Marketing Research Glossary - R
R-Language: An open-source, free software that can manipulate, analyze, and graph large datasets with complex variables. The R-software is free (http://www.r-project.org/) but does require some knowledge of its programming language.
Race: A characterization usually based on skin color, facial or bodily features, and/or country of origin. Race is often used in U.S. surveys, but rarely in other countries (it is viewed as an offensive question outside of the U.S.). Race is slowing fading as a meaningful term, as populations mix.
Raking: Sample marginals are forced to match the known population marginals (from a census or other source) by an iterative procedure. Also known as RIM-Weighting, first proposed by W.E. Deming during the 1940s.
Random: A condition in which each person, event, or object has an equal (nonzero) probability of selection.
Random Error: Error that results from chance variation.
Random Forest Regression: A regression technique which can be used in key driver analysis and classification/segmentation analysis. Random forest regression is one of several modeling techniques using ensemble methods that combine regression models from a number (often thousands) of models using bootstrap samples of the data set. Learn More
Random Variable: A variable whose value is determined by chance. For example, "heads" or "tails" is a random variable for a coin toss. The answer to a given survey question by a person randomly selected from the population is a random variable.
Random-Digit Dialing (RDD): A selection of telephone numbers where the digits in the phone numbers are chosen by random methods (chance). Random-digit samples include unlisted and nonlisted phone numbers.
Random-Digit Sample: Creating phone numbers using random methods, so that unlisted numbers and new connections are included.
Randomized Block Design: 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 groups are matched on that variable.
Range: The maximum value for a variable minus the minimum value for that variable (i.e., the highest number minus the lowest number).
Rank-Order Scale: A scale in which the respondent compares one item with another (or a group of items) and ranks them on some attribute or dimension.
Rate: A fixed ratio between two things.
Ratio: A measure that expresses the relative size of two numbers.
Ratio Scale: An Interval scale with a meaningful zero point so that values can be compared arithmetically. Also known as Metric Scale.
Raw Data: Survey data before quality-assurance processing, and before any tabulation, analysis, or interpretation.
Reach: The percentage of a target population exposed to a commercial or advertisement at least once within some time frame.
Readership: The number of people who read a particular publication. The term is also used to measure the percentage of magazine readers who read a particular ad.
Recall Measurement (Recall Test): A type of posttest that investigates respondents' ability to recall something they may have read, heard, or seen. Recall measurements can be taken without or with the benefit of some form of stimulus material. See Spontaneous or Unaided Awareness and Prompted or Aided Awareness.
Recontact: To go back to a survey respondent and ask additional questions after the initial survey is completed.
Recruiting: The inviting of selected participants (those who qualify) to take part in a research project or survey.
Recruitment: The process of recruiting participants for focus groups or other research studies.
Refusal: A respondent who refuses to participate in a survey. Refusals are tracked at various stages within a research project (initial refusals, qualified refusals, etc.).
Refusal Rate: The percentage of people contacted who decline to participate in the research study.
Regression Analysis: A multivariate technique that attempts to explain changes in a dependent variable based on changes in one or more independent variables.
Regression Coefficients: Values that indicate the relative importance of each of the independent variables in a regression equation.
Regression To The Mean: Tendency for people (or events) to move toward the average during the course of an experiment or over time.
Rejection Region: The region of the distribution of the test statistic whose values would lead to the rejection of the null hypothesis.
Related Recall: A percentage of respondents who can recall a brand's advertising and remember at least one message or idea from that advertising and relate it to the advertised brand.
Related Samples: Samples in which the measurement of a variable in one population may influence the measurement of the variable in the other.
Reminder Email: An email sent to Decision Analyst's online panel members at least two days after the initial email invitation to participate in a survey to encourage panel members to take the survey. The reminder email is sent only to people who have not responded to the survey invitation. Reminder emails in an online survey play the same role as a Callback in a telephone survey.
Repeat Group: A questionnaire programming term. Repeat group allows for the building of repetitive pages or questions with only one markup definition.
Repeat Purchase Rate: The proportion of first-time purchasers of a product who purchased the product again (a second purchase, etc.).
Repeated-Pairs Technique: A product-testing procedure in which respondents are asked to choose between two products, and then repeat the task with an identical pair of products. Learn More
Repertory Grids: Also known as Lelly Triads. Used especially by advertising agencies to elicit consumer language for the products in question. Products (or services) are described or pictured on cards that are dealt three at a time. The respondent is invited to pick the "odd one out" and explain why it is different from the other two cards. The consumer's language and key discriminators are noted.
