have not understood the question, they do not have the information the question requires, they cannot remember the circumstances to which the question refers or they are unable to articulate certain types of responses.

Incidence (aka Strike Rate)

is the proportion of respondents contacted in a survey who qualify for the survey.

Independent Samples

are samples that are not linked experimentally and the measurement of one sample has no effect on the others.

Independent variable (aka Causal Variable)

is a variable that exerts some influence on another (dependent) variable. Research experiments usually involve some manipulation of independent variables and measurement of dependent variables to investigate the relationship between them.

Indirect Approach

is one where the true purpose of a research project is disguised. Indirect approaches are used when revealing the true purpose of the research would bias the responses given by participants.

Indirect Observation (aka Trace Analysis)

is an observation technique where some record of past behaviour is used to deduce what happened during an event, eg looking at packaging usage in a fast food restaurant to understand what flavoured drinks sell the most.

Inferential Analysis

is the analysis of data to test a specific hypothesis.

Information

involves the analysis and interpretation of data to describe something about a market.

Informed Consent

is the principle that researchers should try to avoid both uninformed and misinformed participation by subjects in research.

In-home Interview

is where participants are asked survey questions face-to-face in their own homes.

In-street Interview

is where participants are asked survey questions face-to-face in the street.

Instrument Error

is a type of non-sampling error caused by the survey instrument (or questionnaire) itself, such as unclear wording, asking respondents for information they are unable to supply or the instrument being changed in some way during the course of the research.

Intended Sample

is the ideal sample for a particular research project (which may be different to the resulting sample).

Interaction Effect

is the effect of all the factors working together, which is greater than the sum of the separate effects.

Interactive Testing Effect

see pre-test manipulation interactive bias.

Interdependence Techniques

are types of multivariate analysis techniques that are used where no distinction is made as to which variables are dependent or independent.

Interdependency

is when variables each influence each other to some degree.

Internal Secondary Data

are data that have already been generated and/or collected by a business during its normal course of activity.

Internal Validity

is a measure of the accuracy of an experiment in terms of the degree to which changes in the dependent variable can be attributed to manipulations of the independent variable.

International Code of Marketing and Social Research Practice

is the code of conduct of ESOMAR and sets out the rights and responsibilities of those involved in marketing and social research.

Interquartile Range

is a measure of variability and it is the range between the upper and lower quartiles (ie the middle 50% of a distribution) and it is equal to the difference between the 75th and the 25th percentile.

Interval Scale

is a scale where equal intervals in the scale correspond to equal changes in the characteristic being measured, eg temperature scales (Celsius/Fahrenheit). Differences between objects can be analysed statistically (which is not possible with nominal or ordinal scales).

Interviewer Bias

see interviewer error.

Interviewer Cheating

is when interviewers knowingly do not follow their instructions.

Interviewer Error (aka Interviewer Bias)

is a type of non-sampling error caused by mistakes made by the interviewer. These may include influencing the respondent in some way, asking questions in the wrong order, or using slightly different phrasing (or tone of voice) than other interviewers. It can include intentional errors such as cheating and fraudulent data entry.

Intra-cultural Analysis

is the breakdown of data into individual countries (or cultural units) and analysis within each individual country.

Inverted Funnel Approach

is a way of ordering questions in a questionnaire where the specific questions are asked before the general questions. The approach can be useful where respondents do not have strong feelings or a general view on a subject.

Itemised Rating Scale

is a type of non-comparative scale where each choice category has some form of description (as opposed to a semantic differential scale, where only the extreme categories are labelled).

J   Top

Judgement Sample

is a type of non-probability sample where the selection of units is based on the judgement of the researcher.

Kelly Repertory Grid

is a qualitative research technique that aims to describe how potential consumers perceive products. Participants are presented with three attributes and they are asked to think of ways in which two are similar to each other but different from the third.

Key Informant Technique

see Expert Opinion Survey.

Key Word Recording

is a method used by interviewers to record answers to open-ended questions where only the important words and phrases are recorded.

Kolmogorov Smirnov One Sample test

is a non-parametric goodness-of-fit test that compares the cumulative distribution function for a variable with a specified distribution, based on one sample.

Kolmogorov-Smirnov Two Sample Test

is a non-parametric statistical test to determine whether two distributions are the same. The test takes into account differences in means, dispersion and skewness.

Kurtosis

is a measure of the relative peakedness or flatness of a distribution curve compared with a normal distribution. A positive Kurtosis value indicates that the distribution is more peaked than a normal distribution. A negative Kurtosis value indicates that the distribution is flatter than a normal distribution.

Laboratory Experiment

is one that is conducted in an artificial setting (where the internal validity is usually higher than the external validity).

Laddering

is a technique for conducting depth interviews where questions progress from product characteristics to user characteristics.

Latin Square Design

is a type of statistical experimental design where the aim is to remove the error associated with variations in two non-interacting external variables in order to understand the effect of each variable in addition to the effect of manipulating an independent variable. Experimental units are allocated in such a way that variations in experimental treatments occur once in each row or column (in a table of treatments).

Leading Question (aka Loaded Question)

is one that suggests an answer by the way in which the question is worded.

Level of Significance

see significance level.

Likert Scale

is a type of categorical, non-comparative scale that determines respondents’ levels of agreement to a series of statements relating to an attitude being measured.

Line Chart

is a chart where a series of data points are connected by a continuous line.

Linear Regression Analysis

is a type of regression analysis between variables that are believed to have a linear relationship.

