A Thing like a Human? A Mixed-Methods Study on Sex Doll Usage

Abstract

Sex dolls have evolved over the last decades and are getting more popular. Whether sex doll ownership should be protected as a sexual right that facilitates sexual health or whether its usage might be harmful to women and society is vividly discussed. In this polarized debate, empirical research is scarce and possible differentiations in the psychological functions of sex dolls remain nearly unknown. To further explore sex doll ownership, this study conducted five in-depth interviews with sex doll users and collected 131 written reports on sex doll ownership in an online survey with open and closed questions. In a mixed-methods approach, the data was then analyzed using Mayring’s Qualitative Content Analysis and quantitative measures. Although the demographics of sex doll users are homogenous (single heterosexual males), results indicate that the use is divers. Sex dolls are mostly used as masturbatory aid, as substitutes for human partners, and as synthetic partners. We explore the similarities and differences between these kinds of usage and propose that a differentiation of sex doll use can enrich current research on social robotics, anthropomorphism, and misogyny.

Introduction

Sex dolls have changed immensely over the last decades. They have evolved from dolls made from clothes and blankets by sailors to cheap inflatable rubber dolls to high quality silicon or TPE dolls (Ferguson, 2010). Today’s high-quality sex dolls are sold all over the world and, although it is hard to scientifically estimate the actual prevalence of sex doll ownership (Döring et al., 2020, p. 10), it seems that more and more sex dolls are sold (Butler, 2020; Frank, 2014). After Chinese companies started to offer inexpensive products, sex dolls might be on the verge of becoming a mass-market product (Zhou, 2020).

Scholars estimate the effects of sex dolls on individuals and society very differently. Some authors examine sex dolls and robots rather uncritically, others stress the risks of sex dolls and their potential to foster sexual violence (for an overview of the argument see Harper & Lievesley, 2020). From the perspective of the World Association for Sexual Health (WAS) declaration of sexual rights, the first mentioned authors see sex doll ownership as an expression of one’s sexuality and an individual way of seeking sexual pleasure, while the critics consider sex with dolls a violation of the rights of others (Ford et al., 2021; Kismödi et al., 2017).

Critical scholars emphasize that sex dolls and robots might reinforce traditional beauty standards, increase the objectification of women, and lead to more sexual violence (e.g. Cassidy, 2016; Facchina et al., 2017; Sparrow, 2017). In consequence, some feminist activist groups call for the prohibition of sex robots and dolls such as the Campaign Against Sex Robots in the UK (Campaign Against Sex Robots, n.d.; Walker, 2018 for a similar argument in German). In the US, a similar movement was founded by a mother after a picture of her 8-year-old daughter had been used as template for a child-like sex doll (Guirola, 2020). Advocates of such organizations voice concern about the safety of women and children, suspecting sex dolls to cause an increase in male violence (for an overview of such arguments see Danaher, 2017; Richardson, 2016).

In opposition to such organizations, others argue that sex with dolls is a specific way of expressing one’s sexuality. They state that in the future sex dolls and robots might belong to the human family as dogs and cats do today. Combined with AI and robotics, sex dolls could complement or even replace human relationships (Levy, 2007, 2017) and potentially revolutionize the sex work industry (Yeoman & Mars, 2012). From this perspective, sex dolls and robots are technological progress that is not problematic in itself. Thinking about sex doll ownership as a way to gain sexual pleasure, sex dolls might even improve the user’s sexual health and the related positive health outcomes (Gianotten et al., 2021).

What is striking is that, apart from the vivid ethical debate, empirical accounts on sex dolls and robots remain scarce. Between 1993 and 2019, only 29 articles on sex dolls and 98 on sex robots were published, of which merely five work with empirical data on sex doll use and effects (Döring et al., 2020, p. 8). Most publications are of conceptual nature and—owing to the lack of empirical knowledge—remain at least to some degree speculative. With more empirical data on sex doll ownership and usage, the ethical considerations would be more valid. Furthermore, empirical investigations are also crucial for lawmaking, as some countries such as Germany passed laws against sex dolls to prevent child abuse (§ 184 l StGB/German Criminal Code). It is unclear, however, whether the laws will be effective, as data on how dolls are used, and what the effects of that usage might be, are scarce. It is essential to know how sex dolls are used today, to evaluate and affect the influence they will have on our society in the future. Consequently, many scholars call for empirical studies on sex doll usage (Döring et al., 2020; Harper & Lievesley, 2020).

Among the articles that study sex doll users empirically are Valverde (2012) and Langcaster-James and Bentley (2018). They find that sex dolls are mostly used by “single, middle-aged, relatively well-off, employed, heterosexual male[s]” (p.13). Existing research challenges the stereotype of the sex-hungry pervert, who uses dolls for sexual gratification only and emphasizes the emotional aspects of doll ownership (Ciambrone et al., 2017; Knafo, 2015; Knafo & Lo Bosco, 2017; Langcaster-James & Bentley, 2018; Su et al., 2019). The use of sex dolls is more than solely sexual in nature. Many owners attach to their dolls emotionally, some even seem to declare their dolls as wifes (Lievesley et al., 2022; Stern, 2020). Unfortunatally, none of the previous accounts on sex doll ownership explores the mechanisms involved in anthropomorphizing dolls. Sex dolls are both: a sex toy which function is merely to aid masturbation and a highly anthropomorphized synthetic companion.

To bond with inanimate objects is nothing unusual, but a general human feature. Children for example use stuffed animals as transitional objects (Winnicott, 1953), for adults, objects are much more than mere tools but play a significant role in identity construction (Csikszentmihalyi & Rochberg-Halton, 1981), the material engagement theory shows how important objects are for cognition in general (Malafouris, 2020), and studies on anthropomorphism describe how people perceive objects as human-like (Epley et al., 2007, 2008). These examples emphasize that, although doll use might appear unusal at first sight, it is very typical for humans to attribute meaning to objects that extends the object’s merely functional pupose giving the object human-like characteristics. How these theories might apply to sex doll owners remains unclear in the absence of empirical data. We assume different degrees of object attachment and anthropomorphism in sex doll usage. This variance in attachment and anthropomorphism might signify different groups of users.

After an extensive and systematic literature review, Döring et al. (2020, p. 13) hypothesize six different groups of sex doll owners: passionate doll lovers, sadistic owners craving for domination, pedophiles, owners substituting for a lost or out-of-reach partner, photographers, and couples wanting to enrich their sexuality. During the initial search through online sex doll forums, we found men stating that they use their dolls for sexual purposes only. In addition, we found that sex dolls are often advertised as high quality sex toys for men, implying that one could buy one merely for the sex. Therefore, we assume another group of users, namely, owners using sex dolls as masturbatory aid only.

