Journal of Psychopathology and Behavioral Assessment

https://doi.org/10.1007/s10862-018-9687-1

Psychometric Properties and Clinical Utility of the Specific Phobia

Questionnaire in an Anxiety Disorders Sample
Melina M. Ovanessian 1 & Nichole Fairbrother 2 & Valerie Vorstenbosch 1 & Randi E. McCabe 3 & Karen Rowa 3 &
Martin M. Antony 1

# Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract
Despite an abundance of self-report measures that screen for the presence of specific phobias, there is a lack of comprehensive, well-
validated screening tools for identifying a wide range of fears based on DSM-5 specific phobia types. The current paper reports on the
psychometric properties and clinical utility of the Specific Phobia Questionnaire (SPQ), a new screening tool for assessing fear of a
broad range of phobic stimuli, and the extent to which fear interferes with daily life. An exploratory factor analysis revealed five factors
with internal consistency (Cronbach’s α) ranging from .64–.92. The SPQ also demonstrated good convergent and discriminant validity
with measures of worry, depression, and other specific phobias, and good test-retest reliability. Results also suggest that SPQ scores are
useful for discriminating individuals with specific phobias from those without specific phobias, and for identifying specific phobia
types. Overall, preliminary results suggest that the SPQ can serve as a useful tool in both research and clinical settings, and inform
intervention and prevention efforts.

Keywords Specific phobia . Factor analysis . Measurement

Specific phobias are characterized by excessive and unreasonable 2010). Furthermore, without treatment, specific phobias can be
fear and anxiety in the presence of particular objects or situations. chronic and disabling, and thus have a substantial impact on
The Diagnostic and Statistical Manual for Mental Disorders, 5th social, occupational and physical functioning (e.g., avoidance
edition (DSM-5; American Psychiatric Association 2013) clas- of important medical procedures; Wolitzky-Taylor et al. 2008).
sifies specific phobias into five types, including (1) animal type Specific phobias may be assessed in various ways. Clinician-
(e.g., spiders, insects, dogs), (2) natural environment (e.g., administered interviews, such as the Anxiety and Related
heights, storms, water), (3) blood-injection-injury (e.g., needles, Disorders Interview Schedule for DSM-5, (ADIS-5; Brown and
invasive medical procedures), (4) situational (e.g., airplanes, ele- Barlow 2014) and the Structured Clinical Interview for DSM-5
vators, enclosed places), and (5) other (e.g., situations that may (SCID; First et al. 2015) are considered the gold standard for
lead to choking or vomiting). Prevalence rates for specific pho- establishing a diagnosis, but are costly and time consuming,
bias have been estimated to fall between 2 to 12.5% (Kessler et and therefore are often impractical screening tools. Self-report
al. 2005a, b; Stinson et al. 2007), and 17-month full remission measures are a low cost, time-efficient method for screening,
rates, without treatment, are estimated at 19% (Trumpf et al. and can be used to supplement diagnostic interviews. Self-report
measures can effectively and efficiently screen for the presence
of specific phobias, as well as provide additional information on
* Martin M. Antony symptom severity, frequency, and other facets of the disorder.
[email protected] Moreover, given their sensitivity to change, self-report measures
1
can be useful in monitoring treatment progress. In the past, the
Department of Psychology, Ryerson University, 350 Victoria Street,
Fear Survey Schedule II (FSS-II; Geer 1965) and Fear Survey
Toronto, ON M5B 2K3, Canada
2
Schedule III (FSS-III; Wolpe and Lang 1977), were often used to
Department of Psychiatry, University of British Columbia Island
screen for phobias, in both clinical or research settings. These
Medical Program, 2400 Arbutus Road, Room #002 Perkes Building,
Victoria, BC V8N 1V7, Canada scales consist of a broad list of anxiety-provoking stimuli, includ-
3 ing some that are feared by individuals with specific phobias
Anxiety Treatment and Research Clinic, St. Joseph’s Healthcare
Hamilton, 100 West 5th St., Level 1, Block B, Hamilton, ON L8N (e.g., animals) and others that are not (e.g., ugly people, criti-
3K7, Canada cism). Each item is rated with respect to fear, and a total score
 J Psychopathol Behav Assess

is generated to indicate overall fear level. However, as has been Treatment and Research Clinic (ATRC), a specialized outpa-
noted elsewhere (Hood and Antony 2012), these scales typically tient anxiety clinic at St. Joseph’s Healthcare Hamilton, in
consist only of a list of stimuli, therefore they do little other than Hamilton, Ontario. These data were used for factor analysis,
identify specific objects or situations that an individual may fear, receiver operating characteristic analysis, and to assess the in-
and they perform poorly when used to discriminate individuals ternal consistency of the SPQ; (2) an undergraduate sample of
with specific phobias from nonfearful individuals (Beck et al. introductory psychology students (n = 182) was recruited
1998), or to predict avoidance of feared stimuli (Klieger and through a Participant Pool at Ryerson University in Toronto,
Franklin 1993). For these reasons, these scales have been used Ontario, with students receiving course credit for participation;
much less frequently in recent years. these participants completed measures online, and data were
To address these shortcomings, the Phobic Stimuli used to assess the convergent and discriminant validity, as well
Response Scales (PSRS; Cutshall and Watson 2004) was de- as the test-retest reliability of the SPQ. Informed consent was
veloped. The intent was for the PSRS to assess the underlying obtained from all individual participants included in the study.
focus of individuals’ fears by assessing cognitive and emo- Data from the clinical sample were included in the factor anal-
tional factors of five fears: social, animal, physical confine- ysis if the individual was between 18 and 65 years of age and
ment, bodily harm, and blood-injection. However, as is the had a principal diagnosis of a DSM-IV anxiety disorder (total
case with the FSS-II and FSS-III, the PSRS items do not n = 875; social anxiety disorder, n = 303; panic disorder with
provide coverage for each of the DSM-5 specific phobia agoraphobia, n = 164; panic disorder without agoraphobia, n =
types. Furthermore, given that the FSS-II, FSS-III, and 24; generalized anxiety disorder, n = 181; obsessive-
PSRS were developed to measure affective responses (e.g., compulsive disorder, n = 135; anxiety disorder not otherwise
fear) only, they do not assess other specific phobia symptoms specified, n = 27; posttraumatic stress disorder, n = 14; agora-
(e.g., impaired functioning), and thus are not ideal for screen- phobia without history of panic disorder, n = 6; specific phobia,
ing DSM-5 specific phobias (Hood and Antony 2012). n = 21). A number of participants in the clinical sample also had
Other self-report phobia measures assess for one particular an additional diagnosis of a specific phobia (n = 265).
type of specific phobia, for example the Fear of Spiders Diagnoses were established using the Structured Clinical
Questionnaire (Szymanski and O’Donohue 1995), the Interview for DSM-IV (SCID-IV; First et al. 1995),1 adminis-
Blood-Injection Symptom Scale (Page et al. 1997) and the tered by trained graduate-level clinicians. These clinicians re-
Fear of Flying Questionnaire (Bornas and Tortella-Feliu ceived SCID-IV training by observing at least three interviews
1995). However, self-report measures assessing one specific conducted by experienced interviewers, as well as completing
type of phobia are not suitable for screening for a wide range at least three SCID-IV interviews under the observation of an
of phobias. experienced interviewer. Weekly meetings were held to review
In response to the lack of a comprehensive, well-validated and confirm diagnoses with an experienced psychologist (over
screening tool consistent with DSM-5 specific phobia types, 10 years of experience in SCID administration). Evidence of
we developed a new measure to assess fear of a broad range of strong interrater reliability for the principal diagnosis (ĸ = .89)
phobias, and how much the fear interferes with daily life. The was established on a subset (N = 13) of the SCID-IV interviews.
Specific Phobia Questionnaire (SPQ; Fairbrother and Antony The clinical participants’ mean age was 35.78 (SD = 11.88,
2012; see Appendix A) was designed to assess the extent of range = 18–65), whereas undergraduate participants’ mean
fear and interference for a broad range of objects and situa- age was 21.61 (SD = 4.89, range = 18–44). Table 1 provides a
tions. The SPQ consists of 43 items, each rated using a 5-point summary of demographic features of each of the samples.
Likert scale on two dimensions: (1) level of fear, with scores
ranging from 0 (no fear) to 4 (extreme fear) and (2) extent to
which the fear interferes with one’s daily life, with scores Measures
ranging from 0 (no interference) to 4 (extreme interference).
The primary aim of the current study was to evaluate the Specific Phobia Questionnaire (SPQ; Fairbrother and Antony
psychometric properties and clinical utility of the SPQ for 2012; Appendix A) Each of the samples (i.e., clinical, under-
the screening of specific phobias in a clinical adult population. graduate) completed the SPQ. The SPQ (described earlier) as-
sesses the extent of fear and associated interference for a range
of objects and situations typically feared by individuals with
Method specific phobia. The SPQ is scored by combining fear ratings
of each DSM-5 specific phobia type to create five different
Participants and Procedure composite scores: Animals (sum of items 3, 6, 8, 11, 30, 31,
1
Data were collected prior to the release of the DSM-5. Changes in diagnostic
Data were collected from 2 samples: (1) an adult, treatment- criteria for anxiety disorders from DSM-IV to DSM-5 likely do not impact the
seeking sample (n = 1031) was recruited from the Anxiety findings of the current study.
J Psychopathol Behav Assess