Loaded Question

see leading question.

Location Studies

are research projects that aim to identify the best position (usually for a retail outlet).

Longitudinal Design

is a type of research design where a fixed sample of population units is measured repeatedly.

Mail Survey

is one where respondents complete a questionnaire (unaided) and return it to the sender either by post or e-mail.

Main Testing Effect

see pre-testing bias.

Mall Intercept Interviewing

is a common form of central location interviewing where participants are selected at random in busy shopping malls.

Mann-Whitney U Test

is a test that compares the location of two populations, based on samples from each population. The variables used are measured on an ordinal scale.

Marketing Decision Support System

is a type of marketing information system where decision makers can make particular requests for information that may not be part of the existing on-going reports.

Marketing Information System

is a set of formal procedures for collecting and analysing data from all sources and disseminating information regularly to marketing decision makers.

Marketing Research

(as defined by the American Marketing Association) is the systematic and objective identification, collection, analysis and dissemination of information for the purpose of improving decision making related to the identification and solution of problems and opportunities in marketing.

Matching

is the organising of experimental units into groups so that they share some particular characteristics that are relevant to the research.

Maturation Error

occurs in experiments when there is a gradual change in the dependent variable over time that is not caused by the independent variable, eg participants’ knowledge levels increasing over the period of an experiment.

Mean (aka arithmetic mean)

is a measure of central tendency that takes the sum of a set of values divided by the number of values in the data.

Measurement Scale

is a device that assigns numbers to objects, events or people according to a set of rules.

Measurement Timing Error

occurs in experiments when there are changes in the dependent variable that are caused by taking measurements at different times.

Measures of Central Tendency

are those that describe the centre of a distribution. Examples of measures of central tendency are: mean, median and mode.

Measures of Location

are statistics that describe the location within a data set. Examples of measures of location are the 25th percentile or the largest value. The mean, median and mode are also examples of measures of location (in addition to being measures of central tendency).

Measures of Shape

are skewness and kurtosis that describe the outline of a distribution.

Measures of Variability (or dispersion)

are those that indicate the spread of a distribution. Examples of measures of variability are: range, interquartile range, variance, standard deviation and coefficient of variation.

Mechanical Observation

is when some form of mechanical device records the behaviour of interest, eg a people-meter recording who watches what TV programmes.

Median

is a measure of central tendency that identifies the middle-point value (or 50th percentile) in a set of values when they are arranged in order of magnitude.

Methodology

is a description of the way in which the data is collected for part or all of a research project.

Metric Data

is data that can be analysed statistically, such as that from an interval or ratio scale.

Mini Group Discussion

is a focus group with fewer participants (usually 4-6) than the normal 8-12.

Mode

is a measure of central tendency that identifies the most frequently occurring value in a set of values.

Moderator

is someone who leads (but does not influence the outcome of) group discussions and/or in-depth interviews.

Monadic Test (aka Single Product Test)

is a product test where there is no comparison with other products.

Mortality Error

occurs in experiments when changes in the dependent variable are caused by experimental units no longer being part of the experiment.

Motivational Research

is qualitative research designed to uncover participants’ inner-most feelings and motives that they either may be unaware of and/or they would not normally reveal when questioned directly.

MRA

is the Marketing Research Association (based in the US) and it is a professional society for those who are involved or concerned with marketing and opinion research. Its mission is to promote excellence in marketing and opinion research by providing members with a variety of opportunities for advancing and expanding their marketing research and related business skills and to act as an industry advocate with appropriate government entities, other associations and the public. The address is 1344 Silas Deane Highway, Suite 306, Rocky Hill, CT 06067-0230.

MRS

is the Market Research Society (based in the UK) and it is a professional society for those who are involved or concerned with marketing and opinion research. The address is 15 Northburgh Street, London, EC1V 0AH.

Multi-collinearity

is a state of high intercorrelations among independent variables.

Multi-dimensional Scaling

is a perceptual mapping technique that represents perceptions and preferences of respondents as a spatial map. The axes of each map are the underlying dimensions that respondents use to form their preferences and perceptions.

Multi-item Scale

is a measurement scale that gathers opinions about an object on a number of dimensions and the data can be collated to produce a combined rating. The dimensions used can come from secondary sources and/or qualitative research. The intended use of the resulting data will also determine which dimensions are included in the scale.

Multiple Choice Questions

are those that ask respondents to select one or more alternatives from a set.

Multiple Cross-sectional Design

is a type of research design where two or more samples are drawn from a population of interest, each sample being only drawn once.

Multiple Regression Analysis

is a technique for developing mathematical relationships between two or more independent variables and an interval-scaled dependent variable.

Multiple Response Question

is one where respondents can provide more than one answer, eg by checking more than one item on a list.

Multiple Time Series Design

is a type of quasi-experimental design where a series of periodic measurements is taken from two groups of test units (an experimental group and a control). The experimental group is exposed to a treatment and then another series of periodic measurements is taken from both groups.

Multi-Stage Sample

is a sample that is selected in stages, where the sampling units at each stage are sub-samples from the previous stage.

Multi-variate Techniques

are forms of statistical analysis that are used where there are two or more dependent variables to be analysed simultaneously.

Mutually Exclusive

Categories are mutually exclusive when objects can be placed into one category and no other.

Mystery Shopping

is a type of observation study where someone is sent into a business location to act in the role of a customer to evaluate the performance of a business or an employee.  

