Chap 3

Front 1

abstract system

Back 1

The theoretical constructs used to model marketing phenomena and to make predictions; these can then be compared with empirical data on the phenomena under study.

Front 2

affective component

Back 2

In assessing consumer attitudes toward a product, service, or brand, the component related to emotional response and subjective feelings, such as overall liking and preference.

Front 3

alternative-forms reliability

Back 3

A method to test the reliability of a measure by giving each subject two distinct forms of the measure believed to be equivalent and comparing the results.

Front 4

attitude scaling

Back 4

Operational definitions made to help measure, using specific scales, consumers’ attitudes, beliefs, intentions, and other subjective mental states.

Front 5

behavioral component

Back 5

A consumer’s intention to take action, particularly toward the purchase of a product or service, as well as their eventual behavior.

Front 6

central tendency

Back 6

Any of a number of estimates of the most common of, or the center of, a variable’s values; most common measures are the mean, median, and mode.

Front 7

cognitive component

Back 7

In attitude research, a person’s beliefs about the object of concern; typically focuses on the elements that can be deliberately thought about, not on feelings or behaviors.

Front 8

concurrent validity

Back 8

Correlating two different measurements of the same (marketing) phenomenon administered at the same point in time, primarily to determine the validity of new measurement techniques by correlating them with established ones.

Front 9

consideration set

Back 9

A group of alternatives that a consumer evaluates, thinks about, or makes implicit references to when making a decision.

Front 10

construct

Back 10

Any abstraction used by the researcher based on perceptions of a phenomenon, meant to serve as building blocks for more complex marketing; typical examples include sales, product positioning, demand, attitudes, and brand loyalty.

Front 11

construct validity

Back 11

Relating a construct of interest to other constructs in order to develop a comprehensive theoretical framework for the (marketing) phenomenon being studied; enhanced when the correlation between the construct of interest and the related constructs increases in the predicted manner.

Front 12

content validity

Back 12

Often called face validity, the relation of the appropriateness of the measurement to subjective judgments, typically by experts.

Front 13

contingency tables

Back 13

A table containing numerical observations keyed to two or more (usually categorical) variables that is frequently analyzed using a chi-square test for independence.

Front 14

convergent validity

Back 14

Measuring a construct with independent measurement techniques and demonstrating a high degree of correlation among the measures.

Front 15

counterbalancing

Back 15

A means of preventing the order of questions or listed alternatives from affecting resulting responses by rotating their order among respondents; counterbalancing well can be difficult, so it is usually handled by computer, using an orthogonal design.

Front 16

cross-scale comparability

Back 16

When answers to different questions using the same scale are comparable.

Front 17

discrete choice model

Back 17

A class of statistical models used to analyze the choices made by consumers, or any dependent variable of the “choose exactly one from many” type.

Front 18

discriminant validity

Back 18

The degree to which concepts or variables that a theory says should not be related differ in reality; often assessed by correlation, when the variables involved are interval-scaled.

Front 19

dummy variable

Back 19

A variable that takes on binary values (0 or 1), usually used to indicate whether a condition holds or not.

Front 20

empirical system

Back 20

The set of marketing phenomena under study, often including such quantities as consumer behavior, ad expenditure, competitive tactics, and sales.

Front 21

face validity

Back 21

A method of validating measures, by noting its relationship to another independent measure and whether it accords with a theory’s predictions.

Front 22

heterogeneity

Back 22

A general term for when a group of items, people, variables, or statistically estimated quantities differ; often applied to differences in consumers, particularly in their reactions to the same marketing stimulus.

Front 23

hierarchy-of-effects model

Back 23

A hypothetical model of buyer response, moving from awareness to knowledge, liking to preference, and intention-to-buy to purchase, that is often applied to assess response to advertising.

Front 24

interval scale

Back 24

A measurement scale for which differences in consecutive numbers assigned correspond to constant values in the phenomenon under study (e.g., the difference in the strength of a respondent’s views between 1 and 2 on a seven-point interval scale should be the same as the difference between 5 and 6).

Front 25

linear regression

Back 25

The most common form of statistical model, in which a dependent variable, Y, is related to a group of independent variables, {X1, X2, …, Xk}, as follows: Y = b0 + b1X1 + b2X2 + … + bkXk + ε, where ε represents error, usually normally distributed; the coefficients, {b0, b1, b2, …, bk}, are estimated based on data from a sample.

Front 26

linear transformation

Back 26

Any transformation that involves taking variables or a group of numbers and either multiplying them by or adding to them a constant (e.g., y = a + bx).

Front 27

mode

Back 27

In a list of values, the one that appears the most often, or, equivalently, the highest point on a frequency or probability distribution; if there is more than one such value, the distribution is called multimodal.

Front 28

nominal scale

Back 28

A measurement scale for which numbers are assigned to objects only as labels.

Front 29

operational definition

Back 29

Specifies how a construct is to be measured; defines or gives meaning to a variable by spelling out what the investigator must do to measure it.

Front 30

ordinal scale

Back 30

A measurement scale for which numbers are assigned to objects based on their order (i.e., greater than or less than) or direction.

Front 31

perceptual maps

Back 31

A two- or higher-dimensional visualization of an entire market or category that can be created by working with an explicit set of attributes (e.g., durability and effectiveness) or by implicit or latent scaling techniques (e.g., multidimensional scaling).

Front 32

predictive validity

Back 32

A form of construct validity, also called pragmatic or criterion-related validity, that occurs when a measure can accurately predict what theory says it should.

Front 33

rank-order technique

Back 33

When the respondent is asked to place various objects or items in numerical order with regard to the criterion (attitudes, opinions, beliefs, etc.) in question.

Front 34

ratio scale

Back 34

A measurement scale in which the values assigned to choices or objects possess an absolute zero point, so that both differences and quotients of scale points make sense (e.g., it makes sense to say someone is twice as old as someone else, or lives half as far away, indicating that age and distance are ratio-scaled variables).

Front 35

segmentation

Back 35

The ability to break a market, consumers, or sets of objects into groups; these groups should be relatively homogeneous in terms of the criteria chosen by the researcher (e.g., demographics, behavior, prior experience) and differ substantively across the groups.

Front 36

semantic differential scale

Back 36

A question type in which respondents assess an object in terms of where it falls between two opposed (bipolar) adjectives (e.g., “male–female” or “costly–inexpensive”).

Front 37

split-half reliability

Back 37

A method of testing reliability where a multi-item measurement device is divided into equivalent groups and the item responses are compared among the groups.

Front 38

test–retest reliability

Back 38

Repeatedly measuring the same person or group, using the same scaling device, under similar conditions, and comparing results; the greater the discrepancy, the greater the random error in the measurement process and the lower the reliability.

Front 39

unidimensional measures

Back 39

Measures of some variable that result in a single score on a predefined numerical scale.

Front 40

valence

Back 40

Property of a scale that everyone would agree is ordered in some meaningful way (e.g., everyone would agree that a lower price is more attractive than a higher one, all else being equal, however, people might differ as to whether “traditional” or “contemporary” was superior, so this would not be a valenced scale).