The multifactorial nature of theory of mind: A structural modelling study Larry Cashion Rachel Dryer Michael Kiernan School of Social Sciences & Liberal Studies Charles Sturt University Bathurst NSW Australia Presented at the 14th Australasian Human Development Association Biennial Conference Perth Western Australia July 2005
Presentation Plan Theory of Mind and Classification Current research study method Age, gender, and the multifactorial nature of theory of mind Conclusions and implications
Theory of Mind The ability to attribute mental states, such as thoughts, beliefs, intentions, desires, and feelings, to others and oneself The ability to perform social and laboratory tasks requiring theory of mind has also been called mentalising and mindreading
Classification in Theory of Mind First-order theory of mind Second-order theory of mind Higher-order or advanced theory of mind
First-Order Theory of Mind Unexpected locations “Where will X look for the object?” Unexpected contents “What does X think is in the box?” Appearance-reality “ What is this object really?”
Second-Order Theory of Mind Ice-Cream Van “Where will X look for Y?” Unexpected locations “Where does Y think X will look for the object?”
Higher-Order Theory of Mind Understanding mental states in motivating actions “Does X mean what she says?” “Why did Y do that?” Reading complex mental states “What is X thinking or feeling?”
Theory of Mind Modularity Theory of Mind Module (ToMM) Leslie (1987; Leslie & Roth, 1993) ToMM neurologically separate from other cognitive and brain systems Minimalist modularity Baron-Cohen (1999) Sub-modules of eye direction detection, intentionality detector, shared attention mechanism
False belief & Theory of Mind False belief unrepresentative of theory of mind in general Bloom & German (2000) False belief as a highly complex cognitive function Bloom & German (2000)
Competing Theory of Mind Models 3-factors 1st-, 2nd- & higher-order ToM Common use in literature 2-factors False belief tasks & other tasks Bloom & German 1-factor Theory of mind module Leslie
Method I Participants 216 school-aged children Recruited from State Schools in NSW & Victoria Years 1, 3, and 5 Screened using a modified version of the Social Communication Profile (Coggins & Olswang, 2001) 2 children eliminated from sample prior to testing No adverse incidents Ethics approval from CSU, and NSW & Victorian Departments of Education
Method II First-order tests Sally-Anne Task (unexpected locations) Smarties Task (unexpected contents) Second-order tests Ice-Cream Van Task Second-Order Sally-Anne Task Higher-order tests Strange Stories Test Faux Pas Test Eyes Test – Children’s Version
Methodological Issues Memory prompts No memory prompts or hints were provided to participants Justification questions Often absent from previous first- and second-order ToM research Makes lower-order tasks more consistent with higher-order tasks Ensures understanding, not just recognition
Hypotheses Significant group differences Older children will perform better than younger children Significant gender differences Females superior to males 3-factor model superior Better fit than 1- and 2-factor models
Data Analysis Categorical data Chi-square ( χ 2 ) Continuous data ANOVA + Tukey HSD Structural Modelling Mplus confirmatory factor analysis
Results I TaskYear 1Year 3Year 5 Sally-Anne Interpretation Justification Smarties Interpretation Justification Ice-Cream Van Interpretation Justification Sally-Anne 2nd-Order Interpretation Justification
Results II TaskYear 1Year 3Year 5 Strange Stories (/8) Interpretation Justification Faux Pas (/10) Total Eyes Test (/28) Total
Results III No gender differences for any task
Results IV Model χ 2 pdf CFI TLIWRMR No correlated terms 3-factor factor factor Sally-Anne Tasks correlated 3-factor factor factor N = 216; all models use WLSM estimation & Santorra-Bentler scaled χ 2
Smarties Sally-Anne (1st-order) Sally-Anne (2nd-order) Ice-Cream Van Strange Stories Faux Pas Eyes 1st Order ToM 2nd Order ToM Higher Order ToM e1 e2 e3 e4 e5 e6 e
Summary of Results Hypothesis 1 – age group differences supported For all theory of mind tasks Hypothesis 2 – gender differences not supported For all theory of mind tasks Hypothesis 3 – 3-factor model significant superiority supported
Implications I Support for the multifactorial nature of theory of mind Fits with current theory and use of ToM Challenge to ‘male brain’ theory of Baron-Cohen No gender differences detected No interaction effects Possible that gender effects were not evident because of prepubescent sample – but still fails to fit theory
Implications II Challenge to current orthodoxy in theory of mind research Assumptions of age – ability development of theory of mind were not supported Knowledge that ‘something’ is going on is different from understanding what that ‘something’ is Instruction sets and ‘memory prompts’ affect the ecological validity of ToM tasks and artificially inflate passing rates
Where Now? Further examination of ‘memory prompts’ and instruction sets Further research into the multifactorial nature of theory of mind using a larger array of tasks Using the 3-factor model to examine the relationship with executive functioning
Contact Details Larry Cashion