Content uploaded by Glyn V. Thomas
Author content
All content in this area was uploaded by Glyn V. Thomas
Content may be subject to copyright.
Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Behau. Res. Thu. Vol. 30, No. 4, pp. 381-394,
Printed in Great Britain. All rights reserved
ORIGINS OF FEAR OF DOGS IN ADULTS AND
CHILDREN: THE ROLE OF CONDITIONING PROCESSES
1992 0005-7967/92 $5.00+ 0.00
Copyright 0 1992 Pergamon Press Ltd
AND PRIOR FAMILIARITY WITH DOGS
SHARON DOOGAN and GLYN V. THOMAS*
School of Psychology, The University of Birmingham, Edgbaston, Birmingham 915 2TT, England
(Received 13 September 1991)
Summary--One hundred adults and 30 children completed questionnaires to investigate fear of dogs. Dog
fearful adults asked to recall the origins of their fear reported classical conditioning experiences more
frequently than vicarious acquisition or informational transmission. Overall, however, there was no
difference in the frequency of attacks reported by the fearful and non-fearful groups. Significantly more
fearful than non-fearful adults reported little contact with dogs prior to the onset of their fear which
suggests that early non-eventful exposure to dogs may prevent a conditioning event from producing a dog
phobia. Most adults reported that their fear began in childhood, and dog fears were more frequently
reported by children than by adults. In the aggregate, however, dog-fearful adults and children differed
in several ways: children were more likely than adults to report having received warnings about dogs, but
also to recognize the potential attractiveness of a friendly dog. Unlike dog-fearful children, dog-fearful
adults reported many other fears in addition to their fear of dogs. A better understanding of fear of dogs
in adults may depend on discovering why some dog-fearful children, but not others, apparently lose their
fear of dogs as they become older.
Although it is now widely recognized that there are several possible ‘pathways’ to fear (see
Rachman, 1977, 1990) Pavlovian conditioning is still considered the most likely cause of many
phobias. According to this account, previously neutral stimuli are transformed into the objects of
phobic fear by chance pairings with frightening or painful incidents.
Apparently strong support for the importance of conditioning in many phobias comes from the
patients who have reported that conditioning experiences were the initial cause of their fear (see
Merckelbach, de Ruiter, Van Den Hout & Hoekstra, 1989; Ost, 1985; Ost & Hugdahl, 1981, 1983).
The significance of these reports of conditioning experiences, however, is challenged by the
discovery that non-fearful Ss have also often had similar painful or frightening experiences, but
without subsequently developing a phobia. Thus, Di Nardo, Guzy, Jenkins, Bak, Tomasi and
Copland (1988) found that histories of painful and/or frightening encounters with dogs were
equally often reported by Ss with low- and high-fear of dogs respectively.
Even with refinements, such as the concept of prepared learning (Seligman, 1971), a conditioning
model cannot easily explain why a bad experience with a dog should produce fear of dogs in some
people but not in others. If conditioning experiences do in fact play a causal role in the origins
of these particular fears, some additional factor or factors must determine whether or not a
conditioning experience will give rise to longlasting fear in a given individual.
The present study was performed to investigate further the role of conditioning experiences in
the origins of fear, and to identify possible co-factors that might influence the likelihood of a painful
or traumatic incident giving rise to longlasting fear of stimuli associated with that incident. Fear
of dogs is an example of a simple phobia and was chosen for investigation because painful or
frightening encounters with dogs are not uncommon.
As in previous studies (e.g. Di Nardo et al., 1988; McNally & Steketee, 1985), potential
conditioning experiences were divided into S-R experiences (encounters with dogs producing fear
but no physical injury or pain) and S-S experiences (encounters producing injury and pain). We
also asked about experiences relating to vicarious learning or information transmission as possible
routes to fear.
Our search for possible co-factors that might influence whether or not conditioning experiences
produce enduring fear was focussed primarily on the role of prior familiarity (see Rachman, 1990).
*Author for correspondence.