However, humans are “moving targets,” who react to categorization and change over time (Hacking, 1995, 2007). The relationship with and significance of one’s doll can change drastically over time. Thus, in this study we intent to categorize the type of usage instead of the user. The goal of this study is to explore the characteristics of different kinds of sex doll usage and to understand how sex dolls can be, in one case, an instrument for sexual stimulation and, in another case, be perceived as a human-like companion.

Our research questions are:

1. Social factors, for example: What are the demographics of sex doll owners? Do owners disclose their doll ownership to others? Do they experience prejudice? How are doll owners connected to the doll community?

2. What is the context in which doll owners buy dolls? What motivates doll ownership?

3. How and for what purpose are sex dolls used? How can we categorize the usage?

4. How do doll owners attach to their dolls? How do they anthropomorphize them?

During the investigation of the above mentioned research questions we follow a Positive Sexuality Framework (Williams et al., 2015) which is in line with Döring and Poeschel’s work on sex dolls (2019). Sexuality—even one involving sex dolls—is not pathological per se as different people use different measures to gain sexual pleasure (Ford et al., 2021; Rye & Meaney, 2007) and supporting that view recent studies suggest that sex doll ownership does not correspond with offending proclivities (Harper et al., 2022). At the same time, we take conceptional criticism on sex dolls seriously—sex dolls potentially reinforce sexism on a societal level (Cassidy, 2016). Cassidy (2016) examines the potential of sex dolls and robots and concludes that, although they are currently reproducing sexist clichés, fundamental changes in their development and distribution could alter their sexist appearance. Consequently, the possible negative effects of sex dolls depend on how they are used and who is involved in their development. For this study, we follow Su et al. (2019) in “neither celebrating nor censuring [doll user’s] practices” (p. 13:2). In our approach, we try to understand how it is for our participants to own dolls while at the same time remaining critical.

There is still very little empirical research on sex dolls and their users; therefore, we conducted five interviews and collected data from an online questionnaire with 131 participants featuring mainly open-ended questions. We first analyzed the sample’s demographics. Secondly, we analyzed the open-ended questions and interviews using Mayring’s (2015) Qualitative Content Analysis. Lastly, we statistically examined the quantitative measures and combined them with the qualitative findings.

Methods

Data collection

To approach the topic, the authors conducted an extensive review of the existing literature on sex dolls and engaged in participatory observations in online sex doll communities. Half a year before data collection, the first author introduced himself in two online forums, had informal conversations with forum members and regularly browsed the forums unsystematically. Building on these insights, the first author, in correspondence with the second author, developed a semi-structured interview guideline. The questions concerned demographics, experiences of discrimination, public disclosure of dolls, community connectedness, sexual satisfaction and health, the user’s relationship with his or her dolls, and the context of life in which the dolls were purchased. For each question, the guideline provided optional follow-up questions. To engage with the sex doll community, the first author introduced himself in English-speaking and German-speaking online forums and recruited five sex doll users for telephone interviews. The interviews were held between 2020, November 22 and 2021, January 21 and on average lasted 102 min (SD = 68). The interviewer was in all cases the first author, a trained and experienced qualitative researcher. In the beginning of the interviews, the interviewer informed the respondents about the study’s aims, procedures, and anonymity measures, and the interviewees gave informed consent. The interviews were recorded and transcribed verbatim, except for one participant who allowed to take notes only.

While collecting more interview data, it became evident that many sex doll users would not consent to being interviewed. They either did not want to talk from person to person or feared a lack of anonymity. Instead of an interview, some users offered to give written insights. We soon realized that a survey with open questions would be most convenient for the target audience. Thus, we developed a questionnaire with nine open-ended and 17 closed questions.

The first seven items of the online survey concerned the users’ demographics (age, gender, country of residence, living arrangement, income, sexual orientation, relationship status) followed by the questions “For how many years do you use sex/love dolls?” and “What kind of sex dolls do you own?” Questions 10–18 were open questions about the context in which the owner’s first doll was purchased, about community connectedness, prejudice, occupations related to dolls, sex with dolls, relatives’ attitudes toward sex doll ownership, deviant sexual behavior with dolls, and the role of dolls in one’s life. Following the open questions, the participants were asked, on a Likert scale, if they agree to the following sentences: “I feel emotionally attached to my doll(s),” “For me, a sex doll is just a high-quality sex toy,” and “I would rather like to have a real partner than a doll.” The final items concerned sexual satisfaction, sexual dysfunctions, and frequency of doll sex. The survey was available in English and German.

Except for the question concerning deviant behavior with dolls, we did not explicitly ask for violent, sexists and pedophilic fantasies and behaviors to ensure high compliance in the field. The resulting limitation of this decision is discussed in the limitations section.

Between the 2021, January 12 and 2021, January 25, the survey was accessible online in three online forums, two subreddits¹, and one private Discord² server. Over the course of the data collection, the first author remained in close contact with the community, answering questions and concerns regarding the survey at least twice a day. This proved useful, as several users raised concerns about the neutrality and objectivity of the study. Additionally, the higher traffic caused more people to see the survey link. Traffic varied between 1 and 63 comments and 11 private chat conversations about the survey and related topics.

Participants

The interview partners were five men between the ages of 33 and 63 (M = 46.8, SD = 12.61); one of them owning one sex doll, the others three or more. The interviewees have owned the dolls between 2.5 and 10 years (M = 6.25, SD = 3.8). For further information see Table 1.

Table 1.

Interview Demographics.

	Georg	Werner	Fabian	Karsten	Tobias
Age	37 years	63 years	33 years	56 years	45 years
Sexual orientation	Heterosexual	Bisexual	Heterosexual	Heterosexual	Heterosexual
Relationship status	Single	Divorced	Single	Divorced	Married
Living arrangement	Alone	With family	Alone	Alone	With wife
Profession	Precision mechanic	Casual jobs	Sex toy salesman	Legal administrative officer	Merchant
Length of doll ownership	3.5 years	10 years	2.5 years	9 years	Unknown
Types of possessed dolls	More than 10 female dolls	6 female dolls	1 female doll	3 female dolls	More than 4 female dolls

The online questionnaire was answered by 159 respondents, 28 of which were excluded having answered less than two items. The remaining 131 participants were on average 41.49 years old (min = 20, max = 74, SD = 11.93) and bought their first sex doll between 0 and 28 years ago (M = 3.59, SD = 4.92). For further information see Table 2.

Table 2.