National Proportions

are the characteristics of the country where a research project is being conducted.

Natural Observation

is the observation of behaviour in a natural setting.

Neutral Value

is usually the mean of all responses to a question and it can be used as a substitute for a missing response.

Nominal Scale (aka Classified Scale)

is a scale where the numbers act only as data labels (eg 0=male, 1=female or a social security number). The only analysis that can be performed is to observe how frequently each of the scale members occurs in the survey.

Non-comparative Scale

is one where each object is measured independently of the other objects in the same test and absolute results are obtained.

Non-metric Correlation

is a correlation measure for two non-metric variables that relies on rankings to compute the correlation.

Non-metric Data

is data that cannot be analysed statistically, such as that from a nominal or ordinal scale.

Non-metric Multidimensional Scaling

is a perceptual mapping technique that is based on an analysis of non-metric data such as rank ordering.

Non-parametric Test

is a test that involves non-metric data (ie data that comes from nominal or ordinal scales).

Non-probability Sample (aka Non-random Sample)

is a sample in which the selection of units is based on factors other than random chance, eg convenience, prior experience or the judgement of the researcher. Examples of non-probability samples are: convenience, judgmental, quota and snowball.

Non-random Sample

see non-probability sample.

Non-response Error

is a type of non-sampling error caused by some sub-groups of the sample responding less than the rest of the sample.

Non-sampling Error

is any error caused by factors other than sampling error. Examples of non-sampling error are: selection bias. population mis-specification error, sampling frame error, processing error, respondent error, non-response error, instrument error, interviewer error and surrogate error.

Normal Distribution

is a symmetrical bell-shaped statistical distribution where the mean, the median and the mode all have the same value.

Normative Data

see measures of central tendency.

Null Hypothesis

is a statement to be tested that is usually expressed in a negative (or null) way and suggests that no difference or effect is expected. If the statement is disproved, then the null hypothesis is rejected and the alternative hypothesis is accepted.  

Observation Bias

refers to changes in the behaviour or events under study that are caused by the observation process itself. Personal observation is usually more likely to have observation bias than mechanical observation.

Observation Study

is a research study where data is collected by watching consumer behaviour or events taking place.

Omnibus Study

is a periodic study that asks questions on a number of unrelated subjects. The results may be completely or partially syndicated among clients.

One Group Pre-test - Post-test

is a type of pre-experimental design where a single group of test units is measured, exposed to an experimental treatment and then measured again.

One Stage Sampling

is a sampling procedure that involves collecting data from all available units in selected sub-groups of a population of interest.

One-on-ones

see depth interview.

One-shot Case Study

is a type of pre-experimental design where a single group of test units is exposed to an experimental treatment and a single measurement is taken afterwards.

One-sided Question

is a form of leading question that presents only one aspect of an issue being considered by respondents.

One-tailed Test

is a test where the null hypothesis is expressed directionally, eg the proportion having a characteristic is more than x%. One-tailed tests are more popular than two-tailed tests because the directional result makes the findings more actionable.

Open-ended Questions (aka Unstructured Questions)

are questions that do not provide any ready-made alternatives for respondents, so they answer in their own words.

Opening Questions

are the questions at the beginning of a questionnaire and they should be interesting, simple and non-threatening to gain the confidence and co-operation of respondents.

Opinion

is the verbal expression of an attitude and is not directly verifiable by research data.

Opinion Poll

is a study that collects views of the public on matters of broad interest.

Optimum Allocation

is the allocation of sample units to strata so as to maximise the statistical efficiency of the survey.

Order Bias (aka position bias or sequential bias)

occurs when respondents tend to favour objects because of their position in a list or sequence. The objects at the beginning and at the end of a list can be remembered more than those occurring in the middle. Usual practice is to rotate a list to eliminate this type of bias.

Ordinal Scale (aka Ranked Scale)

is a scale where the numbers assigned represent relative amounts of the characteristic being measured, eg first or second in a ranking. The distances between intervals in an ordinal scale are usually not equal - ie the difference between the values of first and second is not the same as the difference between third and fourth in a ranking.

Osgood Scale

see semantic differential scale.

OTS - Opportunities to See

see frequency.

Outliers

are extremely small or extremely large values in a set, compared with the mean of all values in the set.

Paired Comparison Scale

is a type of comparative scale where respondents are presented with two alternatives and they are asked to choose which they prefer (according to a criterion).

Paired Depth

see friendship pair interview.

Pairwise Deletion

is a treatment of survey data where only the completed questions are analysed, ie partially complete responses are not excluded from the analysis.

Pan-cultural Analysis

is the collection of data from different countries (or cultural units) and analysis of the aggregated data.

Panel

see consumer panel.

Panel Conditioning

is a form of systematic error that occurs when panel participants change their behaviour (that is being observed) as a result of being part of the panel.

Pantry Check

see home audit.

Parameter

is a summary measure used to describe a group of people or objects that is based on data from all units in the group.

Parametric Test

is a test that involves metric data (ie data from interval or ratio scales that can be analysed statistically).

Partial Correlation Coefficient

is a statistic that is calculated to measure the association between two variables after controlling (or adjusting) for the effects of one or more additional variables.

Participant

is a general term covering anyone who is involved in a research study and not just someone who is interviewed, eg in an observation study or a group discussion.

Pearson Coefficient

see product moment correlation coefficient.

Penetration

is the proportion (usually expressed as a percentage) of a population of interest that has accepted a product or an idea in some way.