387
388 SHARON DOO~AN and GLYN V. THOMAS
Laboratory studies have shown that prior exposure to a stimulus reduces the ease with which fear
(or any other response) can be subsequently conditioned to that stimulus (Lubow, 1973). This effect
of prior exposure has been termed latent inhibition, and is thought to be a consequence of reduced
processing of familiar stimuli in a familiar setting (Wagner, 1978). If latent inhibition occurs to
dogs as a class of stimuli, we might expect that a painful or frightening experience with a dog is
more likely to produce a dog phobia in S’s unfamiliar with dogs than in Ss with a prior history
of uneventful or pleasurable contacts with dogs. Furthermore, Ss with prior experience of dogs
are likely to know how dogs behave and thus find their behaviour predictable and controllable.
Such individuals might be more likely than inexperienced Ss to discriminate genuinely threatening
behaviour from playful behaviour in dogs, and thus be less likely to acquire fear of dogs in general.
We also investigated the possibility that dog-fearful individuals are generally more prone to be
fearful by asking about fear of other things besides dogs, thus testing for a possible personality
factor responsible for differences in susceptibility to fear conditioning (see Eysenck & Rachman,
1965).
Our sample of adult Ss was made up of undergraduate students because they provide an
accessible pool of Ss in which fears of animals are relatively common (Geer, 1965). As in earlier
studies of animal phobics by Di Nardo et al. (1988) and by Murray and Foote (1979) fear was
assessed by means of Geer’s (1965) ‘Fear Survey Schedule 2’ (FSS-2). We devised a questionnaire
to explore Ss’ fear of dogs, using two pilot studies (n = 8 in each) to check the wording of the
questions and the response categories.
We planned from the outset to supplement the adult study with a survey of fear of dogs in
children, because previous studies of adult phobic patients have found that they often cannot recall
the circumstances from which their fears arose. In view of this memory limitation, Martin (1973)
has argued that children may be better than adults as a source of data on conditioning histories,
because they are closer in time to relevant early experiences. We chose 8-9 yr olds for our survey
of children, because our pilot work indicated that younger children had difficulty answering our
questions about their past experiences with dogs.
METHOD
Subjects
The Ss in the adult sample consisted of 100 undergraduate student volunteers from the
University of Birmingham, mostly studying a social science or one of the humanities. Twenty-eight
were male and 72 were female. Their ages ranged from 18 yr 1 month to 30 yr 8 months with a
mean age of 20 yr 1 month. When recruited to the study Ss were simply told that the survey
concerned fear of dogs.
The child sample comprised 30 children (15 boys and 15 girls), recruited from a school in South
Birmingham. Most were from middle-class English families and their ages ranged from 8 yr 2
months to 9 yr 1 month with a mean age of 8 yr 7 months.
Test materials
Fear in the adult sample was assessed by means of Geer’s FSS-2, consisting of 51 items to which
Ss were asked to rate their fear on a 7-point scale ranging from ‘none’ to ‘terror’. A subset of 8
questions relating to fear of animals, and a 4-point rating scale were derived from the FSS-2 for
use with the child sample.
The questionnaire exploring fear of dogs administered to the adults consisted of three sections.
The first section asked for details of dog-ownership and of dog attacks. The second section
comprised 38 questions asking about other dog related experiences and reactions to dogs. Questions
in this second section required responses on a &point scale ranging from ‘none’ to ‘exceedingly’;
the six response categories were collapsed into two for the purposes of analysis. The third and final
section of the questionnaire was given only to Ss who had ever been afraid of dogs, and asked
about the origins and resolution of their fear.
The questionnaire given to the children consisted of a shortened and simplified version of
Sections 1 and 2 of the adult questionnaire, with the addition of two initial ‘warm-up’ questions
about favourite and disliked animals.
Fear of dogs in adults and children 389
Procedure
Adult Ss were recruited individually on the University campus. They were first presented with
the fear assessment and questionnaire and asked to read through the instructions. The investigator
(Sharon Doogan) then checked that they understood what they were required to do and left them
to complete the answers on their own, The completed forms were then collected later by the
investigator.
The children were questioned individually in their own school by Sharon Doogan in a quiet, and
otherwise unoccupied, room separate from the children’s classrooms. All the children were first
given a modified version of the FSS-2, and then a simplified version of the adult questionnaire.
The response categories were explained to them and probe questions checked their understanding
of the procedure, before the main part of the questionnaire was administered. All the questions
were read out aloud by the interviewer who wrote down the children’s replies.
