“On the conflict between logic and belief in syllogistic reasoning” Research


In this report we will investigate partial replication of Evans study (1983) of “belief – bias effect”.  This project attempts to identify whether people show a tendency to be influenced by the believability of the conclusion when attempting to solve a problem of syllogistic reasoning. Data was put together to the 2×2 ANOWA design within participant. The report concludes that there can be interaction between believability & logicality.


Do people really rely on belief even if something seems to be illogical? In this study, Evans (1983) uses syllogism to investigate the belief bias effect. Syllogism is a scheme of logical deductions, which is sometimes recognised as superficial statement. The name superficial statement was given because from statement separate it that the deduction from syllogism is consistent and is based on assumption. Often but not always in the syllogism scheme occur in two assumptions – mega premise and mind premise and conclusions – if the two assumptions are true then the conclusion is true. In this study the experimenter provides the participants with two pieces of information; the mega and mind premises and conclusion and asked participants if this conclusion followed logically from these two things.

  • Mega premise: A general statement.
  • Mind premise: A specific statement.
  • Conclusion: based on the two premises.

Evans (1983) in his original study “On the conflict between logic and belief in syllogistic reasoning”, used syllogism to investigate the belief bias by manipulating the conclusion. Example:

(1) Valid argument, believable conclusion (NO CONFLICT)


No police dogs are vicious

Some highly trained dogs are vicious

Therefore, some highly trained dogs are not police dogs

(2) Valid argument, unbelievable conclusion (CONFLICT)


No nutritional things are inexpensive

Some vitamin tablets are inexpensive

Therefore, some vitamin tablets are not nutritional

(3) Invalid argument, believable conclusion (CONFLICT)


No addictive things are inexpensive

Some cigarettes are inexpensive

Therefore, some addictive things are not cigarettes

(4) Invalid argument, unbelievable conclusion (NO CONFLICT)


No millionaires are hard workers

Some rich people are hard workers

Therefore, some millionaires are not rich people. (Evans 2003).

As Evans (1983) predicted, that by manipulating the conclusion there will be interaction between believability of concussions and logicality of conclusions. Also there will be a main effect of “belief bias” (participant had a tendency to accept more believable than unbelievable conclusions) and tendency to accept more valid than invalid arguments. The interaction between both effects will be substantial and statistically significant.

The belief bias or confirmation bias is a tendency to search for new information or explanation of the new information in a way that prove one’s assumption and stop information and understanding which disagree with earlier beliefs.

Furthermore in this study we are going to partially replicate Evans (1983) study and we going to give participant consent form (see appendix A). We show them the instruction sheet (see appendix B) and we will give them the data sheet (see appendix C) with thus four different forms of syllogism. We predict that participant will relay on the belief system and there will be interaction between believabity and logicality. The main four variables in this study are believable & logical, unbelievable & logical, believable & illogical and unbelievable & illogical.


Participant answer 8 syllogism (conclusion)

  • Believable & Logical
  • Believable & Illogical
  • Unbelievable & Logical
  • Unbelievable & Illogical

The study design is 2×2 ANOWA within participant.


One hundred and ninety eight participants took part in this study. Each experimenter recruited two participants – (opportunity sample) male and female from the circle of friends or family.

  • Number of female 100
  • Number of male 98
  • Average age 16- 60


Participant sing the consent form see (appendix A) we show them the instruction sheet see (appendix B) the response sheet see (appendix C).


A copy of the consent form (see appendix A) was given to participant and experimenter enquired if they did not have any question and if they were happy with it. If so they were asked to sign the form and start the experiment.

Participants were given an Instruction sheet (see appendix B) and were asked to respond to a series of eight logic problems called syllogisms. After that they were given a short explanation of the definition of syllogism and were asked to state whether a conclusion logically follows from those two pieces of information. At the end of the instruction participants were asked to provide an honest conclusion which should be based solely on the two pieces of information they were given and were asked if they were happy with the instruction. If yes they could start to circle their response. Participants were informed that they should rely solely on the information provided by experimenter.

Furthermore participants were given the data sheet (see appendix C) with four different forms of syllogism, and participants were asked to give an honest answer for each of the following syllogisms which will indicate whether the conclusion follows logically from the premises or not.

Example :

No police dogs are vicious

Some highly trained dogs are vicious

Some highly trained dogs are not police dogs

The conclusion follows logically from the premises YES / NO


The responses were added together to the different questions in order to come up with two main variables. The responses to the questions 1 & 2 was put together to produce 1 variable – believable & logical and the 3 & 4 was place together to produce the 2 variable – unbelievable & logical and the response to the question 5 & 6 was set together to produce the 3 variable and finally the response of the questions 7 & 8 was set together to produce the 4 variable – unbelievable & illogical.

That gives as four variables: Believable & Logical; Unbelievable & Logical; Believable & Illogical; Unbelievable & Illogical.

Descriptive statistic.

Believable & Logical 1.55 Believable & Illogical 1.40
Unbelievable & Logical .98 Unbelievable & Illogical .31

Mean age is 33.70

The result shows that there was significant interaction between believabity and logicality of conclusions and the result support the hypothesis that the participant will rely on the believe system and there will be interaction between believabity and logicality. Also the result match with Evans (1986) result from the original paper and original research of study.
The result of the partial replication of Evans (1986) study shows that people have a tendency to accept any and all conclusions that match up with their systems of belief, without competition or any deep contemplation of what they are actually agreeing with.
At the same time people also tend to reject a statement that does not match with their belief systems, even though these statements may be perfectly logical and perhaps can be possible.
Exceptionally this is true. People ignore the premises/ principle and focus exclusively on the conclusions being drawn.
Also people, who are not knowledgeable, well-informed and well-read in logic as tend to reason by their experience and not by logic.


  1. Evans, J. ST. B. T. The psychology of deductive reasoning. London: Routledge &: Kegan Paul, 1982.
  2. Evans, J. et al. (2001) Necessity, possibility and belief: a study of syllogistic reasoning. Q. J. Exp. Psychol. 54, 935–958
  3. Evans, J. ST. B. T. Linguistic factors in reasoning. Quarterly Journal of Experimental Psychology, 1977, 19, 297-306.
  4. Evans, J.St.B.T. (1999) The influence of linguistic form on reasoning: the case of matching bias. Q. J. Exp. Psychol. 52, 185–216
  • Calendar

    • July 2018
      M T W T F S S
      « Oct    
  • Search