Web of Words

What is the first thing that comes to your mind when you think of Tide?  

Eliciting and analyzing associations people have with brands is a research practice that goes back decades. Usually, these associations are probed during in-person interviews using a range of projective techniques. These interviews yield rich data, but they take a long time to complete, a luxury we don’t always have. You decide to see what happens if you collect this data in an online survey.

The survey consists of three questions.

  • First, you show the respondent a stimulus, such as a brand logo, and ask “What is the first word that comes to mind when you think of this brand?” You collect the open-ended response with a text box. Let’s say the response is “money”.
  • Then, you pipe in “money” into the next question: “And what is the first word that comes to mind when you think of money?  Once again, you collect the open-ended response. Let’s say the first association with “money” is “travel”.
  • Finally, you pipe in “travel” into the next question: “And what is the first word that comes to mind when you think of travel?”
  • Let’s say the respondent types in the word “freedom”.

As a result, you now have a chain of associations for one respondent: brand – money – travel – freedom.  As you poll more people, new chains emerge: some people may associate the brand with “money”, but for them “money” means “work”, which in turn means “fatigue”. Occasionally, the chains will intertwine: some respondents may product a direct association between “brand” and “travel”, or they will first respond with “freedom” and then with “money”.

The Assignment

Here is an Excel file with a set of such association chains from about 100 people.  The initial stimulus each respondent reacted to was a brand logo. Your assignment is to visualize the connections between words.

You might want to use Gephi, a free open-source software created just for these kinds of jobs, but the choice of tool is up to you. To make the job easier, you might want to clean up the data first. Send us an image or a pdf file with your visualization, or a link to a web page.