Marketing Research Glossary - P
P Value: The numerical probability of getting a computed test statistic value that is due to chance, or random variation, under the assumption that the null hypothesis is true. The smaller the p value, the smaller the probability that the observed result would occur by chance if the null hypothesis were true. When an arbitrary significance criterion is set a priori to an experiment, the p-value may be expressed inexactly as either greater than, or less than, the criterion value.
PackageCheck®: Decision Analyst's proprietary, Internet-based package-testing system that provides diagnostic feedback on new package designs. Learn More
PackageScreen®: Decision Analyst's proprietary Internet-based system that identifies the early-stage package concepts and ideas that resonate with consumers. Learn More
PackageTest®: Decision Analyst's proprietary, Internet-based package-testing system designed to evaluate finished (or near-finished) package designs. Learn More
Packaging Research: Surveys or in-store experiments that attempt to measure consumer reactions to different packages or packaging concepts. Learn More
Page: In Logician®, Decision Analyst's state-of-the-art questionnaire programming system, a page is a set of questions that are grouped together.
Paired-Comparison Scales: Measurement scales that ask the respondent to pick one of two objects in a set, based on some stated criteria.
Paired-Comparison Test: A product test in which respondents are asked which one of two products they prefer. The term can also be applied to any research choice between two alternatives (two ads, two concepts, two packages, etc.) Learn More
Pan-Cultural Analysis: The aggregation of data from different countries (or cultural units) and analysis of the combined data.
Panel: People who have agreed to become members of a group for the purpose of participating in surveys. Sometimes the word "panel" is loosely (and incorrectly) used as a synonym for "sample." Learn More
Panel Cleaning: At Decision Analyst, the term refers to the continuous and ongoing process to identify undesirable panel members and delete them. Special programs are run daily against panel databases to identify and eliminate duplicate panelists, for example. Many other daily processes and systems are used to ensure that all Decision Analyst panels are as clean as possible. Learn More
Panel Guidelines: Decision Analyst's rules, regulations, and standards that govern how each of its panels is used, managed, and compensated. Learn More
PAPI (Paper and Pencil Interviewing): Survey in which the respondent fills out a traditional paper questionnaire. Term also refers to paper interviews administered by an interviewer.
Parameter: A measure of an entire population or universe, as contrasted with a Statistic (which is based on a sample taken from the universe).
Parametric Test: A statistical test that involves data that are assumed to be distributed according to some distribution, usually normal, whose parameters are known, such as data from interval or ratio scales.
Parent (Underlying) Distribution: The distribution of the measurements in the original population or universe.
Pareto Distribution: The Pareto distribution is sometimes described as the 80-20 rule. For example, 20% of a country's population might own 80% of its wealth. This type of skewed distribution is common in business, economic, and scientific data.
Partial Correlation Coefficient: A statistic that is calculated by measuring the association between two variables after controlling (or adjusting) for the effects of one or more additional variables.
Partial Data: Interim data provided to the contracting research company before all data is collected, for the purpose of checking the quality of the data and the data map. Normally such data is provided during the first few days of data collection.
Past Participation: Researchers may want to exclude respondents who have taken research studies during a certain time frame and/or within a certain topic to reduce the potential for result biases due to project or process familiarity.
Pearson's Product Moment Correlation: Correlation analysis technique for use with metric data. The resulting coefficient is a standardized measure of the extent to which the variables covary.
Penetration: The proportion (usually expressed as a percentage) of a target population that has tried and/or used a product. Sometimes penetration is used as a synonym for Market Share.
Penetration Analysis: The study of the market share (or other markers of marketing success) held by a given firm or product within various geographic areas, market segments, and/or demographic groups.
People Meter: An electronic system that attempts to measure TV audiences by physically measuring who is watching a TV program.
People Reader: A machine that links a respondent's reading material to his/her eye movements.
Per Capita Income: The average income per person in a population. It is calculated by dividing total U.S. income by the total population.
Percentile: A value on a scale of 100, or one percentage point. For example, a person who scored in the top-10 percent on a test would be in the 90th percentile or above.
Perceptual Mapping: The process of plotting elements (such as brands) and descriptors on a two-dimensional space. Perceptual maps show the relationships of the elements and descriptors. Learn More
Periodicity: 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.
