A
AAU (attitude, awareness and usage) study
Tracking study that measures consumer attitudes, awareness and usage levels for a product category or brand.
[go to glossary index]
Accuracy
The ability of a sample statistic to match the population.
[go to glossary index]
Ad positioning statement test
Test that determines target audience reactions to positioning statements that may be used in advertisements.
[go to glossary index]
Ad tracking research
Research that measures the effectiveness of advertising over time.
[go to glossary index]
Analysis of variance (ANOVA)
A method of analysis for determining whether two or more groups are statistically different with respect to the mean response of one or more variables.
[go to glossary index]
ASCII (American standard code of information interchange)
Data code in which each column is a letter or specific character.
[go to glossary index]
Awareness
Familiarity with a product, brand name or trademark.
[go to glossary index]
B
Balanced incomplete block (BIB)
An experimental design procedure for rotating a large number of products or items in a test. Each item is evaluated an equal number of times, but all respondents do not evaluate each item.
[go to glossary index]
Banner
Crosstab breaks that run horizontally across the top of a cross tabulation table. These breaks describe the category in which the data belongs. They are also referred to as columns on a table.
[go to glossary index]
Base
The number of respondents used to calculate the percentages in a table.
[go to glossary index]
Benchmark
A control source against which you compare the data you're studying. For example, you may compare the results of a study completed before an ad campaign (the benchmark) to the data collected after the ad campaign.
[go to glossary index]
Bias
Statistical sample error that will cause a misrepresentation of the population. Improper representation of the population in the sample, interviewing techniques, wording of questions, data entry, etc. may all be a cause of this error.
[go to glossary index]
Brand equity
All of the distinguishing qualities of a brand that result in the brand's value either positive or negative.
[go to glossary index]
C
Callback
An attempt to reach a respondent who could not be reached previously.
[go to glossary index]
CAPI (computer-aided personal interviewing)
Interviewer-administered survey using a computer-based questionnaire.
[go to glossary index]
CATI (computer-aided telephone interviewing)
Interviewer-administered telephone survey using a computer-based questionnaire.
[go to glossary index]
Census
A study where the sample includes the entire population.
[go to glossary index]
Census divisions
The nine census divisions are:
- New England
- Middle Atlantic
- South Atlantic
- East South Central
- West South Central
- East North Central
- West North Central
- Mountain
- Pacific
[go to glossary index]
Census regions
The four census regions are:
- Northeast
- South
- Midwest
- West
[go to glossary index]
Central-location study
A survey conducted at one or more conveniently located sites to which respondents come to be interviewed in person.
[go to glossary index]
Chi-square
A statistical test used to detect whether the observed distribution of data from the research is different than the theoretical expected distribution.
[go to glossary index]
Clarifying
A follow-up method for obtaining clear and complete responses to open-ended questions. By doing this the number of unclear or ambiguous responses is reduced. Same as Probing.
[go to glossary index]
Closed-end question
Questions asking respondents to choose from pre-determined answers.
[go to glossary index]
Coding
The process of grouping similar open-ended responses and quantifying them based on similar ideas.
[go to glossary index]
Completion rate
The percent of qualified respondents completing an interview or study.
[go to glossary index]
Concept statement
A written description for a new product or service.
[go to glossary index]
Confidence interval
The range around a survey result for which there is a high statistical probability that it contains the true population parameter.
[go to glossary index]
Confidence level
The probability of accepting the null hypothesis when it is true. Identified as 1-α.
[go to glossary index]
Conjoint analysis
A multivariate technique used to quantify the value that people associate with different levels of product/service attributes. Respondent’s trade product attributes against each other to establish product (brand) preference and the relative importance of attributes.
[go to glossary index]
Co-op payment
The payment to respondents for participating in a research study. Same as Incentive.
[go to glossary index]
Correlation
A measure of the linear association between two variables that says nothing about a cause and effect relationship.
[go to glossary index]
Cross tabulation
Table showing the frequency and/or percentage of responses to one question relative to responses to one or more other questions. Often abbreviated to cross-tabs.