Online Survey Demographics.

	n	M	Min	Max	SD
Age	124	41.49	20	74	11.93
Length of doll ownership (years)	89	4.23	0	60	7.72
	n		%
Gender
Female	5		4
Genderfluid/non-binary	2		2
Male	124		95
Sexual orientation
Heterosexual	99		76
Bisexual	14		11
Homosexual	3		2
Heteroflexible	2		2
Others	3		2
Unknown	10		8
Relationship status
Single	70		53
Married	17		13
Divorced	14		11
Female partner	8		6
Male partner	3		2
Separated	3		2
Doll partner	3		2
Widowed	1		1
Unknown	12		9
Country of residence
USA	41		31
Germany	28		21
UK	6		5
Canada	5		4
Netherlands	2		2
Others	11		8
Unknown	38		29
Living arrangements
Alone	74		56
With partner	20		15
Shared housing	9		7
With parents	7		5
With family	6		5
With mother	2		2
Single parent	1		1
Unknown	12		9
Income
Less than 500€	3		2
500€–1000€	10		8
1000€–1500€	6		5
1500€–2000€	18		14
2000€–2500€	10		8
2500€–3000€	10		8
3000€–3500€	6		5
3500€–4000€	4		3
4000€–4500€	8		6
4500€ or more	16		12
Prefer not to say	22		17
Unknown	18		14
How many participants own how many dolls	Amount
	0	1	2	3	4+
Female dolls	0	36	26	16	26
Female dolls with a penis	49	2	0	0	0
Male dolls	46	5	1	1	0
Penis insert	43	8	1	0	0
Other dolls (e.g., Phicena)	45	1	2	1	6

^aPhicen is a brand that produces 1:6 scale figures for collectors (http://www.phicen.com, 2022, April 6).

Out of 131 participants, 81 answered the open-ended questions. The number of letters varied between 2 and 26,587 (M = 3296, SD = 3975). Three respondents reported not having a doll yet and were excluded from analysis. Probably due to the time-consuming nature of open-ended questions, we observed the highest drop-out after the first open-ended question.

Analysis

The mixed methodology in this paper prioritizes qualitative analysis and consults quantitative measures to generate “additional coverage” of the phenomenon (Kuckartz, 2014, p. 69). The procedure can, thus, be described as an “QUAL + quan” technique (Schoonenboom & Johnson, 2017, p. 119) in which the qualitative and quantitative data was examined at the same time independently. The “point of integration” (p.115) was after the analysis when the quantitative measures were integrated into the qualitative results.

Qualitative analysis

The qualitative data was analyzed using Qualitative Content Analysis (Mayring, 2015), which is a well-established qualitative approach in German-speaking countries. The analysis was conducted using MAXQDA Plus 2020 (VERBI Software, 2019, release 20.4.0).

In the first step of the analysis, the first author coded the first two interviews inductively in order to summarize the interviews and to create major themes (Mayring, 2015, pp. 85–90). Following the analysis of the initial interviews, the first author developed an online questionnaire with open-ended and closed questions. The questions were derived from both the analysis of the first two interviews and the literature. The questionnaire was then reviewed by the second author and altered where he deemed appropriate. The questionnaire is available in English and German in the OSF online repository.³

After collecting the survey data and conducting all interviews, the first author coded the open-ended questions of the online survey and the interview transcripts. Two of Mayring’s approaches were used. Using Content Structuring (2015, p. 103), the categories community, public disclosure, stigma, appealing factors of dolls, non-sexual usage, and sexual usage were theoretically defined (Langcaster-James & Bentley, 2018) and then enriched with codes and subcodes inductively (see Table 3). The second approach—Structuring by Type (2015, pp. 103–106)—allowed us to code the respondent’s type of usage. Each complete case was coded with synthetic partner use, sex toy use, substitute use, couple use, use due to disability, as pedophile use, or esthetic use (see Table 4). These codes were defined theoretically from Döring et al. (2020, p. 13), except the code sex toy usage which came up inductively during coding. Throughout the coding process the, at first, very loose definitions of the theoretically defined codes were reassessed and the final definitions are described in more detail in the results section.

Table 3.

Coding Table.

Category	Code	Sub-code	N a	n b	%c
Community			71
	Contact			65	92
		Contact in person		11	15
		Contact online		64	90
	No contact			6	8
Outing			65
	Closet			46	71
	Partly outed			11	17
	Outed			8	12
Stigmad			64
	Positive or neutral reactions			35	55
	Negative reactions			20	31
		Unspecific negative reactions		6	9
		Avoiding reactions		6	9
		Rejection online		5	8
		Verbal rejection		5	8
	Anticipated prejudice			30	47
	Known stereotypes			25	39
Appealing factors of dollsd			67
	Company and intimacy			30	45
	Sexually appealing			41	61
		Sexual relief		36	54
		Enact sexual fantasies and preferences		8	12
		Alternative to prostitution		2	3
		More sex drive than partner		8	12
	Comparing to human partner			25	37
		No pressure to perform sexually		5	7
		Devotion and loyalty		18	27
		No disagreement		8	12
	Well-being			26	39
		Somebody to care for		12	18
		Emotional stability		19	28
		Therapeutic		7	10
	Others			17	25
		Beauty and perfection		10	15
		Curiosity		9	13
		Explore femininity		1	1
non-sexual usaged			67
	Doll styling			39	58
		Dress and buy dresses		37	55
		Posing the doll		8	12
		Do the doll’s hair		9	13
		Make-up		8	12
		Change iris		1	1
	Company			30	45
		Others		14	21
		Eating together		3	4
		Working with doll present		1	1
		Reading		3	4
		TV		19	28
		Relaxing		3	4
		Listening to music		3	4
		Gaming		4	6
	Intimacy			51	76
		Speaking		19	28
		Cuddling and hugging		38	57
		Sleeping		28	42
		Bathing		3	4
		Kissing		16	24
		Dancing		2	3
		Caressing		3	4
	Creative occupation with or about dolls			35	52
		Photography		34	51
		Cosplay and Roleplay		6	9
		Drawing		1	1
		Writing		3	4
	Maintenance of dolls			32	48
		Repairing		14	21
		Washing and caring		28	42
	Others			18	27
		AI Chat		1	1
		Drug use		4	6
		Looking at dolls		9	13
		Outing or holiday		5	7
		Have guests		1	1
Sexual usaged			62
	Porn			9	15
		Watching porn with doll		8	13
		Watching doll porn		2	3
	Showering			2	3
	Ejaculation			1	2
	Fingering			3	5
	Breasts			5	8
	Masturbation with doll			7	11
	Feet fetish			2	3
	Anal			7	11
	Virtual reality sex with doll			1	2
	BDSM and kink			6	10
	Vaginal intercourse			6	10
	Fantasizing human sex partners with doll			4	6
	Foreplay			27	44
	Afterglow			2	3
	Sexual roleplay and cosplay			9	15
	Rub on doll			7	11
	Kissing			13	21
	Quicky and Free use fantasies			6	10
	Group sex with dolls and humans			3	5
	Group sex with several dolls			7	11
	Oral (doll on user)			6	10
	Oral (user on doll)			6	10
	‘Vanilla Sex’			6	10

^aNumber of cases that provide enough information for coding in this category.

^bNumber of cases that involve the code at least once.

