Symposium: Mental Architecture Model Identifiability Approaches

We will present the Modular Serial-Parallel Network (MSPN) model, a comprehensive and unified theoretical framework for cognitive and perceptual processes across various behavioral domains. MSPN has the potential to generalize to cognitive neuroscience modeling and offers a detailed mechanistic analysis of mental processes involved. In the back end, MSPN synthesizes several perceptual and cognitive approaches, including memory representations, signal detection theory, rule-based decision-making, mental architectures, random walks, and process interactivity. The MSPN model has been applied to two domains to explore the hierarchical nature of mental representations. Firstly, in face perception, MSPN proposes a hierarchical organization of visual processing with low-level features processed first, followed by higher-level features, which is consistent with the two dominant approaches in facial perception: holistic and analytic facial encoding. Also, this is consistent with the idea that mental representations of faces are organized hierarchically. Secondly, in decision-making involving preferential gamble choices, MSPN proposes a similar hierarchical organization of processing, with low-level object attributes processed first, followed by higher-level integration of these properties, which is consistent with the so-called Heuristic- and Utility based approaches to decision making. Using the joint analysis of choice response time distributions, we compared several candidate stochastic models. The MSPN has shown impressive abilities in fitting choice response time distributions over other models in tested tasks. Thus, implying that MSPN can be used as a tool for further development and refinement of theoretical constructs, with the analysis of the model's parameter values providing insights into distinct properties of perceptual and cognitive processes.

This is an in-person presentation on July 20, 2023 (09:00 ~ 09:20 UTC).

We examined the impact of automation accuracy and task difficulty on human decision-making. We hypothesized that highly accurate aids would improve performance only under difficult conditions, and this effect would be influenced by individual selection history. Using a categorization task, we manipulated automation accuracy (high/low) and task difficulty (easy/difficult) with three types of aids presented in separate blocks or randomly intermixed to 36 participants. We used a capacity measure based on the single-target self-terminating (STST) rule within the framework of Systems Factorial Technology (SFT) to assess decision efficiency. Results showed that high-accuracy aids reduced accuracy and increased RTs compared to unaided decisions, regardless of automation accuracy and task difficulty. Notably, high-accuracy aids provided incorrect answers under difficult conditions, leading to a significant decline in performance. However, the STST capacity results showed that high-accuracy aids had supercapacity processing under difficult conditions in the block design, but not in the mixed design. These findings suggest that effective top-down control is essential to utilize high-accuracy aids to improve decision efficiency when the task is relatively difficult. Our study challenges the resource hypothesis and suggests that individuals may rely more on high-accuracy aids as task demands increase. Furthermore, these capacity differences may imply that participants utilize different decision strategies in terms of mental architecture to integrate current percept and aided information. Our research provides novel insights into the potential benefits and limitations of automated aids for information processing efficiency.

This is an in-person presentation on July 20, 2023 (09:20 ~ 09:40 UTC).

Classical work by Bousefield & Sedgewick in the 1940s and that by McGill in the 1960s applied what amounted to stochastic death processes with exponential interarrival times to the inter-retrieval times of times from long-term memory in free recall of items from a category. The exponential models used predicted increasingly longer inter-retrieval times over time and/or number of retrievals. We were interested in the generality of this phenomenon. Our mathematical investigations employing hazard functions, found that although this type of behavior does indeed, follow from a broad class of death processes, there exist intriguing, if perhaps unusual-in-nature, classes of hazard functions (underpinning the parallel systems) which violate this seemingly natural kind of behavior.

This is an in-person presentation on July 20, 2023 (09:40 ~ 10:00 UTC).

Previous work has demonstrated that any joint model of choice and response time can be represented with a Grice model, that is, a race model with deterministic accumulation functions for each choice and random thresholds. Our research is on framing the space of possible choice-RT distributions in terms of their Grice model representations and particularly leveraging differential geometry to examine parametric models in that space. In this talk, we will examine the concept of selective influence through the lens of the Grice representation and highlight connections with related frameworks, particularly the coactive model.

This is an in-person presentation on July 20, 2023 (10:00 ~ 10:20 UTC).

Researchers suggest emergent features are fundamental to visual processing. Earlier work examined the perception of combined local information in terms of the emergent features (i.e., orientation and proximity). Our current study investigated how those emergent features combine together. To examine this question, we use change detection task. We applied systems factorial technology (SFT), a framework for measuring cognitive processes across multiple sources of information. Findings of coactive indicated people coordinate orientation and proximity together to make decisions. Results in line with parallel or serial processing indicated people process multiple sources of information simultaneously or sequentially before making decision.

This is an in-person presentation on July 20, 2023 (10:20 ~ 10:40 UTC).