[go to glossary index]
Customer satisfaction research
Research study conducted to measure overall satisfaction with a product or service and satisfaction with specific elements of the product or service.
[go to glossary index]
D
Data
Observations collected during interviewing.
[go to glossary index]
Data processing
Method used to count, tabulate, or analyze raw data.
[go to glossary index]
Demographics
The characteristics of human populations and population segments (e.g. age, marital status, income, household size, occupation).
[go to glossary index]
Discriminate analysis
A multivariate technique for analyzing the predictive value of a set of independent variables.
[go to glossary index]
Dyad
Non-quantifiable method of questioning two people simultaneously to understand why behavior and opinions are so.
[go to glossary index]
E
Editing
The process of checking questionnaires to be sure they were filled out properly, completely and accurately.
[go to glossary index]
Executive summary
Part of a research report that explains why the research was done, key findings, what those findings mean, and what action, if any, should be taken as a result of the research.
[go to glossary index]
F
Factor analysis
Procedure for data simplification through reducing the many rating scales (or set of variables) used by the researcher to a smaller set of factors or composite variables.
[go to glossary index]
Focus group
Non-quantifiable method of questioning a number of respondents in a small group to understand why behavior and opinions are so.
[go to glossary index]
Frequency
Number of people who gave a particular response.
[go to glossary index]
F-test
Test of the probability that a particular calculated value could have been due to chance.
[go to glossary index]
G
H
Home use test
Test where participants are given one or more product samples to use at home under normal usage conditions. Reactions to the products are obtained in a follow-up survey. Same as a Product placement study.
[go to glossary index]
I
IDI
In-depth interview. A non-quantifiable method of extensively questioning an individual to understand why a behavior or opinion is so.
[go to glossary index]
Incentive
The payment to respondents for participating in a research study. Same as Co-op payment.
[go to glossary index]
Incidence
A figure referring to the percentage of people in a particular category that meet the study’s qualifying criteria. It is used to measure the level of effort required to identify qualified study participants.
[go to glossary index]
Interval data
Ordinal scale with the additional property that the distance between observations is meaningful. An example would be the temperature.
[go to glossary index]
Interviewer
The person conducting an interview.
[go to glossary index]
Interviewer's instructions
Written directions to the interviewer on how to conduct the interview.
[go to glossary index]
J
K
L
Level of significance
Probability of making a Type I error, that is, rejecting the null hypothesis when it is true. Identified as α.
[go to glossary index]
Longitudinal study
A long-term survey based on repeated analysis of either the same sample (called a panel study) or new samples chosen at regular intervals over time.
[go to glossary index]
M
Mall intercept interviewing
Respondents are intercepted in public areas of malls and interviewed in person.
[go to glossary index]
Marginal
A frequency count of the number of people giving each answer to questions in a questionnaire.
[go to glossary index]
Market segmentation
Marketing strategy where the total market for a product or service is divided into sub-groups of consumers who exhibit certain common characteristics.
[go to glossary index]
Marketing research
An objective approach to developing studies to answer marketing management decision-making questions.
[go to glossary index]
Mean
A summary statistic that represents the average of all responses calculated by adding all sample values and dividing by the total number of respondents.
[go to glossary index]
Median
Midpoint of a group of numbers.
[go to glossary index]
Methodology
A detailed explanation of how a study was executed.
[go to glossary index]
Mode
Most frequently occurring value in a group of numbers.
[go to glossary index]
MOSTER
Acronym for MOnitoring Satisfaction To Ensure Retention. MOSTER is a research methodology model that measures customer satisfaction with a company or brand.
[go to glossary index]
Multiple regression analysis
Statistical procedure that relates multiple independent variables simultaneously to a single dependent variable in order to identify a pattern or patterns.
[go to glossary index]
Multivariate analysis
Any statistical procedure that simultaneously analyzes several measurements (variables).
[go to glossary index]
N
Nominal data
Data where numbers serve as labels only.
[go to glossary index]
Non-random
Occurrences which do not have an equal probability of occurring.