^cThe percentage is calculated with the number of cases in each category (N) and the number of cases that involve the code at least once (n; e.g., out of the 71 cases that supply information about the community connectedness 64 cases describe online interactions with other users, thus, 92%).

^dThe number of the codes in this category can only be interpreted as the minimal prevalence as the questionnaire did not ask specifically for each code. e.g., 48% report the maintenance of their dolls unprompted, however, the remaining 52% may maintain their dolls as well but were not asked about it and did not report it themselves.

Table 4.

Kinds of Usage.

			N	n	%
Kinds of usagea			69
	Synthetic partner usage			31	45
	Sex toy usage			20	29
	Substitute usage			13	19
Subtypes of usageb			69
	Couple usage
		Male partner uses doll		9	13
		Both partners use doll		4	6
	Disability usage			5	7
	Pedo-sexual usage			2	3
	Esthetic usage			3	4

^aCodes in this category are exclusive, no case is double coded.

^bCodes in this category are not exclusive, some cases are double coded (e.g., couple use and sex toy use).

After coding five cases, the coding system was revised and reapplied to the data. When the coding system proved consistent, five more cases were added to the analysis until all data was analyzed. Readers can view the final coding system in the online repository.

In this paper, we will state the number of participants providing meaning units to a code (n) and the number of cases that provide enough information that allows coding in this category (N). Cases that did not provide enough information for coding were neither included when calculating the percentage of the code’s appearance nor when calculating correlations.

Once the final coding system had been developed, we calculated the intercoder agreement. 25 coded segments were given to two independent coders for a second coding. The agreement between the author and coder 2 resulted in Cohen’s kappa of κ = 0.75, the agreement between the author and coder 3 was κ = 0.83 (Cohen, 1960, 1968). The calculation of Cohen’s kappa is provided in the OSF online repository.

Quantitative analysis

The quantitative items were analyzed using RStudio (RStudio Team, 2021, version 1.4.1103) and Microsoft Excel (version 16.0.14026.20270). We used descriptive measures to report the prevalence of codes (Table 3), the demographics (Tables 1 and 2), and the items concerning sexual dysfunction and sexual satisfaction (Table 5). The co-occurrence of codes was examined using Fisher’s exact test. To measure the correlations between codes and continuous items, we calculated point-biserial correlations. For correlations between two continuous items, we used Kendall’s τ. Mann-Whitney-U-Tests were calculated to compare non-parametric continuous variables.

Table 5.

Items.

	n	M	SD
I am satisfied with my sex life.a	64	2.07	1.14
I reach orgasm during sex.a	65	1.43	0.72
I experience premature ejaculation.a	55	4.05	1.11
I do not have an erection during sex.a	55	4.09	1.07
I feel emotionally attached to my doll(s).b	69	2.31	1.25
For me, a sex doll is just a high-quality sex toy.b	68	3.77	1.50
I would rather like to have a real partner than a doll.b	63	3.06	1.40

^a1 = always, 2 = Often, 3 = Sometimes, 4 = Rarely, 5 = never.

^b1 = strongly agree, 2 = agree, 3 = neither agree nor disagree, 4 = disagree, 5 = strongly disagree.

Respondent validation

To evaluate the validity of the analysis, we reported the results of the study back to the participants (Birt et al., 2016; Mayring, 2016, p. 147). We summarized our findings, posted the report online, and sent it to our participants for verification.

In general, the respondents validated our findings. The forum post sparked a brief discussion between the first author and some forum members, and some respondents gave feedback via e-mail. Some participants emphasized again that they chose dolls over women not because they could not find a human partner but because they preferred dolls. In one instance, the participant responded to our summary with various misogynist arguments. Others were concerned about young-looking dolls—currently a hot topic in the community. They stated that the communities have clear policies to ban young-looking dolls and that some use small dolls because they are lighter and not because they prefer young-looking dolls.

The largely positive feedback implies that our understanding of the participants’ reports is mostly coherent with the doll owners’ point of view. From a methodological perspective this coherence indicates a good validity of results: The results are close to the data and provide a good understanding of doll owners’ perspectives. On the other hand, it is likely that at least some doll owners try to use this survey for impression management. This issue is further discussed in the limitations section.

Ethical approval

Ethical approval was acquired from the ethic’s committee of the University of Bamberg (dossier number 2021-05/27).

Results

Applying the theoretically defined kinds of sex doll usage (Table 4), three kinds were most prominent: Sex toy usage (n = 20), substitute use (n = 13), and synthetic partner use (n = 31). For users who use the doll as a sex toy, the doll serves a primarily sexual purpose. Accordingly, such users do neither humanize the dolls to a strong degree nor are they deeply attached to them. The criterion for sex toy usage was, thus, the use of dolls primarily as masturbatory aid. Users utilizing dolls as a temporary solution were coded as substitute usage. Although in substitute usage owners humanize and attach to their dolls, too, the relationship with them is of a different quality because they still prefer human partners over dolls. Lastly, participants coded as synthetic partner use humanize their dolls, are deeply attached to them, or consider them an important part of their lives.

In addition to the three main usages, four further subgroups were found (Table 4). Thirteen users use a sex doll in a relationship (couple usage). Couples either use dolls as sex toys together (n = 4) or the male partner uses it alone with or without his partner’s knowledge (n = 9). In five cases, physical and mental disabilities make it difficult for the respondents to find a human partner and, thus, users substitute with a sex doll (disability usage). Further, three participants use their dolls for photography or decoration only (esthetic usage). Lastly, two participants describe pedo-sexual fantasies with their young-looking dolls, and another claims to be attracted to adolescent girls (pedo-sexual usage).

The additional four usages could for the most part be double coded with substitute use, sex toy use or synthetic partner use. A person who is in a relationship and uses a doll for sexual pleasure might be coded as sex toy usage or couple use as well. As the four additional subgroups occurred rarely, this analysis focusses on the three main uses (sex toy, substitute, and synthetic partner usage). The subgroups will not be discussed in further detail.

On the following pages, we will first report our findings on the sexual and non-sexual usage, community connectedness, and stigma. We will then describe the different kinds of sex doll usage and compare them in the end of the section.

Community and stigma

Forty out of fifty-seven (70%) doll owners keep their dolls a secret, ten told one or two relatives about them (18%), and only seven owners talk about their sex dolls openly (12%). All seven owners who disclosed their ownership are synthetic partner users.

The fact that many owners keep their dolls a secret matches the owners’ concern of stigma against doll ownership. Thirty-one percent of the participants report stigma related to doll use (20 out of 64). Negative experiences range from comments online to denunciations at the workplace. No physical violence against doll owners was reported. Although most do not report experiences of discrimination, many anticipate rejection and prejudice against sex dolls. Lucas⁴ (Survey participant 38) describes the anticipated stigma on the one hand and the low occurrence of actual discrimination on the other hand as follows:

I don’t think anyone is judging ME [emphasis in original] for having dolls, but I know that doll owners are judged. (Lucas)

He believes that doll owners are stigmatized, although he as an individual is not. Due to the anticipated rejection, most owners keep their dolls secret.