Table 1 Participant
characteristics Treatment-seeking sample (n = 875) Undergraduate sample (n = 150)

Sex – frequency (%)a

Male 329 (37.6) 23 (15.3%)
Female 545 (62.3%) 124 (82.7%)
Relationship status (%)b
Single 403 (46%) 125 (83.3%)
Married 301 (34.4%) 5 (3.3%)
Cohabiting 96 (10.9%) 12 (8%)
Separated 25 (2.9%) –
Divorced 42 (4.8%) 1 (0.7%)
Widowed 3 (0.3%) –
Ethnocultural background – frequency (%)c
Aboriginal 10 (1.1%) –
Black /Afro-Caribbean/African 9 (1%) 10 (6.7%)
White/European 467 (53.4%) 66 (44%)
Hispanic/Latin American 7 (0.8%) 5 (3.3%)
Asian 361 (41.3%) 48 (32%)
Biracial/Multiracial 5 (0.6%) 7 (4.7%)
Other 9 (1%) 14 (9.3%)
Highest level of education
Did not attend high school 14 (1.6%) –
Some high school 96 (10.97%) –
Completed High School 100 (11.4%) 31 (20.7%)
Some college/university 147 (16.8%) 97 (64.7%)
Completed college/university 213 (24.3%) 19 (12.7%)
Some graduate school 14 (1.6%) –
Completed graduate school 52 (5.9%) 3 (2%)
Annual family income d
< $19,000 223 (25.5%) 21 (14%)
$20,000–$39,000 132 (15.1%) 20 (13.3%)
$40,000–$59,000 110 (12.6%) 16 (10.7%)
$60,000–$79,000 100 (11.4%) 27 (18%)
$80,000–$99,000 83 (9.5%) 17 (11.3%)
> $100, 000 132 (15.1%) 25 (16.7%)
a
One individual in the treatment-seeking sample and three in the undergraduate sample did not provide data for
sex; b Seven individuals in the treatment-seeking sample did not report relationship status. c Seven individuals in
the treatment-seeking sample did not report ethnocultural background. d Twenty-four individuals in the treatment-
seeking sample did not report annual family income

36, 37, 39, 43) Natural Environment (sum of items 1, 2, 13, 14, from 0 (did not apply to me at all) to 3 (applied to me very
23, 27, 32, 33, 38); Situational (sum of items 4, 7, 16, 22, 26, 29, much, or most of the time). The depression, anxiety, and stress
34, 40); Blood-Injection-Injury (sum of items 5, 9, 10, 15, 17, scales have been shown to have good internal consistency
18, 19, 21, 24, 25, 28, 35, 41, 42); Other (sum of items 12, 20). (α = .94, .87 and .91, respectively; Antony et al. 1998), and
good construct, convergent and discriminant validity, compa-
Depression Anxiety Stress Scales, 21-item version (DASS-21; rable to the original 42-item version (Antony et al. 1998;
Lovibond and Lovibond 1995) Both of the samples (i.e., clin- Henry & Crawford 2005). Cronbach’s α for the depression,
ical, undergraduate) completed the DASS, and data were used anxiety, and stress scales in the current sample were, .92, .83,
to assess discriminant validity of the SPQ. The DASS-21 and .87, respectively.
measures depression, anxiety and psychological distress/ten-
sion. Participants rate the extent to which each item applies to Penn State Worry Questionnaire (PSWQ; Meyer et al. 1990)
them over the past week, using a 4-point Likert scale, ranging The undergraduate sample completed the PSWQ, and data
 J Psychopathol Behav Assess

were used to assess discriminant validity of the SPQ. The situations on two subscales: anxiety and avoidance. The scale
PSWQ is a 16-item self-report measure that assesses a gen- has good psychometric properties; Cronbach’s alpha for the
eral tendency to worry excessively. Items include BI know I anxiety and avoidance subscales is .96 and .89, respectively
should not worry about things, but I just cannot help it,^ (Baker et al. 1973). Cronbach’s α for the anxiety and avoid-
and BMany situations make me worry.^ Items are rated on a ance subscales in the current sample were .92 and .81,
5-point Likert scale, ranging from 1 (not at all typical of respectively.
me) to 5 (very typical of me). The PSWQ has found to have
high internal consistency (α = .88 to .95) (Molina & Claustrophobia Questionnaire (CLQ; Radomsky et al. 2001)
Borkovec, 1994), and good test-retest reliability (r = .92; The undergraduate sample completed the CLQ, and data were
Metzger et al. 1990). Cronbach’s α for the PSWQ in the used to assess convergent and discriminant validity of the SPQ
current sample was .80. subscales. The CLQ includes 26 items that measure (1) fear of
suffocation and (2) fear of restriction. Individuals are asked to
Fear Survey Schedule, Second Edition (FSS-II; Geer 1965) The rate items on a 5-point Likert scale (0 = not at all anxious to
undergraduate sample completed the FSS-II, and data were 4 = extremely anxious), to indicate how anxious they would
used to assess convergent validity of the SPQ. The FSS-II feel in each situation. The CLQ has high internal consistency
assesses the amount of fear associated with various stimuli (α = .95; Radomsky et al. 2001), excellent test-retest reliabil-
and situations. The FSS-II consists of 51 items that partici- ity (r = .89), and strong convergent, discriminant, and predic-
pants rate using a 7-point scale measuring fear of water, death, tive validity. Cronbach’s α for the CLQ in the current sample
illness, injury, objects, organisms, violence, social interaction was .94.
and negative social evaluation (Bernstein & Allen, 1969;
Weiss et al. 1968). The FSS-II has demonstrated high internal
consistency reliability (r = .94; Geer 1965). Cronbach’s α for Results
the FSS-II in the current sample was .95.
Psychometric Properties
Phobic Stimuli Response Scale (PSRS; Cutshall and Watson
2004) The undergraduate sample completed the PSRS, and Convergent and Discriminant Validity
data were used to assess convergent validity of the SPQ. The
PSRS assesses the cognitive and emotional aspects of five Prior to conducting the main analyses, data were screened for
types of fears: social, animal, physical confinement, bodily missing values. Of the 182 participants in the undergraduate
harm and blood-injection. The PSRS consists of 46 items sample, analyses were conducted only for participants who
measured on a 4-point Likert scale (1 = strongly disagree to had complete data on measures being used to assess conver-
4 = strongly agree). The scale has acceptable reliability and gent and discriminant validity (n = 150). Pearson’s correla-
validity, and good internal consistency (α = .88; Cutshall and tions were used to assess the convergent and discriminant
Watson 2004). Cronbach’s α for the PSRS in the current sam- validity of the SPQ with the DASS and PSWQ in the under-
ple was .88. graduate sample. Convergent and discriminant validity of
SPQ subscales, according to DSM-5 derived specific phobia
Blood Injection Symptom Scale (BISS; Page et al. 1997) The types, with the FSS, BISS, PSRS, AQ and CLQ and respec-
undergraduate sample completed the BISS, and data were tive subscales were also assessed. These results are presented
used to assess convergent and discriminant validity of the in Table 2 for the undergraduate sample. As Table 2 shows, the
SPQ subscales. The BISS measures the presence of physical SPQ was more strongly associated with other phobia ques-
symptoms when facing situations involving blood or injec- tionnaires than it was to measures of depression, anxiety or
tions. The scale consists of 17 items that participants rate as worry; these differences were significant, p < .001.
Byes^ or Bno^ to indicate whether they have experienced that On average, participants with specific phobias (principal or
symptom during one of their worst experiences involving additional diagnoses) had higher SPQ total scores (sum of fear
blood or injections. Possible scores range from 0 to 17. The and interference) (M = 97.63, SE = 4.22) than participants
scale has good internal consistency (α = .86; Page et al. 1997). with other anxiety disorders (with no comorbid specific pho-
Cronbach’s α for the BISS in the current sample was .86. bia diagnoses) (M = 63.47, SE = 1.79). This difference, 34.16,
BCa 95% CI [25.50, 43.62], was significant t(808) = 7.93, p
Acrophobia Questionnaire (AQ; Cohen 1977) The undergrad- < .01. Participants with specific phobias had both higher SPQ
uate sample completed the AQ, and data were used to assess Fear (M = 57.47, SE = 2.15) and SPQ Interference
convergent and discriminant validity of the SPQ subscales. (M = 40.16, SE = 2.22) scores than participants without spe-
The AQ is a 40-item Likert-type measure of height phobia. cific phobias (M = 38.47, SE = 0.93), (M = 24.89, SE = 0.89),
Individuals rate their anxiety pertaining to height-relevant respectively. The SPQ Fear difference, 18.89, BCa 95% CI
J Psychopathol Behav Assess