RESULTS
Fear assessmen
Completed questionnaires were collected from all the adult Ss who had agreed to participate.
Question 45 of Geer’s (1965) Fear Survey Schedule (FSS-2) asked Ss to describe the amount of
fear they experienced towards a strange dog. Ss who answered ‘much’, ‘very much’ or ‘terror’ to
this question were classified as Ss with high fear of dogs. Those who answered ‘none’, ‘very little’,
‘a little’ or ‘some’ were classified as Ss with low fear of dogs. Out of 100 adults who completed
the questionnaire and FSS-2, 75 were classified as having low fear of dogs, and 25 were classified
as having high fear of dogs. The low-fear group of 75 Ss included four Ss who currently reported
no fear of dogs, but who indicated that they had been afraid of dogs in the past.
Each child was asked twice to what extent they were afraid of strange dogs, once in a question
in the abbreviated and simplified FSS-2 at the start of testing, and a second time toward the end
of the questionnaire. Nine children gave conflicting responses to these two presentations of the
question about fear of strange dogs, and were excluded from further analyses. Of the remaining
children, 11 indicated that they had ‘quite a lot’ or ‘a lot’ of fear of strange dogs and were classified
as high-fear Ss. The remaining 10 children reported that they had ‘little’ or ‘no’ fear of strange
dogs and were classified as having low fear.
Fear of dogs and other fears. To compare fear of dogs with other fears, adults’ responses to the
FSS-2 were further analysed into animal and non-animal fears. By assigning numerical values to
the response categories, 0 for ‘no fear’ through to 6 for ‘terror’, we derived a mean fear score for
animal and non-animal fears. Questions 8, 10, 35, 39 and 48 asked about fear of animals other
than dogs (worms, rats/mice, spiders, snakes, insects). The mean animal and non-animal fear scores
for Ss with high fear of dogs were 2.23 (SD = 0.77) and 2.45 (SD = 0.72) respectively. The
corresponding mean animal and non-animal fear scores for adults reporting low fear of dogs were
1.583 (SD = 0.83) and 1.98 (SD = 0.58). A two-way ANOVA on mean fear scores found that adults
with high fear of dogs reported significantly more fear of other situations and things and of other
animals than did adults with low fear of dogs; F(l,98) = 18.02, P < 0.01. Furthermore, there was
overall a significant difference between levels of animal and non-animal fears, F(1,98) = 12.49,
P -C 0.01. The interaction between fear group and fear type, however, was not significant.
To compare fear toward dogs with that toward other animals in the child sample, the children’s
responses to the questions about other animals in the simplified FSS-2 were further analysed by
assigning numerical values to the response categories: 0 for ‘no fear at all’ through to 3 for ‘a lot
of fear’. The mean fear score of the dog-fearful children was 0.78 (SD = 0.45) and that of the
non-fearful children was 0.64 (SD = 0.33). Thus, the dog-fearful children reported on average a
slightly higher level of fear to other animals than did the children with low fear of dogs. A r-test
indicated that this difference was not statistically reliable.
Experiences concerning dogs. In the following analyses of questionnaire responses, chi-square
tests of significance were performed on the numbers of high- and low-fear Ss making each category
of response. Fisher exact probability tests were substituted where expected frequencies in any cell
fell below 5.
390 SHARON DOOGAN and GLYN V. THOMAS
Table I. Numbers of Ss reporting specified past experiences relating to dogs
Adults Children
High-fear Low-fear High-fear Low-fear
n =25 n = 75 ?Z=ll 0 = 10
Bitten by a dog at least once7
Chased by a dog at least once
Frightened by a dog at least once
At least one painfu1 or frightening encounter with a dog
Little or no direct contact with dogs prior to first frightening/painful
experience with one
Mother dislikes dogs
Father dislikes dogs
Observed fear in others
Scared by dog stories and films
Distressed by reports in press of dog attacks
Warned about dogs when a child
Warned about does bv oarents
4 25 1 0
16 32 4 6
12 9’ 8 3’
21 49 10 I
20 out of 21 15’ out of 49 9 out of 10 5 out of 7
15
14
11
8
20
I
34
29
22
4
35*
IIt 7
*Difference between high- and low-fear Ss significant at P < 0.05.
tDifference between low-fear adults and low-fear children si~ific~t at P < 0.05.