Personal Income: A person's money income plus certain noncash benefits, as defined by the Bureau of Labor Statistics in the U.S.
Personalization Technique: A projective technique where participants are asked to ascribe "personality-type" traits or characteristics to an object or idea. See also Anthropomorphization.
Persuasion: A copy-testing measure, a pre-post purchase intent question sequence. Respondent states planned brand purchase intent before seeing an ad for that brand, and then again after seeing that brand's ad. The difference between the "before" purchase intent and the "after" purchase intent is commonly called persuasion. Some also use the term persuasion to refer to the purchase intent after exposure to an ad without a premeasure.
Photo Sort: Respondents sort photos (of people, brands, settings, etc.) to indicate correspondence to various attributes or answer-specific questions.
Physical Control: Holding the value or level of extraneous variables constant throughout the course of an experiment.
Physicians Advisory Council®: Decision Analyst's proprietary global Internet panel of 25,000 physicians and surgeons, including general practitioners and specialists who are willing to take part in online surveys. Learn More
Pictorial Scale: A type of scale in which answer choices are represented by pictures or illustrations (as opposed to numbers or words).
Pilot Test: A small number of interviews to evaluate the questionnaire and proposed methods before the conduct of a large scale study.
Pii® (Product Improvement Index): Pii® is a mathematical model to help guide product development efforts for new products and the reformulation of existing products. Learn More
PIN (Personal Identification Number): A code provided to survey respondents (or other Internet users) that allows people to log in to a website or a survey.
Pixel Graphic: Pixel is an abbreviation of picture element. It is a single point in a graphic image.
Placement Interview: A screening interview to determine if a respondent is qualified to participate in a home-use product test. The test product is "placed" in the homes of qualified respondents.
PMSA (Primary Metropolitan Statistical Area): Component of a larger area known as a Consolidated Metropolitan Statistical Area (CMSA) as defined by the U.S. Census Bureau.
Point Estimator: Statistic whose value should be a close approximation to the true value of the parameter. The actual numerical value that the point estimator assumes from the collected data (the sample) is called the point estimate.
Poisson Distribution: The Poisson distribution is sometimes referred to as the "distribution of rare events." Examples of Poisson distributed variables are the number of accidents in a random selection of 1,000 people, the number of sweepstakes won in 100,000 people, or the number of catastrophic defects found in a production run. More formally, the distribution describes the frequency of a given event occurring in a fixed number of opportunities (e.g., time periods or trials), where the probability of an event at any opportunity is very small, but the number of opportunities is large.
Poll: A study that collects views of the general public on matters of broad interest.
Polytomous Logistic Regression: An extension of logistic regression to cases where the criterion (outcome) variable has more than two levels. Also known as multinomial logistic regression. Learn More
Population: Entire group of people about whom information is needed. Also called Universe or Population Of Interest.
Population Centroid: According to the U.S. Census Bureau, the median central point of the population distribution.
Population Distribution: A frequency distribution of all the elements of a population (by age or income or geography, etc.).
Population of Interest: See Population.
Population Or Universe: The total group of people or things under study (i.e., the total group from which a sample will be pulled).
Population Pyramid: The graphic representation of a population's age-sex composition. It is a bar graph with the population divided into ages or age groups, represented from the youngest at the bottom to the oldest at the top, with males on the left and females on the right (or vice versa).
Population Specification Error: Error that results from an incorrect definition of the universe (or population) from which the sample is pulled.
Population Standard Deviation: The standard deviation of a variable for the entire population.
Pop-Up Survey: A questionnaire that appears on the screen of a website visitor. The term can also be used to refer to an invitation to participate in a survey. Pop-up surveys are usually triggered by some mechanism, such as the nth visitor, or a visitor who shows evidence of interest in a topic, or a visitor who goes to a particular section of the website, etc.
Positioning: The way a brand is described or characterized in relation to competitive brands and/or in relation to a target audience. For example, a brand could be positioned as the most durable product in a category, or it could be positioned as the technology leader in a category, etc.