Table 2 Convergent and discriminant validity of the specific phobia questionnaire

Measure r

SPQ SPQ SPQ SPQ SPQ SPQ SPQ situational SPQ

fear interference total BII animals situational fear interference situation total

DASS-21
(Treatment-seeking/undergraduate)a
Anxiety .43*/.38* .44*/.43* .45*/.43* – – – – –
Stress .30*/.36* .32*/.36* .33*/.39* – – – – –
Depression .20*/.23* .22*/.27* .22*/.27* – – – – –
Penn state worry questionnaire .32* .30* .33* – – – – –
Fear survey schedule .72* .58* .69* – – – – –
Blood injection symptom scale – – – .44* .06 – – –
Phobic stimuli response scales
Animals – – – .32* .56* – – –
Total .63* .42* .57* – – – – –
Acrophobia questionnaire
Anxiety – – – – – .59* .50* –
Interference – – – – – .43* .46* –
Claustrophobia questionnaire – – – – – – – .64*

SPQ Specific Phobia Questionnaire, DASS-21 Depression Anxiety Stress Scales, 21-item version, BII Blood-Injection-Injury
*p < .001
a
DASS-21 data were available for both the treatment-seeking (n = 875) and undergraduate (n = 150) samples. Remainder of the questionnaires in table
were only completed by the undergraduate sample (n = 150)

[14.39, 23.78], was significant t(808) = 8.35, p < .01. The Meyer-Olkin (KMO) test of sampling adequacy was 0.94,
SPQ Interference difference, 15.27, BCa 95% CI [10.73, exceeding the acceptable standard of 0.5. All individual items
20.37], was also significant t(195.34) = 6.39, p < .01. also exceeded the recommended KMO value of 0.5 (range =
0.87–0.97). A total of 3% of the nonredundant residuals be-
Factor Structure of the SPQ: Exploratory Factor tween observed and reproduced correlations was found, well
Analysis below the 50% maximum acceptable standard (Field 2013).
Since it was expected that items would correlate with one
Prior to investigating individual factor loadings, internal con- another, an oblique (direct oblimin) rotation was used and
sistency was examined for the full 43-item scale, and was factor scores using the regression method were obtained.
found to be high (Cronbach’s α = .95). In line with the primary The initial EFA (without specifying number of factors be-
aim of the current study, which was to evaluate the psycho- forehand) resulted in 10 eigenvalues over 1, suggesting 10
metric properties of the newly developed SPQ, we conducted individual factors, for which the total variance accounted for
an exploratory factor analysis (EFA) using principal axis fac- was 54.41%. However, Kaiser’s criterion of retaining eigen-
toring to investigate its underlying variable structure. Given values over 1 was potentially inappropriate for factor extrac-
the large sample size (n = 875), participants with missing tion as individual communalities after extraction did not all
values on any SPQ items were excluded from the factor anal- exceed 0.7 (Field 2013). Investigation of the scree plot inflec-
ysis; the pattern of missing values appeared to be random. tions justified retaining five factors, however, the scree plot
Data from the clinical sample (n = 811) were used for the tends to overdetermine number of factors (Hayton et al.
EFA. This sample size yields an item-to-participant ratio of 2004). Thus, four, five and six factor solutions were exam-
1:19, which is considered very good for factor analysis ined, using oblimin rotations of the factor loading matrix; item
(Osborne and Costello 2004). To analyze the feasibility of loadings were compared for the multiple scree tests. The five
EFA, Bartlett’s test of sphericity and the Kaiser-Meyer-Olkin factor solution, accounting for 51.59% of the variance was
tests were performed. Bartlett’s test of sphericity was signifi- preferred due to being theoretically linked to several specific
cant, χ2(903) = 18,321.25, p < 0.001, indicating that the R- phobia constructs, the leveling off of eigenvalues on the scree
matrix was significantly different from an identity matrix, or plot after five factors, and difficulties interpreting the other
that the relationship between factors was not 0. The Kaiser- factor solutions.
 J Psychopathol Behav Assess