$Difference between high-fear adults and high-fear children significant at P c: 0.05.
The first section of the questionnaire sought detailed information about Ss’ experiences with
dogs (see Table 1). Within the child and adult samples there were no significant differences in the
numbers of Ss with low and high fear of dogs respectively reporting having ever been bitten or
chased by a dog. Low-fear adults were significantly more likely than low-fear children to report
having been bitten by a dog (P < 0.05).
Significantly more high-fear than low-fear adults reported a frightening experience with a dog
[x’(l) = 22.32, P c 0.011, but this result may reflect only that any encounter with a dog will be
frightening to a dog-phobic person.
There were no significant differences in the frequency with which high- and low-fear Ss (both
adults and children) reported at least one aversive encounter with a dog (either a painful or
frightening experience). When asked specifically about their experience with dogs prior to their first
aversive encounter with one, significantly more high-fear than low-fear adults reporting such
encounters claimed little or no previous direct contact with dogs [x*(l) = 24.56, P < 0.011. There
were no significant differences between high- and low-fear children in this regard.
Overall, there was little evidence for imitation or vicarious learning of fear of dogs: that is to
say, there were no significant differences in frequencies of high- and low-fear Ss reporting that they
had observed fear in others, or that their parents disliked dogs, or that they had been scared by
dog stories and films. With regard to informational transmission, however, significantly more high-
than low-fear adults and children reported being distressed by media reports of dog attacks
[x*(l) = 13.15, P c O.OlJ. There was no significant difference between the numbers of high- and
low-fear adults recalling warnings about dogs given in childhood. In contrast, dog-fearful children
were significantly more likely than dog-fearful adults to report having received warnings about dogs
(P < 0.05), and also significantly more likely to report having received such warnings than
non-fearful children (P < 0.05).
Behaviour and reactions toward dogs. Data on the numbers of Ss reporting various responses
toward dogs are presented in Table 2. High-fear Ss were si~ificantly more likely than low-fear
Ss to report the typical phobic behaviours of watching for dogs when outside (P < O.Ol), and
avoiding them when possible (P ~0.01).
Furthermore, high-fear adults were significantly more likely than low-fear adults to report fear
of being bitten, a dog jumping up, loud barking, sudden movements by dogs, and dogs snapping,
approaching, or snarling/growling [x2( 1) = 28.127, 38.115, 14.694, 39.56, 50.636, 44.457, 35.295,
respectively; all Ps -c 0.01 J. The child sample resembled the adult sample in that high-fear children
were more likely than low-fear children to report fear of snapping, snarling/growling and sudden
movements (all Ps < 0.05). There were no significant differences, however, between high- and
low-fear children in the numbers reporting fear of biting, barking and of a dog jumping up.
Furthermore low-fear children were significantly more likely than low-fear adults to report fear of
these categories of dog behaviour (P -c 0.05). Among the adults, no-one reported fear of a dog’s
furry/hairy body.
Fear of dogs in adults and children 391
Table 2. Numbers of Ss reporting specified behaviour and reactions toward dogs
Adults Children
High-fear Low-fear High-fear Low-fear
n = 25 n =75 n = II n = 10
Watching for dogs outside 7 2*
Worrying about meeting dogs 8 2’
Making detours to avoid dogs 16 If
Avoiding dogs 4 0
Afraid of being bitten 19 14’ II 7*
Afraid of being infected 3 6
Afraid of dog jumping up 17 6* 1 3
Afraid of barking I 2’ 2 0
Afraid of dog making sudden movements 15 6’ 10 2’
Afraid of dog snapping 22 9’ IO 41
Afraid of dog snarling/growling 20 12; 9 21
Afraid of dog’s furry/hairy body 0 0
Happy to approach friendly dog 10 68* 11 lot
Happy to stroke friendly dog 5 65’ 10 9$
*Difference between high- and low-fear Ss significant at P < 0.05.
TDifference between low-fear adults and low-fear children significant at P < 0.05.
fDifference between high-fear adults and high-fear children significant at P < 0.05.