[go to glossary index]
Nonresponse bias
Error that results from a systematic difference between those who do and do not participate in a survey.
[go to glossary index]
Norm
The standard or hypothesized value against which a sample statistic is compared in hypothesis testing. Typically the “norm” is set at 50 in these tests and indicates parity or equality among values being tested.
[go to glossary index]
Normal approximation to the binomial test
A statistical test used to detect differences between two proportions or one proportion to a norm.
[go to glossary index]
Normal distribution
A continuous distribution that is bell shaped and symmetrical about the mean.
[go to glossary index]
O
Objectives
The specific questions that the research study should address.
[go to glossary index]
One-tailed test
Statistical test used to detect differences with regard to direction (either greater than or less than not, but not just different). If no prediction regarding the direction of differences is made beforehand, a two-tailed test should be used. For example, if a test was being conducted to determine if advertising had increased ad awareness it would be appropriate to use a one-tailed test.
[go to glossary index]
Open-ended question
Question where respondent can answer in their own words.
[go to glossary index]
Ordinal data
Ranked data but the distance between observations is not necessarily equal. An example is the following scale: Excellent, Very Good, Good, Fair, Poor.
[go to glossary index]
Over quota
Extra respondents who are interviewed on a study after the representative quota has been satisfied. These respondents are added in order to “beef up” their numbers because they represent a low incidence group.
[go to glossary index]
Overrecruit
The extra respondents who are recruited for a survey to compensate for the possibility that some will not complete the survey.
[go to glossary index]
P
Packaging test
Test used to obtain reactions to different packages and label designs.
[go to glossary index]
Participant
A person included in a survey. Also called respondent.
[go to glossary index]
Pearson's correlation coefficient
The most common measure of the strength of the association between variables.
[go to glossary index]
Placement interview
An interview in which a respondent is recruited and given the product to use in a home use test, or a questionnaire to self-complete.
[go to glossary index]
Population
The complete set of individuals, items, or data from which a statistical sample is taken. Also called universe.
[go to glossary index]
Power of the test
Likelihood of rejecting the null hypothesis when it is false. Identified as 1-β.
[go to glossary index]
Predictive dialing
Telephone numbers are automatically dialed by a computer and connected calls are routed to available interviewers. DRI does not take part in this practice.
[go to glossary index]
Pre-test
Initial test to discover any potential problems with a study design.
[go to glossary index]
Probability sample
A sample in which each unit has an equal, non-zero, and independent chance of being selected. Same as Random sampling.
[go to glossary index]
Probing
A follow-up method for obtaining clear and complete responses to open-ended questions. By doing this the number of unclear or ambiguous responses is reduced.
[go to glossary index]
Product concept testing
Early stage testing of new product ideas.
[go to glossary index]
Product placement study
Test where participants are given one or more product samples to use at home under normal usage conditions. Reactions to the products are obtained in a follow-up survey. Same as a Home use test.
[go to glossary index]
Product positioning research
Research used to determine how brands or products are perceived relative to others on key dimensions.
[go to glossary index]
Product pricing research
Research used to measure consumer sensitivity to different prices for a product.
[go to glossary index]
Projectability
The capability of research results to be extrapolated to the target population on the assumption that the sample is representative of the total.
[go to glossary index]
Proportion
The number or percent of respondents answering in a manner compared to the total sample. Same as Study proportion.
[go to glossary index]
Q
Qualitative research
Non-quantifiable method of questioning respondents in small groups or individually to understand why behaviors and opinions are so. Includes: focus groups, dyads, triads or in-depth personal interviews.
[go to glossary index]
Quantitative research
Questioning respondents in large enough groups that data is quantifiable and, if sample has been drawn correctly, projectable to the target population.
[go to glossary index]
Questionnaire
Set of questions to be asked of respondents.
[go to glossary index]
R
Random sampling
A sample in which each unit has an equal, non-zero, and independent chance of being selected. Also known as probability sample.
[go to glossary index]
Random-digit dialing
A telephone sampling procedure that generates random combinations of telephone numbers in order to include listed and unlisted numbers in a survey sample.
