Condi ional Reasoning: Scena io o con ex e ec s*
Ma ía Dolo es Valiña, Glo ia Seoane, Sonnya Geh ing,
Ma ía José Fe aces & José Fe nández-Rey
Uni e si y o San iago de Compos ela (Spain)
INTRODUCTION
The abili y o sol e p oblems in ol ing condi ional ela ionships ( g. “i
p... hen q”) has a cen al ole no only in he e e yday easoning bu also in
he scien i ic in e ence. The aim o his pape is conce ned wi h con ex
e ec abou condi ional human easoning. The unde s anding o his e ec
is a p incipal phenomenon o be discussed i we a e in e es ed o in es iga e
how people consis en ly do eason in he daily li e.
The ques ion o wha is he impo ance o scena io o he con ex in
e e yday easoning may be answe ed in many di e en ways and om many
di e en pe spec i es. The e a e se e al heo ies ha a gue people use
in e nalized in e ence logical ules (B aine, 1978; Rips, 1983, 1984; B aine
& cols., 1984). These app oaches ha e p oposed a syn ac ic iew o human
easoning.
_____________
* This pape was p esen ed a he FIFTH CONFERENCE OF THE EUROPEAN
SOCIETY FOR COGNITIVE PSYCHOLOGY - ESCOP, celeb a ed in Pa is, Socie é
F ancaise de Psychologie-UNIVERSITÉ DE PARIS-SUD (Sep embe , 12-16, 1992).
In con as , o he au ho s ha e p oposed ha people eason using
p agma ic easoning schemas o abs ac knowledge s uc u es induced om
o dina y li e expe iences, as pe missions, obliga ions, and causa ions
schemas (Cheng & Holyoak, 1985, 1989; Cheng, Holyoak, Nisbe , &
Oli e , 1986).
Cheng & cols' in es iga ions p o ide suppo o he iew ha people
o en eason using knowledge s uc u es o in e media e le el o abs ac ion
and con ex sensi i es, a he han pu ely syn ac ic ules o domain-speci ic
expe ience ha cued ele an solu ions in memo y (G iggs y Cox, 1982;
G iggs, 1983).
Mos o he esea ch o he pas wen y yea s examined p agma ic
easoning in Wason's selec ion ask (Wason, 1966, 1968; Wason & Johnson-
Lai d, 1972). The p e ious s udies examine many expe imen al
manipula ions, and his expe imen al ask has gene a ed a la ge li e a u e
(see G iggs, 1983, E ans, 1989, Johnson-Lai d, 1991 o a e iew). A
acili a ion con en e ec was i s epo ed (Wason & Shapi o, 1971;
Johnson-Lai d, Leg enzi, & Sonino Leg enzi, 1972) bu subsequen esea ch
ailed o eplica e he hema ic e ec (see G iggs, 1983 and Wason, 1983,
o a e iew). Du ing he pas decade he conclusion abou he "elusi e
hema ic ma e ial e ec s" in Wason's selec ion ask (G iggs y Cox, 1982)
was ha he pe o mance on hema ic p oblems can be acili a ed, bu his
acili a ion was "mo e limi ed han o iginally hough and b ough abou by
a e y di e en mechanisms han o iginally p oposed" (Jackson & G iggs,
1990, p. 354).
Du ing he 90's Jackson & G iggs (1990) also ound he p agma ic
easoning schemas e ec o be e y "elusi e". They examined he p agma ic
easoning schema heo y o deduc ion easoning and es ablished ha he
inding o acili a ion on abs ac e sions o he selec ion ask is no he
esul o e oking p agma ic easoning schemas, bu i was dependen upon
wo p esen a ion ac o s. They conclude " he esul s o he p esen
expe imen s mi o hose obse es h oughou he pas 20 yea s o esea ch
on he selec ion ask in ha a esul can be changed d ama ically by only a
sub le change in p oblem p esen a ion" (p. 371).
In con as , his s udy explo es he p agma ic easoning schemas e ec in
e e yday condi ional easoning, bu eal-wo ld easoning p oblems, ha
hold in e es o he subjec s, we e used. The condi ional a gumen s we e
included in na a i e ex s; hese ex s ep esen ed scena ios o eal wo ld.
