JUST PUBLISHED: Not Just Pineapple and Water: How do People Integrate Information from Multiple Sources?

When choosing a restaurant for a dinner with friends we need to combine information prior to decision, concerning the location, menu, and price range. Similarly, when crossing a busy road, we sometimes need to integrate information from multiple sources, such as horn sounds and the sight of approaching cars. A recent paper published by myself and colleagues does not tell you which restaurant to choose for your party or how to safely cross the road. Rather, it provides a means for evaluating how people combine information from various sources prior to decision, and implements these tools in the context of a simple laboratory experiment of visual detection.

We can refer to the above means as ‘Cognitive Modeling Tools,’ as they tell us something meaningful about the underlying model that subserves our cognitive system. We cannot observe the cognitive system directly; it is a hypothetical entity, and to be honest I have never seen it myself despite studying it for several years. Thus, our understanding of the kind of information-processing that happens in the brain comes from observing overt behavioural measures. Indeed, the current paper focuses on measures of response times and accuracy that anyone can observe and record.

Our response-time measures (based on the work of Jim Townsend and his colleagues) use the entire distribution and are thus quite powerful. Our measure of accuracy is a novel expansion of Shaw’s No Response Probability Contrast. For brevity, I often refer to this paper as the ‘NRPC paper’.

Cognitive modelling requires some level of mathematical understanding. Luckily, much of the analyses nowadays can be implemented as readily available and easy-to-use computer codes. Relevant Matlab codes are available for free on my website. So, if you are interested in questions of information integration, such as how to combine the wits and the looks of your date, or the taste of water with pineapple please consult the paper, or contact me directly at Ami.Eidels@newcastle.edu.au

For more information, please see the following paper:

Eidels, A., Townsend, J., Hughes, H., & Perry, L. (2014). Evaluating perceptual integration: Uniting response-time- and accuracy-based methodologies Attention, Perception, & Psychophysics DOI: 10.3758/s13414-014-0788-y

JUST PUBLISHED: Decision Making - Monkey See, Monkey Do (But Not Like a Human)

A great deal is known about how we make simple decisions, right down to the way neurons in our brains connect to translate the things we sense into the responses we make. Some of the most important neural studies of decision-making have used monkeys as an analogue for humans. The broader scope of methodology which can be used with primates has provided information far beyond that obtainable from human experimentation. However, conclusions based on animal experiments may not always translate to humans.

Recent work using single cell recordings from monkeys (Bonnet macaques) drew a particularly strong conclusion about human decision making – that all current theories about how decisions are made were wrong, in spite of these theories having decades of support from empirical studies in humans. This conclusion was supported by analysis of how the monkeys responded to instructions to change their decision-making strategy. There was an apparent disconnect between the way the neural recordings from the monkeys’ neurons changed between strategy conditions and the way theories of human decision-making suggested those changes should occur.

In a re-analysis of the monkey data, Peter Cassey, Andrew Heathcote, and Scott Brown came to a different conclusion; the disconnect is not between cognitive and neural accounts of decision making, but between the decision-making strategies employed by humans and those employed by monkeys, even in the same – and very simple! – task. When we allowed both behavioural and neural data to inform our analysis, it was apparent that the monkeys from these experiments were just not making decisions in the same manner as humans.

There is a cautionary tale here - the failure to properly account for inter-species differences, even for processes assumed to be analogous, such as simple perceptual decisions, can lead to incorrect conclusions about the nature of humans. The commonly-used analogy between human and monkeys in simple decision-making appears to break down once strategic choices enter the picture.

For further details, please see the following open-access article:

Cassey, P., Heathcote, A., & Brown, S. (2014). Brain and Behavior in Decision-Making PLoS Computational Biology, 10 (7), 1-7 DOI: 10.1371/journal.pcbi.1003700

JUST PUBLISHED: Context Effects in Decision-Making - Do You Really Want that Extra Large Popcorn?

For those of you have children who watch the "Brain Games" series on Foxtel you may remember an episode about how our decisions can be manipulated by marketing. They described a scenario where movie goers were offered a small popcorn for $2 or a large one for $7. Everyone chose the cheaper option until a third one was introduced, a medium popcorn for $6.50, which caused people to prefer the large popcorn with comments like "We may as well get it because it is only 50c more".

This is an example of a decision context effect, where the introduction of a third option (which doesn't get chosen) causes a change in preference between the original two options. Our paper shows that this context effect, and two others, occur not only in consumer decisions but also in simple perceptual decisions (about the size of rectangles) and so tell us something fundamental about most, if not all, human decision processes.

This work arose out of a visit to the Newcastle Cognition Laboratory in 2012 by Jennifer Trueblood, a graduate student at Indiana, now faculty at UC Irvine. We presently are revising a paper that has been invited for resubmission at Psychological Review, describing a new theory of these effects, the MLBA model, hopefully something we can describe in more detail in a future post!

-----------------------------------------

For more information, please see the following journal artcile:

Trueblood, J., Brown, S., Heathcote, A., & Busemeyer, J. (2013). Not Just for Consumers: Context Effects Are Fundamental to Decision Making Psychological Science, 24 (6), 901-908 DOI: 10.1177/0956797612464241

or email Andrew.Heathcote@newcastle.edu.au