Effects of Schema Expectancy on Recognition
Memory for Real and Virtual Environments

Rena L. Walles
Saint Anselm College


Acknowledgments
Abstract
Introduction
Methods
Results
Implications
References
Relevant Links

 

Acknowledgments

    I would like to take this opportunity to thank all the people who have helped me to stay sane and complete this work.  Professor Flannery, thank you for your help over the past few years in the projects I have done.  I could not have written this without your support and guidance.  Professor Krauchunas, thank you for taking the time to help me with all my virtual reality questions.  I would also like to thank my family for supporting me in so many ways.  And where would we be without good friends?  Thanks to Beth, Angel, Nikki, Mira, my lovely roommates, and my Experimental II class.  Jami, thank you for supporting me and helping me when everything seemed hopeless.  You are my best friend and I could not have asked for anyone better.  To everyone else who has helped me, thank you.  I know I could have never done this alone.

top
 
 

Abstract

Schemas are cognitive shortcuts that enable us to rapidly and efficiently process all the information we have seen before, alleviating the need to reprocess an object each time we see it (Hess & Flannagan, 1992).  The purpose of this experiment was to examine whether or not schemas operate differently in real versus virtual environments.  Twenty-four undergraduate students from a small liberal arts college in the northeast participated for course credit.  Each participant completed a recognition task and a short questionnaire regarding memory strategy.  Contrary to the initial hypothesis, there was no significant interaction between testing environment and recognition scores by question type.  There were, however, main effects for both variables.  Participants in the real world condition performed significantly better on the recognition task than participants in the virtual world condition.  Despite this difference, follow up analyses showed results that strongly support schema theory irrespective of environment type.  This study helps to clarify contradictions introduced by previous research (Brewer & Treyens, 1981; Pezdek, Whetstone, Reynolds, Askari, & Dougherty, 1989), and gives insight into the future of testing in virtual reality environments.  Future studies should aim to replicate this paradigm of multiple environments with additional focus on certain variables such as the type of memory being tested.

Key terms: schema theory, memory, virtual reality

top
 
 

Introduction

    Whether we are consciously aware of it or not, memory and schemas are part of our everyday existence.  From the time we wake up in the morning, until the moment we fall asleep, our brains are hard at work sorting out the continuous streams of information we receive.  When you consider the amount of information we encounter at any given moment, it is amazing we can still function.
    One main reason we are able to continue processing this information is through the help of schemas.  Schemas are cognitive shortcuts that enable us to rapidly process all the information we have previously encountered (Hess & Flannagan, 1992).  For example, each time you enter a classroom you do not have to examine and encode the features of the room such as desks, chairs and blackboards.  We come to expect these things because throughout our lives we have acquired schemas of them.
    Out of the current research have come two conflicting views of how schemas affect memory.  First is consistency theory, which states that we will focus our attention on inconsistent objects and therefore remember them more accurately; and the opposing view, schema theory, which asserts that we will encode and therefore better recognize items that are consistent with our schema (Pezdek et al., 1989).  Consistent items are items we would normally associate with a certain environment.  Inconsistent items are items that do not appear to belong in a certain place.
    Brewer and Treyens (1981) hypothesized that schemas determine which objects are encoded into memory, and therefore believed schema consistent objects would be more accurately recalled and recognized in subsequent memory tasks.  The researchers designed a task in order to determine if schemas would guide memory for an office when participants were not specifically told their memory would be tested (i.e. incidental learning).  The researchers found both schema expectancy and saliency were positively correlated with recall and recognition measures.  Brewer and Treyens’ results show strong support for their hypothesis, thus supporting schema theory in real world environments.
    Conversely, Pezdek and his colleagues (1989) hypothesized that participants would spend more time focusing on schema inconsistent objects and therefore recall and recognize more of these objects in later memory tasks.  This phenomenon is labeled the consistency effect.  Participants were brought into one of the environments, either an office or a preschool classroom, and instructed to pay careful attention to their surroundings because they would be tested at a later time.  In both conditions, participants correctly recognized more inconsistent items than consistent items.  The researchers posited several explanations as to why this phenomenon occurred, including the theory that inconsistent items would "stick out" more and may even be held in working memory longer.  This research supports the premises of the consistency effect, and provides evidence for the veracity of consistency theory.
    Virtual reality was chosen as a testing medium in order to examine the efficacy of VR as a psychological testing tool.  Gamberini (2000) focuses on the use of virtual reality technology to assess specific memory abilities.  Using virtual reality, experimenters are able to bring their participants to environments that are best suited to the research at hand and that may be impossible to create in the real world.  Furthermore, new technology allows participants to feel as if they are truly in the virtual environment.  These new products have had a tremendous impact on the field of psychology, and have taken experiments to a new level of control and sophistication.  If this experiment shows real and virtual worlds to be similar in their effects on memory, then researchers will be able to have more confidence using VR and generalizing their results to real world settings.

top
 
 

