You are on page 1of 17

Human-Animal Interaction Bulletin

2017, Vol. 5, No. 1, 41-57

Personality Differences between

Dog People and Cat People

Andrea D. Guastello1, Denise D. Guastello2,

& Stephen J. Guastello3

University of Florida1, Carroll University2, & Marquette University3

Understanding the relationship between human personality and preferences for pet species
is important for bettering human-animal relationships, supporting animal welfare, and
supporting pet therapy. The present study examined personality differences between dog
people and cat people with the objective of addressing some discrepancies in previous
research that could be traced to the use of broadly defined versus narrowly defined traits.
Participants were 418 undergraduates who completed the Sixteen Personality Factors
Questionnaire (16PF) and a Pet Survey that classified them as dog or cat people. The 16PF
contains 15 narrowly defined primary personality traits, a brief measure of general
intelligence, and 5 global factors that are comparable in meaning to those of the Five Factor
Model. The principal differences were that the 352 dog people scored higher on warmth,
liveliness, rule consciousness, and social boldness compared to the 66 cat people. The latter
scored higher on general intelligence, abstractedness, and self-reliance. Overall, primary
traits corresponded to a pattern known to be associated with creative personalities and
produced better discrimination between the two groups than traits corresponded to the Five
Factor Model. Some traits were more salient for females than they were for males.

Keywords: cat, dog, personality, pet owner

Correspondence concerning this article should be addressed to Stephen J. Guastello, Ph.D.,

Department of Psychology, Marquette University, P. O. Box 1881, Milwaukee, WI 53201-
1881. phone: (414) 288-6900. email:

Introduction loneliness and higher scores on social

interaction (Connell & Lago, 1984; Gold-
In the United States, about 70-80 meir, 1986; Garrity, Stallones, Marx, &
million dogs and 74-96 million cats are Johnson, 1989; Zasloff & Kidd, 1994).
owned as pets. Approximately 37-47% of Additionally, Paden-Levey (1985)
all households in the United States own a concluded that the negative correlation
dog and 30-37% own a cat (American between pet-ownership and neuroticism
Society for the Prevention of Cruelty to indicates that emotional stability could be
Animals [ASPCA], 2016; American supported by pet ownership. Given the
Veterinary Medical Association numerous psychological benefits
[AVMA], 2016). Pets allegedly provide associated with pet ownership, it appears
numerous benefits to their owners. Most that pets may be an important coping
notably, pets can help their owners to mechanism for millions of people,
love and feel loved, contributing to the although the causative relationship
owner’s mental health and emotional between the two variables is not
wellbeing (Kidd & Kidd, 1980). conclusive.
Compared to non-owners, pet-owners Unfortunately, many human-animal
have lower scores on depression and relationships end in failure. Every year,

41 | H A I B

7.6 million pets enter shelters in the US, found no significant difference between
and 2.7 million pets are euthanized pet owners and non-owners, and they
(ASPCA, 2016). Understanding the concluded that pet-owners and non-
relationship between human personality owners may be victims of stereotyping.
and preferences for pet species is Other studies have examined a
important for selecting a pet that could variety of personality variables and their
provide physical and psycho-social relationship to self-identification as a cat
benefits, for limiting the number of person or a dog person (Gosling et al.,
surrendered or abandoned animals, and 2010; Reevy & Delgado, 2015). For
ensuring animal welfare (Gosling, Sandy, instance, it has been found that self-
& Potter, 2010; Kidd & Kidd, 1980). identified dog people rated themselves as
significantly less hostile and less
Pet Owners submissive than cat people (Woodward &
The general population subscribes to Bauer, 2007). Additionally, dog people
the idea of there being cat people and dog were higher on social dominance
people who differ on basic personality orientation and competitiveness (Alba &
characteristics, and many people are Haslam, 2015). These results are similar
generally comfortable categorizing to those of Bagley and Gonsman (2005)
themselves by these groups (Alba & who found that dog-loving men were
Haslam, 2015; Gosling et al., 2010; high in dominance and aggression, and
Woodward & Bauer, 2007). Some dog-loving women were high in
specialists in human-animal relationships dominance but low in aggression. A
assert that the preferences for dogs or sample of male and female dog people
cats, and particular breeds of dogs, are also rated themselves as more masculine
associated with underlying human on the abbreviated Bem Sex-Role
personality differences (Coren, 1998). Inventory than cat people. Dog people
Unfortunately, the current body of also rated themselves as more
research is disjointed and occasionally independent than did cat people (Perrine
contradictory, as is the case for research & Osbourne, 1998). Dog people are
pertaining to pet ownership generally. reportedly more extroverted (Gosling et
For example, Cameron and Mattson al., 2010; Reevy & Delgado, 2015), and
(1972) found that pet owners liked other agreeable (Reevy & Delgado, 2015;
people less than non-owners did, tended Zasloff, 1996). Gosling, Sandy, and
to prefer their pets to other people, and Potter (2010) found that dog people
thought themselves as less well-regarded scored higher on conscientiousness (i.e.,
by others. However, Jouber (1987) found self-discipline) than cat people, but
that pet owners reported spending more Reevy and Delgado (2015) did not
time with other people, and Miller and support that finding. Finally, people who
Lago (1990) noted a positive correlation indicated a dog as their favorite pet
between friendliness of pet dogs and the tended to express a stronger attachment
friendliness of their owners. As another than those who reported a cat to be their
example, Hyde, Kurdek, and Larson favorite pet (Johnson, Garrity, &
(1983) found that college-aged pet Stallones, 1992).
owners tend to have higher empathy and Conversely, cat people were high in
interpersonal trust scores than non-pet autonomy and low in dominance and
owners, but Johnson and Rule (1991) nurturing (Kidd & Kidd, 1980). Higher