People Meter (aka Audimeter)

is an electronic device attached to a TV set that monitors which programmes are watched by whom.

Perceptual Mapping

is an analysis technique that is designed to demonstrate how consumers view a concept or brand relative to the alternatives. Perceptual maps can be constructed from different types of analysis, eg Factor analysis, correspondence analysis and multi-dimensional scaling.

Periodicity

is a cyclical variation in a characteristic being measured. Periodicity can lead to a bias in systematic sampling when the period of the cycle coincides with the sampling interval being used.

Personal Interviewing

is when an interview is conducted in a face-to-face setting.

Personal Observation

is when human beings record the behaviour of interest or events taking place.

Personalisation Technique

is a projective technique where participants are asked to ascribe “personality-type” traits or characteristics to an object or idea.

Pictograph

is a type of graph that uses pictures or symbols to display the data.

Pilot Testing

see pre-testing.

Placement Test

see home use test.

Population Mis-specification Error

is a type of non-sampling error caused by incorrectly defining the population of interest for a survey and either excluding or including certain types of respondents, eg limiting a survey to previous consumers of a product (and not asking the views of those who have not tried it).

Population of Interest (aka Target Population or Ideal Population)

is the group about whom the researcher wants to know more and from whom a sample will be drawn.

Position Bias

see order bias.

Post-test

is an evaluation of advertising after it has been run in the media.

Post-test Only Control Group Design

is a type of true experimental design where test units are randomly allocated to an experimental group and a control group. The experimental group is exposed to a treatment and both groups are measured afterwards.

Power

is the probability that a hypothesis testing procedure will lead to a rejection of a null hypothesis that is false. The power is one minus the probability of committing a Type II error, ie 1-*.

Precision Level

is a measure of the spread of a confidence interval. The narrower the interval, the higher the level of precision.

Pre-coding

is the assignment of codes (coding) before the fieldwork of a survey is carried out.

Predictive Validity

is the degree to which a measured marketing phenomenon at one point in time is able to predict another marketing phenomenon at a future point in time.

Pre-experimental Design

is an experimental design where randomisation procedures are not used to control for external variables. Examples of pre-experimental designs are: one shot case study, one group pre-test/post-test and a static group.

Pre-measurement Error

see pre-testing bias.

Pre-recruited Participants

are selected participants who have been contacted before the research fieldwork takes place and they have agreed to participate in the study.

Pre-teen (aka Tweenager or Tween)

is a young person aged around 8/9 to 12, who does not consider themselves to be a child, but they are not yet a teenager.

Pre-test - Post-test Control Group Design

is a type of true experimental design where test units are randomly allocated to an experimental group and a control group. Both groups are measured before and after the experimental group is exposed to a treatment.

Pre-test Manipulation Interaction Bias (aka Interactive Testing Effect)

is present when the pre-testing in an experiment heightens participants’ sensitivity and makes them behave in a systematically different way during the experiment that prevents the results being generalised to a population of interest (ie it reduces external validity).

Pre-testing (aka Pilot Testing)

is when the questionnaire is tried on a (statistically) small group of respondents to identify any unforeseen problems such as the wording or flow of the questions.

Pre-testing Bias (aka Pre-measurement Error or Main Testing Effect)

is when the pre-testing in an experiment has a systematic effect on the main experimental results, regardless of what happens between the measurements.

Primary Data

are data that are collected specifically for a current research project.

Probability Proportionate to Size

is an approach used in cluster sampling where the clusters have widely differing numbers of units in them. Clusters are selected with a probability proportional to the number of units in them. The units within each selected cluster are then sampled with a probability inversely proportional to the number of units, so that the probability of selecting each unit from the selected clusters is equal.

Probability Sample

is a sample where all units in the population of interest have a known and non-zero chance of being selected. Examples of probability samples are: simple random, Systematic, Stratified and Cluster.

Probing

is the asking of additional questions to encourage a respondent to enlarge on a particular answer or opinion so that their answer can be further understood by the researcher.

Processing Error

is a type of non-sampling error caused during the processing of the data. Examples include the incorrect entry of data and incorrect assignment of values.

Product Moment Correlation Coefficient (r) (aka Correlation Coefficient)

is a statistic that is calculated to determine whether a linear relationship exists between two metric variables and it takes values between -1 and +1 (depending on the degree of the relationship). A negative value indicates that the variables move in opposite directions and a positive value indicates that they move in the same direction. A value of 0 indicates that there is no linear relationship (although there may be a non-linear relationship).

Product Placement Test

see home use test.

Professional Respondent

is one who regularly attends group discussions, whose views may well not be typical of the intended respondents.

Profile

is a general term that covers the description of a population of interest (or a sub-group) according to certain characteristics. For example a demographic profile describes consumers in terms of their age, gender, life-cycle stage and occupation.

Projective Techniques (aka Enabling Techniques)

are a form of disguised questioning that encourage participants to attribute their feelings, beliefs or motivations to another person, object or situation. Examples of projective techniques are word association, sentence completion and thematic apperception tests.

Prompt Material

is any material that is shown to participants in research projects.

Prompted Awareness (aka Aided Awareness)

is the percentage of respondents who claim to have seen something (eg a brand or an advert) after having been shown some form of stimulus material.

Proportionate Stratified Sample

is a type of probability sample where the probability of a unit being selected from a stratum is proportional to the number of units in the stratum.