Methods

    Twenty-four students from an introductory psychology course at a small liberal arts college in the northeast participated in this study in exchange for course credit.  Twelve participants were in the virtual reality condition, and 12 were in the real world condition.  Participants were randomly assigned to groups.
    Participants in both conditions were initially told they would be participating in a virtual reality maze navigation experiment.  The real world participants were then instructed to wait in an office while another student purportedly finished the navigation experiment.  After 20 seconds, the researcher returned and took the participant into a separate room.  The researcher then informed the participant that the real purpose of the experiment was to test for his or her memory of the objects in the office, and gave the participant the recognition task.  They received one point for each correct response in the respective category.  Participants were also required to rate their confidence level for each item on a three point Likert scale ranging from (1) pure guess to (3) complete confidence.  After completion of the recognition task, a demographic questionnaire created by the experimenter was given to the participant.
    Participants in the virtual reality condition followed much the same course of events as those in the real world condition, except they were allowed to practice navigating in a blank virtual world before beginning the actual experiment.  Participants were told that this was practice for the supposed virtual navigation experiment.  The virtual worlds were created using SuperScape technology.
    The real and virtual worlds were identical, containing the same objects and scaled to the same dimensions.  Before any participants were run in the experiment, each object was rated on a Likert scale of one to five for schema expectancy and saliency by a group of independent raters.  Items chosen for the experiment were derived from these ratings.

top
 
 

Results

     A 2 (world) x 4 (question type) mixed model analysis of variance (ANOVA) was conducted for recognition scores.   The results for the ANOVA indicated a significant main effect for world, F(1,22)=6.06, p<.05, as well as a significant main effect for question type, F(3,20)=29.57, p<.001.  Participants in the real world condition (M=73.44, SD=10.01) were significantly more accurate than those in the virtual world condition (M=64.06, SD=8.58).  There was no significant world by question type interaction effect, F(3,20)=1.12, p>.05.
    Follow-up paired sample t-tests for accuracy based on item type indicated significance in four out of six pairwise comparisons after controlling for error rate using the Bonferroni procedure.  Recognition scores were significantly higher for inconsistent distracters (M=92.71, SD=8.17) than for consistent items (M=74.48, SD=14.50), t(23)=5.05, p<.001, inconsistent items (M=60.42, SD=23.22), t(23)=6.14, p<.001, and consistent distracters (M=47.40, SD=29.25), t(23)=7.48, p<.001.  Scores for consistent items were significantly higher than scores for consistent distracters, t(23)=3.74, p<.001.

top
 
 

Implications

     With respect to the recognition scores, the hypothesis that there would be an interaction effect was not confirmed.  There were, however, main effects for both group and object typeThese findings, though not expected, have several important implications.
     First, the real group performed significantly better overall than the virtual group irrespective of item type.  This finding has important implications for the future of VR as a testing tool.  If memory tasks are found to be more difficult in virtual environments, then it will be inappropriate for researchers to generalize findings to real world environments.  Wilson (1999) admits there are differences between real and virtual worlds in testing, but says if measures are taken to make the worlds as similar as possible then VR can still be utilized and results may be generalized.  For example, it may be true that people perform better in real environments, but the overall pattern of results is similar in both real and virtual worlds.  In this case the results from virtual experiments could be generalized with the understanding that accuracy may have been slightly reduced due to the nature of the testing environment.
    Despite main group differences, the results from both environments supported schema theory.  Schema theory proposes our schemas are activated when we enter a well known place, and therefore guide our memory for these places (Pezdek et al., 1989).  The similar findings may be due to the fact that the real and virtual worlds were specifically created to be identical (Wilson, 1999).  More research must be done in order to explore the contradictory findings reported by Brewer and Treyens (1981) and Pezdek et al. (1989).
    Some ideas for future research include conducting an experiment which utilizes more than one testing environment, more than one type of memory task, and compares and contrasts incidental and intentional memory.  Also, additional experiments should be conducted in order to further explore the differences and similarities between testing in real and virtual environments.

top
 
 

Selected References

Brewer, W., Treyens, J. (1981).  Role of schemata in memory for places.  Cognitive Psychology, 13,
    207-230.
Gamberini, L. (2000).  Virtual reality as a new research tool for the study of human memory.
    Cyberpsychology & Behavior, 3, 337-342.
Hess, T., Flannagan, D. (1992).  Schema-based retrieval processes in young and older adults.  Journal of
    Gerontology: Psychological Sciences, 47, 52-58.
Lampinen, J., Copeland, S. (2001).  Recollections of  things schematic: Room Schemas Revisited.
    Journal of Experimental Psychology: Learning Memory & Cognition, 27, 1211-1222.
Pezdek, K., Whetstone, T., Reynolds, K., Askari, N., Dougherty, T. (1989).  Memory for real-world
    scenes: The role of consistency with schema expectation.  Journal of Experimental Psychology:
    Learning, Memory & Cognition, 15, 587-595.
Wilson, P. (1999). Active exploration of a virtual environment does not promote orientation or memory
    for objects.  Environment & Behavior, 31, 752-763.

top
 
 

Relevant Links

Schema Related Information: Background information on the function of schemas
Schema Theory: Introductory explanation of schema theory
The Memory Page: Games and tutorials concerning memory
Virtual Reality Resources: Listing of VR resources, guides, and software developers

top










  E-Mail Me!
 
 

      Return to Psychology Department Homepage         Return to Senior Thesis Homepage

Return to Top