42 | H A I B

ratings of neuroticism were associated that are more narrowly defined than the
with cat people rather than dog people FFM traits. The primary traits are
(Gosling et al., 2010; Reevy & Delgado, hierarchically organized into global traits
2015). Finally, males and females who that are similar in meaning to those of the
disliked cats rated themselves lower in FFM. On the 16PF, the global trait of
femininity on the abbreviated Bem Sex- extroversion is composed of narrowly
Role Inventory (Perrine & Osbourne, defined primary traits including warmth
1998). and social boldness. Someone who scores
high on warmth but low on social
Personality Constructs boldness may have the same extroversion
Gosling et al. (2010) proposed that score as someone who scores low on
some of the disagreement and holes in the warmth and high on social boldness. The
current body of literature are the result of: first person is more likely to be
(a) the studies using a broad range of experienced by others as warm, modest,
concepts and scales that make it difficult and concerned for others, while the
to compare findings across studies, and second person is more likely to be viewed
(b) the studies failing to systematically as bold, talkative, and attention seeking.
chose traits that represent the breadth of Thus the ways these two individuals
the personality spectrum. To resolve interact with their environments are
these issues, Gosling et al. (2010) qualitatively different (Cattell & Mead,
advocated the use of the Big Five or Five 2008). Additionally, some studies have
Factor Model (FFM; McCrae & Costa, found null or only small predictive
1985) to evaluate personality factors. relationships between FFM traits and
The FFM resulted from factor behavior, but have found stronger
analyzing trait terms from numerous predictive relationships between
personality tests that were thought to narrowly defined traits and the same
contain primary traits. The procedure behaviors. Such findings are noteworthy
isolated five traits: neuroticism, because it is possible that important
extroversion, openness, agreeableness, predictive relationships could be
and conscientiousness. The FFM overlooked if research only examines the
personality traits are broadly defined, and FFM traits (Dudley, Orvis, Lebiecki, &
each trait encompasses several facets of Cortina, 2006; Guastello, 2009;
personality. Although useful for Guastello, Guastello, & Guastello, 2014;
understanding general patterns, relying Szymura, 2010). The advantages for
only on the broadly defined FFM traits detecting meaningful relationships and
may obscure other important individual interpreting findings are possible because
differences. For instance, if two the 16PF primary factors were
individuals have the same extroversion constructed to be factorially independent,
score, one might assume that the two but potentially correlated, which was not
individuals interact with their social the case with the development of the FFM
environment in similar ways. However, (Cattell & Mead, 2008).
this assumption may be premature. The present study seeks to
The Sixteen Personality Factor incorporate the recommendation by
Questionnaire (16PF; Cattell, 1994; Gosling et al., but also to improve
Cattell, Eber, & Tatsuoka, 1970) offers an predictive accuracy, by using the 16PF
alternative taxonomy of primary traits (Cattell, Cattell, & Cattell, 1994) to

43 | H A I B

measure personality factors. The 16PF I) (Conn & Rieke, 1994, p. 45).
makes use of narrowly defined Therefore, evaluating gender differences
personality traits and measures 16 in the analysis is important for
primary factors of personality including: determining the generalizability of the
warmth, reasoning, emotional stability, findings (Hergovich, Mauerer, & Riemer,
dominance,liveliness, rule- 2011).
consciousness, social boldness, The following hypotheses were
emotional sensitivity, vigilance, investigated in the present study:
abstractedness, privateness, app- • Hypothesis 1: There are global factor
rehension, openness to change, self- personality trait differences
reliance, perfectionism, and tension. The (consistent with the FFM) between
reasoning factor is a brief measure of self-identified cat people and dog
general intelligence. Although general people. Previous research suggests
intelligence is not usually considered a that dog people would be more
personality trait, Cattell et al. (1970) extroverted, agreeable, and possibly
argued that one must know the general more conscientious, and that cat
intelligence level of a person in order to people would be more neurotic and
interpret the personality profile in a show greater levels of openness.
meaningful way. The 16PF primary • Hypothesis 2: There are primary
factors can be interpreted in a profile, personality trait (of the 16PF)
used individually in a prediction model, differences between self-identified
or used to calculate secondary factors that cat people and dog people. The
are similar to FFM personality traits. The factors that were most likely to be
16PF is unique because it ensures the important, as indicated by previous
precision of the primary personality traits research, were friendliness (16PF
but also allows researchers to investigate Factor F), dominance (Factor E),
the FFM traits and compare the results to emotional sensitivity (Factor I), rule-
other studies that have used FFM traits consciousness (Factor G), and self-
(Cattell & Mead, 2008). sufficiency (autonomy, Factor Q2).
• Hypothesis 3: The differences in the
Gender primary or global personality
There is reason to consider that differences of cat people and dog
gender is related to preferences for either people are moderated by gender. It is
cats or dogs. Several studies found that possible that some traits could be
gender moderates the relationship relevant for one gender but not the
between personality factors and pet other.
preferences (Bagley & Gonsman, 2005;
Kidd & Kidd 1980; Perrine & Osbourne Methods
1998). Additionally, men and women in
the general population differ on the FFM Participants
traits (Schmitt, Realo, Voracek, & Alik, The 600 participants in the study
2008). Gender differences are also were undergraduate students who were
observed on two of the primary traits of enrolled in psychology courses at a small,
the 16PF; women tend to score lower than private, Midwestern US University. The
men on dominance (Factor E) and higher sample was 25.8% male and 74.2%
than men on emotional sensitivity (Factor female. Self-reported racial/ethnic

44 | H A I B

identities are as follows: 89.8% White, month test-retest reliability coefficients

