The role of variability in the property listing task

Abstract

It is generally believed that concepts can be characterized by their properties (or features). When investigating concepts encoded in language, researchers often ask subjects to produce lists of properties that describe them (i.e., the Property Listing Task, PLT). These lists are accumulated to produce Conceptual Property Norms (CPNs). CPNs contain frequency distributions of properties for individual concepts. It is widely believed that these distributions represent the underlying semantic structure of those concepts. Here, instead of focusing on the underlying semantic structure, we aim at characterizing the PLT. An often disregarded aspect of the PLT is that individuals show intersubject variability (i.e., they produce only partially overlapping lists). In our study we use a mathematical analysis of this intersubject variability to guide our inquiry. To this end, we resort to a set of publicly available norms that contain information about the specific properties that were informed at the individual subject level. Our results suggest that when an individual is performing the PLT, he or she generates a list of properties that is a mixture of general and distinctive properties, such that there is a non-linear tendency to produce more general than distinctive properties. Furthermore, the low generality properties are precisely those that tend not to be repeated across lists, accounting in this manner for part of the intersubject variability. In consequence, any manipulation that may affect the mixture of general and distinctive properties in lists is bound to change intersubject variability. We discuss why these results are important for researchers using the PLT.

How to carry out conceptual properties norming studies as parameter estimation studies: Lessons from ecology

“…In previous work (Chaigneau, Canessa, Barra, & Lagos, 2018), we have argued that weeding out low-frequency properties reduces data variability. Because much information in CPNs is carried by variability, rather than cleaning noise from data, weeding out may in actuality be throwing away relevant information.…”

“…However, there are two related reasons that suggest this is not a good practice. First, there is evidence that there is substantial variability in the PLT data between individuals, and within the same individual across time (Barsalou, 1987;Chaigneau, Canessa, Barra, & Lagos, 2018). When feature overlap is researchers' main goal, and property frequency distributions are pruned, part of this variability is lost (a similar argument in favor of retaining the long tails of property frequency distributions can be found in De Deyne, Navarro, Perfors, Brysbaert, & Storms, 2019).…”

Developing And Calibrating An ABM Of The Property Listing Task

“…Given that Equation (6) reflects only part of the relation between s1 and k1, i.e., other processes may also intervene in that relation, we can use Equation (6) as the guiding functional form in a regression analysis to see whether it empirically holds. Indeed, regression equations like Equation (6) explain on average almost 50% of the variance in k1, and parameters a and b are statistically significant across many concepts (Chaigneau, Canessa, Barra and Lagos 2017). Thus, if our ABM models the PLT, the corresponding ABM´s outputs should also show that relationship.…”

Modeling stereotypes and negative self‐stereotypes as a function of interactions among groups with power asymmetries

“…In our model, this occurs because agents that interact more, have higher chances of experiencing lack of connectedness, such that their stereotypes and self-stereotypes become progressively less valuable in communication. Given that we know that concepts in general, and stereotypes in particular, vary across individuals (Barsalou, 1993(Barsalou, , 1987Chaigneau et al, 2017;Lyons & Kashima, 2001), we believe this mechanism should also be at work in the actual social groups. However, our model also predicts that these effects will be modulated by the probability of connectedness and by the differences in power, which means that contact by itself will not abolish stereotypes.…”

“…In CTS, because there is variability in the way people represent concepts in general (i.e., there is inter-subject variability in conceptual content; Barsalou, 1987Barsalou, , 1993Chaigneau, Canessa, Barra, & Lagos, 2017), the optimal strategy to achieve connectedness is to produce semantic content that is thought to be shared (i.e., the most stereotypical content). This strategy would reduce the probability of reaching disagreement and losing connectedness Lyons & Kashima, 2003).…”