Proposal

is an outline (usually in writing) of how marketing research data could be collected and used to solve a specific problem. Proposals are normally written by research suppliers and they usually are divided into the following sections: background, objectives, methodology, costing and timings.

Proprietary Techniques

are research techniques that are considered to be the intellectual property of a research supplier.

Pseudo Research

is research that is carried out for reasons other than to reduce the risk in a resource-allocation decision.

Psychodrawing

is a projective technique where participants are asked to attach abstract notions such as colours, shapes or symbols to objects.

Psychographics

are quantified psychological profiles of individuals, based on their attitudes and behaviour.

Pupilometer

is a device used to measure the dilation of a participant’s pupil in response to a visual stimulus.

Purchase Intentions

are a measure of respondents’ attitudes towards buying a particular product or service.

Purposive Sample

see quota sample.  

Q-Sort Scaling

is a type of comparative scale where respondents are asked to sort between 60-90 objects (usually) into piles according to some criteria. It is a useful technique for sorting relatively large numbers of objects quite quickly.

Qualitative Research

involves the use of unstructured exploratory techniques (such as group discussions and in-depth interviews) that are based on statistically small samples in order to understand a problem further.

Quality Control

refers to a set of procedures to ensure that interviewers follow the instructions provided by the sampling plan.

Quantitative Research

involves the collection of (statistically) large samples of quantitative data and usually some form of statistical analysis. Quantitative research is often used to substantiate the findings from qualitative research.

Quasi-experimental Design

is a type of experimental design that applies some but not all of the procedures of a true experiment and lacks full experimental control. Examples of quasi-experimental designs are: time series and multiple time series.

Questionnaire

a structured technique for collecting data consisting of a series of questions. Questionnaires can be self-completion or administered by an interviewer, they can be completed orally or in writing.

Quintiles

are a set of four values that divide the total frequency into five equal parts.

Quota Sample

is a type of non-probability sample where the required number of units with particular characteristics are specified.

Quota Sample (aka Purposive Sample)

is a type of non-probability sample where the required numbers of units with particular characteristics are specified.

 Random Digit Dialling

is a method of reducing sampling frame error and involves the use of randomly generated numbers for a telephone survey, instead of relying on telephone directories or other lists of numbers that may exclude certain types of consumers.

Random Error

is an error that affects measurements in an unpredictable way due to random chance changes or differences.

Random Sampling Error

is the error caused by a particular sample not being representative of the population of interest due to random variation.

Randomisation

involves the use of random chance to assign experimental units to different treatments.

Randomised Block Design

is a type of statistical experimental design where units are blocked (or grouped) on the basis of one external variable to ensure that the experimental and control group are matched on that variable.

Range

is a measure of variability that is the difference between the largest and the smallest value in a set of values.

Rank Order Scaling

is a type of comparative scale where respondents are presented with a set of objects and they are asked to rank them according to a criterion.

Ranked Scale

see ordinal scale.

Ratio Scale

is a type of interval scale with a zero point, which represents an absence of the characteristic being measured. The data from ratio scales can be analysed statistically (which is not possible with data from nominal or ordinal scales).

Reach

is the proportion of a specified television audience or universe who have been contacted by an advert at least once.

Reactive Bias

is a type of experimental error where participants in an experiment change their normal behaviour as a direct result of the experimental conditions.

Recall Test

is a type of post-test that investigates respondents’ ability to recall the advertising unaided and aided.

Recommendations

are suggested courses of action or an assessment of the relative strengths and weaknesses of particular courses of action.

Reconvened Group Discussion

refers to two group discussions with the same participants that are separated by a period of time.

Recruiter

is the person who locates and invites selected participants to attend group discussions.

Refusal Rate

is the percentage of contacted people who decline to co-operate with the research study.

Regression Analysis

is the analysis of the relationship between a dependent variable and one or more independent variables. Its purpose is to determine whether a relationship exists and the strength of the relationship. It is also used to determine the mathematical relationship between the variables, predict the values of the dependent variable and control other independent variables when evaluating the effect of one or more independent variables.

Reliability

is the extent to which a research process can be repeated and produce consistent results (ie it is free from random error).

Representative Sample

is a sample that contains units in the same proportion as the population of interest.

Research Design

is the framework for conducting a market research project that specifies how the research will be carried out.

Research Effect

is a general term that suggests the research itself may be contributing to a change in the data.

Residual

is the difference between an observed and a predicted value.

Respondent

is the person who is interviewed by a researcher.

Respondent Error (aka Response Bias)

is a type of non-sampling error caused by respondents intentionally or unintentionally providing incorrect answers to research questions. Possible sources of respondent error can be: inability error, best light phenomenon, social group norms or selection bias.

Respondent Fatigue

is when respondents’ are disinclined to continue participating in a research project and it can lead to invalid responses (usually towards the end of the research project).

Respondent Level Data

is data from an individual respondent.

Response Bias

see respondent error.

Response Latency

is the speed with which a respondent provides an answer.

Response Rate

is the percentage of all attempted interviews that are completed.

Resulting Sample

is the sample that has been achieved in a particular research project (which may be different to the intended sample).

Role Playing

is a projective technique where participants are asked to play the role or assume the behaviour of someone else.

 Sagacity Segmentation

is a means of segmenting a population of interest using life-cycle stage, income and occupation variables.

Sample

is a sub-group of the population of interest that has been selected for study.

Sample control

is the efficiency and the effectiveness of the sampling approach at reaching the units specified in the research brief.

Sample Definition

is a description of the ideal (or the intended) sample for a research project.

