1 of 1

Interactions (con’d)

Item x Instruction

F(2, 82) = 4.13, p = .020

mixed

results

Verbatim & Verbatim = .48 Gist & Verbatim = .37

Verbatim & Gist = .49 Gist & Gist = .47

Verbatim & Mixed = .55 Gist & Mixed = .48

Truth x Item x Instruction F(2, 82) = 5.05, p = .009

mixed

results

​

True & Verbatim & Verbatim = .78 False & Ver & Ver = .18

True & Verbatim & Gist = .79 False & Ver & Gist = .19

True & Verbatim & Mixed = .91 False & Ver & Mixed = .18

True & Gist & Verbatim = .71 False & Gist & Ver = .03

True & Gist & Gist = .91 False & Gist & Gist = .03

True & Gist & Mixed = .91 False & Gist & Mixed = .05

Results

Fuzzy-Trace Theory

​

  • People simultaneously encode verbatim representations of new information (e.g., “People with a BMI of 30 or greater are considered obese”) as well as qualitative gist interpretations (e.g., “A high BMI equals obesity”).1
  • Gist memories are more enduring and easier to retrieve than verbatim memories.2
  • Fuzzy-Trace Theory-based Conjoint Recognition models differentiate between gist-based and verbatim-based memory processes.3
    • Conjoint Recognition builds on--but goes substantially beyond--recollection/familiarity and global memory models conceptually, mathematically, and methodologically.3

Study Design

​

    • Sample
    • N = 84 healthy female adults (♀ = 84 (100%), Mage = 19.22, SDage = 1.53)
    • 68.2% Caucasian, 19.3% Asian, 5.3% African American, 8.3% Hispanic

​

    • Intervention
      • GistFit4 is a Fuzzy-Trace Theory-based obesity prevention intervention that teaches users about nutrition, exercise, and gist comprehension.
      • Sharable Knowledge Objects were used to create a web-based Intelligent Tutoring System that actively engages users in back-and-forth dialogue.

​

    • Memory Test (Within-Subjects Factors)
    • Instruction Type (3)
    • Accept verbatim only
    • Accept gist only
    • Accept verbatim or gist
    • Item Type (4)
      • True and verbatim
      • True and gist-consistent
      • False and verbatim
      • False and verbatim

Selected References

Acknowledgments

Preparation of this poster was supported in part by the National Cancer Institute of the National Institutes of Health under Award Number R21CA149796, awarded to Valerie F. Reyna and Chrstopher Wolfe (Miami University in Oxford, Ohio). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.

1 Reyna, V. F. (2008). A theory of medical decision making and health: fuzzy trace theory. Medical Decision Making, 28(6), 829–833.

2 Reyna, V. F., Corbin, J. C., Weldon, R. B., & Brainerd, C. J. (2016). How fuzzy-trace theory predicts true and false memories for word

sentences, and narratives. Journal of Applied Research in Memory and Cognition, 5(1), 1-9.

3 Brainerd, C. J., Wright, R., Reyna, V. F. (2001). Conjoint recognition and phantom recollection. Journal of Experimental Psychology: Learning Memory and

Cognition, 27 (2), 307-327.

4 Brusr-Renck, P. G., Reyna, V. F., Wilhelms, E. A., Wolfe, C. R., Widmer, C. L., Cedillos-Whynott, E. M., & Morant, A. K. (2017). Active engagement in a

web-based tutorial to prevent obesity grounded in fuzzy-trace theory predicts higher knowledge and gist comprehension. Behav Res Methods, 49(4), 1386-1398.

Purpose

  • Health-centered interventions aimed at promoting healthy decision making require that individuals adequately understand and remember the intervention content, especially information that is related to healthcare risks and benefits.
  • GistFit, a 30-minute computer-based health information intervention, is a prime example of an intervention that conveys decision-relevant health information.
  • Health-centered interventions are designed to inform people about healthcare risks and benefits, but modern memory theories have not been applied to assess what is encoded and retained from these interventions.

