Roger Vilardaga, Ana Estévez, Michael E. Levin and Steven C. Hayes University of Nevada, Reno

Empirical Support for the Utility of a Three-Step Perspective Taking Model for the Development

of Psychosis Proneness in College Students

Roger Vilardaga, Ana Estévez, Michael E. Levin and Steven C. Hayes

University of Nevada, Renoand

Universidad de Deusto

WC ABCS, Reno NV, June 22, 2010

Key points of presentation• Study conducted in colaboration with Ana Estévez from

the Universidad de Deusto in Bilbao, Spain

• Contextual behavioral approach• DV: Social anhedonia• IVs: Experiential Avoidance, Empathy, Perspective taking• Design: Cross-sectional• Results suggest three processes that together can help us

understand psychosis pronenness and that can be targetted in future ACT protocols for psychosis

Background

Psychosis proneness

• Social anhedonia has been shown to predict psychosis (Chapman et al, 1994; Kwapil et al, 1997; Gooding et al, 2005)

• And it relates to features that are characteristic of schizophrenia:• Social adjustment (Mishlove et al,

1985)

• Lack of friends (Kwapil, 1998)

• Working memory (Gooding et al 2003)

Relational Frame Theory

• Behavioral account of language and cognition

• Verbal stimuli have an impact on human behavior through their participation in relational frames

• Relational frames allow individuals to interact with the world more effectively without the need to experience it directly

• There are multiple kinds of relational frames (i.e., coordination, hierarchy, opposition)

Betrangsour

salivation

citrus

bumpy

lemonade yellow

Limoo

Betrang

1. Mutual Entailment

2. Combinatorial Entailment

3. Transformationof Functions

soursalivation

citrus

bumpy

lemonade yellow

Defining properties of relational frames

Note: Slide borrowed from Ian Stewart, 2008, June.

attention

rushian?

what?

RFT preliminary studies (Villatte et al, 2008, 2009, 2010)

• Deficits in a specific type of relational framing among individuals diagnosed with psychosis and/or high social anhedonia

• RFT can provide:– A more fine-grained analysis of psychosis

proneness– Key targets for the remediation of deficits in

individuals with high social anhedonia and schizophrenia

Perspective-takingDeictic framing

Experiential avoidanceContextual control over deictic functions

Social Anhedonia

I/YouHere/There

Now/ThenI/You Exciteme

ntAffection

Joy

I/You Sadness

Fear

Hate

Contextual control

EmpathyTransformation of functions of deictic relations

3 step model and predictions

+ +

- - +

DRT (Vilardaga et al, 2009) IRI (Davis, 1983) AAQ (Hayes, 2004)

rSAS (Eckblad, 1982)

Development of new deictic protocol with more ecological validity

• We made the wording of trials more appropriate for an adult population: each single trial added a new content:

• We eliminated simples and simplified reversals: – i.e., instead of “if I were you and you were me”, we asked “If you were me” or “If I were you”

and balanced itTRIALS STRUCTURE (total trials = 375)

Complexity REVERSALS DOUBLE REVERSALS

Trial types I-YOU HERE-THERE NOW-THEN HERE-THERE/ NOW-THEN

I-YOU/ HERE-THERE

# trials 75 75 75 75 75

TRIALS CONTENTNOW-THEN “in a month,” “last week,” “ten years ago,” etc.HERE-THERE “Tahoe,” “Los Angeles,” “Singapur,” etc. I-YOU “Marc,” “Maria,” “John,” etc.

Duke is watching the sunset on the rooftop, and Dafney is watching TV in the living room. If the rooftop was the living room, what would Duke be watching?

The sunsetThe TV

18 to go

Example of deictic assessment trial

Now Floyd is digging a hole in Death Valley, and next winter he will be making snow angels in the Alps. If Death Valley was the Alps and now was later, what would Floyd be doing now?

Making snow angels

Digging a hole

1 to go

Example of deictic assessment trial

Improvements in procedure

› Thanks to Ruth Anne Rehfeldt we elaborated on her basic VBA code and created a new automated procedure

› Data was automatically written in a text file:› Accuracy› Fluency (response time)› Mistake latency

› Less social desirability effects

› We solved ceiling effects of previous empathy ratings by adding more empathy questions

Participants

• College students from University of Deusto (Bilbao, Spain): N=110

• Criteria:– Being fluent in Spanish

• Sample characteristics:– 88.2% female– Mean age: 20 (range: 18-32)– Caucasian– Participant’s father: 9% college degree, 20% high school,

25% school diploma, 13% professional school

Predictors of social anhedonia; Sequential multiple regression

Social AnhedoniaR2 ∆F β 95% CI

Step 1 .036 1.865 Gender -.17 [-.308, .018] Age -.09 [-.042, .016]Step 2 .10* 6.914 Gender -.13 [-.270, .052] Age -.10 [-.043, .014] Deictic ability -.26* [.004, .029]Step 3 .15* 5.172 Gender -.06 [-.218, .113] Age -.04 [-.035, .023] Deictic ability -.23* [.003, .027] Empathic concern -.23* [.010, .148]Step 4 .26** 14.599 Gender -.02 [-.172, .139] Age .05 [-.020, .035] Deictic ability -.18† [.000, .023] Empathic concern -.26* [.024, .154] Experiential Avoidance .35** [-.015, -.005]

Note: *p< .05, **p≤ .001, †p< .10

Perspective-takingDeictic framing

Experiential avoidanceContextual control over deictic functions

Social Anhedonia

EmpathyTransformation of functions of deictic relations

Significant (p=.000)26% varianceMedium effect size

Marginally significant in predicted direction

r = .13†

Non significant but in predicted direction

r = .03ns

Significant at 2nd and 3rd step (β = -.23*) but not last (p=.056)in predicted direction

Significant at 3rd and last step (β = -.26*) in predicted direction

Significant at last step (β = .35**)in predicted direction

10%

15%26%

Limitations and future directions

• Early study• Non-clinical population• Cross-sectional design / need stronger methods

---• Test the same model longitudinally and experimentally• Behavioral measures of empathic concern and experiential

avoidance are highly needed• Further basic research to understand the contextual factors

that give rise to each of these behavioral processes• We have collected data from 162 participants responding to

the same measures

ConclusionThis data suggests that in addition to

psychological flexibility, future ACT protocols for psychosis might consider targeting individuals’ ability to engage in deictic framing and their ability to transform its functions.

We recommend this is done with the development of behavioral tasks to train/measure each of these processes.

The essence of the ACT model

is a Contextual Behavioral Strategy

Thanks for your attentionContact information/resources:

A ppt/audio version of this presentation will be posted in the following website:

Association for Contextual Behavioral Science: http://contextualpsychology.org/

Roger Vilardaga, M.A.: [email protected]