In conjunction with the scree test, parallel analysis (PA) 1999). We also report the chi-square value and the associated
was also used to support the number of factors to be retained, degrees of freedom, but this was not used to assess model fit,
as it provides the most accurate factor-retention method given this statistic is dependent on sample size (e.g., Bryant et al.
(Hayton et al. 2004; Henson & Roberts, 2006). Given the 1999).
eigenvalues of the five factors of the scree plot were all greater The five-factor DSM-5 model of the SPQ appears to be
than the average of the eigenvalues of the parallel factors, all a good fit to the data, as was evidenced by all indices
were retained (Hayton et al. 2004). (SRMR = 0.088; RMSEA = 0.069 [90% CI = 0.061,
The five interpretable factors that emerged comprised 0.074], CFI = 0.916); χ 2 = 8954.70, df = 850, p < .001.
eigenvalues of 13.39, 3.49, 2.21, 1.56, and 1.52. Based on standardized estimates, all items loaded on to
Together, these factors accounted for a total of 51.59% of their respective factors, all ps < 0.001 (See Table 4, for a
the variance in item responses (Factor 1 = 31.14%, Factor summary of DSM-5 factor loadings onto corresponding
2 = 8.13%, Factor 3 = 5.14%, Factor 4 = 3.63%, Factor 5 = questionnaire items).
3.54%). The five emerging factors were labeled as follows: The EFA derived five-factor model also fit the data
Factor 1 (12 items) - Situations and Natural Environment; well (SRMR = 0.08; RMSEA = 0.078 [90% CI = 0.067,
Factor 2 (10 items) – Blood-Injection-Injury; Factor 3 (9 0.083], CFI = 0.927); χ 2 = 6206.01, df = 655, p < .001;
items) – Animals; Factor 4 (4 items) – Health-related; goodness of fit values were acceptable on all indices;
Factor 5 (3 items) – Driving. Standardized factor loadings ranged from 0.31 to 0.87
Although some items cross-loaded on more than one (all ps < 0.001).
factor, the primary loading was linked to its theoretically-
relevant factor and was therefore included as part of the Internal Consistency
theoretically-relevant factor (e.g., elevators loaded primar-
ily on to BSituations and Natural Environment,^ but also Each of the five factors retained were assessed for reliabil-
loaded on BHealth-related^). Seven items (items 2, 6, 8, 18, ity. Cronbach’s alpha for each factor obtained in 1) the
19, 20, 33) did not load substantively (loadings below 0.3) factor analysis, and 2) per DSM-5 scoring, is displayed in
on the interpretable factors, but were not discarded from Table 5. Factor/subscale intercorrelations and correlations
the questionnaire, given that removing these items did not with the total scale, and item-total correlations were com-
improve the internal consistency of the scale. Items 18 and puted to assess the internal consistency of the scale.
20 loaded together on one meaningful and interpretable Tables 6 and 7 show the subscale intercorrelations for the
factor (Health-related), and were thus retained as part of SPQ, for both factor analytically derived and DSM-5 de-
this factor (See Table 3, for a summary of each factor and rived subscales, respectively. Intercorrelations between
the corresponding questionnaire items, and factor subscales ranged from 0.31 to 0.60, indicating that sub-
loadings). scales are measuring distinct but related constructs. High
correlations, ranging from 0.60 to 0.88, were found be-
Confirmatory Factor Analysis tween subscale scores and the total scale score. Item total
correlations for items with their respective subscale and
Next, the EFA model was compared with the DSM-5 model with the total scale were typically in the 0.53–0.99 range.
using Confirmatory Factor Analysis (CFA), to assess for the With the exception of one item, all items showed item-total
best model fit. CFA was conducted on the full dataset and run correlations with both subscale and total scale greater than
with LISREL 9.30 (Jöreskog & Sörbom, 2012). 0.30. These results suggest that the SPQ has good internal
The DSM-5 specific phobia conceptual framework was consistency.
used to guide the model specification, in which items were
grouped to load onto the following latent variables: Animal, Test-Retest Reliability
Natural Environment, Situational, Blood-injection-injury,
Other. Two-week test–retest data for the SPQ were available from 26
Given the data were normally distributed, maximum like- participants (undergraduate sample). This sample was used to
lihood estimation was used (Kline, 2005). To evaluate good- compute the test–retest reliabilities for both DSM-5 and em-
ness of fit between the model and observed data, we used the pirically derived scores, to determine the stability of SPQ
standardized root mean residual (SRMR < .10; Hu & Bentler, scores over time.
1999), Comparative Fit Index (CFI > .90), and the Root mean Given that SPQ scores violated the assumption of normal-
square error of approximation (RMSEA < .08; Hu & Bentler, ity, Pearson correlations were used, as they are robust against
J Psychopathol Behav Assess

Table 3 Summary of exploratory factor analysis results for specific phobia questionnaire (N = 811)

Factor loadings

Item Situation/ Natural Blood-injection- Animals Driving Health- Do not load

environment injury related substantively

High open places (1) .79 . . . . .

Looking out the window on the top floor of a tall building (27) .78 . . . . .
Bridges (40) .59 . . . . .
Tunnels (4) .56 . . . . .
Elevators (29) .53 . . . .41 .
Standing on a ladder (13) .47 . . . . .
Boating in deep water (32) .45 . . . . .
Enclosed places (16) .45 . . . .31 .
Flying in an airplane (22) .46 . . . .39 .
Swimming in a swimming pool (14) .35 . . . . .
Swimming in a lake or ocean (23) .35 . . . . .
Choking (12) .30 . . . . .
Watching someone else give blood (21) . −.93 . . . .
Watching someone else get an injection (5) . −.89 . . . .
Watching someone else get stitches (10) . −.86 . . . .
Receiving an injection (35) . −.78 . . . .
Giving blood (17) . −.77 . . . .
Blood tests (42) . −.66 . . .32 .
Watching surgery on television (24) . −.64 . . .
Attending to someone else’s cut (25) . −.57 . . . .
Attending to your own cut (28) . −.46 . . . .
Getting minor surgery (9) . −.40 . . .31 .
Bugs (37) . . .83 . . .
Spiders (43) . . .67 . . .
Moths or butterflies (30) . . .62 . . .
Bees or wasps (11) . . .59 . . .
Rodents (e.g., mice, rats) (31) . . .59 . . .
Worms (39) . . .58 . . .
Snakes (36) . . .50 . . .
Birds (3) . . .39 . . .
Heavy rain (38) . . .34 . . .
Visiting a hospital (15) . . . . .37 .
Developing an illness (41) . . . . .37 .
Having your blood pressure taken (18) . . . . .28 .
Vomiting (20) . . . . .28 .
Driving on highways (7) . . . .86 . .
Driving in bad weather (26) . . . .85 . .
Driving in new places (34) . . . .84 . .
The dark (2) . . . . . .26
Visiting the dentist (19) . . . . . −.25
Cats (6) . . . . . .23
Thunder and lightning (33) . . . . . .20
Dogs (8) . . . . . .17
 J Psychopathol Behav Assess

Table 4 Summary of confirmatory factor analysis results for specific phobia questionnaire (N = 875)

Factor loadings

Blood-injection- Animals Other Situation Natural

injury environment

Watching someone else give blood (21) .84 . . . .

Watching someone else get an injection (5) .75 . . . .
Receiving an injection (35) .74 . . . .
Watching someone else get stitches (10) .72 . . . .
Giving blood (17) .72 . . . .
Blood tests (42) .67 . . . .
Watching surgery on television (24) .50 . . . .
Attending to someone else’s cut (25) .49 . . . .
Attending to your own cut (28) .43 . . . .
Visiting a hospital (15) .31 . . . .
Having your blood pressure taken (18) .28 . . . .
Visiting the dentist (19) .26 . . . .
Developing an illness (41) .18 . . . .
Bugs (37) . .72 . . .
Worms (39) . .58 . . .
Rodents (31) . .57 . . .
Snakes (36) . .55 . . .
Spiders (43) . .55 . . .
Moths or butterflies (30) . .48 . . .
Bees or wasps (11) . .47 . . .
Cats (6) . .23 . . .
Dogs (8) . .21 . . .
Choking (12) . . .65 . .
Vomiting (2) . . .44 . .
Tunnels (4) . . . .60 .
Bridges (40) . . . .60 .
Elevators (29) . . . .54 .
Enclosed spaces (16) .52
Driving in bad weather (26) . . . .45 .
Flying in an airplane (22) . . . .43 .
Driving on highways (7) . . . .40 .
Driving in new places (34) . . . .36 .
Boating in deep water (32) . . . . .54
Swimming in a lake or ocean (23) . . . . .47
Looking out the window on the top floor of a tall building (27) . . . . .46
Thunder and lightning (33) . . . . .40
Heavy rain (38) . . . . .40
Standing on a ladder (13) . . . . .39
Swimming in a swimming pool (14) . . . . .37
High open places (1) . . . . .35
The dark (2) . . . . .33

violations of normality (Havlicek and Peterson 1977). The 2- scores at time 1 and time 2 were significantly positively asso-
week test-retest reliability coefficient for the total scale was ciated. The test-retest reliability of both DSM-5 and empiri-
0.95. As indicated in Table 8, SPQ Fear and Interference cally derived SPQ subscales is also presented in Table 8.
J Psychopathol Behav Assess