Positioning Research: Studies designed to help identify the optimal positioning for a brand or service. Learn More
Post Hoc Segmentation: The process of segmenting a market or markets empirically, allowing the data to define the segments rather than basing the segmentation on predefined (a priori) segmentation variables. Cluster analysis is an example of post hoc segmentation. Learn More
Postal Survey: Also called Mail Survey. A questionnaire administered by mail. Respondents are sent the question by mail, asked to complete it unaided, and then return the questionnaire to the research company by mail.
Postwave: A survey conducted after an event (such as an advertising campaign). This type of surveying is normally used in conjunction with a Prewave survey to measure the changes that occur due to some stimuli (advertising, promotion, price change, etc.).
Poverty: A term that characterizes the lowest socioeconomic stratum within a society. The official definition of poverty income levels is revised annually in the U.S. based on changes in the cost of living, as reflected in the Consumer Price Index.
Pre/Post: A type of experimental design, involving a survey before the introduction of a stimulus and then an identical survey after the introduction of a stimulus. All variables other than the stimuli must be the same between the prewave and the postwave.
Precision: The degree of accuracy in a statistic based on sampling and surveys. In general, the larger the sample size, the more precise the resulting measurement, assuming a random sample.
Pre-Coded Questions: Questions that ask a respondent to choose from a list of precoded answers. Also called Closed-Ended Questions.
Predictive Analytics: The general term for using a variety of statistical, mathematical, modeling, etc., techniques on multiple data sets in order to deliver forward-looking marketing information and models. Learn More
Predictive Dialing: A computer system that dials telephone numbers and transfers connected calls to available interviewers. Predictive dialers are used to improve telephone-interviewing productivity.
Predictive Function: The predictive function of research is the specification of how to use descriptive and diagnostic research to predict the results of a planned marketing decision.
Predictive Validity: The degree to which the future can be forecast by a current measurement instrument or scale.
Predictor Variables: Another name for an Independent Variable in multivariate statistics. Predictor variables are those that explain or predict differences in a dependent variable.
Pre-Experimental Design: An experimental design that offers limited or no control over extraneous factors (i.e., environmental, economic, or competitive influences that might distort or bias the experiment).
Preference Test: A survey in which participants are asked to choose a preferred product or ad from two or more alternatives.
Prefix: The first three digits of a phone number. Also known as the Exchange. It represents the town, community, or neighborhood in which a telephone number is located, or at least used to before the advent of cell phones and overlapping prefixes.
Prerecruited Central-Location Test: A survey conducted at a conveniently located site to which screened and qualified respondents come to participate in the research project. Typically, the respondents are screened and recruited by telephone or via the Internet.
Presentation Software: Personal computer software, such as PowerPoint, that provides easy-to-use platforms for creating presentations.
Pretest: A small number of interviews that is administered to determine if a questionnaire is understood by respondents in order to ensure that the data collected is valid and free of errors or defects.
Pretest Questionnaire: A rough-to-final version of the questionnaire that is used in the pretest. Sometimes, a pretest questionnaire poses every question as an open-end. The answers are valuable in determining if respondents understood the question, and in identifying the full range of answer choices for the final questionnaire.
Pre-Walk-Through: A meeting scheduled by the Control Center at Decision Analyst to discuss all the details of the project. A pre-walk-through is recommended for all difficult and complex projects. The Operations employees working on the project and the Client Service team in charge of the project are required to attend.
Prewave: A survey conducted before a marketing stimuli (such as an advertising or promotion campaign). This type of survey is normally used in conjunction with a Postwave survey to measure the changes related to the stimuli.
Price Segmentation: A segmentation technique where the market is segmented based on price of the product offerings. Variation in household incomes creates an opportunity for creating products at multiple price points. Learn More
Pricing Research: Various types of research used to explore the role of price, optimal pricing, and pricing demand curves. Learn More
Primary Data: This is new, original data based on the conduct of a survey or research project. Secondary data is previously collected and published data.
Primary Grocery Purchaser: A person in the household responsible for at least 50% of that household's grocery purchases.
Primary Grocery Shopper (PGS): or the household's major grocery purchaser. Generally defined as someone who purchases at least 50% of the household's groceries.
Primary Metropolitan Statistical Area (PMSA): A metropolitan area that is adjacent to another, as defined by the U.S. Census Bureau.