7.3% Hispanic, 1.8% African American, ranged from .64 to .79 (Conn, 1994).
0.7% Asian, and 0.3% reported belonging The correspondence between the
to another racial or ethnic identity. Their 16PF global factors and the FFM was
ages ranged from 18 to 22 years. determined by a factor analysis of global
traits and FFM facets (Conn & Rieke,
Materials 1994). There is a direct relationship
16PF. The 16PF measures 16 between three of the 16PF global factors
primary factors and 5 global factors of and three of the FFM traits. The global
personality. The 16 primary factors are: factor of introversion/extroversion
warmth (Factor A), reasoning (Factor B), corresponds with the FFM trait
emotional stability (Factor C), dominance extroversion (factor loading = .67). The
(Factor E), liveliness (Factor F), rule- global factor of anxiety corresponds to
consciousness (Factor G), social boldness the FFM trait neuroticism (factor loading
(Factor H), emotional sensitivity (Factor = .85). The global factor of self-control
I), vigilance (Factor L), abstractedness corresponds to the FFM trait of
(Factor M), privateness (Factor N), conscientiousness (factor loading = .72).
apprehension (Factor O), openness to The other two global factors of the 16PF
change (Factor Q1), self-reliance (Factor are inversely related to the FFM traits.
Q2), perfectionism (Factor Q3), and The global factor of tough-mindedness is
tension (Factor Q4). Factor B contains 15 inversely related to the FFM trait
items, Q1 contains 14 items, and the other openness: higher scores in tough-
primary scales contain 10 or 11 items mindedness correlate to lower scores in
each. The five global factors are openness (factor loading = -.70). The
calculated as weighted combinations of global factor of independence is inversely
primary scales: introversion/ related to the FFM trait agreeableness:
extroversion, low anxiety/high higher scores on independence correlate
anxiety, receptivity/tough-mindedness, to lower scores on agreeableness (factor
accommodation/independence, and lack loading = -.72; Conn & Rieke, 1994,
of restraint/self-control (Cattell, 1994, p. p. 134).
16). All 16PF scales are normed and Pet Survey. The Pet Survey was a
calibrated with means of 5.5 and standard researcher-designed questionnaire, which
deviations of 2.0. assessed participants’ history of pet
The 16PF measurements have high ownership, liking of specific types of
construct validities and strong test-retest pets, and attitudes regarding pet
reliabilities (Conn & Rieke, 1994). Two ownership. Demographic information
general population samples (N = 820 and (i.e., gender and racial/ethnic identity)
2500) and one sample of college was also collected on the Pet Survey.
undergraduates (N = 1340), yielded Examples of history of pet ownership
internal consistency reliability questions include: “Did you have a pet
coefficients for the 16PF primary factors growing up?” and “If you answered yes,
ranging from .68 to .87 (Conn, 1994, p. check all that apply: dog, cat, rabbit, bird,
81), and from .70 to .86 when all three fish, ferret, tortoise, snake, horse, pig,
samples were combined. In a sample of lizard, turtle, duck, frog, hamster/gerbil,
undergraduate students (N = 159), two- other.”

45 | H A I B

For the questions assessing the liking Data Analysis

of specific types of pets, participants were All statistical analyses were
asked to rank their liking of different performed using the Statistical Package
types of pets on a 1-5 Likert scale ranging for the Social Sciences (SPSS). For all
from “dislike very much” to “like very significance testing, an alpha of .05 or
much.” Sample items include: “cats,” greater was used to designate an effect as
“dogs,” and “tortoise.” Participants were not statistically significant.
also asked to rate on a scale from 1 Data were analyzed in three steps.
(strongly disagree) to 5 (strongly agree) First, personality differences between cat
the amount that they agreed with the people and dog people were examined
statements: “I am a cat person,” and “I am including gender as a covariate in a series
a dog person.” of ANCOVAs. The five global traits and
For the questions assessing attitudes the 16PF primary traits were entered as
regarding pet ownership, participants the independent variables and the
were asked to select their response on a categorical variable as a cat person or a
1-5 Likert scale ranging from “strongly dog person was the dependent variable.
disagree” to “strongly agree.” Sample The ANCOVA analyses treated gender as
items include: “I would like to own a pet a variable that could obscure or confound
in the future,” “a pet requires a lot of work differences between dog and cat people if
and care,” and “the benefits of owning a the males and females in the study scored
pet outweigh the drawbacks.” differently on the traits for reasons other
than pet preference.
Procedure The second step used a stepwise
The study was approved by the discriminant analysis to identify the
Institutional Review Board before any smallest number of traits that could
data collection began. Participants were maximally distinguish dog people and cat
given about five minutes to review and people. Separate analyses were
sign an informed consent form. conducted for primary and global traits.
Participants who signed the informed Males and females were included
consent form then completed the 16PF together in these analyses. The third step
and the pet survey, which also collected repeated the stepwise discriminant
demographic information. The analyses for males and females
questionnaires were completed in a small separately. These analyses would
group setting and took about 45 minutes determine whether traits that distinguish
to complete by paper and pencil cat and dog people apply to both genders
administration. No identifying or just one.
information was collected on the surveys.
Additionally, the informed consent forms Results
and the surveys were collected and kept
in separate locked cabinets to ensure The responses from the Pet Survey
confidentiality. Participants were given were analyzed to assess whether
course credits for participation in participants were cat people or dog
psychology research studies, including people. On the Pet Survey participants
this study; however, other study options were asked to rate on a scale from 1
and alternative assignments were also (strongly disagree) to 5 (strongly agree)
available for equivalent course credit. the amount that they agreed with the

46 | H A I B

statements: “I am a cat person,” and “I am Construct validity for self-report of

a dog person.” People who ranked dog person or cat person was established
themselves as a 4 (agree) or 5 (strongly as follows. Participants who reported a
agree) for either “I am a cat person” (N = strong liking for cats identified
352) or “I am a dog person” (N = 66) were themselves as a cat person (r = .815,
classified in the corresponding category p < .001) whereas, participants who
for subsequent analyses. People who reported a strong liking for dogs
rated themselves as a 4 or 5 for both identified themselves as a dog person
questions were classified as “both” (N = (r = .780, p < .001).
140) and were eliminated from The majority of dog people (81.8%)
subsequent analyses because the overlap had a dog as a family pet while growing
between dog people and cat people could up, and many reported having a cat or
obscure differences between the two other type of pet (e.g. gerbil, bird, fish).
primary groups. Participants who ranked A similar majority of cat people (80.3%)
themselves as 1 (strongly disagree), had a cat as a family pet while growing
2 (disagree), or 3 (neutral) for both up and also reported having dogs or other
questions were classified as “neither” (N pets. A breakdown of previous and
= 42) and were eliminated from current pet ownership at the family or
subsequent analyses. A total of 418 permanent residence, current residence,
participants were retained in the or both places appears in Table 1.
subsequent analysis: 84.2% of whom
self-identified as dog people, and 15.8%
of whom self-identified as cat people.