Online forums allow doll owners to share the secret use of sex dolls with others. As the online communities are the only place in which doll users can openly converse about dolls, the participation in online communities is very high, as 65 out of 71 participants (92%) are active members of online forums. It is noteworthy that most respondents were connected to the doll community in the digital sphere only. The involvement in an offline community was rare (11 out of 78, that is, 14%, attended an offline meeting in their lives at least once). Reported reasons for participation in doll communities are technical information about dolls and conversation with likeminded people. However, this finding might be an artifact of the recruitment procedure, as we almost exclusively recruited in online communities.

Doll related activities

Respondents have sex with their dolls 3.91 times per week on average (M = 3.91, SD = 5.48, n = 55). Only in six cases, the respondents do not engage in sexual intercourse at all. Of these six participants, two are women and two use their dolls for esthetic purposes only. Twenty-three codes emerged describing specific sexual activities ranging from vanilla sex to group sex to BDSM and other kinks (see Table 3 for a detailed list of activities).

Non-sexual usage, too, was common among all users (the codes in kind of usage and non-sexual usage are unrelated: p = .11, two-tailed Fisher’s exact test). Cleaning, dressing, and cuddling are typical activities in non-sexual usage (see Table 3 for a detailed list of activities).

Nearly all doll owners use dolls for sexual as well as non-sexual purposes. Nevertheless, the importance of sexual and non-sexual activities differs. In the next section, we will describe sex toy use which emphasizes the sexual aspects of doll use.

Using sex dolls for sex

Twenty users report to use their dolls for mainly (but not always exclusively) sexual purposes. Respondents engaging in sex toy usage report a variety of reasons for using a doll for sex: higher sex drive than their partner, rare sexual preferences, high costs for sex workers, and the desire for a new sensation. Fabian (Interview partner 3) states:

I keep it very simple. But merely because the whole process is too tiring. So I’m not the user who tries different positions cause in the end it’s not a real woman. This is how I see it and I’m realistic in that sense. It’s simply missionary and it’s not about the body itself it’s simply about the feeling. It’s definitely next level compared to masturbating with hands. So yeah, really simple […] sometimes I don’t even put the head on [laughing]. (Fabian, German in original)⁵

For him, it is simply about the physical sensation of the doll that is better in comparison to masturbation. Once Fabian is aroused, he uses the doll as a masturbator. The doll itself does not excite him. Fabian, therefore, neither engages in foreplay with his doll nor does he prepare it for sex. On the contrary, he tries to keep its maintenance to a minimum. He is using his doll exclusively the way in which he would use a sex toy.

Owners using dolls as masturbatory aid do not have more sex with them per week compared to other usages (r_pb(55) = .05, p = .682). Fabian, for example, uses his doll occasionally, whereas Jim (Survey participant 51) has sex with his doll daily. While both own their dolls primarily for sex, Jim uses his for other activities as well: at night he enjoys the doll as a sleeping companion, during the day he sets “her” up in his flat for decoration, and sometimes he takes pictures of “her.” The dolls are always available:

Whilst the doll is not as good as a real woman (sex worker), it is much more economical, e.g. I cannot afford a sex-worker [sic] every day, every night. By being a constant feature rather than an occasional treat that is constantly craved, it is more satisfying and leaves me more content. (Jim)

The dolls relieve Jim of his sexual desires without being too expensive.

Peter (Survey participant 109) has a human partner and owns a doll, nonetheless. Due to a severe illness of his wife, they do not have sex anymore. He describes himself as supportive of her, yet he fears that he might eventually cheat on her. To substitute for the lack of sex with his wife, he bought a sex doll:

The role of my dolls is to help me control my sexual desires. I was really worried […] before I got my dolls that I would eventually cheat on my wife with another woman. The dolls have changed my life I no longer have the pressure of “what if I cheat on my wife”. (Peter)

The dolls serve the purpose of having sex without cheating.

Oliver (Survey participant 54) does not long for more, but for different sex. He has a leather fetish but his partners have not. For him, using sex dolls is the only way to enact his sexual preferences:

Theoretically, most things [I do with my dolls] I could do with real women, too – If you find the needle in a haystack, who likes that. But I’m 50+ now and not as attractive as I used to be. And I always searched for, but never found the woman who shares my passions [for leather]. (Oliver, German in original)

Oliver, similarly to Peter, is not satisfied sexually and tries to compensate with dolls. Both men claim to be successful so far. They share the narrative that the male sex drive must be acted out and, thus, cannot be controlled. In their understanding the dolls helps them to channel these sexual urges.

While the reasons for having sex with dolls differ in sex toy usage, the dolls serve primarily a sexual purpose. In the following section we will introduce doll owners who use dolls as substitutes for human partners.

Using sex dolls as substitutes

Thirteen participants use their dolls to substitute for a human partner. None of them is in a relationship, and most of them live alone. The respondents have been unable to engage in a human relationship for different reasons and now use sex dolls instead. The dolls help to overcome loneliness and the owners often attach to the dolls. However, the doll is a proxy for human partnership only; substitute users would prefer a human partner over a synthetic one. The doll can be understood as a transitional object between being single and finding a human partner. The preference for human partners is reflected in the correlation between substitute use and the item “I would rather like to have a real partner than a doll” (r_pb(56) = −.51, p < .001). Until a human partner is found, the doll will simulate intimacy and company.

Substitute users do not engage in human relationships due to social phobia, circumstances such as the Covid-19 lockdown, and sexual dysfunction. Yet, the latter was barely present in the sample (see Table 5) and substitute usage correlated insignificantly with items concerning ability to orgasm and lack of erection (ability to orgasm: r_pb(59) = .14, p = .291; lack of erection during sex: r_pb(49) = .19, p = .175). Premature ejaculation was weakly correlated with substitute usage as the three respondents who report to always have a premature ejaculation are substitute users (premature ejaculation and substitute use: r_pb(49) = −.26, p = .064).

The reason why Jonathan (Survey participant 53) substitutes a doll for a human partner is the Covid-19 lockdown. The restrictions force him to stop dating and, thus, he cannot find a human partner. In such lonely times, the doll keeps him company:

Proxy for human partner in bed (sleeping together and sex). [….] Replacement for past partner/relationship. The doll has helped me overcome loneliness and isolation. I've slept better and feel less anxiety/despair when sleeping next to a partner; the doll has been a stand in for that. (Jonathan)

Jonathan’s doll not only offers sexual pleasure but emotional support as well. Despite interacting with “her” daily, he still neither names “her,” nor would he prefer “her” over a human partner.