Replicate: Geographically representative subsample that is systematically selected from the entire sample.
Reporting Directory: An internal Decision Analyst tool for Internet studies showing the number of completes, the number of "Do Not Qualifies" (DNQs), and topline reports.
Representative Sample: A sample in which each unit has an equal and independent chance of selection. Also known as Random Sampling.
Rescreening: A brief interview conducted with potential participants when they arrive at a focus group facility to ensure that they really qualify for the session. In the rescreening process, the respondent is asked some of the key questions that were originally asked when the person was first recruited.
Research Design: The plan of research that is used to answer the research objectives; the structure or framework to solve a specific problem.
Research Hypothesis: What is believed to be true if the null hypothesis is false. Also known as the Alternative Hypothesis.
Research Management: Overseeing the development of excellent communication systems, data qualifications, time schedules, cost controls, client profitability, and staff development.
Research Panel: A group of individuals who agree to participate in a series of research studies over time. Decision Analyst proprietary Internet panel, American Consumer Opinion?, has over eight million members worldwide. Learn More
Residual Error: What still cannot be explained in a regression equation, after estimating the coefficients of the independent variables.
Respondent Fatigue: The term used when during the process of answering questions in a long survey, respondents can become tired, bored, and disinterested.
Respondent ID: A unique code (usually a numeric sequence) assigned to each respondent to identify that respondent during the course of a study and to identify that respondent's answers in the study data file.
Respondent Qualifications: See Eligibility Criteria.
Response Bias: Error that results from the fact that people who respond to a survey might not have the same opinions as those who do not respond.
Retail Audit: Also called Store Audit or Audit. The measurement of a product’s sales in a store for a time period by counting beginning inventory, adding new shipments, and subtracting ending inventory.
Retail Depletions: Retail depletions refer to products purchased in a store and carried out the front door. Retail depletions are the same thing as retail sales measured by the checkout cashier or checkout scanner data. Retail depletions, as a measure, exclude all product inventories in warehouses, the back of the store, or on retail shelves. Decision Analyst’s Conceptor? Volumetric Forecasting Model predicts retail depletions. Learn More
Retail Site Selection: Location analysis applied to retail sites—a study that evaluates the business opportunity offered by a specific retail site. Learn More
Return Link: A URL that is used to direct traffic to an originating place. If Decision Analyst launches its panel members to another website to participate in a survey, a return link is attached to bring the respondent back to Decision Analyst for recording and incentives.
Return On Investment (ROI): The ratio measure of profit (return) to investments (expenses) used to produce that return, achieved by a firm through its basic operations. ROI is an indicator of management's general effectiveness and efficiency. Learn More
Return On Marketing Investment (ROMI): The ratio measure of profit (return) to marketing investments (advertising and promotion expenses) used to produce that profit achieved by a firm through its basic operations. ROMI is an indicator of marketing management's general effectiveness and efficiency. Learn More
RFP/RFQ (Request For Proposal/Request For Quotation): The process of soliciting proposals and bids for a marketing research project.
Rim-Weighting: A process first proposed by W.E. Deming in the 1940's. Sample marginals are forced to match the known population marginals (from a census or other source) by an iterative procedure. Also known as Raking.
ROI: See Return On Investment.
Root Node: Also referred to as a Survey Node; a marked-up survey document (questionnaire) must always have a root survey node.
Rotate: A questionnaire instruction to rotate the brand list, attribute list, or answer choices. Rotate performs the same function as Randomization.
Routing: The automated distribution of a questionnaire to Control Center and Operations managers before a walk-through to ensure that the questionnaire is easy to understand and consistent with Decision Analyst's procedures, guidelines, and quality standards.
Rural Population: A term used by the U.S. Census Bureau referring to people in the U.S. who are not classified as Urban Population. The U.S. Census Bureau defines urban as a town, community, or city with at least 2,500 inhabitants. All population not in these urban areas is classified as rural population.
Rustbelt: States in the Great Lakes region of the U.S. and/or the Ohio River Valley.
Contact Decision Analyst
If you would like more information on Marketing Research, please contact Jerry W. Thomas by emailing email@example.com or calling 1-817-640-6166.