UX Magazine

Defining and Informing the Complex Field of User Experience (UX)
Article No. 499 March 4, 2010

Organized Approach to Emotional Response Testing

Most user experience designers will have heard of the Product Reaction Cards (doc), a set of 118 words and phrases developed for Microsoft by Joey Benedek and Trish Miner in 2002 that can be deployed in a user testing workshop to help people articulate their emotional responses to a product.

The Product Reaction Cards are part of the Desirability Toolkit (doc) that suggests facilitators ask users to choose the cards that "best describe the product or how using the product made them feel" and then ask them to narrow their selection to just five cards. The cards selection process is then followed by an interview where the participant explains why they selected those five cards.

Whilst the 118 card deck seems to work for the creators of the PRC, some people think it's too much—I posted a question on UX Exchange a few months ago about and received responses like "unnecessarily fiddly" whilst another said they use a subset of the cards. Donna Spencer, author of Card Sorting, commented:

…at the end of the test the last thing a participant wants to do is go through this big pile of cards. It takes quite a lot of time, but I don't think the gain is worth the pain.

Whilst I support the goals of the cards to prompt people and provide a full vocabulary than might otherwise come to mind during workshop sessions I've been wondering if there might be a different approach.

For example, in the book People Skills, Robert Bolton talks about using adverbs to describe the level of intensity as well as grouping feeling words into "families":

By preceding feeling-word adjectives with appropriate adverbs, you can communicate with some accuracy the degree or intensity of feeling.

You could select adverbs appropriate to the adjective, as in the example Bolton uses:

  • You feel a little sad because your dog died
  • You feel quite sad over your dog's death
  • You feel very sad that your dog died
  • You feel deeply sad since your dog died

Or you could opt for a normalised Likert scale approach that could be applied to any adjective; although that would require participants to explicitly state their opinion of every feeling word, phrased as questions like, "This product makes me feel stressed: Strongly Disagree, Disagree, Neither Agree or Disagree, Agree or Strongly Agree."

It's intensive but it is a more analytical and thorough approach.

The "families" that Bolton refers to is a matrix of categories of feeling words grouped by levels of intensity for example in the category of emotional feeling words for "sadness":

Strong:

  • Desolate
  • Anguished
  • Despondent
  • Depressed

Mild:

  • Glum
  • Blue
  • Sad
  • Out of sorts

Weak:

  • Below par
  • Displeased
  • Dissatisfied
  • Low

This grouping of feeling words by level of intensity, the use of adverbs or a Likert scale, coupled with the Production Reaction Cards authors' recommendation to maintain a 60/40 ratio of positive to negative words should provide you with a better framework should you wish to alter or reduce the list of 118 words and phrases whilst ensuring you still cover the full range of emotional responses.

Think of it like a paint palette where the type of emotion is the hue and the intensity is the brightness. You might not need your 16.7 million colours but if you're going to cull your palette at least take a sensible and logical approach to it.

This is especially important if you want to follow a quantitative approach to reporting on research conducted using the PRC as mentioned in the book Measuring the User Experience by Thomas Tullis and Bill Albert.

What are your experiences with using the Product Reaction Cards—specifically if you culled the list of words and phrases or came up with your own? What technique did you use for developing your custom set of cards and how do you think your choices affected the quality and thoroughness of the emotional response inquiry?

ABOUT THE AUTHOR(S)

User Profile

Nathanael has been working in web app development for 13 years with a background in development before specialising in UX. He has worked mostly in government but has several years experience in e-commerce and e-learning.

Add new comment

Comments

24
22

Great article - I have never used this Product Reaction Card method before, but it reminds me of the Bipolar Emotional Response Testing (BERT) method. With BERT, participants are asked to express their reaction to a product, Web page, etc. using a Likert scale between a series of opposing adjective pairs (i.e. 'Formal'-'Casual'). At the end the participant is asked to think about the 'perfect' product or page and rate that also. The facilitator follows up to ascertain motivations and details and the analysis compares the tested with the 'ideal' ratings.

This method is fairly easy to conduct and analyze and does not tax the participant much or take a great deal of time - I usually add it to the beginning of usability testing sessions as a way to supplement behavioral data with perceptual.

For details on how the BBC used BERT analysis in a recent redesign, go to: http://www.liamdelahunty.com/tips/usability_bbc_redesign_the_glass_wall.php

A variation on this is called the Repetory Grid method: http://www.uxmatters.com/mt/archives/2007/12/the-repertory-grid-eliciting-user-experience-comparisons-in-the-customers-voice.php

Here is a recent related UX Matters site article on this topic:
http://www.uxmatters.com/mt/archives/2010/02/rapid-desirability-testing-a-case-study.php

Jeff