When asked about their willingness to approach and stroke a friendly dog, there were no
significant differences between high- and low-fear children, nor between low-fear adults and
low-fear children. High-fear adults, however, were significantly less likely to be willing to approach
a friendly dog than were either low-fear adults [x’(l) = 28.05, P < 0.011 or high-fear children
(P < 0.01). High-fear adults were also significantly less likely to be willing to stroke a friendly dog
than were low-fear adults [x*(l) = 39.68, P < 0.011 or high-fear children (P < 0.01).
History of dog fear in adults. Questions on the history of their fear were put only to the 25
high-fear adults and the four currently low-fear adults who reported having had a fear of dogs in
the past. Out of this total of 29 Ss, 23 (79%) considered that their fear was ‘not particularly
irrational’; and only six Ss (20%) considered their fear to be ‘definitely irrational’.
Fifteen of the 29 Ss reported that their fear of dogs commenced before they reached school age,
and ten more reported that their fear started when they were still attending primary school (i.e.
aged under 8 yr). Only four Ss reported that their fear commenced when in their teens or older.
As noted above, only four of the 29 Ss reported that their fear had now completely subsided. Of
these four people, one stated that the fear disappeared of its own accord and three said it
disappeared when they got a dog of their own.
Twelve of the 29 Ss could not recall the circumstances in which their fear of dogs first arose.
Of the 17 who could recall the details, four recalled being bitten, and seven reported incidents in
which a dog either jumped up at them or chased them. A further five Ss traced their fear of dogs
to the influence of their parents, and one S reported that news coverage of a dog attack had initiated
her fear.
DISCUSSION
In this discussion we will address three issues, (1) methodological considerations which will
constrain our interpretations of the questionnaire replies, (2) the implications of the results for our
understanding of the etiology of fear of dogs, and (3) similarities and differences between adults
and children in relation to fear of dogs.
Methodological considerations
Answers to questions about events long in the past are vulnerable to failures and distortions of
recall. The child sample was included in the present study partly to meet this difficulty, but may
itself have produced a different set of problems of interpretation. It is possible, for example, that
the children and adults may have perceived questions differently; and the adult Ss-with their
greater experience and self-knowledge-may have been more able than the children to judge
accurately how they would actually behave in real life.
A further constraint on interpretation concerns the representativeness of the samples. In the child
sample, the unclassifiable children who responded inconsistently to the two questions asking about
392 SHARON DOOGAN and GLYN V. THOMAS
fear of strange dogs were excluded from the study, and this selection may conceivably have
introduced a bias. From inspection of their replies to the rest of the questionnaire, however, it was
not readily apparent what that bias might be.
In the adult sample, the preponderance of social science and humanities students reflects the
campus locations from which Ss were recruited. The sex ratio in the sample was similar to that
for the student population in those disciplines. While the sample is biased by the inclusion only
of students who were willing to give up their time to answer the questions, it is not obvious how
willingness to help with the study might bias the sample with regard to fear of dogs.
Aetiology of fear of dogs
Consistent with previous findings (McNally & Steketee, 198.5; Merckelbach et al., 1989), almost
all the dog-fearful adults reported that they acquired their fear during childhood, but only half the
dog-fearful adults were able to recall the onset of their fear. The evidence of those who could recall
implicated conditioning experiences in the majority of cases, with social learning or info~ation
transmission playing a major role in the remainder. This pattern of results, implicating conditioning
processes in the origins of fear in approx. 50% of fearful Ss is also broadly consistent with results
obtained in other studies, including those by Di Nardo et al. (1988) of dog-phobics, and by Ost
and Hugdahl (1981) of a mixed sample of social-, animal- and claustrophobics. Despite this wide
agreement, however, it would be unwise to assume that Ss have accurately identified the actual
cause of their fear. It is possible, for example, that individuals may have wrongly attributed their
fear to a conditioning event because it was salient and provided a plausible explanation of their
fear. The true causes may have been overlooked because they were not accessible to consciousness,
or simply less obvious or less memorable (see Nisbett & Wilson, 1977).
A clearer picture of the likely origins of fear can potentially be obtained by comparing the
frequencies with which fearful and non-fearful Ss reported different dog-related experiences. As
in the study by Di Nardo et al. (1988), the present survey found no significant difference in the
numbers of high- and low-fear Ss who reported having at least one aversive encounter with a dog.
Aversive encounters involving only extreme fear but no physical pain (S-R conditioning)
were signi~cantly more frequently reported by high- than low-fear Ss in the present study.