Demonstrated by the case of Harald (Survey participant 4), most substitute users would stop using dolls, if they had a human partner:

If I find a partner again, then the girls will lounge around or will go back to their boxes. (Harald, German in original)

This doll owner would dispose of his dolls if he had a human partner again. As all substitute users prefer human partners, dolls are perceived as a temporary solution. Users are not single by choice and, thus, substitute use is often accompanied by self-doubt:

Sometimes I think: “shit, now I’m stuck … here is a doll … was that it? Never again a relationship? … life sucks … why did I buy it? … Disposal? I’m in conflict with myself: Do I want to die with a doll? … But currently I have nothing else … I’m happy that at least someone [the doll] is there for me … what shall I do? I don’t know. I simply hope that one day we can separate […] He [the male doll] is simply my temporary companion so I don’t have to feel all alone. (Kilian, German in original)

The doll helps Kilian (Survey participant 94) in difficult times while at the same time it reminds him of his solitude. Kilian’s case illustrates an ambivalence in substitute use: the dolls alleviate the negative consequences of being a single while simultaneously reminding the users of their loneliness.

Although substitute users do emotionally bond with their dolls, most of them prefer human relationships. Even if users are intimate with their dolls, the dolls remain temporary solutions to cope with loneliness.

Using sex dolls as synthetic partners

In 31 cases, sex dolls do not serve as sex toys nor as substitutes for human partners. Instead, in synthetic partner use users prefer dolls over human partners in general. Synthetic partner use is characterized by humanization of and emotional attachment with dolls. Synthetic partner users consider their dolls an important part of their lives and they report falling in love with them. A relationship with a human partner is no longer desired. In this section we decided not to use quotation marks anymore when referring to a doll as she/her, because we want to respect the point of view of the synthetic partner users for whom the dolls are human-like and she/her seems the appropriate way of addressing them.

For Karsten (Interview partner 4), his main doll Sarah—next to a second doll that he bought as a companion for Sarah—is an integral part of his life:

Aside from work I spend most of my time with my dolls, they’ve become an essential part of my life, [my doll Sarah] supports me tremendously, I couldn’t imagine a life without her. (Karsten, German in original)

Karsten is strongly attached to his doll Sarah. He is in love with her and considers her his wife. Having Sarah, he does not want a human partner any longer.

It takes time, however, to develop a strong bond such as that. Harrison (Survey participant 55) bought his doll initially for sexual purposes but soon realized that the doll means more to him:

I ordered “her” [quotation marks in original] shortly after COVID hit and the bars shut down, initially intending to only use the doll for sexual gratification while hooking up carried a risk of infection. My career was stable, and I had broken off a three-year engagement a year ago (at that time, 2 years from present). At that time, I suffered from long-running depression and passive suicidal ideation, as well as alcoholism. When “she” first arrived, I didn’t think much of it, not reacting to the doll while unpacking her (the doll itself wasn’t arousing, and I saw “her” as merely an object). However, I began sleeping beside her, and perhaps a month or two later I started caring for her more than I would [for] an object and emotionally valuing her more like I would a person, something that caught me completely off guard. At the very least, I no longer felt loneliness, and my outlook on life improved within months. (Harrison)

Harrison never intended to use his doll for more than sex yet over time the doll became an important part of his life. This is emphasized by the quotation marks he uses when talking about the doll when “her” arrived and after the doll became an emotionally valued partner, he has no need for quotation marks anymore.

But it is not only the time one spends with dolls that causes synthetic partner use (r_pb(55) = −.02, p = .863), neither does this kind of use correlate with the age of the user (r_pb(65) = .22, p = .105). Thus, it is not only time or age that determines falling in love with dolls. To have a human-like relationship with synthetic partners, users create a personality for them and establish daily routines. Tobias (Interview partner 5) for example created an avatar for his doll Kazane on a popular social media platform. It appears as if Kazane herself created all content. To keep “her” personality coherent, Tobias created a complex backstory. Through inventing a life story and creating an avatar Kazane comes to life—at least for Tobias. Similarly, Robert (Survey participant 15) describes the process in which he invented a detailed personality for his doll:

When she arrived, I spent a week just looking at her, gently moving her limbs and extremities just to see how it all worked. Soon after, a name popped into my head. Once she had a name my relationship with her intensified rapidly and is a worldly 35 years old. In my head, she now has preferences for everything ranging from wardrobe, jewelry, hair, and makeup to music. Appearance-wise, she is on what I'd call the Scarlett Johansson-Elisha Cuthbert spectrum, but in my head she has Elizabeth Hurley’s vocal tone and accent. (Robert)

Robert worked out a complex personality for his doll that allows his interactions with “her” to be increasingly human-like. He is willing to sacrifice most of his time and to risk his health by carrying “her” around (his doll weighs 50 kg) to keep “her” as vivid as possible.

In addition to the intellectual creation of a personality, users establish routines such as dressing up, moving around, and caring for their dolls. In the next quote, Karsten describes the routines he established with his dolls:

She sleeps next to me every night, during the day she waits for me to come home from work, in the evenings I watch TV with her on my side. I quickly developed a routine that I like very much. Every morning we cuddle extensively, afterwards I help her into her wheelchair, I put on her wig (at night she has a wig with short hair and during the day long hair), I put on her lipstick, and apply some perfume, then I bring her to the living room. When I go to work, I kiss her goodbye, when I come back, I do the same. When I go to bed, I change her wig, remove her make-up, and lay her next to me. (Karsten, German in original)

Karsten treats his dolls as he would a real person, moves them around, and cares for them deeply.

In synthetic partner usage, sex plays a different part in the relationship between user and doll than it does in substitute and sex toy usage. Synthetic partner users do not use dolls to merely release sexual tension; instead, sex with a doll expresses one’s love and affection. Georg (Interview partner 1) for example often has sex with his dolls after he interacted with them and perceived the interaction as intimate:

Well, the sexual aspect [of doll use] arises from the presence, from interacting with dolls. I mean that’s how people imagine it, that’s how it should be, that sexuality between two humans is not only about relieving sexual tension. […] Instead, time spent together and proximity, say when the doll sits on my lap, and you watch a nice movie, then it can happen that you get in the mood for sex. Actually, a very natural and beautiful process. Because that’s how it should be between humans, too. (Georg, German in original)

Here, sexual arousal is the consequence of an interaction. In George’s interpretation, the sexuality emerges from the relationship he has with his dolls. He experiences the interaction as mutual teasing, which in turn creates his sexual desire.

Daniel (Survey participant 9) never intended to engage in sex with his dolls until after owning “her” for some time a dream changed his mind:

I didn’t have sex with [Zandy] initially but I did have her lay in the bed with me and found that I slept better. One night I had a dream that she was alive and then I found myself having sex with her while waking from the dream, which by itself was quite an experience. After this I developed a bond with [Zandy] that remains to this day, I feel protective of her and this is enhanced by all the photoshoots I've done with her. [….] The fact that I actually care about all my gals is an important factor here so, say when I'm with [Zandy], it’s not just a sexual experience but also an emotional one and that’s the bit that I was not expecting back in 2016 and which has been a revelation! (Daniel)

Daniel is surprised by the emotional weight that intercourse with his dolls carries. Aside from sharing a bed and enjoying photography with his dolls, sex is yet another activity that strengthens the bond between doll and user.

