Results of the Knox College Creepiness Study

First of all, Sara Koehnke and myself (Frank McAndrew) would like to thank all of you who took the time to complete our Creepiness Survey about a year ago.  Sorry that it took so long to get the results out to you!  

We have just submitted a paper to a peer-reviewed psychology journal and hope to get it published.  Please do not repost this anywhere on the intenet, and if you are a journailist or science writer, please do not write about any of this without contacting me first.  

It would be too boring and tedious to give you every scrap of what we found, but I will try to provide a summary of the high spots without all of the statistical mumbo-jumbo.  If anyone want more details, feel free to email me at :

Here are some of the details:

WHO WAS IN THE STUDY? (besides you):

Participants were recruited through invitations to Facebook events that were created by the researchers, through campus-wide emails distributed to students, faculty, and staff at a liberal arts college in the American Midwest, and through the “Social Psychology Network” website. Volunteers were encouraged to forward the link to the online survey to their friends and acquaintances. A brief description of the study and a link to the survey were posted on the invitation page. 
This resulted in a final sample of 1,341 individuals (1,029 females, 312 males) ranging in age from 18 to 77 with a mean age of 28.97 (SD = 11.34).  We did not ask participants to report their country of origin, but in an unrelated study using an identical recruitment strategy, respondents from 54 different nations were acquired.  Thus, although our sample was primarily American, we are confident that there was significant international representation.  Participants had to check a box confirming that they were at least 18 years of age before they could access the survey.


Given how frequently the concept of “creepiness” is invoked in everyday life, it is surprising that it has not been studied by psychologists. This study attempted to uncover the cues that are used to label someone as “Creepy” and to identify the basic elements of creepiness.  An international sample of 1,341 individuals responded to an online survey about creepiness.  The results revealed that males are perceived as creepier than females and that females are more likely to associate sexual threat with creepiness. Behaviors and characteristics associated with unpredictability are also predictors of creepiness, as are some occupations and hobbies.  The results are consistent with the hypothesis that being “creeped out” is an adaptive emotional response to uncertainty about the presence of threat that enables us to maintain vigilance during times of uncertainty.


Tests of Hypotheses. 

The first prediction was that creepy individuals would be expected to be males more often than females.  This prediction was assessed directly via the question that asked people to choose whether a creepy person was more likely to be a male or a female.  95.3% of our respondents thought that creepy people were much more likely to be males than females, a finding that was highly significant.  This perception was equally likely to be held by male participants (95.5% vs. 4.5%) and female participants (95.2% vs. 4.8%).  Thus, our first prediction was supported: males are creepier than females.

The second prediction was that females are more likely to perceive a sexual threat from a creepy person than are males.  This hypothesis was tested by comparing male and female responses to two items: The degree to which steering a conversation toward sex was perceived as a probable characteristic of a creepy person and the degree to which the respondent agreed with the statement that the creepy person “has a sexual interest in me.”  The prediction was supported by both of these items.  Females were more likely than males to think that steering a conversation toward sex was characteristic of a creepy person, and they were also more likely to think that the creepy person had a sexual interest in them.

The third prediction was that occupations would differ in their level of creepiness according to how threatening or strange the “subject matter” of the occupation is.  The means and standard deviations of the creepiness ratings for the 21 stimulus occupations are displayed in the Table below.  The analysis revealed that the differences in how occupations were rated was highly significant; all of the occupations except two (construction workers and computer software engineers) were significantly different from each other.  However, only four occupations were judged to be significantly higher than the neutral value of “3” on the creepiness rating scale: Clowns, Taxidermists, Sex Shop Owners, and Funeral Directors.  Therefore, it appears that occupations associated with death (taxidermy and funeral directors) or reflective of a fascination with sex (sex shop owners) are perceived as creepy; clowns were the creepiest of all.

Table 1

Creepiness Ratings of Occupations

Occupation                    Mean                SD       
Clown                            3.71                1.24
Taxidermist                    3.69                1.19
Sex Shop Owner             3.32                1.30
Funeral Director             3.22                1.23
Taxi Driver                    2.86                1.19
Unemployed                  2.83                1.29
Clergy                           2.57                1.28
Janitor                           2.51                1.19
Garbage Collector          2.25                1.12
Guard                           2.18                1.08
Writer                           2.14                1.08
Actor                            2.13                1.02
Construction Worker*    2.09                1.09
Computer Software Engineer*   2.09                1.11
Cafeteria Worker           2.08                1.06
Financial Adviser          1.78                0.98
Doctor/Physician           1.77                0.96
College Professor           1.67                0.86
Farmer                          1.65                0.90
Teacher                         1.57                0.82
Meteorologist                1.53                0.83       

Note: Occupations marked with an asterisk are not significantly different from each other (Tukey HSD = .01).  Ratings were made on a “1” (not very creepy) to “5” (Very creepy) scale.
The fourth prediction was that things that make a person unpredictable also predict creepiness.  One item among the ratings of creepy individuals (“I am uncomfortable because I cannot predict how he or she will behave”) and one item among the items assessing beliefs about creepy people (“Even though someone may seem creepy, I usually think that I understand his or her intentions”) allowed a direct test of this prediction.  Collectively, the results of the analyses of these two items indicate that unpredictability is indeed an important component of creepy behavior.