Apply Fuzzy-Trace Theory to assess true and false memory for

decision-relevant health information

Conclusions

​

  • Educated, young females encoded both true and “false” (gist-based) memories of decision-relevant health information communicated through

a web-based obesity prevention curriculum called GistFit.

  • Recognition was dominated by gist-based similarity judgments rather than correct identification of presented, verbatim intervention content.

Discussion

​

  • Conjoint Recognition Model
    • The Conjoint Recognition model with B1 = B3 fit the data well. Reducing the number of parameters allows the model to be tested, but did not appreciably affect parameter estimates.
    • Results revealed distinct verbatim and gist memories for health information, with gist-based similarity judgments dwarfing verbatim recollection.
    • Participants produced false (but gist-consistent) memories.

​

    • ANOVA
      • Predominately, participants accepted true (83%) and rejected false items (89%). Acceptance rates were highest in the “Mixed” condition.

Decision-Relevant Memory for Health Information:

A Conjoint Recognition Model Based on Fuzzy-Trace Theory

Valerie F. Reyna*, Julia Nolte*, Robert Rong*, David M. N. Garavito*, Priscila G. Brust-Renck**, & Charles J. Brainerd*

* **

Limitations

​

  • The sample consisted of healthy, young adults. Our results may not generalize to older age adults (as aging influences verbatim and gist memory) or people with prior knowledge about obesity prevention (e.g., patients).

Table 3: Conjoint Recognition Model Parameter Estimates for B1 = B3

Parameter

Parameter Explanation

Estimate

Lower CI

Upper CI

B1

probability of accepting a distractor due to response bias in the verbatim condition

.699

.606

.792

B2

probability of accepting a distractor due to response bias in the gist condition

.786

.623

.948

B3

probability of accepting a distractor due to response bias in the mixed (verbatim + gist) condition

.699

.606

.792

E

erroneous recollection rejection (rejects target based on gist-cued verbatim traces of different target)

.052

–.002

.107

I

identity judgment (accepts a target based on verbatim traces of that target)

.021

–.046

.088

R

probability of rejecting a distractor based on gist-cued verbatim traces of a target

.081

.003

.159

P

probability of distractor acceptance due to phantom recollection

.094

.032

.156

S1

similarity judgment for target

.590

.425

.755

S2

similarity judgment for related distractor

.662

.624

.699

Fit a Conjoint Recognition model to participants‘ memory data

Table 1: Mean Acceptance Rates for Memory Test Items (M and (SD))

​

Table 2: Model Fit for B1 = B3

Instruction Type

Verbatim Items

Gist Items

False Items

​

Goodness-of-Fit

Information Criteria

Verbatim

.78

.71

.10

​

Log Likelihood = –1,671.41

AIC = 16.03

Gist

.79

.90

.11

​

G2 = .03, df = 1, p = .869

BIC = 65.78

Mixed

.91

.86

.12

​

The model does not significantly deviate from the data.

combined into one “false items” type for the Conjoint Recognition model

target

related distractor

Main Effects

Truth

F(1, 83) = 2274.62, p < .001

​

False = .11

​

​

True = .83

​

Item Type

F(1, 83) = 30.06, p < .001

​

Gist = .44

​

​

Verbatim = .50

​

Instruction Type

F(2, 82) = 15.02, p < .001

all

p < .05

Verbatim = .43

​

Gist = .48

Mixed = .51

Interactions

Truth x Item

F(1, 83) = 62.54, p < .001

​

​

​

n. s.

False & Gist = .04

False & Verbatim = .18

True & Gist = .84

True & Verbatim = .83

Truth x Instruction

F(2, 82) = 9.05, p < .001

​

n. s.

False & Verbatim = .10

False & Gist = .11

False & Mixed = .12

True & Verbatim = .75

True & Gist = .85

True & Mixed = .91

Memory Test Items Acceptance Rate – ANOVA (Truth (2) x Item Type (2) x Instruction Type (3))