Table 1: Number and Percentages of Participants Reporting Previous and Current Pet Ownership.
Owned a pet Cat People Dog People
Number (%) Number (%)
While growing up Total 64 (97.0) 331 (94.0)
Dog 36 (54.5) 288 (81.8)
Cat 53 (80.3) 102 (29.0)
Other pet 50 (75.8) 261 (74.1)

Current ownership At family or

permanent residence 34 (51.5) 184 (52.8)
Current residence only 7 (10.6) 14 (4.0)
Both of above 3 (4.5) 67 (19.0)
Neither 14 (21.2) 86 (24.4)

Pet at family residence Dog 11 (16.7) 224 (63.6)

Cat 39 (59.0) 44 (12.5)
Other 16 (24.2) 78 (22.2)

Pet at current
residence Dog 2 (3.0) 39 (11.0)
Cat 15 (22.7) 14 (3.9)
Other 7 (10.6) 49 (13.9)

47 | H A I B

Cat People versus Dog People: sensitivity (Factor I), abstractedness

Identifying differences (Factor M), and self-reliance (Factor Q2);
The ANCOVA analyses identified see table 2. The effect sizes for the
two statistically significant differences statistically significant variables were
among the global traits: dog people were small, with dog versus cat people
more extroverted and more tough-minded accounting for 1-4% of the variance in a
than cat people. There were statistically single personality trait. The covariate for
significant differences between cat gender was not significant (p > .05) for
people and dog people on eight primary any of those variables.
personality factors: warmth (Factor A),
reasoning (Factor B), liveliness (Factor
F), rule-consciousness (Factor G), social
boldness (Factor H), emotional

Table 2: ANCOVA for the 16 Primary and 5 Secondary Factors.

Mean (SD) Mean (SD) F Partial Eta

Cat People Dog People Squared
Primary Factors
Warmth (A) 5.73 (2.01) 6.18 (1.64) 6.54* 0.02
Reasoning (B) 5.89 (1.66) 5.05 (1.49) 16.47** 0.04
Emotional Stability (C) 4.38 (1.52) 4.72 (1.60) 2.22 0.01
Dominance (E) 4.59 (1.61) 4.95 (1.77) 2.00 0.01
Liveliness (F) 5.83 (1.97) 6.59 (1.79) 9.00** 0.02
Rule-Consciousness (G) 4.15 (1.61) 4.76 (1.39) 11.21** 0.03
Social Boldness (H) 4.86 (1.81) 5.60 (2.02) 7.64** 0.02
Emotional Sensitivity (I) 6.20 (1.54) 5.42 (1.74) 10.36** 0.02
Vigilance (L) 6.33 (1.54) 6.43 (1.65) 0.109 0.00
Abstractedness (M) 6.44 (1.61) 5.71 (1.66) 13.37** 0.03
Privateness (N) 5.52 (1.83) 5.55 (1.88) 0.01 0.00
Apprehension (O) 6.52 (1.37) 6.25 (1.55) 1.02 0.00
Openness to Change (Q1) 5.62 (1.90) 5.33 (1.84) 1.56 0.00
Self-Reliance (Q2) 5.86 (1.67) 5.22 (1.72) 8.07** 0.02
Perfectionism (Q3) 4.85 (1.75) 5.05 (1.82) 0.94 0.00
Tension (Q4) 5.88 (1.63) 5.74 (1.46) 0.34 0.00
Secondary Factors
Introversion/ 5.43 (1.92) 6.12 (1.75) 9.62** 0.02
Low Anxiety/ 6.71 (1.52) 6.44 (1.53) 1.29 0.00
High Anxiety
Receptivity/ 4.81 (1.47) 5.48 (1.66) 7.99** 0.02
Accommodation/ 4.97 (1.57) 5.33 (1.72) 2.24 0.01
Lack of Restraint/ 4.30 (1.58) 4.70 (1.43) 5.23* 0.01
*p < .05, **p < .01

48 | H A I B

Cat People versus Dog People: and dog people. The difference between
Predictive Models dog people and cat people was
A stepwise discriminant analysis was statistically significant (Wilks Λ = .90,
conducted to determine if the five global χ2 = 42.43, df = 4, p < .001). The model
traits of the 16PF could predict group accounted for 9.7% of the variance
membership of cat people and dog people between groups (canonical correlation =
without separating males and females. .312). Using the leave-one-out cross-
The difference between dog people and validation function in SPSS, the model
cat people was statistically significant correctly classified 84.9% of cross-
(Wilks Λ = .952, χ2 = 20.631, df = 2, validated cases. The stepwise model
p < .001). The model accounted for 4.8% revealed that reasoning (Factor B),
of the variance between groups liveliness (Factor F), rule consciousness
(canonical correlation = .22). The model (Factor G), and emotional sensitivity
correctly classified 84.0% of cross- (Factor I) were driving the differences
validated cases. The stepwise model between cat people and dog people. Cat
revealed that dog people scored higher people scored higher on reasoning
than cat people on tough-mindedness (Factor B) and emotional sensitivity
versus receptivity and extroversion (Factor I) than dog people (Wilks Λ = .96,
versus introversion (Wilks Λ = .98, F(1,416) = 17.02, p < .001; Wilks Λ =
F(1, 416) = 9.22, p < .001; Wilks Λ = .95, .90, F(4, 413) = 11.14, p < .001). Dog
F(2, 415) = 10.58, p < .001); see table 3. people scored higher on liveliness (Factor
A stepwise discriminant analysis was F) and rule consciousness (Factor G) than
conducted to determine if the 16 primary cat people (Wilks Λ = .92, F(3, 414) =
personality factors of the 16PF could 12.63, p < .001; Wilks Λ = .94, F(2, 415)
predict group membership of cat people = 13.10, p < .001); see table 3.

