Kirsty Kitto | Research · Attitudes

I am currently seeking postgraduate students to work in this area!
See my Research Students page for more details.

Attitude Change in a Social Context

The opinions of individuals are frequently driven by their internally held attitudes, a construct essential to the field of social psychology. Privately held attitudes play a critical role in people's personal choices about their health, education, social groups, and housing, as well as the importance they attribute to national issues such as the environment, immigration and state security. As a result, attitudes help to determine a wide variety of highly consequential outcomes at a social scale. However, people's attitudes are not static immutable objects, but change in response to persuasion, and attempts to maintain cognitive consistency. We often express different attitudes in accordance with the social context we find ourselves in, and it is frequently the case that an explicitly expressed attitude is quite different from an internally held one. This makes the problem of predicting the attitudes of a society of individuals seem intractable, and yet such predictions would greatly assist with the development of policy solutions acceptable to a population.

I am developing theoretical and computational models of attitude change, and of the manner in which the social context of an individual will affect their expressed attitudes and opinions to many key social concepts. The subsections below describe some of the current areas in which I am working here.


Models of Evolving Attitudes and Self-organising Ideologies

This is the more theoretical aspect of my research in this area. I am trying to model the manner in which the attitudes that form within a society can evolve and change during time. Frequently, we see many of the members in that society adopting extreme stances, or ideologies. Thus, a set of pro vs anti euthanasia opinions might emerge, or liberal and conservative voters, or a gun lobby that argues against a set of people who would like to see tighter gun control laws. Social psychology has learned much about such phenomena, but very few mathematical and computational models of this process of attitude change exist. Together with my collaborator Fabio Boschetti, I have been trying to develop such a model. Essentially we:

  1. Make use of Quantum Decision Theory to define the notion of an agent who has a cognitive state that represents their attitude towards some social issue.
  2. Define a multi-agent system that considers a collection of these agents - a society!
  3. Introduce a notion of different local and global framings of an issue to define two different vector bases (as specified by the frame). These are defined such that:
    • Local frames represent each agent's personal understanding of the issue.
    • Global frames represent an aggregated subset of a number of agent's local frames.
  4. Let the agents make a decision within a particular frame. This represents them expressing an attitude within some social context.
  5. Utilise a notion of cognitive dissonance to motivate the update in agents frames and cognitive states. Essentially, we claim that an agent who has chosen to express a certain attitude within a specific framing of an issue will find themselves in a situation where their cognitive state does not correspond with this attitude. This dissonant state will motivate them to update their cognitive state, and their understanding of that issue (as is represented by the frame).
  6. Keep the agents evolving in time.
Over time we see the agents self organise into groups, all of who share a common set of cognitive states and understandings of the issue at hand. We designate these groups ideologies. Some example scenarios are below, with a video showing the dynamics.
INSERT VIDEOS HERE
You can find the Matlab code that produces these initial toy models here.

Of course we now need to scale up to more complex scenarios, with much higher dimensionality, and more complex structure...

Indicative Papers

More details can be found in these indicative papers. The first in particular gives many more details than the above brief description:

  1. Kitto, K., Boschetti, F. (In Press). Attitudes, Ideologies and Self-Organisation: Information Load Minimisation in Multi-Agent Decision Making. Accepted by Advances in Complex Systems.
  2. Kitto, K., Boschetti, F. (In Press). The Effects of Personality in a Social Context. Accepted for CogSci2013.


Applications

1. The Attitudes of Academics to Multi-disciplinary Research & Collaboration

I am currently running a project that is trying to understand how the underlying personalities of academics relate to their attitudes about research, collaboration, and environment. For example:

  • If you have a workforce that has a high proportion of introverts, then how will they feel about being placed within large open plan environments?
  • Are there styles of collaboration that depend upon personality?
  • Does risk aversion correlate with certain research career choices?

This project will provide me with a large data set which will hopefully help me to start answering questions like these. I am then hoping to start modelling the manner in which a set of academics adjust and evolve within their changing organisational context. At present I am constructing a survey instrument that will be used to gather an initial set of baseline data, not just about the above attitudes, but also about the underlying personalities of the academics who answer my survey. Thus, with a bit of luck I might be able to build a model of the way in which these attitudes change in time using the approach sketched out above.

Now I just have to get a bunch of academics to fill out my surveys :(


2. The Evolving Political Attitudes of the Australian Public

I am also working with Iain Walker & Fabio Boschetti to try and model the evolving political attitudes of the Australian public. We are especially interested in trying to model the attitudes of populations to political issues like climate change. More details soon.