Marketing Scales xLevel of Scale xCentral Tendency x Types of Scales xAssessing Goodness of Scale
Levels of Measurement of Scales zNominal – scale which partitions objects into mutually exclusive and collectively exhaustive categories; numbers simply used as labels; numbers used to categorize objects.
Levels of Measurement of Scales zOrdinal – scale in which order of objects is known; usually rank-order scales.
Levels of Measurement of Scales zInterval – scale in which distance between each descriptor is known.
Levels of Measurement of Scales zRatio – scale in which a true zero origin exists; the value of zero means that the entity possesses none of the characteristic
Marketing Scales zLikert - assesses degree of agreement/disagreement with statement.
Marketing Scales zSemantic Differential - scale anchors are bipolar adjectives.
Marketing Scales zGraphic Rating Scale - pictures represent descriptors.
Marketing Scales zStapel Scale - uses range of numbers, from –x to +x to indicate intensity of response.
Marketing Scales zPercentage Scale - uses percentages to indicated intensity of response.
Marketing Scales zConstant Sum Scale - respondents need to divide numerical responses in such a way that total score is constant
Construct (Concept) Characteristic of an entity that varies. zManifest – concrete (age, gender, income) (one indicator is usually enough to measure the construct.)
Construct (Concept) Characteristic of an entity that varies. zLatent - abstract concepts that you can’t easily see (variety-seeking, intelligence) (usually need multiple indicators to measure the construct.)
Reliability Degree to which measures are free from random error and therefore yield consistent results.
Measures of a Scale’s Reliability zRepeatability - administer the same scale to the same respondents at two separate points in time.
Measures of a Scale’s Reliability zInternal consistency - if a scale is internally consistent, it should generate roughly the same result by dividing the scale into parts.
Validity Ability of a scale to measure what it was intended to measure.
Validity zFace (Content) Validity – degree to which items appear to measure what they are intended to measure.
Validity zPredictive validity – degree to which some future event can be predicted by a current measurement scale.
Validity zConvergent validity – degree of association among different measurement instruments that purport to measure the same concept.
Validity zDiscriminant validity – refers to a lack of association among constructs that are supposed to be different.
Validity zConcurrent validity – degree to which a variable, measured at the same point in time as the variable of interest, can be predicted by the measurement instrument
If a measure is valid, it has to lack both random and systematic error. Reliability is a necessary but not sufficient condition for validity.