P e ious in es iga ions abou syllogis ic in e ence (Henle, 1962; E ans,
Ba s on & Polla d, 1983; Valiña & De Vega, 1988) and condi ional
easoning (Pipe , 1985) used na a i e ex s.
The expe imen epo ed he e ha e been designed o explo e he e ec s o
he ype o con ex (causal and p omises/ h ea s), he cong uence o he ex
and he ac ual ela ionship be ween an eceden and consequen on he
condi ional e e yday easoning.
METHOD
Subjec s
120 s uden s (75 emales and 45 males; mean age 17 yea s 1 mon h) a he
college le el, in San iago de Compos ela (Spain) pa icipa ed in his
expe imen .
The subjec s ac ed as unpaid olun ee s. They ha e no ecei ed any
ins uc ion in logic.
Design
A 2 x 2 x 3 x 4 ( ype o con en x cong uence x deg ee o ac ual
ela ionship x ype o ule) ac o ial design wi h epea ed measu es on he
las wo ac o s was used. The ype o con ex (causal and p omises/ h ea s)
and he cong uence be ween he ac ual cong uence explici in he s o y and
he logic conclusion (cong uen s. incong uen ) we e manipula ed be ween
subjec s, wi h 30 subjec s andomly assigned o each expe imen al g oup.
The las wo ac o s we e deg ee o ac ual ela ion be ween an eceden
and consequen o majo p emise (de e minis ic, p obabilis ic and wi hou
ela ion) and ype o condi ional ule (MP, MT, AC and DA).
The dependen measu es we e: (a) co ec answe s o p oblems o
condi ional in e ence and (b) subjec s’ ce ain y in he answe selec ed.
Ma e ial
Eigh a ia ions o he 24 condi ional easoning es we e used. Tes s
bookle s we e cons uc ed, each con aining one ins uc ion page, and he
condi ional easoning p oblems p esen ed on sepa a e pages.
The e we e 24 condi ional a gumen s o each o wo di e en con ex s:
(a) Causal ( g. “i you a e on die hen you will lose weigh ”) o (b)
P omises/Th ea s ( g. “i she is no ag ee wi h his p essu es hen he will lose
he wo k”). Each bookle was p epa ed con aining condi ional a gumen s
acco ding o he ype o logical ule (MP, MT, AC and DA) and he
p obabili y o ac ual ela ionship be ween an eceden and consequen o
majo p emise (de e minis ic, p obabilis ic o wi hou ela ion). These
condi ional sen ences we e based on p e ious expe imen al wo k (see Valiña
Seoane, Ma ín, Fe nández-Rey & Fe aces, 1992).
In his expe imen he condi ional easoning p oblems we e p esen ed in
he con ex o na a i e ex s. The p ocedu e was he same as ha used in
p e ious expe imen al in es iga ion abou syllogis ic easoning (Valiña &
De Vega, 1988). 48 sho s o ies abou scena ios o he daily li e we e
desc ibed in hese ex s.
The s eps in cons uc ing an e e yday scena ios can be ou lined as
ollows: (1) Fi s he p incipal pe sons and hei si ua ion we e p esen ed
( g. “Ma a, ha is chemis y eache , eaches labo a o y p ac ical classes
o he s uden s”), (2) The “i p … hen q” a e ha condi ional sen ence we e
included in he con ex o hese p e ious si ua ion ( g. “Be o e beginning he
class he no iced he s uden s hey we e o wo k wi h e y dange ous
subs ances, in o de ha hey mus ca e ully pay a en ion he ins uc ions,
and also hey we e ad ised o pu hey labo a o y coa , bu as someone o
hem did no wan do i she no iced hem ha i sulphu y acid spilled on hei
clo hes hen hose we e bu n ”), (3) The “p”, “no -p”, “q” o “no -q” ac ion
is p esen ed ( g. “Du ing he p ac ice he clo hes o a s uden gi l, who had
no wo e labo a o y coa , we e splashed wi h sulphu y acid”). Then he MP,
MT, AC o DA condi ional ules we e de e mined. Finally (4) he
cong uence be ween he ac ual consequence explici in he s o y and he
logic conclusion is manipula ed. A cong uen ending ( g. “The opinion o
ha s uden was changed by he inciden , and she esol ed o wea he
labo a o y coa in o de o he e en did no happen again”) o an
incong uen ending ( g. “The opinion o ha s uden was no changed by he
inciden and she wen on wi h one's wo k wi hou o wea he labo a o y
coa ”) inished he s o y.