Table 5 Scale reliability for factors Table 6 Empirically derived subscale intercorrelations for the specific
phobia questionnaire
Factor α
Subscale Situation/ Blood- Animals Health-
Factor analysis Natural injection- related
1 (Situations/Natural environment) .89 environment injury
2 (Blood-Injection-Injury) .92
Situation/Natural . . . .
3 (Animals) .84 environment
4 (Health-related) .64 Blood-injection-injury .52* . . .
5 (Driving) .85 Animals .60* .48* . .
DSM-5 Health-related .57* .58* .46* .
1 (Situations) .90 Driving .52* .31* .41* .36*
2 (Natural Environment) .88
*Correlations are significant at p < .001
3 (Blood-Injection-Injury) .93
4 (Animals) .90
5 (Other) .72
with Animal phobia types from individuals with other
α = Cronbach’s Alpha
anxiety disorders or other specific phobia types. The
95% confidence interval of the AUC ranged from .66 to
.81.
Receiver Operating Characteristic Analysis The analysis revealed a good ROC curve for the SPQ
Blood-Injection-Injury subscale fear scores (AUC = .86; p
Predicting Presence or Absence of Specific Phobias < .001), that was significantly better than chance for discrim-
inating individuals with BII phobia types from individuals
To determine how well the SPQ distinguished between with other anxiety disorders or other specific phobia types.
individuals with versus those without specific phobias The 95% confidence interval of the AUC ranged from .81 to
(principal or additional diagnoses on the SCID), we .90.
performed receiver operating characteristic (ROC) anal- The analysis revealed a fair ROC curve for the SPQ
yses using the sum of fear scores. 2 Area under the Natural Environment subscale fear scores (AUC = .79; p
curve (AUC) for the SPQ was significant in predicting < .001), that was significantly better than chance for discrim-
specific phobias (AUC = 0.71, 73.1% sensitivity/60% inating individuals with Natural Environment phobia types
specificity). from individuals with other anxiety disorders or other specific
phobia types. The 95% confidence interval of the AUC ranged
Predicting DSM-5 Specific Phobia Types from .74 to .84.
The analysis revealed a fair ROC curve for the SPQ
To predict DSM-5 specific phobia types, composite fear Situational subscale fear scores (AUC = .74; p < .001),
scores were created according to DSM-5 derived specific that was significantly better than chance in discriminat-
phobia types: Animals (items 3, 6, 8, 11, 30, 31, 36, 37, ing individuals with fears within the Situational phobia
39, 43); Natural Environment (1, 2, 13, 14, 23, 27, 32, 33, type from individuals with other anxiety disorders or
38); Situational (4, 7, 16, 22, 26, 29, 34, 40); Blood-
Injection-Injury (items 5, 9, 10, 15, 17, 18, 19, 21, 24,
25, 28, 35, 41, 42); Other (12, 20). Table 9 summarizes
Table 7 DSM-5 subscale intercorrelations for the specific phobia
the results of the ROC analyses (per subscale).3 The anal- questionnaire
ysis revealed a fair ROC curve for the SPQ Animal sub-
scale fear scores (AUC = .73; p < .001), that was signifi- Subscale Situation Natural Blood- Animals Other
cantly better than chance in discriminating individuals environment injection-
injury

Situation . .72* .58* .52* .53*

2
ROC Analyses were also conducted for the interference subscale; AUC for Natural . . .58* .60* .61*
the interference subscale was also significant in predicting specific phobias environment
(AUC = .69, 67% sensitivity/60% specificity) Blood-injection- . . . .52* .52*
3
ROC Analyses were also conducted for the interference subscales; AUCs for injury
the interference subscales (AUC = .72–.83, all ps < .001) were significantly Animals . . . . .53*
better than chance for discriminating individuals with DSM-5 specific
phobias. *Correlations are significant at p < .001
 J Psychopathol Behav Assess

Table 8 Test-retest reliability of

the SPQ Subscale Time 1 Time 2 r
Mean (SD) Mean (SD)

SPQ fear 27.28 (20.96) 20.13 (20.24) .92*

SPQ interference 16.46 (21.86) 12.35 (22.61) .95*
DSM derived
Animal 11.84 (10.87) 8.54 (9.81) .89*
Blood injection injury 13.56 (22.41) 12.14 (23.86) .96*
Situation 6.60 (8.37) 5.12 (8.18) .91*
Natural environment 10.00 (8.37) 6.76 (6.99) .77*
Other 1.32 (2.01) 0.92 (1.81) .92*
Empirically derived
Animal 11.76 (11.41) 8.16 (9.14) .88*
Blood injection injury 9.88 (17.53) 9.47 (19.82) .94*
Situation/Natural environment 11.77 (10.63) 8.76 (10.92) .83*
Health 3.68 (4.95) 1.63 (3.63) .85*
Driving 2.84 (3.37) 2.04 (3.69) .75*

SPQ Specific Phobia Questionnaire

*Correlations are significant at p < .001

other specific phobia types. The 95% confidence interval of AUC of the EFA derived Blood-Injection Injury subscale was
the AUC ranged from .68 to .79. not significantly different than the AUC of DSM-5 derived
The analysis revealed a fair ROC curve for the SPQ Other subscale. The AUC of the EFA derived Animal subscale was
subscale fear scores (AUC = .77; p < .05), that was significantly not significantly differently than the AUC of the DSM-5 de-
better than chance in discriminating individuals with Other pho- rived subscale. The AUC of the EFA derived Situation/
bia types from individuals with other anxiety disorders or other Natural Environment subscale was not significantly different-
specific phobia types. The 95% confidence interval of the AUC ly than the AUC of the DSM-5 derived subscales. The EFA
ranged from .68 to .86. derived Driving and Health subscales could not be compared
to any DSM-5 subscales.

Significance of the Difference Between the Areas Under Two

Independent ROC Curves
Discussion
To compare against the DSM-5 derived subscales, ROC anal-
yses were also done on the EFA derived SPQ subscales. This study reported on the psychometric properties and clini-
Table 10 summarizes the results of these ROC analyses. The cal utility of the SPQ, a self-report measure of specific

Table 9 Sensitivity and

specificity of the SPQ subscales DSM-5 specific phobia type SPQ subscale cut-off values Sensitivity Specificity
in the detection of individuals for SPQ fear scores
with DSM-5 specific phobia types
within an anxiety disorder sample Animal (n = 38) 8 .68
.63
Natural Environment (n = 46) 15 .80
.69
Blood-Injection-Injury (n = 44) 20 .82
.79
Situational (n = 85) 4 .66
.65
Other (n = 25) 6 .72
.72

SPQ Specific Phobia Questionnaire

J Psychopathol Behav Assess

Table 10 Sensitivity and

specificity of the SPQ subscales EFA derived specific phobia type AUC SPQ subscale cut-off values Sensitivity Specificity
for the detection of specific for SPQ Fear scores
phobia types within an anxiety
disorder sample Animal (n = 38) .74* 12 .75 .65
Natural Environment/Situational (n = 131) .81* 15 .80 .70
Blood-Injection-Injury (n = 44) .86* 20 .82 .79
Health (n = 17) .91* 11 .80 .94
Driving (n = 32) .80* 10 .75 .76