Although users create a personality for their dolls and establish routines that “animate” them, the doll’s personality never exists independently. The doll’s imagined thoughts and feelings are a projection, which the respondents are all aware of. Robert states:

I've projected an entire personality into her [, although] I know she’s just 105lbs of steel and thermoplastic elastomer […] (Robert)

He knows that it is he who invented “her” personality and that in fact the doll is a lifeless object.

Jeffrey (Survey participant 29) explicitly reflects on his relationship with his doll:

She is ready for me whenever I need her, and infinitely patient when I am not available. My wants are her wants, my needs are her needs, and my tastes are her tastes. Whatever I come to her for at any given time (sex, comfort, fun), that is what she is in the mood for. She is forever eager to please and to be pleased. I never have to worry about whether what she really thinks or feels is different from what she says. I know exactly what’s going on in her head, because she is merely a part of me. None of this would be possible with a real person, nor would it be desirable or morally acceptable. [….] But from my doll, these things feel wonderful to me. (Jeffrey)

Jeffrey experiences the relationship with his doll as symbiotic. He knows everything about “her”, as he is the creator of its mind. Moreover, he reports being able to experience through the doll what it is like to be “her:”

I have also realized that I am in a way projecting myself onto her, experiencing through my doll the cuteness and beauty that I can’t experience with my own body. (Jeffrey)

For many doll owners it is the first time doing extensive shopping, painting fingernails, and designing a style. With their dolls, they can experience by proxy what it is like to be a young woman—or what they believe it is like to be a young woman.

The non-responsiveness and passivity of the doll, however, is not considered a disadvantage. Differing from substitute and sex toy usage, synthetic partner users embrace the dolls with all their shortcomings and reframe their passivity with devotion: the doll is always available and always faithful. The dolls are—at least in the users’ eyes—eternally young and, if treated right, eternally beautiful. Some owners emphasize that such preferences apply to dolls only and not to women in general.

Owning and caring for dolls becomes a constituent part in synthetic partner users’ personality, the community of doll owners their family. As the doll is of such importance for the users, being a doll enthusiast becomes a pivotal part of their identity. Georg, for example, is a well-known doll enthusiast. He not only provides the community with technical instructions on doll use, but also serves as a spokesperson for the whole community and is one of the few willing to give public interviews. His whole life revolves around sex dolls: he has worked for the doll industry, spends most of his time on doll maintenance, and all his friends are doll lovers, too.

The emotional attachment with dolls increases throughout the three kinds of usage. In sex toy usage, users are not strongly attached; in substitute use, users do sometimes emotionally attach but do not consider the dolls a longtime relationship; and in synthetic partner use, owners attach to their dolls as if they were humans.

These findings are reflected in the quantitative measures as well. Respondents who use dolls as synthetic partners tend to agree to the sentence “I feel emotionally attached to my doll(s)” (r_pb(61) = .72, p < .001). In comparison, participants who were coded with sex toy usage (r_pb(61) = −.59, p < .001) tend to disagree. Accordingly, sex toy users agree to the item “For me, a sex doll is just a high-quality sex toy” (r_pb(61) = .85, p < .001), whereas synthetic partner users likely disagree (r_pb(61) = −.60, p < .001). Substitute users answer the two preceding items ambiguously (emotional attachment item: r_pb(65) = −.22, p = 0.074; item “just a high-quality sex toy”: r_pb(61) = .10, p = 0.447).

In the next section, we will discuss our findings in relation to the literature. We will then consider the implications of the kinds of sex doll usage for debates in ethics and anthropomorphism.

Discussion

The demographics of the sample confirm previous research: Sex doll owners (that engage in online communities) are single, middle-aged, and heterosexual (Langcaster-James & Bentley, 2018; Valverde, 2012). Langcaster-James and Bentley (2018) find that sex doll owners are “relatively well off” (p. 13) and suggest that sex dolls are more prevalent in wealthy men, as they dispose of the time and money to buy and maintain a sex doll. Our data suggests a variety of income level among doll users. Dolls have become rather inexpensive (one participant stated that 800€would be sufficient for a new sex doll) and a second hand market emerged. This makes dolls more affordable for a less wealthy population as well.

The data mostly confirms previously assumed categorizations of sex doll usage (Döring et al., 2020, p. 13) and adds first estimates of their prevalence. The most common kinds of sex doll usage are synthetic partner usage, sex toy usage, and substitute usage; female users, homosexual users, exclusively non-sexual use, and couple use, however, are rare. Pedophilic doll owners and sadistic doll use were also seldom found (see the limitations section for a discussion on possible biases that might cause an underrepresentation of pedophilic and sadistic users). It must be noted that these findings are not generalizable as the study is explorative and the sample size rather small.

Intimate human relationships have many facets among which are love, attraction, sexuality, and attachment (Brehm et al., 2002). While a doll is not equivalent to a human being, a variety of characteristics found in human relationships are reported by doll owners as well. In synthetic partner use, the relationship between doll and owner is emotional and satisfies basic human needs for companionship. Langcaster-James and Bentley (2018) even suggest calling sex dolls “allodolls” (p. 14) to capture this non-sexual part of doll use. Our findings, however, add that nearly all users engage in sex with dolls. Thus, in nearly all cases they do satisfy sexual needs. The usage of sex dolls is more diverse than previous empirical research suggests and neither calling them sex dolls nor allodolls or love dolls can account for the variety of psychological functions they serve.

The different kinds of usage relate to different neighboring fields of research. The use of sex dolls as masturbatory aid for example is similar to the use of sex toys (Döring et al., 2020, pp. 1–2; Döring & Poeschl, 2020), whereas synthetic partner usage shares similarities with the Reborn baby dolls community⁶ (Döring et al., 2020, p. 14; Knafo & Lo Bosco, 2017). At the same time, sex doll owners do have a lot in common: Nearly all owners are single, heterosexual men, who engage in sex with their dolls and are not open about their preference.

Benefits of differentiating kinds of usage

Differentiating kinds of sex doll usage can enrich current debates and guide future research. We will briefly demonstrate the advantages of differentiating sex doll usage for debates on misogyny and for the study of anthropomorphism.

Sex doll use and misogyny

In a variety of publications, scholars discuss the ethical implications of sex dolls (for an overview see Döring et al., 2020, p. 9; Ray, 2016). Most critics of sex dolls follow the symbolic-consequences argument (for an overview see Danaher, 2017). Proponents of such claims suggest that “ethically problematic sexual norms” (p.9) will have negative effects on society, even though the ethically problematic actions creating these norms are only acted out on dolls. Harper et al. (2022) are the first to statistically explore whether sex doll owners indeed score higher on related measures such as rape proclivity. They conclude that although sex doll owners do have more fantasies about coercive sex, they are less likely to act on those fantasies. Depending on some moderating variables, sex doll ownership can even be a protective factor for sexual aggression. While these results have to stand the test of replication, they run against the symbolic-consequences argument.

