Sampling and assessment accuracy in mate choice: A random-walk model of information processing in mating decision

Abstract

Cite this article as: Sergio Castellano and Paolo Cermelli, Sampling and assessment accuracy in mate choice: a random-walk model of information processing in mating decision., Journal of Theoretical Biology, doi:10.1016Biology, doi:10. /j.jtbi.2011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ABSTRACTMate choice depends on mating preferences and on the manner in which mate-quality information is acquired and used to make decisions. We present a model that describes how these two components of mating decision interact with each other during a comparative evaluation of prospective mates. The model, with its well-explored precedents in psychology and neurophysiology, assumes that decisions are made by the integration over time of noisy information until a stopping-rule criterion is reached. Due to this informational approach, the model builds a coherent theoretical framework for developing an integrated view of functions and mechanisms of mating decisions. From a functional point of view, the model allows us to investigate speed-accuracy tradeoffs in mating decision at both population and individual levels.It shows that, under strong time constraints, decision makers are expected to make fast and frugal decisions and to optimally trade off population-sampling accuracy (i.e. the number of sampled males) against individual-assessment accuracy (i.e. the time spent for evaluating each mate).From the proximate-mechanism point of view, the model makes testable predictions on the interactions of mating preferences and choosiness in different contexts and it might be of compelling empirical utility for a context-independent description of mating preference strength.4

Computational mate choice: Theory and empirical evidence

“…on the decision variable) aim at understanding the relationship between signal variation and mating success, and adopt a 'descriptive' approach. In contrast, studies that focus on mating decision rules aim at understanding how females gather and use information on prospective mates for optimal mating decision and, thus, they adopt a 'normative' approach (Castellano and Cermelli, 2011).…”

“…Recently, Castellano and Cermelli (2011) presented a sequential-sampling model of mate choice that can be viewed as a logical extension of the signal detection model of Phelps et al (2006). Sequential-sampling models of decision making are a dynamic variant of signal detection models (Pleskac and Busemeyer, 2010), in that they drop the assumption that decision makers use a fixed sample size of evidence and assume that decision is made as soon as the accumulating noisy evidence reaches a fixed threshold (Castellano, 2009a;Castellano and Cermelli, 2011).…”

“…Sequential-sampling models of decision making are a dynamic variant of signal detection models (Pleskac and Busemeyer, 2010), in that they drop the assumption that decision makers use a fixed sample size of evidence and assume that decision is made as soon as the accumulating noisy evidence reaches a fixed threshold (Castellano, 2009a;Castellano and Cermelli, 2011). Unlike signal detection models, sequential-sampling models incorporate time in the decision process and, thus, they could be used for investigating speed-accuracy tradeoffs in decision making (Chittka et al, 2009;Sullivan, 1994) and for making testable predictions on the relationship between choice probability and time response (Kacelnik et al, 2011;Shapiro et al, 2008).…”

“…Unlike signal detection models, sequential-sampling models incorporate time in the decision process and, thus, they could be used for investigating speed-accuracy tradeoffs in decision making (Chittka et al, 2009;Sullivan, 1994) and for making testable predictions on the relationship between choice probability and time response (Kacelnik et al, 2011;Shapiro et al, 2008). Although these models have been only recently introduced in behavioural ecology (Castellano and Cermelli, 2011;Kacelnik et al, 2011;Shapiro et al, 2008;Trimmer et al, 2008), they have a long history in cognitive psychology (review in Ratcliff and Smith, 2004) and in neurobiology (review in Bogacz, 2007;Gold and Shadlen, 2007;Smith and Ratcliff, 2004) and they could represent a useful theoretical tool for building bridges between neurobiology and behaviour (Busemeyer et al, 2006). Sequential sampling models have been developed to account for simple two-choice decisions, although they have been also used for modeling multi-alternative decisions (Roe et al, 2001).…”

“…Two issues were investigated using this model (Castellano and Cermelli, 2011). The first was optimality in mating decision.…”