This is a new approach developed by J. T. Blacklege and Joe Ciarocchi at the University of Wollongong.
In an August 2005 email J. T. said this:
Joseph Ciarrochi & I (with invaluable suggestions from Steve) have just finished designing two new ACT values questionnaires that borrow elements from Sheldon's Personal Striving assessment form (Joseph discovered Kennon Sheldon's work a while back and it pointed us in a direction we felt might enhance ACT values assessment). One is a full-length form called the Personal Values Questionnaire (which measures all 9 ACT values domains); the second is called the Social Values Survey (which measures only social, family, and couples relationships) that we tailored for a brief intervention with young adolescents.
There is currently no psychometric data for either (they are changed enough that Sheldon's Personal Striving data is largely irrelevant), though we will be validating the SVS on a sample of 8th graders in a few weeks, and validating the PVQ on a university student sample early next year. Please feel more than free to validate these questionnaires on any samply you see fit (just let us know--we'd love to see the data). The format of the questionnaire is close enough to Sheldon's for us to expect the measures to have similarly reasonable psychometric properties, but, of course, who knows until we see data.
We had two primary purposes in mind while we were designing these instruments. First, we wanted to describe each values domain in a way likely to influence subjects to write relatively ACT-consistent values--even if these subjects had not been exposed to ACT therapy. As we all know, ACT talks about values in a different way than the term is usually used--and it's thus hard to expect someone not familiar with ACT to state a value in an ACT-consistent way without interacting with a therapist. We wanted to make it clear to subjects that by value, we are referring to unilateral actions that are likely to lead to increased vitality, meaning, purpose--not static end states that appear implicitly out of one's control. In other words, to avoid getting responses like "I value close friendships", we included prompts like, "What kinds of friendships would you most like to build? If you were able to be the best friend possible, how would you behave toward your friends? For example, you might value building friendships that are supportive, considerate, caring, accepting, loyal, or honest—but choose for yourself which qualities you would most like to bring to your friendships.
Some subtle changes from wording used on previous versions of values questionnaires, but we felt the 'build' theme, along with examples, seemed to provide the kind of prompts that might be more helpful. Second, we wanted to include Likert-scale questions that assessed things like how much each stated value might be a function of things like pliance or experiential avoidance. Steve oriented us back toward RFT/rule governed behaviour terms that capture what we were trying to assess: as it stands now, question one under each values domain on the SVS and PVQ assesses pliance, question 2 assesses avoidant tracking, and questions 3 & 4 assess augmentals. There are also items that get at importance of each value, effectiveness in moving toward it, etc. As it stands (using Sheldon's scoring algorithms and common sense), subtracting the sum of items 1 & 2 from the sum of items 3 & 4 would yield a sort of 'value purity' score that tells us to what degree a subject/client values the stated value for the reasons we'd hope for from an ACT perspective (higher positive score = greater 'purity' of the value; negative score means the 'value' is actually a function of pliance and/or avoidance).