Sample Selection Error

occurs in experiments when a bias is introduced into the way in which experimental units are assigned to groups.

Sample Size

is the number of sample units to be included in the sample.

Sample Unit

an individual member of the sample.

Sampling Error

is the error in a survey caused by using a sample to estimate the value of a parameter in the population.

Sampling Frame

is a list of the population of interest that is used to draw the sample in a survey, eg a telephone directory or a list of members of a profession.

Sampling Frame Error

is a type of non-sampling error in a survey caused by a sampling frame containing either more or less of a particular type of potential respondent, compared with the population of interest.

Sampling Interval

is the size of the step between selected units in systematic sampling.

Sampling plan

is a set of instructions given to interviewers concerning the selection of participants for a research project.

Sampling with Replacement

is when sample units can be included more than once in the sample.

Sampling without Replacement

is when sample units cannot be included more than once in the sample.

Scale Accuracy

is the extent to which a scale is valid and reliable, ie free from bias and random error.

Scale Reliability

is the degree to which a scale is free from random error and can reproduce consistent results.

Scale Transformation

is the manipulation of scale variables to ensure comparability with other scales and enable comparisons to be made. The most frequently used scale transformation procedure is standardisation.

Scale Validity

is whether a scale measures what it is intended to measure. A scale may be biased by the way in which a question is worded and therefore the results it produces are not valid.

Scanner Data

is the data recorded by bar-code scanners.

Scanner Panel

is a type of consumer panel where participants use a bar-code scanner to record purchases.

Scatter Diagram (aka Scattergram)

is a plot of the values of two variables for all the cases or observations.

Screening Questions

are the questions at the beginning of an interview or questionnaire to ensure that a potential respondent is eligible for the survey.

Secondary Data

are data that have already been collected and published for another research project (other than the one at hand). There are two types of secondary data: internal and external secondary data.

Segmentation

is the process of dividing markets into groups of consumers who are similar to each other, but different to the consumers in other groups.

Selection Bias

is a type of non-sampling error where the sample units are selected for treatment in a particular way that produces a different profile to the population. Selection bias can be introduced by researchers and/or by respondents (putting themselves into groups to which they aspire to belong, but they do not currently belong).

Self Administered (aka Self Completion)

is where a respondent completes a survey without the assistance or involvement of a researcher, eg a self-completion questionnaire sent through the mail.

Self Completion

see self administered.

Self-effacing Situation

is where the true answer to a question does not reflect favourably on the respondent or it contravenes a social group norm. In order to avoid bias in the data, one should try to avoid putting respondents in self-effacing situations if possible.

Semantic Differential Scale (aka Osgood Scale)

is a type of categorical, non-comparative scale with two opposing adjectives separated by a sequence of unlabelled categories, eg Good 1 2 3 4 5 6 7 Poor.

Semiotics

is the study of symbols and the meanings they convey. It is often used to discover the meanings of various consumption behaviours and rituals.

Sensitivity

is the extent to which a research technique can identify differences between sub-groups of participants.

Sensitivity Panel

is a type of focus group where the same participants are convened several times over a period of time. The subsequent discussions may or may not be on the same subject.

Sentence Completion

is a projective technique where participants are asked to complete a number of incomplete sentences that are provided by the researcher.

Sequential Bias

see order bias.

Sequential Sampling

is where the sample units are drawn one by one or in groups and the results of the drawing at any stage decide whether sampling is to continue.

Share of Mind

is the extent to which a particular brand will be thought of in relation to a specific product category.

Share of Voice

is the percentage of all promotion in a specific product category that is accounted for by a particular brand.

Show Cards

are a type of prompt material in the form of cards with images that are shown to participants in research studies.

Significance Level

(represented by *) is the maximum probability of rejecting a true null hypothesis (or committing a type I error).

Simple Correlation

see product moment correlation coefficient.

Simple Random Sample

is a type of probability sample where all units in a population of interest have an equal, known and non-zero chance of being selected.

Simulated Test Market

is a type of laboratory experiment that aims to imitate real life, where respondents are selected, interviewed and then observed making or discussing their purchases.

Single Cross-sectional Design

is a type of research design where one sample is drawn from the population of interest only once.

Single Item Scale

is a measurement scale that gathers opinions about an object on a single dimension.

Single Product Test

see monadic test.

Single Response Question

is one that asks respondents for only one answer.

Single Source Data

is data on a variety of behavioural habits from consumers in the same households, such as TV watching, reading and shopping habits.

Skewness

refers to the symmetry of a distribution. A skewed distribution is one where the mean, the median and the mode have different values, whereas in a symmetrical distribution they all have the same value.

Skip Pattern (aka Branching Question)

is a question used to guide an interviewer through a survey to different questions (ie skipping some questions), depending on the answers given.

Smiling Face Scale

is a balanced sequence of pictures of facial expressions that are often used in children’s research to obtain satisfaction ratings.

Smoothing

is the process of removing fluctuations in an ordered series of data.

Snowball Sampling

is a type of non-probability sampling where initial respondents are selected at random and subsequent respondents are then selected by referrals or information from the earlier respondents.

Social Class

is a means of segmenting consumers into groups based on the occupation and income of the head of household.

Social Desirability

is the tendency for respondents to give answers that are socially desirable or acceptable, that may not be accurate.

Social Group Norms

are expected overt modes of behaviour that constitute culturally acceptable ways of behaving in specific situations. Participants in market research projects may well bias their responses in order to conform to expected behaviour.

