Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

 
2007 Annual International CHRIE Conference & Exposition 
376
mental models of the world to evaluate choices and as context to consider their decisions. The concept of mental 
modeling recently has been used to elaborate how people should manage stress and respond to disasters (Paton, 
2003; Paton & Jackson, 2002). This method approaches a piece of text, such as the answer to an open-ended 
question, as a phenomenon composed of concepts, relationships, and statements. A mental model is a map of the 
network of concepts and the relationships between and among them. It has been found to be a good approach for the 
analysis of open-ended survey responses (Jackson & Trochim, 2002). Concepts are symbols that have meaning only 
through their relationship to other concepts. A set of concepts, for example, a set of single words extracted from a 
text, form a vocabulary or lexicon. Relationships are the links that supply direction, strength, sign, and meaning 
between concepts. When two concepts are linked by a relationship they form a statement. A map of the network of 
concepts, relationships, and statements is the mental model (Carley & Palmquist, 1992) or mental model mapping 
(Kolkman, Kok & van der Veen, 2005). For example, in an open-ended response where a respondent reported that 
he or she would be more attentive while traveling after the events of 9/11, felt psychological fear regarding 
terrorists’ attacks, and noted the possible threat to their own life, the conceptual mapping can be re-constructed into 
a series of linear causal relations: “Psychological factors” (Risk assessment) 
→
“Perception of travel risk” 
→
“Reactive response to perceived risk” (Response style and decision) 
→
“Mitigate such possible risk” (Response 
strategy) 
→
“Alter travel behavior by being more attentive while traveling” (Future behavioral change).
This type of mental mapping of concepts has been seen as a way to blend word-based and code-based 
approaches to interpreting open ended survey questions (Jackson & Trochim, 2002). Supporters argue that this 
analysis strategy shares the strengths of both approaches, minimizes some of their weaknesses, and is particularly 
well-suited to analyze exploratory open-ended questions. Additionally, this approach is more feasible today than in 
the past thanks to software products such as ATLAS.ti (Muhr, 2004; Muhr, & Friese, 2004).

---