Nevertheless, given that almost any encounter with a dog is potentially distressing for a dog-phobic
person, it would be surprising if the dog-fearful Ss were not more likely than non-fearful individuals
not only to have frightening experiences with dogs but also to remember and report them. Indeed,
it is more likely that frightening but non-painful encounters with dogs are a consequence of the
development of fear of dogs rather than its primary cause. While such S-R conditioning events
may be important in maintaining fear (see, for example, Eysenck, 1979) they cannot account for
its origins.
In contrast, S-S conditioning events-aversive encounters that involved pain (S-S conditioning)
which could plausibly account for the development of fear of dogs-were not reported significantly
more frequently by high- than by low-fear Ss in the present study nor in that by Di Nardo et al.
(1988). Consequently, consideration of S-S conditioning events alone also cannot provide an
adequate account of acquisition of fear of dogs. Either conditioning processes in fact play no causal
role, or some co-factor determines whether or not a conditioning event will produce lasting fear.
Our search for such co-factors began with a test of whether familiarity with dogs prior to an
aversive encounter with one might prevent the conditioning of enduring fear. In support of this
hypothesis we found that significantly more dog-fearful than non-fearful adults, but not children,
reported that they had had little contact with dogs prior to their first aversive experience with one.
There is an interesting parallel here with fear of snakes; Murray and Foote (1979) found that the
more experience people have with snakes, the less they fear them.
With survey data only it is impossible to prove that exposure prevents or reduces fear of dogs;
there could plausibly be some other unknown factor, such as personality, that both reduces the
likelihood of an individual having contact with dogs and predisposes him/her to develop fear of
dogs. Nevertheless, there is now substantial evidence from laboratory experiments and from clinical
research that prior exposure to a stimulus retards subsequent fear conditioning to that stimulus
(see Lubow, 1973), and that non-eventful exposure (extinction) is also often effective at reducing
already established fear (Rachman, 1990).
Fear of dogs in adults and children 393
The second possible co-factor investigated was the suggestion that conditioning events produce
enduring fear only in certain susceptible individuals. In support of this hypothesis, we found that
dog-fearful adults were more likely than the dog-fearless to report fear of a variety of other
situations and things. It must count against this notion, however, that there was no evidence at
all that dog-fearful children had more fears of other animals than did dog-feariess children.
The data from the adult sample gave little indication that vicarious learning and imitation play
a major role in the origins of dog fears. There were some indications, however, that information
transmission might be important in creating or sustaining fear of dogs in adults, an impression
which was confirmed by the responses of the children. Dog-fearful children were significantly more
likely than either dog-fearful adults or non-fearful children to report having received warnings. This
latter result is consistent with such warnings playing an important role in the establishment of fear
of dogs, but being forgotten by most dog-fearful adults. An alternative hypothesis, that the higher
frequency of reported warnings by children than the adults simply reflects recent heightened mass
media coverage of dog attacks, fails to account for the low frequency of warnings reported by the
non-fearful children. The possible role of warnings about dangerous dogs in creating general fear
of dogs perhaps deserves further investigation.
Only a fifth of the adult phobics considered their fear irrational, which suggests that they are
concentrating mainly on the dangers that dogs present to the exclusion of their affectionate and
playful characteristics. In contrast, McNally and Steketee (1985) found that most animal-phobics
(other than dog-phobics) recognized their fear as irrational. The present results support the
conclusion drawn by Di Nardo et al. (1988), that a strong expectation of harm appears to be a
significant factor in adult’s fear of dogs.
A comparison of fear of dogs in adults and children
We found several major differences between dog-fearful adults and children. Dog-fearful children
were as likely as the non-fearful to recognize the attractiveness of a friendly dog (even if their
questionnaire response overestimated their willingness to play with a dog in reality). In contrast,
dog-fearful adults expressed little willingness to approach or stroke a friendly dog. The fearful
children seemed to disc~minate between threatening and friendly dogs much more than did the
fearful adults.
Dog fears seem to be more frequent amongst children than amongst the adults in our samples;
and the child sample seemed much less clearly differentiated than was the adult sample, even though
many of the children were dropped from consideration because their responses did not clearly
identify them as low- or high-fear Ss. In their response to many of the questions the low- and
high-fear children were much more similar to each other than were the low- and high-fear adults.