Table 3: Significant Factors for Both Genders on the Discriminant Function.

Mean (SD) Mean (SD) F Wilks’ Lambda

Cat People Dog People

Primary Factors
Reasoning (B) 5.89 (1.66) 5.05 (1.49) 17.02** 0.96
Rule-Consciousness (G) 4.15 (1.61) 4.76 (1.39) 13.10** 0.94
Liveliness (F) 5.83 (1.97) 6.59 (1.79) 12.63** 0.92
Emotional Sensitivity 6.20 (1.59) 5.42 (1.74) 11.14** 0.90
Secondary Factors
Receptivity/ 4.81 (1.47) 5.48 (1.66) 9.22** 0.98
Tough- Mindedness
Introversion/ 5.43 (1.92) 6.12 (1.75) 10.58** 0.95
**p < .01

49 | H A I B

Table 4: Significant Factors for Men Only on the Discriminate Function.

Primary Factors Mean (SD) Mean (SD) F Wilks’ Lambda

Cat People Dog People
Rule-Consciousness (G) 3.56 (1.37) 4.50 (1.29) 7.26** 0.95
Emotional Sensitivity (I) 5.06 (1.29) 4.05 (1.54) 7.09** 0.90
**p < .01

Males Only: The stepwise being dog people and 17.2% reported
discriminant analyses were repeated being cat people. For the analysis with the
examining only the male participants’ five global factors using only the female
responses. There were 127 male participants, there were statistically
participants: 87.4% reported being dog significant differences between the cat
people and 12.6% reported being cat and dog groups (Wilks Λ = .94,
people. For the analysis with the five χ2 = 18.54, df = 2, p < .001). The model
global factors, the difference between the accounted for 6.3% of the variance
two groups was not statistically between groups (canonical correlation =
significant (Wilks Λ = .97, χ2 = 3.78, .25). Using the leave-one-out cross-
df = 1, p > .05). For the analysis with 16 validation function, the model correctly
primary personality factors, the classified 82.5% of cross-validated cases.
difference between the two groups was As in the model for both genders, the
statistically significant (Wilks Λ = .897, stepwise model revealed that dog people
χ2 = 13.42, df = 2, p < .001). The model scored higher than cat people on tough-
accounted for 10.2% of the variance mindedness versus emotional sensitivity
between groups (canonical correlation = and extroversion versus introversion
.32). Using the leave-one-out cross- (Wilks Λ = .96, F(1,289) = 12.69,
validation function in SPSS, the model p < .001; Wilks Λ = .94, F(2,288) = 9.58,
correctly classified 85.8% of cross- p < .001); see table 5 (following page).
validated cases. The stepwise model For the analysis with the 16 primary
revealed that rule consciousness personality factors, the difference
(Factor G) and emotional receptivity between the two groups was statistically
(Factor I) comprised the differences significant (Wilks Λ = .893, χ2 = 32.58,
between cat people and dog people. As in df = 3, p < .001). The model accounted
the model for both genders, cat people for 10.7% of the variance between groups
scored higher on emotional sensitivity (canonical correlation = .33). Using the
(Factor I) than dog people (Wilks Λ = .90, leave-one-out cross-validation function,
F(2, 124) = 7.09, p < .01). Dog people the model correctly classified 84.5% of
scored higher on rule consciousness cross-validated cases. The stepwise
(Factor G) than cat people (Wilks Λ = model revealed that reasoning (Factor B),
.95, F(1,125) = 7.26, p < .01); see table 4. rule consciousness (Factor G), and self-
Females Only: The stepwise reliance (Factor Q2) were driving the
discriminant analyses above were differences between cat people and dog
repeated examining only the female people. As in the model for both genders,
participants’ responses. There were 291 cat people scored higher on reasoning
female participants: 82.8% reported (Factor B) (Wilks’ Λ = .94, F(1, 289)

50 | H A I B

=18.65, p < .001), and dog people scored (2015) did not find. Dog people scored
higher on rule consciousness (Factor G) higher on tough-mindedness (lower on
(Wilks’ Λ = .89, F(3, 287) =11.48, FFM trait openness) than cat people,
p < .001), In the women-only model, cat which was consistent with Gosling et al.
people also scored higher on self-reliance (2010) and Reevy and Delgado (2015).
(Factor Q2) than dog people There were no significant differences
(Wilks’ Λ = .91, F(2,288) = 14.91, between cat people and dog people on
p < .001); see table 5. anxiety (i.e., FFM trait neuroticism),
which contradicts previous findings
Discussion (Gosling & Bonnenburg, 1998; Reevy &
Delgado, 2015).
Global Factors The results for the secondary factors
As predicted in the second are also consistent with the interpretation
hypothesis, there were several differences that cats may be more appealing to
between cat people and dog people on the creative people. The FFM trait openness
secondary factors of the 16PF: is often positively associated with
introversion/extroversion, receptivity/ creative behavior, and the FFM trait
tough-mindedness, and impulsivity/self- conscientiousness is negatively
control. These secondary factors correlated with creative behavior (Feist,
corresponded to the FFM traits 1998; Guastello, 2009). This same pattern
of extroversion, openness, and of results was observed here in the
conscientiousness respectively. personality profile of cat people.
Consistent with previous findings Several stepwise discriminate
(Gosling et al., 2010; Reevy & Delgado, analyses were conducted to see which
2015), dog people were more extroverted. factors were the most prominent
Dog people were higher in self-control personality differences and which factors
(higher on FFM trait conscientiousness) could be eliminated due to overlap in
than cat people, which Gosling et al. variance explained. The primary factors
(2010) found, but Reevy and Delgado that were retained were: reasoning, rule-

Table 5: Significant Factors for Women Only on the Discriminate Function.

