Causal network analysis validation using synthetic recall protocols

Abstract

Subjects read and recalled 12short texts in a memory recall experiment. The order in which subjects recalled the propositions in the text was recorded. A causal network analysis of each text was then done in order to determine how the propositions in each text were causally related. In addition, an episodic memory network analysis of each text was done in order to represent the original order of propositions presented to each subject in the experiment. The human text recall data were then ana­ lyzed using a new statistical methodology known as the temporalMarkovfield (TMF)approach, which makes explicit probabilistic predictions about the ordering of propositions in human subject recall pro­ tocols in terms of the causal network and episodic memory network analysis of a given text. Samples from the TMF probability model were then used to generate synthetic protocol data using half of the human subject data. Statistics computed with respect to the remaining half of the human subject data and the synthesized protocol data were qualitatively similar in many respects. Relevant discrepancies between the human protocol data and synthesized protocol data were also identified. The generic model of text comprehension and mem­ ory is based upon the idea that the reader forms a situa­ tion model as a by-product ofthe text comprehension pro­ cess (Van Dijk & Kintsch, 1983). This situation model then plays a dominant role in guiding story recall, summa­ rization, and question-answering processes. The com­ ponents ofa situation model can be modeled as knowledge digraphs (i.e., knowledge directed graphs or semantic networks). One example of such a knowledge digraph is the network generated by a story grammar (Stein & Glenn, 1979). Another closely related text knowledge digraph is a causal network that indicates how the propositions in a text are causally related to one another. It has been shown that statements with more causal connections are more likely to be recalled from memory (Trabasso, Secco, & Van den Broek, 1984; Trabasso & Van den Broek, 1985), included in a summary of a text (Van den Broek & Tra­ basso, 1986), and rated as more important (Van den Broek, 1988). This paper discusses an application of the temporal Markov field (TMF) framework for constructing explanaThe author is grateful to Inah Choi, who assisted in writing the 14 texts used in this study, Dafna Yee,who assisted in collecting the human text recall data and the design ofthe human text recall data experiment, and to Frank Manganaro, who assisted in coding the text recall data. The author would also like to thank Neil Coady and Sean Ragan for their assistance in the development of the causal network analyses of the 12 texts used in this study, and Rebecca Horn and Mark Freeman for their ongoing encouragement, support, and feedback at various stages throughout this project. Also, the author thanks the School of Human Development at the University of Texas at Dallas for the necessary funds and support required to support this research. Correspondence should be addressed to R. M. Golden, Cognition and Neuroscience Pro­ gram, School of Human Development (GR4.1), University of Texas at Dallas, Box 830688, Richardson, TX 75083-0688 (e-mail: golden@ utdallas.edu).