Socio-economic Information

is information that is based on the occupation and income level of the head of household.

Solomon Four-Group, Six-Study Design

is a type of true experimental design where test units are randomly allocated to two experimental groups and two control groups. One of the experimental groups and one of the control groups is measured. Both experimental groups are then exposed to a treatment. Afterwards both experimental and control groups are measured. Six measurements are taken in all and the design aims to account for pre-testing bias and pre-test manipulation interaction bias.

Specialist Research

see expert opinion survey.

Split Ballot Technique

is a procedure where a sample is divided into two halves and each receives a slightly different questionnaire.

Spontaneous Awareness (aka Unaided Awareness)

is a measure of how many respondents can quote a brand name without any assistance on behalf of the interviewer.

Spurious Correlation

is when two variables appear to change together but there is no causal relationship between them.

Standard Deviation (aka Standard error of the mean)

is a measure of variability (or dispersion) of a distribution and it is equal to the square root of the variance.

Standard Error

is a measure of the amount of sampling error present and it is equal to the standard deviation.

Standard Error of the Mean

see standard deviation.

Standard Industrial Classification (SIC)

is a means of classifying businesses based on the products or services they provide.

Standardisation

is a scale transformation procedure that involves manipulating data from different types of scales so that they can then be compared. It consists of subtracting the sample mean from each score and dividing by the standard deviation.

Stapel Scale

is a type of categorical, non-comparative scale that is similar to a semantic differential scale except that instead of having two opposing adjectives, there is only one adjective in between a sequence of positive and negative categories. Respondents indicate how relevant each single characteristic is to an object in question.

Static Group

is a type of pre-experimental design where there are two groups; an experimental group and a control group. The test units are not assigned randomly between both groups. The experimental group is exposed to a treatment and afterwards, both groups are measured.

Statistical Design

is a type of experimental design that allows for the statistical analysis and control of external variables. Examples of statistical design are: randomised blocks, Latin square and factorial design.

Statistical Efficiency

involves a comparison of the sampling errors generated by different sampling procedures. A statistically efficient sampling procedure is one that produces fewer sampling errors for the same sample size than the other procedures.

Stochastic Process

is one where there is an element of random variation.

Story Completion

is a projective technique where participants are asked to provide the conclusion to a number of incomplete stories provided by the researcher.

Stratified Sample

is a type of probability sample where the units in a population of interest are divided into mutually exclusive and collectively exhaustive strata and a (proportionate or disproportionate) random sample is drawn from each stratum.

Stratum Chart

is a set of line charts where the data are successively aggregated over the series. The magnitudes of each variable are represented by the areas between the line charts.

Strike Rate

see incidence.

Structured Observation

is a form of observation study where the behaviour of interest and the method of observation are clearly specified in advance of the study.

Structured Questions (aka closed questions)

are those where the choice alternatives for respondents are all specified.

Stub

is a row heading in banner format tabulations. It is usual practice to have the dependent variables as stubs.

Sub Sample

is a sample of a sample, which may or may not be selected using the same approach as the original sample.

Sugging

is sales under the guise of research and it is one of the reasons why potential participants in market research projects are reluctant to take part.

Surrogate Error

is a type of non-sampling error where someone other that the intended respondent participates in the research and this produces a systematic difference between the resulting sample and the intended sample.

Surveys

involve a large number of interviews with respondents, using pre-designed questionnaires.

Syndicated Research

is where the findings and costs of a research project are shared (partially or fully) among a number of clients.

Synectics

is a type of brain storming where the participants are from a variety of different backgrounds are asked to solve a specific problem. The approach is intended to increase the creativity of ideas produced.

Systematic Error

is an error that affects measurements in a consistent way.

Systematic Sample

is a type of probability sample where every nth unit is included in the sample from a list of the population of interest. The value of n is calculated by dividing the number of units in the population of interest by the required sample size.

T Distribution

is a symmetrical bell-shaped distribution that is used for testing samples smaller than 30.

T Test

is a statistical test that compares a sample mean with a hypothesised mean for a population of interest. The test is appropriate for small sample sizes (less than 30) or when the population variance is unknown. The test is also designed for comparing the differences between two means.

Tab Houses

are specialist suppliers of research data and simple analysis.

Tab Plan

see code and tab plan.

Tabulation

is a frequency count of each question’s answers.

Tachistoscope

is a device that allows a participant to look at some stimulus material (eg packaging or a brand name) for a pre-defined brief period of time. The aim is to investigate the stand out and/or recognisability of the material under investigation.

Target Population

see population of interest.

Teleology

is a consequential approach to evaluating ethics whereby the degree of ethicalness depends on the outcomes or actions that result from the decisions (rather than the decisions themselves).

Telephone Interview

is where participants are asked survey questions over the telephone.

Telescoping

is when respondents mis-remember when an event occurred, usually by stating that it occurred more recently than it really did.

Television Rating (TVR aka Gross Rating Point)

= (reach * frequency) One TVR represents 1% of a specified television audience or universe seeing an advert at least once. (At extremes, television ratings can be made from either many people seeing an advert only once, or just a few people seeing an adve

Temporal Ordering

is when a variable X (or a change in X) must occur before the variable Y (or a change in Y) is observed and it is one of the three conditions that have to be met to infer a causal relationship.

Termination Rate

is the number of eligible respondents who do not complete an interview once started.

Test Marketing

is a type of controlled experiment conducted in a carefully selected geographical area to understand the impact of a marketing programme.