Mean (SD) Mean (SD) F Wilks’ Lambda

Cat People Dog People
Primary Factors
Reasoning (B) 6.08 (1.63) 5.08 (1.46) 18.65** 0.94
Self-Reliance (Q2) 6.16 (1.53) 5.12 (1.74) 14.91** 0.91
Rule-Consciousness (G) 4.34 (1.65) 4.88 (1.42) 11.48** 0.89

Secondary Factors
Introversion/ 5.35 (1.84) 6.34 (1.77) 12.69** 0.96
Receptivity/ 4.68 (1.24) 5.19 (1.61) 9.58** 0.94
**p < .01

51 | H A I B

consciousness, liveliness, and emotional agreeable than cat people. Cat people also
sensitivity. The direction of each scored lower on warmth and social
relationship remained the same as boldness, and thus tend to be more
described above. The secondary factors impersonal, detached, and shy than dog
that were retained were tough- people. Cat people scored lower on rule-
mindedness and extroversion. Again, the consciousness indicating that they tend to
direction of each relationship remained be nonconforming and disregarding of
the same as described above. These social conventions compared to dog
results show that, although cat and dog people, according to the standard
people differ on a variety of personality interpretations of 16PF traits (Cattell,
factors, only a few constructs are needed 1994). Additionally, cat people scored
to predict group membership as a cat lower on liveliness, suggesting that they
person or a dog person. It is important to are more serious and restrained than dog
eliminate factors that overlap in variance people, who are more animated and
explained in order to develop expressive. Finally, cat people scored
parsimonious interpretations of our higher than dog people on abstractedness
findings. In this case, the findings suggest and emotional sensitivity, which suggests
that people who score higher on rule that cat people may be more imaginative
consciousness and liveliness tend to and sentimental than dog people, who
prefer dogs and people who score higher tend to be more grounded and pragmatic.
on reasoning and emotional sensitivity The foregoing implications from the
tend to prefer cats. 16PF trait constructs could open new
avenues of research on dog and cat
Primary Factors people.
As predicted in the first hypothesis, Taken together, these findings
cat people and dog people differed on describe the personalities of the average
several of the primary factors of the cat person as shy, solitary, impersonal,
16PF, some of which had not been serious, and nonconformist, but also
identified in previous research: warmth creative, sentimental, independent, and
(Factor A), reasoning (Factor B), self-sufficient. Conversely, these
liveliness (Factor F), rule-consciousness findings describe dog people as
(Factor G), social boldness (Factor H), grounded, pragmatic, and dutiful, as well
emotional sensitivity (Factor I), as warm, outgoing, sociable, expressive,
abstractedness (Factor M), and self- and group oriented. According to Coren
reliance (Factor Q2). Cat people scored (1998) people tend to have better
higher on self-reliance indicating that relationships with dog companions that
they tend to be more solitary, have personalities similar to their own. It
individualistic, and self-sufficient than is not known whether the principle of
dog people, according to the standard personality matching extends to pets
interpretation of 16PF factors (Cattell, other than dogs; this speculation would
1994). This finding would be consistent need to be verified in future research
with Kidd and Kidd (1980) who found efforts, which would also need to take
that cat people tend to be more into account individual differences in the
autonomous than dog people, and pets themselves.
consistent with Reevy and Delgado Surprisingly, cat people scored
(2015) who found dog people to be more higher on reasoning (i.e., general

52 | H A I B

intelligence) than dog people by a margin Gender

of 0.42 SD. A possible interpretation of The third hypothesis predicted that
the results is that dogs are more appealing gender would moderate the relationship
to a broader range of the population, between personality factors and
meaning that dogs appeal to people preference for cats or dogs. Although
across the reasoning continuum, while gender was not a statistically significant
cats appeal mainly to people on the higher variable in the ANCOVA analyses,
end of the reasoning continuum. Another gender differences appeared when the
interpretation may be that when the stepwise discriminate analyses were
reasoning results are combined with the conducted on men and women separately.
results for the other primary traits, it When considering the primary factors,
appears that the personality profile of cat rule-consciousness appeared to be a
people is similar to the personality profile salient factor for both men and women.
of creative people. The 16PF profile of However, emotional sensitivity was only
creative people includes: higher a salient factor for men (high among cat
reasoning, greater self-reliance, greater people), and reasoning was only salient
abstractedness, lower attention to rules for women (higher among cat people).
(e.g., social conventions), and greater Liveliness, which was a salient factor
emotional sensitivity (Guastello, 2009; when both genders were observed
Rieke, Guastello, & Conn, 1994). together, was not a salient factor for either
Therefore, the personality profiles of gender when the genders were considered
creative people and cat people are separately. Finally, self-reliance was not
remarkably similar, suggesting that observed when both genders were
people who are considered creative may observed together, but was a salient factor
tend to prefer cats as pets. for the women when the genders were
Personality traits only accounted for conducted separately.
1-4% of the variance in pet preferences When considering the global factors,
when the traits were considered tough-mindedness and extroversion were
separately. The stepwise discriminant discovered as significant factors
analysis showed that efficient predicting cat people and dog people
combinations of primary traits accounted when both genders were considered
for 9.7-10.7% of the variance. Secondary together. However, when each gender
or global traits, in contrast, only was analyzed separately, none of the
accounted for 4.8% of the variance for global factors were significant predictors
both genders together, 6.3% for females for men. When women’s data were
only, and no variance for males only. The analyzed separately, the same two
comparative results lend further support factors, tough-mindedness and
to the principle that primary traits are extroversion, were significant predictors.
more closely connected to actual This finding suggests that differences on
behaviors, and that they should be given the global factors between cat people and
greater priority in pet-related and other dog people might only apply to women.
research in which personality traits are of This interpretation is consistent with
interest. findings from Bagley and Gonsman
(2005) who found differences between
cat people’s personalities and dog