[go to glossary index]
Ratio data
Data that can be ranked and ordered with equal distance between each unit. This scale has a natural origin and ratios have meaning, allowing one to say, for example, that a is twice the value of b. Examples include weight, height, age, etc.
[go to glossary index]
Ratio scale
Data that can be ranked and ordered with equal distance between each unit. Examples include weight, height, age, etc.
[go to glossary index]
Regression analysis
A multivariate technique that relates a dependent variable to one or more independent variables.
[go to glossary index]
Replicate
Subsample that has similar characteristics to the total sample.
[go to glossary index]
Representative sample
A sample that is representative of a larger population. Results are projectable to the target population. Also known as probability sample or random sample.
[go to glossary index]
Research design
Basic plan which guides the entire market research process for a particular test.
[go to glossary index]
Respondent database
Group of respondents that have precontacted and screened, then are periodically contacted for survey participation.
[go to glossary index]
S
Sample
A selected subset of a population. Sampling is the process of obtaining a sample.
[go to glossary index]
Sample size
The number of respondents selected to participate in a survey.
[go to glossary index]
Sampling
The process of selecting a portion from a population.
[go to glossary index]
Self-administered questionnaire
A questionnaire completed by the respondent with no interviewer involvement.
[go to glossary index]
Screener
Questions used to identify qualified respondents.
[go to glossary index]
Significant difference
In mathematical terms, difference between tests of two or more variables. The significance difference varies with the confidence level desired.
[go to glossary index]
Skip pattern
Questionnaire logic designed to direct respondents to questions based on previous answers.
[go to glossary index]
Standard deviation
A measure showing how spread out the values are around a particular mean. The larger the value of the standard deviation, the more the values are spread out.
[go to glossary index]
Stub
Responses to the question being tabulated. These descriptions run vertically down the left side of a cross tabulation table, often referred to as rows.
[go to glossary index]
Study proportion
The number or percent of respondents answering in a manner compared to the total sample. Often abbreviated to “proportion”.
[go to glossary index]
Supervisor's instructions
Written directions to the field supervisor on how to conduct the interview along with other pertinent information about the project.
[go to glossary index]
Survey result
Response given to particular questions in a survey.
[go to glossary index]
T
Target population
The population from which information is desired.
[go to glossary index]
Test market
Geographically isolated and representative market used for marketing experimentation.
[go to glossary index]
Test statistic
A statistic calculated from the sample data which is used to test the hypothesis.
[go to glossary index]
Topline
Preliminary results from a project before results are finalized.
[go to glossary index]
Tracking
Studies repeated over time to identify changes in a brand, product category or service.
[go to glossary index]
Triad
Non-quantifiable method of questioning three people simultaneously to understand why behavior and opinions are so.
[go to glossary index]
True population mean
Actual mean response you would get to a given question if the entire population were interviewed.
[go to glossary index]
True population proportion
Response you would get to a given question if the entire population were interviewed.
[go to glossary index]
T-test
A statistical test used to detect differences between two means or one mean and a norm. When testing a mean to a norm and the sample size is greater than 30, a z-test should be performed.
[go to glossary index]
Two-tailed test
Statistical test used to detect differences without regard to direction. If a prediction regarding the direction of differences is made beforehand, a one-tailed test should be used.
[go to glossary index]
Type I error
Rejection of a null hypothesis when it is true. Identified as α.
[go to glossary index]
Type II error
Failing to reject a null hypothesis that is false. Identified as β.
[go to glossary index]
U
V
Validation
Recontacting respondents to verify that an interview was conducted.
[go to glossary index]
Verbatim
A transcript of the actual open-ended comments made by respondents.
[go to glossary index]
W
Weighting
Adjusting data by multiplying it by factors to more accurately represent some total population or sub-group.
[go to glossary index]
X
Y
Z
Z-test
A statistical test used to detect differences between two proprtions (or means) or one proportion (mean) and a norm. If testing a proportion (mean) to a norm and the sample size is greater than 30, a z-test rather than a t-test should be performed.
[go to glossary index]
|