Primary Sampling Units (PSU): The smallest geographic area, or smallest "bucket," from which respondents will be chosen for a survey when the sampling method is an Area Probability Sample. Think of "primary sampling units" as the building blocks of a sample. For example, counties could be the PSUs, or census tracts could be the PSUs, or groups of counties or groups of census tracts could be the PSUs.
Probability Distribution: Describes the range of possible values that a discrete random variable can attain and their associated probabilities. Often summarized in a table.
Probability Of A Simple Event: A number between zero and one that measures the likelihood that an event will occur when an experiment is performed. The probability of all simple events must sum to one.
Probability Of An Event: The sum of the probabilities of the simple events that make up the event.
Probability (Or Density) Function: Gives the probabilities that the random variable will assume each of its possible values. Often summarized in a table.
Probability Samples: Samples in which every member of the population being sampled has a known nonzero probability of being selected. Probability samples can be designed so that statistical inferences from the sample data are generalizable to the population under study without the broad, sometimes gratuitous, assumptions that must be made with nonprobability samples about distributions of the variables of interest within the population.
Probe: The Decision Analyst questionnaire instruction (to the telephone or in-person interviewer) for the interviewer to ask "What else" one time after the respondent's answer to an open-ended question. In qualitative research, the term probe instructs the moderator to use a variety of follow-up questions to better understand the respondent's original answer.
Probe Fully: The Decision Analyst questionnaire instruction (to the telephone or in-person interviewer) for the interviewer to ask "What else" repeatedly until the prompt is nonproductive. In qualitative research, the term "probe fully" instructs the moderator to use a variety of follow-up questions to better understand the respondent's original answer.
Probit Model: A model used in regression analysis with binary response data. Similar to a logit model except it uses a cumulative normal curve rather than a logistic curve.
Processing Error: Any mistake that results from incorrect transfer (or omission) of data, from the data collection phase of a research project to the final statistical tables.
Product Attributes: Words or phrases that express the impressions of or characteristics of a product or service, such as durability, reliability or good taste.
Product Code: This term can refer to the numbers, letters, or names given to products in a product test. The term can also refer to the UPC (Universal Product Code) placed on all consumer packaged goods to help automate the retail checkout process.
Product Concept Testing: The testing of new product ideas at an early stage in concept form. The concept might be communicated by a rough print "ad," or a product illustration, a rough package, or even a product prototype. Learn More
Product Image: Similar in meaning to Brand Image. The mental impression, mental pictures, perceptions, and emotions associated with a product (or brand).
Product Improvement Index (Pii®): Pii® is a mathematical model to help guide product development efforts for new products and the reformulation of existing products. Learn More
Product Optimization: Product optimization refers to the process of improving a product until it reaches a maximum level of consumer satisfaction or acceptability. Learn More
Product Placement Study: The test product is placed in the households of target-market consumers to use the product as they would normally use that type of product. After usage is complete, the respondent answers a series of questions about the product and the usage experience. Also called a Product Test or HUT (Home Usage Test).
Product Prototype Tests: Target-market consumers are shown early-stage prototypes of proposed new products and then asked to answer a series of questions about their interest in and perceptions of the new product prototypes. Learn More
Professional Respondent: A respondent who is a frequent participant in focus groups and/or survey research projects. There is a risk that this "professional respondent" might be nonrepresentative and/or biased in some way. Decision Analyst goes to great lengths to prevent professional respondents from joining its online panels, and performs many ongoing checks and evaluations to identify and delete any online panel members whose behavior or answers trigger suspicions.
Profile: A description of a population or a segment of that population. For example, a demographic profile describes consumers in terms of their age, gender, income, education, occupation, etc.
Programmatic Research: A series of research projects or investigations conducted over a span of time according to a plan, usually conducted by a team of researchers. Term is most commonly used in medical research and the social sciences.
Projectability: The degree to which research results can be extrapolated to the larger universe or population based on the sample surveyed.
Projection: An estimate of a future pattern, trend, or activity (i.e., a sales projection). Forecasts and projections are similar terms and are often used interchangeably.
Projective Techniques: Data collection techniques that elicit respondents' deepest feelings and motivations via "projection" of those feelings and motivations into an unstructured and undefined situation (these techniques include role-playing, storytelling, sentence completion, picture drawing, reactions to ink blots, reaction to pictures, etc.).