53 | H A I B

people’s personalities among women but people in this sample identified as White.
not among men. Therefore, race and ethnicity were not
The differences that were noticed included as variables due to lack of
when the data from the men and women statistical power. It is possible that racial
were analyzed separately suggest that or ethnic differences could be important,
while personality differences between cat however. Future research should use
and dog people exist for both genders, the more diverse samples to explore these
nature of these differences is not differences and make the findings
necessarily the same for both genders. generalizable beyond White Americans.
Previous research has determined that in The present study was confined to an
general men and women tend to differ on undergraduate population between the
some personality traits (Schmitt et al., ages of 18-22. It is not known how
2008) such as emotional sensitivity attachment needs for older adults could
(Conn & Rieke, 1994). Therefore, it is be different or how the connection
important to consider gender in research between pet preferences and personality
examining these traits. Thus the traits could change in later life. Future
observation above, that distinguishing research should thus consider broader age
emotional sensitivity between cat people ranges.
and dog people was important for men In this study, cat people scored
but not for women, is likely related to the higher on reasoning (i.e., general
natural difference between men and intelligence) than dog people. This is a
women on the trait of emotional new finding, and to our knowledge, this
sensitivity. Finally, if this line of research is the first study to examine differences in
is to be used to help individuals select reasoning between cat people and dog
companion animals, a better people. In this study, reasoning was
understanding of gender differences measured using the 16PF, which has a
would be important for helping reasoning factor (Factor B). The
individuals choose animals that will meet reasoning factor has been found to be
their unique needs. valid (Rieke & Conn, 1994). However,
future studies may consider using more
Limitations and Future Research sophisticated measures of general
There were some limitations to this intelligence or multi-factor measures of
study that suggest opportunities for future intelligence to add support and expand
research. For instance, the final sample this finding further. For instance, one
consisted of 84.2% dog people. Although could explore whether non-verbal
there was sufficient power for the intelligence is more relevant than verbal
analyses, future researchers should try to intelligence given that animals do not
collect a more balanced sample. communicate the way humans
Additionally, this study only included communicate with each other.
people who identified as either a cat Previous research has shown that
person or a dog person. People who pets could provide physical and
identified as “neither” or “both” were not psychological health benefits to their
examined in this study; additional owners (Connell & Lago, 1984; Garrity et
research is needed to understand how al., 1989; Gold-meir, 1986; Zasloff &
these individuals compare with the dog Kidd, 1994). Continued research into
and cat people. Moreover, 89.8% of the human-animal relationships is important

54 | H A I B

for optimizing the benefits of companion advantage because it was possible to

animals (e.g., pets, therapy animals, identify the creative personality
emotional support animals) and configuration as the pattern
personality factors related to pet distinguishing dog people from cat
preferences is an avenue of such research. people. Future research is needed to
Further research could be influential in replicate these findings and establish their
helping individuals who are considering generalizability.
getting a companion animal to choose
one that would be compatible with their
own personalities, following Coren References
(1998), and thus provide maximum Alba, B., & Haslam, N. (2015). Dog
benefits. Additionally, 7.6 million people and cat people differ on
companion animals enter shelters in the dominance-related traits.
United States every year (ASPCA, 2016), Anthrozoös, 28, 37-44.
but improving the match between the American Society for the Prevention and
companion animal and owner may limit Cruelty to Animals. (2016). Shelter
the number of surrendered and intake and surrender. Retrieved
abandoned animals and improve animal January 18, 2016 from
Conclusion surrender/pet-statistics.
American Veterinary Medical
In conclusion, this study found that Association (2016). U. S. pet
cat people and dog people differ on ownership statistics. Retrieved
several primary and secondary March 27, 2016 from
personality traits. Additionally, some of
the salient traits seem to differ based on s/Statistics/Pages/Market-research-
gender. The most important personality statistics-US-pet-ownership.aspx
factors for predicating group membership Bagley, D. K., & Gonsman, V. L. (2005).
among males were rule-consciousness Pet attachment and personality type.
and emotional sensitivity. The most Anthrozoös, 18, 28-42.
important personality factors for Cameron, P., & Mattson, M. (1972).
predicting group membership among Psychological correlates of pet
females were reasoning, self-reliance, ownership. Psychological Reports,
rule-consciousness, extroversion, and 30, 286.
tough-mindedness. Cattell, H. E. P. (1994). Development of
The present study added to the the 16PF Fifth Edition. In S. R. Conn
growing list of research findings that & M. L. Rieke (Eds.), 16PF Fifth
show better predictive accuracy for Edition technical manual (pp. 3-20).
primary personality traits than for Champaign, IL: Institute for
secondary traits. The relative accuracy of Personality and Ability Testing.
the primary traits (percentage of variance Cattell, H. E. P., & Mead, A. D. (2008).
accounted for) in the discriminant The Sixteen Personality Factor
analysis with both genders was double Questionnaire (16PF). In G. J. Boyle,
that obtained for the secondary traits. G. Matthew, & D. H. Sakofske
Primary traits also offered a conceptual (Eds.), The Sage handbook of

