The conflicting interaction of belief-bias
and logicality in syllogistic reasoning tasks
The study conducted replicated Evans (1983) experiment to investigate the
presence of believe-bias in syllogistic reasoning tasks, using an equal number of male
and female participants to avoid gender differences in the results. The findings
showed there was an interaction between believability and logicality, suggesting that
dual-processing theories influenced the results. The Implicit and the Explicit systems
interacted together, where there was a conflict between belief-bias and logicality.
Belief bias and confirmation bias are two related phenomena. While belief
bias refers to making a biased evaluation of the evidence that is found, confirmation
bias is a tendency to look for evidence that justifies a prior belief, avoiding conflicting
evidence. There has been a long established connection between belief bias literature
and syllogistic reasoning, where subjects are encouraged to engage in deductive
reasoning, drawing conclusions that follow only from the premises given and
apparently the subjects are not able to distinguish between judgments of validity and
judgments of real world truthful value (Evans & Over, 1996).
Also, there is an idea about human thought that has been around for a long
time, it argues that reasoning is separated into two distinct kinds of cognitive systems
with different evolutionary histories. These systems are referred to as Implicit (system
1) and Explicit (system 2) (Evans & Over, 1996 and Reber, 1993 cited in Evans,
2003), even though some dual-process theorists would rather emphasizing the
functional distinctions between the two systems leaving the relation to consciousness
open (Sloman, 1996 & 2002 cited in Evans, 2003).
Firstly, the Implicit system is explained as a universal form of cognition
experienced by humans and animals. It is a group of sub-systems operating with
certain autonomy (Stanovich & West, 2003 & Stanovich, 2004 cited in Evans, 2003).
This system includes instinctive behaviours, but also includes processes formed
through associative learning produced by neural networks (McLeod, 1998 cited in
Evans, 2003); and its processes are quick, automatic and parallel in nature: the final
product created is the only one posted into consciousness (Evans, 2003).
Secondly, the Explicit system is believed to be uniquely human by most of the
theorists, and is thought to have evolved a lot more recently than the Implicit system.
The Explicit system is slow and sequential, making use of the central working
memory system (Baddeley, 2000 & Gathercole, 2003 cited in Evans, 2003). It has a
limited capacity and operates slowly but it allows abstract hypothetical thinking,
being a distinctly human skill of great importance, which can not be produced by the
Implicit system (Evans, 2003).
In addition, the most extensively aspect studied about deductive reasoning is
the ability to solve syllogisms, the only kind of argument whose properties were well
understood from ancient Greece up to the middle of last century. Syllogisms have two
premises, each of them must be in one of the four forms: A) All A are B, I) Some A
are B, E) No A are B and O) Some A are not B. The letters A, I, E, O, referred as the
“moods of premises”, derive from the first two vowels of the latin words AffIrmo and
nEgO, me (Garnham & Oakhill, 1994).
For example, one of the first biases to be identified derives from studies of
Aristotelian syllogisms using realistic material, where subjects are presented with four
arguments: 1) All the athletes are healthy, some healthy people are wealthy, some of
the athletes are wealthy (believable & logical), 2) All the students are poor, no
Please join StudyMode to read the full document