Since in the last article we discussed the importance of action standards in product tests, let’s remain with product tests. How does one set up a test design?
It’s quite easy, if one is dealing with just two or three products. In this case, you just test one product after the other; bearing in mind, of course, that you need to rotate the order of products from one respondent to the next in order to prevent positional bias. This procedure of exposing a respondent to a product and having him/her evaluate it before proceeding to the next product is known as a sequential monadic test design.
In practice, a simple rotated test design for three products would look like this:
So, when dealing with few test products, every respondent can test and evaluate every product. However, this becomes impractical when dealing with a large number of test products. In a central location test, you could not expect any respondent to taste and evaluate more than two or three different products. And yet, your client insists on having a large number of products evaluated…
The way out of this bind is a round-robin test design with matched cells, which will allow you to include a large number of products in your test.
Here is an example of a round-robin design:
Although you have only 20 respondents per cell; i.e. combination of tested products, in the end each product will have been tasted and evaluated by 80 respondents. And every possible combination of test products would have been accomplished. Your total sample would be 200 respondents.
Round-robin test designs go hand-in-hand with a matched cell approach. You must ensure that the cells are as similar as possible in terms of respondent demographics; e.g. if you have 10 males and 10 females in the first cell, you should have that precise gender distribution in every cell… Other parameters could include socio-economic class or main brand usage, etc.