SPQ Specific Phobia Questionnaire

*p < .001

phobias. The results of this study provide preliminary evi- example, the animal fears that loaded on the Animal factor
dence that the 43-item SPQ has good psychometric properties. (e.g., insects, spiders rodents, snakes, birds) differ from those
An exploratory factor analysis revealed five factors: (1) that didn’t (e.g., cats, dogs) with respect to size (dogs and cats
Situations and Natural Environment, (2) Blood-Injection- tend to be larger than the others), and the extent to which they
Injury, (3) Animals, (4) Health-related and (5) Driving. elicit a disgust response (dogs and cats may be less likely to
Some of these factors (Blood-Injection-Injury, Animals) elicit a disgust reaction than insects, spiders, rodents, and
mapped directly onto the types found in DSM-5, though for snakes). In the literature, disgust is conceptualized by a feeling
others the relationship with the DSM-5 types was less clear. of revulsion toward an unpleasant stimulus (Barlow 2002).
Whereas the DSM-5 includes distinct BSituational^ and Indeed, research suggests that small animal phobias (e.g., spi-
BNatural Environment^ phobia types, these types combined ders, bees/wasps and other insects, rodents, snakes) not only
into a single factor called Situations/Natural Environment. elicit fear, but also elicit disgust responses (e.g., Gerdes et al.
The analysis also revealed two factors that did not directly 2009; Muris et al. 2008). On the other hand, disgust does not
map onto DSM-5 types: Health-related and Driving. In the play a role in other animal phobias (e.g., dogs, cats; Davey
DSM-5, these factors are instead captured by the BOther^ or 1992). Continued investigation of the relationship between
BSituational^ specific phobia types. fear and disgust can facilitate our understanding of the etiolo-
Both the five-factor model of the SPQ that emerged in the gy and maintenance of specific phobias, as well as lead to
EFA and the DSM-5 specific phobia conceptual framework potential revisions of their DSM-5 classification.
were deemed to be a good fit for the 43-item questionnaire. Furthermore, some objects and situations do not easily
The divergence from the DSM-5 specific phobia types in the fit into one DSM-5 type vs. another. For instance, research has
current study is not surprising, given that research on the called into question whether a fear of the dark is better
DSM-5 classification of types has been mixed. The DSM-5 captured by a situational or natural environment fear, or
groupings are based on research suggesting that specific pho- whether a fear of visiting the dentist should be grouped
bia types differ with respect to several variables, including age as a blood-injection-injury or Bother^ type (Antony et al.
of onset, gender, focus of fear, neurobiology/physiology, and 1997). In the current study, this difficulty was evidenced
comorbidity (LeBeau et al. 2010). In line with this, factor by the fact that neither the dark nor visiting the dentist
analytic and correlational studies have shown that DSM-5 loaded onto any factor.
specific phobia types do cluster together based on the above- In summary, the fact that the factor structure of the SPQ
mentioned variables. For example, blood-injection-injury did not map on perfectly to the DSM-5 specific phobia
items tend to cluster together, as do animal fears, natural phe- types may reflect a limitation of the DSM-5 types rather
nomena, and situational fears. However, as was the case in the than a limitation of the SPQ. Not all phobias can be clas-
current study, other studies have also shown different patterns sified easily into a DSM-5 type, and findings on the diag-
of clustering, calling into question the validity of existing nostic reliability of the DSM-5 types have been mixed. As
DSM-5 specific phobia types. such, and as has been suggested elsewhere (e.g., Antony et
For example, whereas some studies have shown inconsis- al. 1997), the DSM-5 typing should be refined and specific
tent patterns of clustering, others have shown that natural en- phobia diagnoses should include naming the specific pho-
vironment and situational types of specific phobias often clus- bia (e.g., storms) rather than the type (e.g., natural envi-
ter together (Fredrikson et al. 1996; Muris et al. 1999). In ronment type). This approach may be more simple, infor-
addition, it may be the case that items that did not load sub- mative, and useful in treatment planning.
stantively onto any factor (e.g., cats, dogs, the dark, visiting Regarding other psychometric properties, the SPQ demon-
the dentist) differ in important ways from some of the other strated high internal consistency, good convergent and dis-
objects and situations from the same DSM-5 types. For criminant validity, and good test-retest reliability. Further
 J Psychopathol Behav Assess

supporting the clinical validity of the SPQ, individuals with diagnostic data, some of which was recorded according
fears that met diagnostic criteria for specific phobias reported only to individual phobia type (e.g., BAnimals,^ without
significantly greater SPQ scores than those whose fears did any record of which animal the individual feared), and
not meet diagnostic criteria. In addition, SPQ subscale scores some by the distinct phobia (e.g., BSnakes^), and 2) there
(calculated according to both DSM-5 types and empirically were several different specific phobias such that breaking
derived types) differentiated specific phobia types. Finally, the down the data into exact specific phobias left an inade-
results of the ROC analyses for SPQ total scores and both quate sample size (e.g., only 4 individuals had a fear of
empirically or DSM-5 derived factor scores demonstrated dogs); this last point may account for why some common
good screening properties to differentiate individuals with fears (e.g., dogs, the dark) may have not mapped on to any
and without specific phobias/types. On the other hand, the of the factors. Therefore, the decision to create composite
utility of the FSS-III or PSRS to screen for specific phobias scores was most conducive to the aims of the current paper,
is questionable. For example, it has been demonstrated that the which were to report on the psychometric properties and
FSS-III scale is unable to correctly classify any individuals clinical utility of the SPQ.
with specific phobias in a specific phobia sample (Beck et Second, whereas the current study compared the SPQ
al. 1998). Moreover, the PSRS was not designed as a specific with other self-report measures of phobias, it did not in-
phobia screening tool; nor does it assess hallmarks of DSM-5 clude any behavioral measures. It would be useful for fu-
specific phobias such as fear and interference (Cutshall and ture studies to examine the extent to which scores on the
Watson 2004), and remains to be validated in a clinical SPQ are associated with subjective fear ratings or physio-
sample. logical measures during a behavioral approach test with the
Based on the findings of the current study, it would make feared phobic object or situation. Third, future studies
sense for the SPQ to be scored according to either the empir- could examine whether an elevation in the composite score
ically derived or DSM-5 derived types. However, given that for any one specific phobia (e.g., an elevated composite
the SPQ was developed to serve as a screening tool of specific score for BHeights,^ a composite score for BBridges,^
phobias consistent with DSM-5 typology, it is suggested that it etc.) can be a useful way to identify and screen for the
be scored according to the DSM-5 types (see Appendix B, for presence of that particular phobia. Lastly, it would be use-
scoring instructions). Specifically, creating composite scores ful for future studies to assess the SPQ’s generalizability to
for fear, consistent with DSM-5 typology, is an efficient and other populations (e.g., older adults, adolescents, culturally
psychometrically sound way to alert the clinician to the pres- diverse samples, etc.). Finally, future research could be
ence of a specific phobia type, and enhances the clinical utility used to further assess the validity and utility of the existing
of the measure. DSM-5 specific phobia types.
This study had a number of strengths. First, we used a Overall, the results of this study provide preliminary evi-
large, clinical sample in our factor analysis, receiver oper- dence of the psychometric properties of the SPQ in a clinical
ating characteristic analysis, and to assess the internal con- sample. Specifically, the present study found high internal
sistency of the SPQ. Second, the use of a clinical popula- consistency for the SPQ total scale and its subscales, good
tion further strengthens the utility of the measure, and sug- convergent and discriminant validity, and good test-retest re-
gests that the SPQ is suitable to use in a clinical population. liability. Preliminary results also suggest that the SPQ has the
In addition, the use of self-report measures to assess spe- ability to discriminate individuals with specific phobias, as
cific phobias was complemented by a diagnostic interview well as specify the type. Therefore, the SPQ can be a useful
(i.e., the SCID). tool in both research and clinical settings, and inform inter-
The study also had several limitations. First, given the vention and prevention efforts.
current study population for the factor analysis was a clin-
ical sample, further research is needed to determine wheth- Compliance with Ethical Standards
er the SPQ can accurately distinguish between clinical and
nonclinical populations with similar sensitivity and speci-
Conflict of Interest Melina Ovanessian, Nichole Fairbrother, Valerie
ficity. For example, the second author is in the process of Vorstenbosch, Randi McCabe, Karen Rowa and Martin Antony declare
assessing the ability of the SPQ to distinguish between that they have no conflict of interest.
clinical and nonclinical populations in a sample of preg-
nant and postpartum women. Furthermore, the decision to Ethical Approval All procedures performed in studies involving human
conduct an ROC analysis using composite fear scores for participants were in accordance with the ethical standards of the institu-
tional and/or national research committee and with the 1964 Helsinki
each DSM-5 type (e.g., Animals), rather than individual declaration and its later amendments or comparable ethical standards.
items, (e.g., snakes) may be viewed as a limitation.
However, the reason for creating composite scores was Informed Consent Informed consent was obtained from all individual
twofold: 1) we wanted to consolidate the available participants included in the study.
J Psychopathol Behav Assess

Appendix A

SPECIFIC PHOBIA QUESTIONNAIRE (SPQ)

Name ______________________________________________

Age____________ Sex: M F Date______________

Below are 45 different situations that people may fear and avoid. For some people, the fear and
avoidance occurs frequently enough to cause them interference in their daily lives. This
questionnaire has two parts:

1. For each of the 45 situations please indicate how fearful you are of each situation. Please
use the following scale as a guide.