Finally, a lis o al e na i e conclusions om which he subjec mus selec
was p esen ed. An a i ma i e, nega i e and non-p oposi ional conclusions
we e p esen ed o each i em.

The leng h and accessibili y o ocabula y o each s o y was con olled.
The o de o he 24 condi ional p oblems was andomized and he i ems we e
p esen ed in his o de and ano he e e se o de .
Task and P ocedu e
30 subjec s we e assigned a andom o each o ou expe imen al g oups.
All pa icipan s we e gi en one e sion o he bookle s. In condi ions (a) and
(b) he subjec s ecei ed 24 easoning p oblems p esen ed as a s o y wi h
causal condi ional p emises (causal con ex ). In condi ions (c) and (d) he
subjec s ecei ed ano he 24 condi ional easoning p oblems wi h
p omises/ h ea s con ex in he p emises (p omises/ h ea s con ex ).
Hal o he subjec s ecei ed a cong uen s o y (g oups 1 and 3) and he
o he hal ecei ed an incong uen s o y (g oups 2 and 4). Wi hin each o he
ou main g oups hal o he subjec s ecei ed he p oblems in a andomized
o de and he emaining ones ecei ed in e e se o de .
Subjec 's ask was assumed ha he wo p emises o each condi ional
a gumen a e ue and o selec om a lis o h ee al e na i es he conclusion
ollowed om such p emises. The subjec s we e equi ed o a e he secu i y
in hei selec ion o he answe ; a i e-poin scale was used.
The expe imen was un in ou expe imen al sessions. The ins uc ions
we e ead o he subjec s, and hey we e asked o ead hem o hemsel es
again. Be o e he es p oblems, subjec s had wo p ac ice i ems using
condi ionals a gumen s.
Ques ions we e solici ed and inally hey we e allowed o wo k a hei
own speed. They we e gi en unlimi ed ime o ca y ou he ask.
Resul s
A) Co ec Answe s
A 2 x 2 x 3 x 4 ( ype o con ex x cong uence x deg ee o ac ual ela ion
x ype o ule) ANOVA wi h epea ed measu es o he las wo ac o s was
compu ed. The e we e signi ican main e ec s o cong uence (F (1, 116) =
21.68); p< .0001); deg ee o p obabili y o ac ual ela ion be ween
an eceden and consequen (F (2, 232) = 36.25; p < .0001), and ype o ule (F
(3, 348) = 29.27; p< .0001). The e we e also signi ican in e ac ion
be ween cong uence x ype o con ex (F (1, 116) = 7.05; p < .0091), ac ual
ela ionship x ype o con ex (F (2, 232) = 7.92; p < .0005), ype o con ex
x ype o ule (F (3, 348) = 6.71; p < .0002), ac ual ela ion x ype o ule (F
(6, 696) = 56.72; p< .0001) and cong uence x ype o ule ( F (3,348) = 14.62;
p < .0001). The e was no main e ec in ol ing ype o con ex .
The pa e n o esul s was he p edic ed by ou hypo hesis. The cong uen
e sion o easoning p oblems p oduced a signi ican ly highe a e o co ec
esponses han he incong uen e sion wi h all con ex s (causal and
p omises/ h ea s).
As i was p edic ed, he condi ional in e ence ask is easie o pe o m
when i in ol es ealis ic ma e ial. Mo eo e , he p obabili y o ac ual
ela ion be ween he wo componen s o condi ional sen ences had e ec s on
subjec s’ pe o mance.