Test Statistic

is a measure (derived from a formula) of how close the sample has come to the null hypothesis.

Test-retest Reliability

is a method of estimating the reliability of a test by repeating the measurement using the same scaling device under conditions that are judged to be similar.

Theatre Test

is an advertising testing procedure where participants are exposed to test and control commercials in the context of a TV programme or a film.

Thematic Apperception Test TAT (aka Cartoon Test or Bubble Test)

is a projective technique where participants are presented with one or more pictures or cartoons that depict a situation. They are asked to suggest what will happen or what one character may be saying or thinking in response to another character or a situation.

Third Person Technique

is a projective technique where participants are presented with a situation and are asked to suggest the feelings and attitudes of other people to the situation.

Time Series Design

is a type of quasi-experimental design where a series of periodic measurements is taken from one group of test units, followed by a treatment, then another series of measurements.

Topline Report

is a brief summary of the main findings of a study.

Total Error

is the difference between the true value of a parameter in the population and a value derived from a survey. Total error is the sum of the sampling and non-sampling errors in a survey.

Trace Analysis

see indirect observation.

Tracking Study

is a study that provides periodic updating of company and competitor performances on criteria other than sales.

Trade Off Analysis

see conjoint analysis.

Transcribing

see data entry.

Treatment

refers to the manipulation of one or more independent variables during an experiment.

Triad

see friendship pair interview.

Triangle Group

is a focus group with three participants who each have differing viewpoints.

Triangle Product Test

is a type of blind test that involves two similar products to determine the degree to which potential consumers can differentiate between them. Two of the products are identical and one is different. Participants are asked to test all three and indicate which two are similar to each other.

True Experimental Design

is a type of experimental design where the researcher randomly assigns test units and treatments to the experimental groups. Examples of true experimental designs are: pre-test - post-test control group, post-test only control group and a Solomon four group, six-study design.

TVR

see television rating.

Tweenager

see pre-teen.

Two Stage Sampling

is a simple case of multi-stage sampling where a sample of units is drawn from selected sub-groups of a population of interest. (Whereas with one stage sampling, data is collected from all available units in the selected sub-groups.)

Two Way Classification

is the classification of a set of observations according to two criteria.

Two-tailed Test

is a test where the null hypothesis is expressed in terms of a characteristic existing or not existing, eg a value is 50%. From the results of the test, it is not possible to determine on which side of the test value the true answer lies, therefore directional (or one-tailed tests) are used more often.

Type I Error

is a mistake that is made when a researcher rejects the null hypothesis and accepts the alternative hypothesis, when the null hypothesis is correct.

Type II Error

is a mistake that is made when a researcher accepts the null hypothesis and rejects the alternative hypothesis, when the null hypothesis is incorrect.

Unaided Awareness

see spontaneous awareness.

Unbalanced Scale

is a scale where the number of favourable and unfavourable categories is not the same.

Unbiased Questions

are those that are worded so that they do not influence respondents’ opinions.

Undisguised Observation

is the observation of behaviour or events with participants’ knowledge.

Unforced Rating Scale

is a scale that allows a neutral or no opinion choice.

Univariate Techniques

are forms of statistical analysis that are used where there is a single measure of each variable or where each variable is measured in isolation of other variables.

Unstructured Observation

is a form of observation study where the behaviours of interest and/or the method of observation are not clearly specified in advance of the study.

Unstructured Questions

see open-ended questions.

Unstructured Questions (aka open-ended questions)

are those that do not have any suggested alternatives and respondents answer in their own words.

Unwillingness Error

is where respondents are reluctant to answer a particular question. This may be because too much effort is required, the situation or context is not seen as appropriate for disclosure, no legitimate purpose or need is seen for the information requested or the information is seen as sensitive.

Usage & Attitude Surveys (U&A)

are research projects that aim to describe users (and non-users) of a product, together with their attitudes towards the product.

Validation (aka Back-checking)

is the subsequent re-contacting of respondents in a survey to check that the proper procedures have been followed, eg the respondent was qualified to be included in the survey, the responses recorded were accurate or even that the interview actually took place.

Validity

is the extent to which a research process is accurate and reflects actual market conditions (ie it is free from systematic error).

VALS (Values and Lifestyle)

is a psychographic segmentation system offered by the Stanford Research Institute.

Variable

is a property that takes on different values at different times.

Variable Respecification

is the transformation of data to create new variables (or modify existing variables) so that they are more consistent with the objectives of the study.

Variance

is a measure of variability (or dispersion) of a distribution and it is equal to the mean of the squared deviations of all values from the mean.

Venn Diagram

is a graphical method of representing operations on sets that is often used to illustrate probabilities.

Verbal Protocols

is a technique used to understand respondents thought processes while they are performing a task or making a decision by asking them to think aloud.

Verbal Rating Scales

are those that require respondents to indicate their position by selecting among orally identified categories.

Verbatim Statement

is a reproduction of all (or a representative part) of a respondent’s opinion of an object or concept.

Weighting

is when the responses from some (or all) sub-groups are assigned a statistical weight to reflect the importance of the sub-group in the population of interest.

Word Association

is a form of projective technique where participants are presented with a list of words and, after each word, they are asked to say the first word that comes into their minds.

Yea Saying

see acquiescence bias.  

Z Test

is a statistical test that is based on the standard normal distribution.

Z Value

is the number of standard deviations a point on a distribution is away from the mean.