In their analysis Harper et al., however, do not differentiate between different kinds of doll use. Future research could explore whether the effects found in the statistical analysis differ between the kinds of usage established in this study. For example, we found that some doll owners from the sex toy usage category shared the narrative of an uncontrollable sex drive. Harper et al. (2022) suggest a relation between this narrative and higher self-reported sexual aggression proclivity. Therefore, one could hypothesize that sexual aggression proclivity might be higher in sex toy use than in other kinds of usage.

While the ethical debates continue and the empirical evidence is unclear, we can only speculate whether sex doll usage will and should be protected as a sexual right and be accepted as one of the many ways to gain sexual pleasure in the future (Ford et al., 2021; Kismödi et al., 2017; Rye & Meaney, 2007).

Robot anthropomorphism, and sex dolls

We argue that social robotics and social cognition are fields that could profit from insights in sex doll usage as well. So far only theoretical connections between doll ownership and social robotics were drawn. Levy (2007) for example claims that sex dolls are the predecessors of sex robots. We follow Royakkers and van Est (2015) in assuming that sex dolls will only very slowly progress to sex robots, yet we suggest that social robotics can still profit from sex doll research.

Anthropomorphism—mainly found in synthetic partner usage—is a vividly discussed issue in social robotics as well. Following Epley et al. (2007), anthropomorphism is “the attribution of human characteristics or traits to nonhuman agents” (p. 865) and is discussed in social robotics (Coeckelbergh, 2021; Spatola et al., 2021; Złotowski et al., 2015) and social cognition (Epley et al., 2007; Waytz, 2013). In social robotics, studies focus on the design of robots to purposefully facilitate or hamper anthropomorphism (Złotowski et al., 2015). Authors stress that the psychological mechanisms that underly anthropomorphism need to be studied as well (Coeckelbergh, 2021). These mechanisms are the subject of inquiries in social cognition (Epley et al., 2007).

Our analysis yields different kinds of sex doll usage which vary between low rates of anthropomorphism, for example in the use of dolls as sex toys and high degrees of anthropomorphism for example in synthetic partnership usage. The Three-Factor Theory by Epley (2007) tries to explain why some users deeply attach to their dolls while others do not. While the morphological similarity between dolls and humans equally increases humanization in all kinds of usage, it is plausible that different degrees of loneliness and social disconnection explain why some humanize their dolls and others do not (Epley et al., 2007, p. 875; Shin & Kim, 2020). In addition, some owners might appreciate the predictability offered by sex dolls (Epley et al., 2007, p. 871).

Haslam et al. (2008) suggest that anthropomorphism is linked to the phenomenon of de-humanization studied in social psychology. While people tend to humanize in-group members, members of an out-group are more likely to be de-humanized. This might also apply to sex doll users and could further reinforce social disconnection. Future research on sex dolls should further investigate how anthropomorphism varies over the different kinds of usage, which psychological determinants are involved, and how anthropomorphism and social isolation are related.

Current debates on anthropomorphism and ethics can benefit from the distinction between kinds of sex doll usage. We have discussed our findings with two related fields of research and hope that such debates can profit from sex doll research and that the study of sex dolls in turn can profit from them. In the next section we will discuss the limitations of our approach.

Limitations

This study is explorative; thus, the findings are not generalizable, but allow a closer description of the phenomenon of sex dolls and help formulating hypotheses for statistical confirmation.

The topics pedophilia and misogyny in sex doll ownership were excluded from the study to ensure high compliance in the field. The sample is therefore biased, as we did not specifically ask about pedophile fantasies and attitude toward women. We only recruited in forums with strict policies against young-looking dolls. For example, one only forum banned a variety of young-looking dolls to prevent legal and moral charges. Thus, our results cannot estimate the prevalence of pedophilia and misogyny in the population. However, the data indicates that both issues are related to sex doll usage and should be studied in more detail (Harper et al., 2022; Harper & Lievesley, 2022).

It is important to involve community members in the research process to avoid stereotyping, increase transparency and allow a better understanding of the community needs. To involve participants in our study, we asked two of the interview partners, who presented themselves as spokespersons of the community to advise us. Unfortunately, both were not willing to help for reasons of time.

We assume that the dropout due to the time-consuming open-ended questions and the recruitment on online platforms with their specific audience created a bias in our sample toward doll owners who strongly identify with the topic such as in synthetic partner usage. Casual users are probably less represented in this study. This assumption is supported by the higher number of characters used in the open-ended questions by synthetic partner users versus others (U(N_{synthetic partner use} = 33, N_{non-synthetic partner use} = 36) = 323, z = −3.05, p = .001). This indicates that respondents who use dolls as synthetic partners were much more willing to answer to open-ended questions than others. Together these factors create an overrepresentation of users for whom sex dolls are more than casual sex toys. Future research could design shorter surveys to avoid such biases.

Lastly, because we analyze self-report data, social desirability effects might confound the data. Doll owners might emphasize the aspect of care and love to counteract the perceived stigma and prejudice against sex dolls. This is especially the case in Fabian, who works as a sex doll salesman and has an interest in describing doll ownership to be perfectly harmless. Thus, as it is the case with all self-reported data, we must assume that the stories are honest and must carefully consider the plausibility of each report.

Conclusion

Sex doll ownership remains ambiguous. On the one hand, sex doll owners are mostly heterosexual single males who live in accordance with hegemonial masculinity (Connell, 2005) and are, thus, not associated with discrimination, instead they are themselves suspected to cause harm to others by using dolls (Richardson, 2016). On the other hand, doll users emphasize how sex doll usage is stigmatized in society, and some even call themselves a sexual minority. Accordingly, while sex doll users call for a normalization of sex dolls similarly to the normalization of vibrators for women (Döring & Poeschl, 2020; Herbenick et al., 2010), others call for a ban (Campaign Against Sex Robots, n.d.; Guirola, 2020; Walker, 2018).

So far little empirical studies on sex dolls exist and little is known about the actual usage of dolls. This study contributes further insights in sex doll ownership with a first differentiation between different kinds of sex doll usages, and we hope that current debates and future research can profit from such a differentiation.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Data availability statement

To ensure transparency of the analysis, readers can find the datasheet including the demographics, quantitative measures, and number of codes, the coding system, the interview guideline in the online repository (https://osf.io/xqgz8/?view_only=1f39551ffe9245d6b9742984d7d3c88e). The interview transcripts and the answers to the open-ended questions are not made public to secure participants’ anonymity.