The e we e mo e co ec answe s o p oblems wi h he de e minis ic
ela ion be ween he wo componen s o majo p emises han hose in ol ing
a p obabilis ic ela ion o wi hou - ela ion be ween an eceden and
consequen . These di e ences we e signi ican (Tukey's es , p < .05).
Inspec ion o pai ed compa isons o he ac ual ela ionship x ype o
con ex in e ac ion showed ha he subjec s made mo e e o s in causal
con en p oblems han hose wi h p omises/ h ea s only he de e minis ic
condi ion (p < .05).
The cong uence x ype o con ex in e ac ion was also explo ed.
Pe o mance on he causal con ex was signi ican ly be e han pe o mance
on he p omises/ h ea s p oblems wi h he cong uen e sion bu he Tukey's
es compa isons (p < .05) indica ed also signi ican ly wo se pe o mance on
he incong uen causal con ex han hose wi h p omises/ h ea s con en .
Fo he hema ic condi ional a gumen s, he mean pe cen age o co ec
esponses o MP, MT, AC and DA ules we e 58.18%, 47.78%, 38.45% and
34.85%, espec i ely. The e we e no signi ican di e ences o p oblems
in ol ing AC and DA ules, bu pos hoc compa isons (Tukey's es ) showed
ha condi ional AC and DA a gumen s we e signi ican ly (p < .05) mo e
di icul han he co esponding MP and MT. On he o he hand, MP
a gumen s we e consis en ly (p <.05) easie han MT a gumen s.
The compa ison (Tukey's es ) o he mean wi hin each ype o ule
indica ed ha a gumen s wi hou ac ual ela ionship we e signi ican ly
easie o AC ules. Fo MP, and MT ules a gumen s in ol ing a majo
p emise wi h de e minis ic ela ion we e signi ican ly (p < .05) easie hose
in ol ing a p obabilis ic o wi hou ela ion be ween he componen s.
Finally, he p obabilis ic ela ion showed be e pe o mance o DA
p oblems.
Inspec ion o pai ed compa isons o he cong uence x ype o ule showed
ha he subjec s made mo e e o s wi h he incong uen e sion on he MP
and MT in e ence ules (p< .05). Bu none o o he compa isons we e
signi ican .
Finally, Tukey's es compa isons indica ed ha subjec s make mo e e o s
wi h causal con ex in MT and DA easoning p oblems, bu hey pa icula ly
ind di icul on he MP and AC p oblems wi h p omises / h ea s con ex .
B) Sujec s´Ce ain y
The e was signi ican main e ec o he cong uence (F(1, 216) = 9.36;
p< .0028), deg ee o ac ual ela ionship be ween he wo componen s o
mayo p emise (F(2, 232)= 9.17; p< .0001) and ype o condi ional ule (F(3,
348) = 6.88); p< .0002). I was also signi ican he in e ac ion ype o con en
x deg ee o ac ual ela ion x ype o ule (F(6, 696) = 6.90; p< .0001).
The incong uen i ems esul ed in signi ican ly wo se con idence han
ei he o o he cong uen p oblems. Tukey's es s compa isons showed
ha judgemen s conce ning de e minis ic ela ion be ween he wo
componen s o condi ional sen ences a e mo e con idence han judgmen
conce ning p obabilis ic ela ion o no ela ion be ween an eceden and
consequen .
Inspec ion o pai ed compa isons o he ype o con ex x p obabili y o
ac ual ela ionship in e ac ion showed ha he subjec i e ce ain y in he
co ec answe s is g ea e o he p omises/ h ea s con ex wi h de e minis ic
ela ion; bu he subjec s ha e mo e con idence wi h he p obabilis ic
sen ences on he causal con ex p oblems.
Finally, a inal a ea o be b ie ly examined conce ned he deg ee o ac ual
ela ion x ype o ule in e ac ion. In MP p oblems, subjec s possessed mo e
con idence wi h he de e minis ic condi ional sen ences han wi h he o he
ela ionships. The p obabilis ic ela ions had g ea e con idence sco es wi h
MT and AC p oblems, whe eas subjec s’ ce ain y wi h NA a gumen s