V4 - Marketing Scales Handbook

VOLUME IV

Volumes I, II, and III of this series covered the years 1980-1989, 1990-1993, and 1994-1997, respectively. The fourth volume covers 654 scales that were reported in articles published from 1998 to 2001. As with the earlier books, this one should not be viewed simply as a revision of the previously published material as scales from the first three volumes were not automatically included in this fourth book. In fact, the contents of this volume are predominately new. Only scales reported in the first three volumes that were used again during the 1998 to 2001 period were included in V. IV. Thus, the first three volumes have hundreds of scales not contained here. Given that, V. IV should be viewed as complementing the preceding volumes rather than superseding them.

A key difference in this volume and the previous three is that the decision was made to publish the two largest sections (Consumer Behavior and Organizational Behavior) separately due to their growing sizes, different targets, etc. The volume with consumer scales was completed first and then disaster struck. In brief, it was one of the many victims of Katrina in that a significant amount of material being readied for the organizational behavior volume was lost in the storm or the looting that followed. Reluctantly, we decided not to spend the years necessary to rewrite the material that was lost. Thus, of the two books that were suppose to comprise V. IV, only the compilation with the consumer-related scales was published. It also contains a separate section of advertising-related scales.

There are a few other differences between V. IV and the previous volumes that should be noted. It became clear rather quickly when working on this new volume that many more scales were being published in our field than could be reviewed in the time available. One way to facilitate the review work and reduce the number of pages that must be bound was to condense the number of sections in each review. Providing the scale items as well as information about psychometric quality are always provided but description of the samples in which a scale has been used and the results of its use are more limited than in previous volumes.

We are also more sensitive in this volume to distinguishing between reflective scales and formative ones. The intention of the series has been to focus on the former though some of the latter have been included at times. Only in recent years has our field begun to appreciate the difference between the two. Our rule in this volume was to be a bit more vigilant; if a measure was described as a formative scale it was not reviewed. If it was treated by the authors as a reflective measure with evidence attesting to its reliability (an important characteristic of a reflective measure) then we probably included it except when it was obvious to us that the scale was formative.

As has been true throughout the series, only multi-item scales were reviewed. The first volume in this series reviewed scales with just two-items but since then scales have only been reviewed if they had three items or more. Further, a minimum amount of information was required for a scale to be included, with particular emphasis on scale items and information about reliability. Beginning in this volume, any scale which otherwise met the requirements but had a reliability of less than .60 was not included. Such low reliability implies that the scale is of such tenuous quality that it should not be used again without substantial improvement. Even with the scales that are included here with reliabilities greater than .60, we have commented for many of them that more developmental work is called for to improve their psychometric quality.

Along those lines, readers are urged to take care in the selection of scales. Those without a basic understanding of psychometrics are encouraged to improve their ability to evaluate alternative measures and make a selection. A suggested reading list is provided at the end of the book from which a rich explanation of psychometric issues can be found.

Our rule since the beginning has been to attempt to describe multiple uses of a scale in the same review. The problem has always been deciding when two scales that are not exactly the same should be included in the same review. The simple answer is that they were reviewed together when they appeared to be measuring the same construct and had about half or more items in common. This means that there are many cases where substantially different scales for the same or similar constructs exist (e.g., #335-#339). This rule was very hard to apply with respect to some scales, most notably the semantic differential versions of Attitude-Toward-the-Product/Brand (#59) and Attitude-Toward-the-Ad (#562). Although these two have been the most popular constructs to measure in scholarly marketing research using multi-item scales there has been little agreement on how to measure them, i.e., they both have been measured dozens of ways with dozens of items. Several years of working with the hodgepodge of Attitude-Toward-the-Ad scales led to an initial grouping. A full explanation of the logic and analysis that led to the grouping has been published. Unfortunately, similar efforts to unravel the jumble of scales for measuring Attitude-Toward-the-Product/Brand have not been as successful. Given that, many measures have been written up together because, at least on the surface, they appear to be measuring the same construct in roughly the same way (multiple semantic differentials) with authors cherry picking from an identifiable pool of items. For those reviews as well as others, the set of items used in each study is identified so readers can determine which is the most appropriate for their purpose.

The Table of Contents can be viewed to see a listing of the scales included in V. IV.

Comments: Dr. Gordon C. Bruner II

Office of Scale Research