siu logo

office of scale research

Department of marketing


Main Content Area

Scale-related Pet-Peeves

Blog #15


Stop Making Scales For These Constructs!

My first pet-peeve lamented the tendency for researchers to recreate the wheel, meaning that far too much effort has been spent making new measures when good ones already exist. In this blog, I will go a step further by giving examples of constructs for which plenty of scales have been developed.

By far, the most multi-item scales in scholarly marketing research have been created for measuring attitude toward the ad and attitude toward the brand/product. It is difficult to estimate the number of alternative measures for each one because so many of them vary just a little, such as by one item or the phrasing of the items. Having said that, there is no doubt that there have been dozens if not hundreds of variations based on my review of 30 years of scales in six marketing journals. Imagine how many there could be if more journals were reviewed!

Beyond those two constructs, here are some others I can readily identify for which enough options are available for most researchers’ purposes. Although this quick sort does not attempt to provide exact counts, I have made two lists. First, some constructs are listed for which at least ten alternative scales exist. Following that are some for which at least five alternatives exist. These are the more generalized measures unless otherwise stated. That means more specific versions of each one can also be found, e.g., website ease of use vs. ease of use (in general). Further, this does not account for measures that have slightly different names but are measuring the same construct, e.g., brand commitment vs. brand loyalty.

Constructs For Which At Least 10 Scales Have Been Developed:

Constructs For Which At Least 5 Scales Have Been Developed:

No doubt, a more thorough review would uncover many more constructs. This is sufficient to illustrate that there are many constructs for which multiple, maybe dozens, of alternative scales have been created. Researchers should NOT create new measures of these constructs until they do two things: 1) examine the available alternatives, and 2) provide compelling reasons for not using one of the available scales. This is part of being a good researcher. Those who do not do this have failed to do their homework and reviewers of their work should take it into account as they consider the papers for publication.