Exploratory Analyses:
The many items in our survey afford ample opportunities for exploration of the elements of creepiness.  Our first step in this direction was to combine items that seemed to be measuring the same thing within the two longest sections of our questionnaire.  The first section contained 44 items assessing the likelihood that a creepy person described by one’s trusted friend would display a particular behavior or possess a particular physical characteristic.  In an attempt to reduce the number of variables to be analyzed, these 44 items were subjected to statistical manipulation called "factor analysis." The factor analysis was able to identify only one factor that connected multiple variables.  This factor included 15 of the 44 items, all of which reflected a nonverbal behavior or physical characteristic of creepy people.  A new variable called Appearance/NVB was calculated by computing a mean based upon the scores of each individual on these 15 items.  The 15 items that comprised this new variable are as follows. 
    The person stood too close 
    The person had greasy hair 
    The person had a peculiar smile 
    The person had bulging eyes 
    The person had long fingers 
    The person had unkempt hair 
    The person had very pale skin 
    The person had bags under his or her eyes 
    The person was dressed oddly 
    The person licked his or her lips frequently 
    The person was wearing dirty clothes 
    The person laughed at unpredictable times 
The person made it nearly impossible to leave the conversation without appearing rude 
The person relentlessly steered the conversation toward one topic

This new composite Appearance/NVB variable along with the remaining 29 items from the first portion of the questionnaire were analyzed via one-sample t tests to determine which of these characteristics was significantly above the neutral point of “3,” and therefore very likely to be a characteristic of a creepy person.  The following elements were thought to be very likely to be found in a creepy person:

The appearance and nonverbal behavior items in the composite variable (Appearance/NVB), being of the opposite sex (probably due to the predominantly female sample in our study), being extremely thin, not looking the interaction partner in the eye, asking to take a picture of the interaction partner, watching people before interacting with them, asking about details of one’s personal life, having a mental illness, talking about his/her own personal life,  displaying too much or too little emotion, being older, and steering the conversation toward sex.

Similarly, the section of the questionnaire consisting of 15 items that reflected beliefs about the nature of creepy people was subjected to a factor analysis. 
The analysis yielded four factors on which at least two items loaded. 

The first factor tapped into how fearful or anxious the person felt while interacting with a creepy person, and it included the following items;  Each statement began with the expression “When I meet someone that seems creepy . . .
I am sure that the person intends to harm me 
I am uncomfortable because I cannot predict how he or she will behave
I feel anxious 
I believe that he or she is intentionally hiding something from me 

The second factor reflected how intimately involved one would be with a creepy person, and it consisted of two items:
People are less creepy if I know I won’t have to speak to them ever again 
People are creepier when I meet them online compared to face-to-face 

The third factor measured the extent to which creepiness is an inherent part of the individual, and it consisted of two items:    
Some people can do the exact same behavior as someone else and one person can be perceived as creepy while the other person is not 
Behaviors often admired in “bad guys” in movies and TV shows are actually really creepy if done in real life

The fourth factor reflected the extent to which people willfully deviate from social norms, and it consisted of two items:
When I meet someone who seems creepy, I expect him or her to follow the usual rules for socially acceptable behavior 
People choose to act in a creepy manner 

New composite variables labeled “fearfulness,” “proximity,” “individual creepiness,” and “non-normativity,” were calculated by computing a mean of the items that loaded on each factor. These four composite variables along with the remaining 5 items from the last portion of the questionnaire were analyzed to determine which of these characteristics was significantly different from neutral point of “3,” and therefore strongly believed to be characteristics of a creepy person. The following things were believed to be true of a creepy person:    

They make us fear fearful/anxious (composite fearfulness variable)
Creepiness resides in the individual more than in his/her behavior (composite individual creepiness variable)
We think they may have a sexual interest in us
They are creepy when they exhibit multiple “symptoms” of creepiness rather than just one
The expected intimacy and frequency of interaction with the person moderates perceptions of creepiness
Creepy people are unable to change, but they do not necessarily have bad intentions
People who follow social rules of behavior are not perceived as creepy

There was also one final item in which participants chose among “yes,” “no,” and
“unsure” in response to the question “Do most creepy people know that they are creepy?”   The responses were 115 “yes” (8.6%), 797 “no” (59.4%), and 429 “unsure” (32%), indicating that our participants did not believe that most creepy people know that they are creepy.

There were many significant correlations between the age of the participant and his/her responses to the items in the survey. The general finding of interest was that older people seemed to be less alarmed by creepy people than are younger people, being less likely to perceive sexual threat or intended harm.  They also expressed less anxiety at the prospect of interacting with a creepy person.

Just for fun, we asked our participants to list two hobbies that they thought of as creepy.  Easily, the most frequently mentioned creepy hobbies involved collecting things (listed by 341 of our participants).  Collecting dolls, insects, reptiles, or body parts such as teeth, bones, or fingernails was considered especially creepy.  The second most frequently mentioned creepy hobby (listed by 108 participants) involved some variation of “watching.”  Watching, following, or taking pictures of people (especially children) was thought to be creepy by many of our participants, and bird watchers were considered creepy by many as well.  A fascination with pornography or exotic sexual activity and taxidermy were also frequently mentioned.