Projective Test: Technique for tapping respondents' deepest feelings by having them project those feelings into an unstructured situation.
Promotion Research: Survey research designed to measure consumer reactions to promotional ideas and concepts. Early-stage promotional ideas and concept boards to the final finished promotion can be tested. Learn More
PromotionCheck®: Decision Analyst's proprietary, Internet-based monadic pretesting system to evaluate early-stage promotional ads and storyboards. The detailed diagnostic feedback helps in refining the promotion concept. Learn More
PromotionScreen®: Decision Analyst's proprietary Internet-based comparative system to evaluate early-stage promotion ideas and concepts, typically in print ad format. Learn More
PromotionTest®: Decision Analyst's proprietary, comprehensive monadic pretesting system designed to predict the effectiveness of the finished (or near-finished) promotion ad or commercial. Learn More
Prompted (Aided) Awareness: The percentage of respondents who claim to have seen or heard something (e.g., brand or advertisement) after verbal or written stimulus. For example, aided brand awareness could be measured by the question, "Which of the following brands are you aware of?" followed by a list of brands.
Prompts: A type of prompt material in the form of cards with images, words, scales, pictures, etc., that are shown to participants during in-person research surveys. Also called Answer Cards or Show Cards.
Propensity Model: A mathematical model of the propensity of members of a population of interest to exhibit a certain behavior. Usually uses some form of logistic regression yielding the probability, based on a set of predictor variables that a given member of the population of interest will exhibit the behavior of interest (e.g., respond to a solicitation by purchasing the product being offered).
Proportional Property Of The Normal Distribution: The feature of normal distributions such that the proportion of observations falling between the mean and a given number of standard deviations from the mean is the same for all normal distributions.
Proportional Sample: A sample in which the number of elements drawn is proportionate to the relative number of elements in the population.
Proportionate Stratified Sample: 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: A proposed plan for a marketing research project, including background, objectives, proposed methods, timing, and cost. Once a proposal is approved by the client, the proposal becomes the roadmap to the conduct of the research.
Proprietary Techniques: Research techniques that are the intellectual property of a research company. Decision Analyst, for example, owns and trademarks over 20 proprietary mathematical models used in the analysis of its research projects.
Protomonadic Monadic: The definition of this term varies greatly from researcher to researcher. The protomonadic design begins as a monadic test, followed by a paired-comparison. Learn More
Pseudo-Depth Interviews: A qualitative research technique discussed as on quantitative survey. This technique uses a survey format and consists of open-ended questions. Learn More
Psychographic Research: A type of research that describes psychological traits and characteristics of consumers (values, perceptions, motivations, etc.) that might be valuable in explaining consumer buying behavior. A similar term is Lifestyle Research. Learn More
Psychographic Segmentation: A segmentation technique where respondents are segmentated based upon multivariate analyses of consumer attitudes, values, behaviors, emotions, perceptions, beliefs, and interests. Also known as Lifestyle Segmentation. Learn More
Pupilometer: A device that measures changes in the human eye's pupil dilation (an indicator of arousal or interest, according to some sources).
Purchase Intent: A measure of the likelihood that a respondent will purchase a product or service. Also called Buying Intent.
Purchase Intent Question: The standard question is similar to the following: "If this product were available where you normally shop, how likely would you be to purchase this product?" The standard answer choices are "definitely buy," "probably buy," "might or might not buy," "probably not buy," or "definitely not buy." There are many other types of purchase intent questions as well.
Purchase-Intent Scales: Any scale used to measure a respondent's intention to buy a product. The most common purchase-intent scale is the "definitely buy," "probably buy," "might or might not buy," "probably not buy," "definitely not buy" scale.
Purchase Lab: See Purchase Laboratory
Push Poll: A political survey undertaken with the purpose of biasing (or pushing) respondents. For example, a survey might ask voters if they are aware of Candidate A's criminal history, when in fact Candidate A has no criminal history.
Contact Decision Analyst
If you would like more information on Marketing Research, please contact Jerry W. Thomas by emailing firstname.lastname@example.org or calling 1-817-640-6166.