FEAR SCALE

No fear 0 I am not at all fearful of this situation.

Mild fear 1 I am a little bit fearful of this situation, but my fear is manageable. I
probably would not avoid the situation, but I might take some minor
precautions to protect myself if I was in the situation.
Moderate fear 2 I am quite fearful of this situation. I would probably avoid the
situation from time to time, and I would take precautions to protect
myself in the situation.
Severe fear 3 I am very fearful of this situation or I would usually avoid it because
of fear.
Extreme fear 4 I’m extremely terrified of this situation or would always avoid it at
all cost.

2. For each of the 45 situations please indicate how much your fear interferes with your life (i.e.
work, social life, family, hobbies, etc.). Please use the following scale as a guide.

INTERFERENCE SCALE

No interference 0 This fear does not interfere with any aspect of my life.
Mild interference 1 This fear interferes with my life, but only in very small ways (e.g., it
only comes up once in a while, and doesn’t stop me from doing
most things that I want to do)
Moderate 2 This fear interferes with some activities that are important to me
interference (e.g., going to my favorite restaurant, visiting friends in the
hospital, taking certain vacations). The fear definitely causes
problems in my life.
Severe 3 This fear prevents me from completing some important life tasks
interference (e.g., getting married, having relationships and friendships,
 J Psychopathol Behav Assess

enjoying hobbies, attending important medical appointments,

going to school, working, caring for my children).
Extreme 4 This fear prevents me from completing many important life tasks
interference (e.g., getting married, having relationships and friendships,
enjoying hobbies, attending important medical appointments,
going to school, working, caring for my children).

LIFE
SITUATION FEAR
INTERFERENCE

Moderate

Moderate
Extreme

Extreme
Severe

Severe
None

None
Mild

Mild
1. High open places 0 1 2 3 4 0 1 2 3 4

2. The dark 0 1 2 3 4 0 1 2 3 4

3. Birds 0 1 2 3 4 0 1 2 3 4

4. Tunnels 0 1 2 3 4 0 1 2 3 4
5. Watching someone else get an
0 1 2 3 4 0 1 2 3 4
injection
6. Cats 0 1 2 3 4 0 1 2 3 4

7. Driving on highways 0 1 2 3 4 0 1 2 3 4

8. Dogs 0 1 2 3 4 0 1 2 3 4

9. Getting minor surgery 0 1 2 3 4 0 1 2 3 4

10. Watching someone else getting
0 1 2 3 4 0 1 2 3 4
stitches
11. Bees or wasps 0 1 2 3 4 0 1 2 3 4

12. Choking 0 1 2 3 4 0 1 2 3 4

13. Standing on a ladder 0 1 2 3 4 0 1 2 3 4

14. Swimming in a swimming pool 0 1 2 3 4 0 1 2 3 4

15. Visiting a hospital 0 1 2 3 4 0 1 2 3 4

16. Enclosed places 0 1 2 3 4 0 1 2 3 4

17. Giving blood 0 1 2 3 4 0 1 2 3 4

18. Having your blood pressure taken 0 1 2 3 4 0 1 2 3 4

19. Visiting the dentist 0 1 2 3 4 0 1 2 3 4
J Psychopathol Behav Assess

LIFE
SITUATION FEAR
INTERFERENCE

Moderate

Moderate
Extreme

Extreme
Severe

Severe
None

None
Mild

Mild
20. Vomiting 0 1 2 3 4 0 1 2 3 4
21. Watching someone else give blood 0 1 2 3 4 0 1 2 3 4

22. Flying in an airplane 0 1 2 3 4 0 1 2 3 4

23. Swimming in a lake or ocean 0 1 2 3 4 0 1 2 3 4

24. Watching surgery on television 0 1 2 3 4 0 1 2 3 4

25. Attending to someone else’s cut 0 1 2 3 4 0 1 2 3 4

26. Driving in bad weather 0 1 2 3 4 0 1 2 3 4

27. Looking out the window on the top
0 1 2 3 4 0 1 2 3 4
floor of a tall building
28. Attending to your own cut 0 1 2 3 4 0 1 2 3 4

29. Elevators 0 1 2 3 4 0 1 2 3 4

30. Moths or butterflies 0 1 2 3 4 0 1 2 3 4

31. Rodents (e.g., mice, rats) 0 1 2 3 4 0 1 2 3 4

32. Boating in deep water 0 1 2 3 4 0 1 2 3 4

33. Thunder and lightning 0 1 2 3 4 0 1 2 3 4

34. Driving in new places 0 1 2 3 4 0 1 2 3 4

35. Receiving an injection 0 1 2 3 4 0 1 2 3 4

36. Snakes 0 1 2 3 4 0 1 2 3 4

37. Bugs 0 1 2 3 4 0 1 2 3 4

38. Heavy rain 0 1 2 3 4 0 1 2 3 4

39. Worms 0 1 2 3 4 0 1 2 3 4

40. Bridges 0 1 2 3 4 0 1 2 3 4
41. Developing an illness 0 1 2 3 4 0 1 2 3 4

42. Blood tests 0 1 2 3 4 0 1 2 3 4

43. Spiders 0 1 2 3 4 0 1 2 3 4
44. Other (specify):
0 1 2 3 4 0 1 2 3 4
45. Other (specify):
0 1 2 3 4 0 1 2 3 4
____________________