55 | H A I B

personality theory and assessment, validity of narrow traits. Journal of

2: Personality measurement and Applied Psychology, 91, 40–57.
testing, (pp. 135–159). Thousand Feist, G. J. (1998). A meta-analysis of
Oaks, CA: Sage. personality in scientific and artistic
Cattell, R. B., Cattell, H. E. P., & Cattell, creativity. Personality and Social
K. W. (1994). The Sixteen Psychology Review, 2, 290-309.
Personality Factor Questionnaire Garrity, T. G., Stallones, L., Marx, M. B.,
5th ed. Champaign, IL: Institute for & Johnson, T. P. (1989). Pet
Personality and Ability Testing. ownership and attachment as
Cattell, R. B., Eber, H. W., & Tatsuoka, supportive factors in the health of the
M. M. (1970). Handbook for the elderly. Anthrozöos, 3, 35-44.
Sixteen Personality Factor Gold-meir, J. (1986). Pets or people:
Questionnaire. Champaign, IL: Another research note. The
Institute for Personality and Ability Gerontologist, 26, 203-206.
Testing. Gosling, S. D., & Bonnenburg, A. B.
Conn, S. (1994). Reliability and (1998). An integrative approach to
equivalency: Comparison of the personality research in
16PF Fifth Edition and Fourth anthrozoology: Ratings of six
Edition (Form A). In S. R. Conn, & species of pets and their owners.
M. L. Rieke (Eds.), The 16PF Fifth Anthrozöos, 11, 148-156.
Edition, Technical Manual (pp. 75– Gosling, S. D., Sandy, C. J., & Potter, J.
99). Champaign, IL: Institute for (2010). Personalities of self-
Personality and Ability Testing. identified dog people and cat people.
Conn, S. R., & Rieke, M. L. (1994). Anthrozöos, 233, 213-222.
Construct validation of the 16PF Guastello, A. D., Guastello, S. J., &
Fifth Edition. In S. R. Conn & M. L. Guastello, D.D. (2014). Personality
Rieke (Eds.), 16PF Fifth Edition trait theory and multitasking
technical manual, (pp. 103-142). performance: implications for
Champaign, IL: Institute for ergonomic design, Theoretical
Personality and Ability Testing. Issues in Ergonomics Science, 15,
Connell, C.M., & Lago, D. (1984). 432-450.
Favorable attitudes toward pets and Guastello, S. J. (2009). Creativity and
happiness among the elderly. In R. personality. In T. Rickards, M. A.
K. Anderson, B.L. Hart, & L.S. Hart Runco, & S. Moger (Eds.),
(Eds.), The pet connection: Its Routledge companion to creativity
influence on our health and quality (pp. 267-278). Abington, UK:
of life (pp. 241-250). Minneapolis, Routledge.
MN: CENSHARE. Hergovich, A., Mauerer, I., & Riemer, V.
Coren, S. (1998). Why we love the dogs (2011). Exotic animal companions
we do. New York: Free Press. and the personality of their owners.
Dudley, N. M., Orvis, K. A., Lebiecki, J. Anthrozöos, 24, 317-327.
E., & Cortina, J. M. (2006). A meta- Hyde, K. R., Kurdek, L., & Larson, P.
analytic investigation of (1983). Relationships between pet
conscientiousness in the prediction ownership and self-esteem, social
of job performance: Examining the sensitivity, and interpersonal trust.
intercorrelations and the incremental Psychological Reports, 52, 110.

56 | H A I B

Johnson, S. B., & Rule, W.R. (1991). Rieke, M. L., & Conn, S. R. (1994). The
Personality characteristics and self- revised reasoning (Factor B)
esteem in pet owners and non- scale. In S. R. Conn & M. L. Rieke
owners. International Journal of (Eds.), 16PF Fifth Edition technical
Psychology, 26, 241-252. manual (pp. 23-31). Champaign, IL:
Johnson, T. P., Garrity, T. F., & Institute for Personality and Ability
Stallones, L. (1992). Psychometric Testing.
evaluations of the Lexington Rieke, M. L., Guastello, S. J., & Conn, S.
Attachment to Pets Scale (LAPS). R. (1994). Leadership and creativity.
Anthrozoös, 5, 160-175. In S. R. Conn & M. L. Rieke (Eds.),
Jouber, C. E. (1987). Pet ownership, 16PF Fifth Edition technical manual
social interest, and sociability. (pp. 183-212), Champaign, IL:
Psychological Reports, 61, 401-402. Institute for Personality and Ability
Kidd, A. H., & Kidd, R. M. (1980). Testing.
Personality characteristics and Schmitt, P., Realo, A., Voracek, M., &
preferences in pet ownership. Allik, J. (2008). Why can’t men be
Psychological Reports, 46, 939-949. more like women? Sex differences in
McCrae, R. R., & Costa P. T. Jr. (1985). Big Five personality traits across 55
Updating Norman’s “Adequate cultures. Journal of Personality and
Taxonomy:” Intelligence and Social Psychology, 94, 168-182.
personality dimensions in natural Szymura, B. (2010). Individual
language questionnaires. Journal of differences in resource allocation
Personality and Social Psychology, policy. In A. Gruszka, G. Matthews,
49, 710–721. & B. Szymura (Eds.), Handbook of
Miller, M. & Lago, D. (1990). Observed individual differences in cognition:
pet-owner in-home interactions: Attention, memory, and executive
Species differences and association control (pp. 231–246). New York:
with the Pet Relationship Springer.
Scale. Antrozoös, 4, 49-54. Woodward, L. E., & Bauer, A. L. (2007).
Paden-Levy, D. (1985). Relationship of People and their pets: A relational
extraversion, neuroticism, perspective on interpersonal
alienation, and diverse incidence complementarity and attachment in
with pet-ownership. Psychological companion animal owners. Society
Reports, 57, 868-870. and Animals, 15, 169-189.
Perrine, R. M., & Osbourne, H. L. (1998). Zasloff, R. L. (1996). Measuring
Personality characteristics of dog attachment to companion animals: A
and cat persons. Anthrozöos, 11, 33- dog is not a cat is not a bird. Applied
40. Animal Behaviour Science, 47, 43-
Reevy, G. M., & Delgado, M. M. (2015). 48.
Are emotionally attached companion Zasloff, R. L., & Kidd, A. H. (1994).
animal caregivers conscientious and Attachment to feline companions.
neurotic? Factors that affect the Psychological Reports, 74, 747-752.
human-companion animal
relationship. Journal of Applied
Animal Welfare Science, 18, 239-

57 | H A I B