Arguments, Conflicts, and Decisions - Conflict and Multimodal Communication: Social Research and Machine Intelligence - page 118

instance, it seems likely that items from D to G determine C and thus impact on A
and B). Nevertheless, this tentative list of criteria provided useful guidance in
the first two studies devoted to investigate experimentally how arguers make the
decision to engage in argument, as outlined in what follows. 4
The first experimental test of the model was conducted on a sample of 509
undergraduates (of which 473 completed the survey) at a large public Mid-Atlantic
university in the United States. The seven factors listed above were studied in
interaction with two personal variables, verbal aggressiveness (Infante and Wigley
1986 ) and argumentativeness (Infante and Rancer 1982 ), and a situational variable,
consisting in three different topics of discussion: public, private, and workplace
topic (Johnson 2000 ). After compiling the personality scales and providing basic
personal information, participants were presented with a short dialogical scenario,
regarding one of the three topics under study, in which a disagreement with a
close friend was described, suggesting (but not necessarily forcing to accept) the
possibility of engaging in argumentation to resolve it. After reading the scenario,
respondents compiled several multi-item scales, to probe their assessment of the
factors listed above with respect to that situation, as well as their willingness
to engage in argumentation under such circumstances (the dependent variable).
Data analysis was performed using a structural equation model (SEM), connecting
personal, situational, and argumentative variables to intention to engage, to verify
the relative weight and role of each factor in determining the willingness to engage
in argument, and to explore the impact of different situational contexts on the
decision to engage (i.e., what factors are more relevant in different situations?).
Overall, this first study provided partial support to the model and in particular
revealed that:

• Personal variables do not affect consideration of argumentative factors: in

assessing costs and benefits of arguing, people seem to be fairly objective — at
least as far as their degree of argumentativeness and verbal aggressiveness is
concerned.

• The intention to engage in argument is strongly predictable (the final equations

explained most of the variance in the data), but the relevant factors differ across
situations.

• Likelihood of success and contextual appropriateness are always present as

relevant predictors of the decision to engage, and success is by far the most
important consideration.

• More generally, manipulating situational variables proved to be instructive, not

only to highlight different minor predictors but also to make some unexpected
and/or counterintuitive findings emerge, potentially spurring more focused
research (for additional details on this point, see Hample et al. 2011 ).

4 Both studies are published (Study 1: Hample et al. 2011 ; Study 2: Cionea et al. 2011 ), and readers
are referred to those articles for further methodological elucidation.