The Validity and Scalability of the Five-Item Theory of Mind Scale for Use with Toddlers and Pre-Schoolers

Background: The theory of mind literature has had a significant impact in the field of developmental psychology. Most notable is the link between this cognitive construct and typical social development, making theory of mind a key cognitive deficit in autism spectrum disorder (Hughes & Leekam, 2004). Despite its potential importance for typical development, two key issues affect our ability to draw conclusions from the extensive theory of mind literature. One is the continued focus on false-belief as the sole measure of theory of mind, despite this ability not emerging until around four years of age. The second is the lack of empirically validated measures of theory of mind, particularly outside false-belief tasks. These are two challenges that also currently limit our ability to accurately identify potential early theory of mind deficits (before the age one would expect false-belief understanding to exist).

Objectives: Wellman and Liu (2004) proposed a five-item scale said to assess first-order theory of mind development in the pre-school years. While this scale has shown promise in allowing the longitudinal assessment of early theory of mind development, it is yet to be used with children younger than 3.5 years of age. Our research aimed to assess the suitability of this scale for use with children from two years of age. Further to this we assessed the scalability of these five tasks with this younger age range, along with the tools validity against the oft-used Sally-Anne false-belief task.

Methods: Sixty-eight typically-developing pre-school children (ranging in age from 24 to 61 months) were assessed using the Australian version of the five-item theory of mind scale (Peterson, Wellman, & Liu, 2005), as well as a sixth Sally-Anne false-belief task. Abilities measured by the five-item scale are (1) diverse desires, (2) diverse beliefs, (3) knowledge access, (4) false belief, and (5) hidden emotion.

Results: All children showed comprehension of the control question presented in the first task, while the vast majority also showed comprehension of the presented scenarios for tasks two and three. There was also evidence of children showing an understanding of diverse desires and diverse beliefs from two years of age. However, our data failed to replicate the strict scalability of these five tasks both overall and with only the older (over 42 months) or younger (under 42 months) age groups. Despite this, the tool showed good internal consistency and convergent validity against the Sally-Anne false belief task (r = .67, p < .001).

Conclusions: Our results provide initial evidence in support for the usability of this tool as a valid assessment of theory of mind for children from two years of age. However, based on our results we propose that the most accurate ability index should be based on the total items scored correct, rather than the currently used highest item scored correct. Results have clear implications for the ability for research to now empirically establish at what age a theory of mind deficit may be considered problematic, thus potentially paving the way for the early identification of this cognitive deficit.