Simulating Heterogeneous Farmer Behaviors
models in the social sciences , Axtell ( 2000 ) describes how agent-based models allow for a better understanding of intractable abstract formulations of social processes particularly those involving individuals changing state over time . Axtell ( 2000 ) concludes that since agent-based computational models are powerful analytic tools with effective presentation technologies , these models may someday be the first line of attack on new problems , with mathematics used later “ only to ‘ tidy up ’ what the agent model has clearly presented to [ be ] a robust feature of the problem .”
Based on all this past work , we deemed an agent-based model to be an ideal way to explore the population-level impact of different outreach programs on citizen response . We chose the TPB put forth by Ajzen ( 1991 ), which we further describe in the section the TPB , as a basis for this exploration as it theorizes that an individuals ’ intention to perform a behavior is a function of their attitude toward the behavior , their perceived normative pressure to perform the behavior , and their perceived behavioral control ( PBC ). Agent-based models have been used quite often in modeling TPB ; however , these models have modeled TPB for very specific behavioral applications . For example , Zhang and Nuttall ( 2007 , 2011a , 2011b ) employed agent-based models ( ABMs ) and TPB to explore consumer purchasing behavior and smart metering technology adoption . Schwarz and Ernst ( 2009 ) used household lifestyle data and an ABM of TPB to explore the social diffusion of water-saving innovations . Richetin et al . ( 2010 ) applied both agent-based modeling and models of goal-directed behavior with TPB to analyze the utility of both types of models in understanding the consumption of soft drinks and the engagement in physical activity . They reached the conclusion that the agent-based modeling technique when used to define agents as psychological processes was more realistic and produced statistically better results than using goal-directed behavior models to study intra-individual decision-making processes . Ceschi , Dorofeeva , Sartori , Dickert , and Scalco ( 2015 ) applied an agent-based model of TPB to examine householders ’ recycling attitudes . Other applications of agent-based models of TPB cover a large range of human behavior , including diffusion of organic farming practices ( Kaufmann , Stagl , & Franks , 2009 ), child maltreatment ( Hu & Puddy , 2011 ), tobacco use ( Verzi , Brodsky , Brown , Apelberg , & Rostron , 2012 ), and energy use ( Mogles , Ramallo-González , & Gabe-Thomas , 2015 ).
There has been some recent criticism of the TPB ( Sniehotta , 2009 ; Sniehotta , Presseau , & Araújo-Soares , 2014 ) that suggest that TPB was not intended to serve as a model of behavior change . With this in mind , and unlike models that have used TPB for studying specific behaviors , our intent is to build a generalized agentbased framework for using TPB to explore any type of citizen behavioral response to various outreach programs . The remainder of this paper describes our approach .
The Theory of Planned Behavior
The TPB was put forth by Ajzen to extend the Theory of Reasoned Action ( TRA ) ( Fishbein & Ajzen , 1975 ). TRA posited that an individual ’ s intention to perform a behavior is a function of one ’ s attitude toward the behavior and one ’ s perceived SN regarding the behavior . To this framework , TPB adds the third dimension of perceived control to perform the behavior . The concept of PBC is similar to the idea of self-efficacy ; that is , an individual ’ s perception of his or her ability to perform the behavior . In theory , all three components of behavior-
200