© 2012 Nichole Fairbrother and Martin M. Antony

 J Psychopathol Behav Assess

Appendix B Brown, T. A., & Barlow, D. H. (2014). Anxiety and related disorders
interview schedule for DSM-5: Client interview schedule. New
York: Oxford University Press.
Bryant, F. B., Yarnold, P. R., & Michelson, E. A. (1999). Statistical
Specific Phobia Questionnaire Scoring Instructions methodology: VIII. Using confirmatory factor analysis (CFA) in
emergency medicine research. Academic Emergency Medicine:
The Specific Phobia Questionnaire (SPQ; Fairbrother and Official Journal of the Society for Academic Emergency Medicine,
6, 54–66.
Antony 2012) was designed to assess the extent of fear and
Cohen, D. C. (1977). Comparison of self-report and overt-behavioral
interference for a broad range of objects and situations. The procedures for assessing acrophobia. Behavior Therapy, 8, 17–23.
SPQ consists of 43 items, each rated using a 5-point Likert Cutshall, C., & Watson, D. (2004). The phobic stimuli response scales: a
scale on two dimensions: (1) level of fear, with scores ranging new self-report measure of fear. Behaviour Research and Therapy,
from 0 (no fear) to 4 (extreme fear) and (2) extent to which the 42, 1193–1201.
Davey, G. C. L. (1992). Characteristics of individuals with fear of spiders.
fear interferes with one’s daily life, with scores ranging from 0 Anxiety Research, 4, 299–314.
(no interference) to 4 (extreme interference). Field, A. P. (2013). Discovering statistics using IBM SPSS statistics: And
The SPQ is scored by combining fear ratings of each DSM- sex and drugs and rock ‘n’ roll (4th ed.). London, UK: Sage
5 specific phobia type to create five different composite Publications.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1995).
scores:
Structured clinical interview for DSM-IV Axis I disorders. New
York: Biometrics Research Department, New York State
Animals (sum of items 3, 6, 8, 11, 30, 31, 36, 37, 39, 43) Psychiatric Institute.
Natural Environment (sum of items 1, 2, 13, 14, 23, 27, First, M. B., Williams, J. B. W., Karg, R. S., & Spitzer, R. L. (2015).
32, 33, 38) Structured clinical interview for DSM-5—research version.
Arlington, VA: American Psychiatric Publishing.
Situational (sum of items 4, 7, 16, 22, 26, 29, 34, 40) Fredrikson, M., Annas, P., Fischer, H., & Wik, G. (1996). Gender and
Blood-Injection-Injury (sum of items 5, 9, 10, 15, 17, 18, ages differences in the prevalence of specific fears and phobias.
19, 21, 24, 25, 28, 35, 41, 42) Behaviour Research and Therapy, 3, 33–39.
Other (sum of items 12, 20) Geer, J. H. (1965). The development of a scale to measure fear. Behaviour
Research and Therapy, 3, 45–53.
Gerdes, A. B. M., Uhl, G., & Alpers, G. W. (2009). Spiders are special:
fear and disgust evoked by pictures of arthropods. Evolution and
Human Behavior, 30, 66–73.
Henson, R., & Roberts, J. (2006). Use of exploratory factor analysis in
References published research: Common errors and some comment on im-
proved practice. Educational and Psychological Measurement, 66,
393–416.
American Psychiatric Association. (2013). Diagnostic and statistical Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention
manual for mental disorders (5th ed.). Arlington, VA: American decisions in exploratory factor analysis: A tutorial on parallel anal-
Psychiatric Publishing. ysis. Organizational Research Methods, 7, 191–205.
Antony, M. M., Brown, T. A., & Barlow, D. H. (1997). Heterogeneity Havlicek, L. L., & Peterson, N. L. (1977). Effect of the violation of
among specific phobia types in DSM-IV. Behaviour Research and assumptions upon significance levels of the Pearson r.
Therapy, 35, 1089–1100. Psychological Bulletin, 84, 373–377.
Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P. Henry, J. D., & Crawford, J. R. (2005). The short-form version of the
(1998). Psychometric properties of the 42-item and 21-item versions Depression Anxiety Stress Scales (DASS-21): Construct validity
of the Depression Anxiety Stress Scales in clinical groups and a and normative data in a large non-clinical sample. British Journal
community sample. Psychological Assessment, 10, 176–181. of Clinical Psychology, 44, 227–239.
Fairbrother, N. & Antony, M. M. (2012). Specific Phobia Questionnaire. Hood, H. K., & Antony, M. M. (2012). Evidence-based assessment and
Unpublished scale. treatment of specific phobias in adults. In T. E. Davis III, T. H.
Baker, B. L., Cohen, D. C., & Saunders, J. T. (1973). Self-directed de- Ollendick, & L.-G. Öst (Eds.), Intensive one-session treatment of
sensitization for acrophobia. Behaviour Research and Therapy, 11, specific phobias (pp. 19–41). New York: Springer.
79–89. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covari-
Barlow, D. H. (2002). Anxiety and its disorders: the nature and treatment ance structure analysis: Conventional criteria versus new alterna-
of anxiety and panic. New York: Guilford Press. tives. Structural Equation Modeling, 6, 1–55.
Beck, J. G., Carmin, C. N., & Henninger, N. J. (1998). The utility of the Jöreskog, K. G., & Sörbom, D. (2012). LISREL 9.10 for Windows
fear survey schedule-III: an extended replication. Journal of Anxiety [Computer software]. Skokie, IL: Scientific Software
Disorders, 12, 177–182. International, Inc..
Bernstein, D. A., & Allen, G. J. (1969). Fear survey schedule (II): Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., &
Normative data and factor analyses based upon a large college sam- Walters, E. E. (2005a). Lifetime prevalence and age-of-onset distri-
ple. Behaviour Research and Therapy, 7, 403–407. butions of DSM-IV disorders in the national comorbidity survey
Bornas, X., & Tortella-Feliu, M. (1995). Descripcion y analisis replication. Archives of General Psychiatry, 62, 593–602.
psicometrico de un instrument de autoinforme para la evaluacion Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005b).
del miedo aVolar [Description and psychometric properties of a Prevalence, severity, and comorbidity of 12-month DSM-IV disor-
self- report assessment measure for fear of flying]. Psicología ders in the national comorbidity survey replication. Archives of
Conductual, 3, 67–86. General Psychiatry, 62, 617–627.
J Psychopathol Behav Assess

Klieger, D. M., & Franklin, M. E. (1993). Validity of the fear survey Osborne, J. W., & Costello, A. B. (2004). Sample size and subject to item
schedule in phobia research: a laboratory test. Journal of ratio in principal factors analysis. Practical Assessment, Research
Psychopathology and Behavioral Assessment, 15, 207–217. and Evaluation, 9, 1–9.
Kline, R. B. (2005). Principles and practice of structural equation Page, A. C., Bennett, K. S., Carter, O., Smith, J., & Woodmore, K. (1997).
modeling (2nd ed.). New York, NY: Guilford. The blood-injection symptom scale (BISS): assessing a structure of
LeBeau, R. T., Glenn, D., Liao, B., Wittchen, H., Beesdo-Baum, K., phobic symptoms elicited by blood and injections. Behaviour
Ollendick, T., & Craske, M. G. (2010). Specific phobia: a review Research and Therapy, 35, 457–464.
of DSM-IV specific phobia and preliminary recommendations for Radomsky, A. S., Rachman, S., Thordarson, D. S., McIsaac, H. K., &
DSM-V. Depression and Anxiety, 27, 148–167. Teachman, B. A. (2001). The Claustrophobia questionnaire. Journal
Lovibond, P. F., & Lovibond, S. H. (1995). The structure of negative of Anxiety Disorders, 15, 287–297.
emotional states: comparison of the Depression Anxiety Stress Stinson, F. S., Dawson, D. A., Chou, P. S., Smith, S., Goldstein, R. B.,
Scales (DASS) with the Beck Depression and Anxiety Inventories. Ruan, J., & Grant, B. F. (2007). The epidemiology of DSM-IV
Behaviour Research and Therapy, 33, 335–342. specific phobia in the USA: results from the national epidemiologic
Metzger, R. L., Miller, M. L., Cohen, M., & Sofka, M. (1990). Worry survey on alcohol and related conditions. Psychological Medicine,
changes decision making: The effect of negative thoughts on cogni- 37, 1047–1059.
tive processing. Journal of Clinical Psychology, 46, 78–88. Szymanski, J., & O’Donohue, W. (1995). Fear of Spiders Questionnaire.
Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Journal of Behavior Therapy and Experimental Psychiatry, 26, 31–
Development and validation of the Penn State Worry Questionnaire. 34.
Behaviour Research and Therapy, 28, 487–495. Trumpf, J., Margraf, J., Vriends, N., Meyer, A. H., & Becker, E. S.
Molina, S., & Borkovec, T. D. (1994). The Penn State Worry (2010). Specific phobia predicts psychopathology in young women.
Questionnaire: Psychometric properties and associated characteris- Social Psychiatry and Psychiatric Epidemiology, 45, 1161–1166.
tics. In G. C. L. Davey & F. Tallis (Eds.), Worrying: Perspectives on Weiss, B. W., Katkin, E. S., & Rubin, B. M. (1968). Relationship between
theory, assessment and treatment (pp. 265–283). Oxford, UK: John a factor analytically derived measure of a specific fear and perfor-
Wiley and Sons. mance after related fear induction. Journal of Abnormal Psychology,
Muris, P., Schmidt, H., & Merckelbach, H. (1999). The structure of spe- 73, 461–463.
cific phobia symptoms among children and adolescents. Behaviour Wolitzky-Taylor, K. B., Horowitz, J. D., Powers, M. B., & Telch, M. J.
Research and Therapy, 37, 863–868. (2008). Psychological approaches in the treatment of specific pho-
Muris, P., Mayer, B., Huijding, J., & Konings, T. (2008). A dirty animal is bias: a meta-analysis. Clinical Psychology Review, 28, 1021–1037.
a scary animal! Effects of disgust-related information on fear beliefs Wolpe, J., & Lang, P. (1977). Manual for the fear survey schedule. San
in children. Behaviour Research and Therapy, 46, 137–144. Diego: EdITS.