If this analysis is correct, then our original aim of using data from dog-fearful children to substitute
for information forgotten by dog-fearful adults cannot be fulfilled, because dog-fearful adults and
dog-fearful children in the aggregate may be quite different with respect to fear of dogs and its
antecedents.
Given that dog fears almost always originate in childhood, the differences between dog-fearful
children and adults seem most readily explained if we assume that dog-fearful adults originate as
a sub-group of the population of dog-fearful children who for some reason do not lose their fear
of dogs. The persistence of fear of dogs in some individuals may be part of a more general and
habitual pattern of responding (such as avoidance) which tends to preserve fear. It is consistent
with this supposition that in our survey dog-fearful adults reported fear of other animals and of
other things more frequently than did dog-fearless adults.
CONCLUSION
The role of conditioning events in producing fear of dogs must be considered as non-proven.
If such conditioning events do play a causal role then it is only in conjunction with some other
factor such as lack of prior uneventful exposure to dogs or in especially susceptible individuals.
The present results from children suggest that information transmission may be more important
in engendering fear of dogs than studies of adults might suggest. Although most fearful adults
report that their fear of dogs began in childhood, it is clear that not all dog-fearful children grow
394 SHARON DOOGAN and GLYN V. THOMAS
up to become dog-fearful adults, which raises the question of why some children, but not others,
eventually lose their fear of dogs. Uneventful exposure to dogs would seem to be important in at
least some cases, but other factors cannot be excluded. In the aggregate dog-fearful adults and
children differ in several other important respects. Consequently, rather less than we had hoped
can be learned about the development of fear of dogs in adults from a cross-sectional study of fear
of dogs in children.
REFERENCES
Di Nardo, P. A., Guzy, L. T., Jenkins, J. A., Bak, R. M., Tomasi, S. F. & Copland, M. (1988). Etiology and maintenance
of dog fears. Behauiour Research and Therapy, 26, 241-244.
Eysenck, H. J. (1979). The conditioning model of neurosis. The Behavioral and Brain Sciences, 2, 155-199.
Eysenck, H. J. & Rachman, S. (1965). The causes and cures of neurosis. London: Routledge & Kegan Paul.
Geer, J. H. (1965). The development of a scale to measure fear. Behaviour Research and Therapy, 3, 45-53.
Lubow, R. E. (1973). Latent inhibition. Psychological Bulletin, 79, 398407.
Martin, B. (1973). Abnormal psychology. Glenview, Ill.: Scott, Foresman.
McNally, R. J. & Steketee, G. S. (1985). The etiology and maintenance of severe animal phobias. Behauiour Research and
Therapy, 4, 431435.
Merckelbach, H., De Ruiter, C., Van den Hout, M. A. & Hoekstra, R. (1989). Conditioning experiences and phobias.
Behauiour Research and Therapy, 27, 657-662.
Murray, E. J. & Foote, F. (1979). The origins of fears of snakes. Behaviour Research and Therapy, 17, 489493.
Nisbett, R. E. & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological
Review, 84, 231-259.
Ost, L.-G. (1985). Ways of acquiring phobias and outcome of behavioral treatments. Behauiour Research and Therapy, 6,
683-689.
Ost, L.-G. & Hugdahl, K. (1981). Acquisition of phobias and anxiety response patterns in clinical patients. Behaviour
Research and Therapy, 19, 439443.
Ost, L.-G. SC Hugdahl, K. (1983). Acquisition of agoraphobia, mode of onset and anxiety response patterns. Behauiour
Research and Therapy, 21, 623-631.
Rachman, S. (1977). The conditioning theory of fear-acquisition: A critical examination. Behauiour Research and Therapy,
15, 375-387.
Rachman, S. (1990). The determinants and treatment of simple phobias. Advances in Behauiour Research and Therapy, 12,
l-30.
Seligman, M. E. P. (1971). Phobias and preparedness. Behavior Therapy, 2, 307-320.
Wagner, A. R. (1978). Expectancies and the priming of STM. In Hulse, S. H., Fowler, H. & Honig, W. K. (Eds), Cognitive
processes in animal behavior. Hillsdale, N.J.: Erlbaum.
