Chapter+7

Chapter 7
 * 3.1 - Distinguish among a Likert scale, a lifestyle scale, and a semantic differential scale.**
 * Likert Scale**
 * A synthetic metric scaled response form commonly used by marketing researchers.
 * Respondents are asked to indicate their degree of agreement or disagreement on a symmetric agree-disagree scale for each of a series of statements.
 * “Flat” or plain statements are used so that the respondents can choose whether they strongly agree, agree, somewhat agree, disagree or strongly disagree.
 * Example of a Likert Scale on page 242 of the textbook, table 7.5


 * Lifestyle Inventory**
 * A special application of the Likert question form.
 * Measures a person’s psychographics, which include, activities, interests and opinions with a Likert scale toward their work, leisure time and purchases.
 * Lifestyle questions measure consumer’s unique way of living.
 * Example of a Lifestyle statements on a questionnaire on page 244, table 7.6


 * Semantic Differential Scale**
 * A specialized scaled-response question format, which has sprung directly from the problem of translating a person’s qualitative judgments into the metric estimates.
 * Contains a series of bipolar adjectives for the varies properties of the object under study.
 * Respondents indicate their impressions of each property by indicating locations along it’s continuum.
 * Focus is the semantic differential is on the measurement of the meaning of an object, concept or person.
 * Some of the bipolar words or phrases are “friendly-unfriendly”, “hot-cold”, and “convenient- inconvenient”.
 * Example on page 245, table 7.7


 * 3.2 - What are the arguments for and against the inclusion of a neutral response position in a symmetric scale?**

Inclusion of a neutral option
 * Some respondents have no formed opinions on that item, and they must be given the opportunity to indicate their ambivalence.

Not including a neutral option
 * Respondents may use the neutral option as a way to hide their opinions.
 * Forces the respondents to indicate their opinions or feelings.


 * 4.2- In conducting a survey for the Equitable Insurance Company, Burke Marketing Research assesses reliability by calling back a small group of respondents to re-administer five questions. One question asks, "If you were going to buy life insurance sometime this year, how likely would you be to consider the Equitable Company?" Respondents indicate the likelihood on a probability scale (0%- 100% likely). Typically this test-retest approach finds the respondents are within 10% of their initial response. That is, if respondents indicated that they were 50% likely in the initial survey, they responded in the 45% to 55% range on the retest. The survey has been going on for 4 weeks, and will be 2 more weeks before the data collection is completed. Respondents who are retested are called back exactly one week after the initial survey. In the last week, reliability measures have been different. Now Burke is finding that the retest averages are 20% higher than the initial retest. //__Has the scale become unreliable? If so, why has its previous good reliability changed? If not, what has happened, and how can Burke still claim that it has a reliable measure?__//**

No the scale has **//__not__//** become unreliable. Reliability means that if one were to measure sample after sample from the same population under study, with the same measuring instruments, the results would be the same **//__within a range of statistical acceptance.__//** This also applies to a single respondent. In this case the question itself is the same, the way the question was asked is still the same, but the only thing that is different is that the question was asked again a week later and the respondent’ s answer was slightly different. This slight change in the respondent’s answer is not enough to make the scale unreliable; this is because data in a re-test scenario does not necessarily always have to be the exact same. It can be slightly different as long as it is within a range that is statistically acceptable. Burke’s re-test averages only changed from being within 10% of the respondent’s initial response to within 20%, which can still be considered an acceptable range, and still allows them to claim that the scale has remained a reliable measure.