1
Trait Sexual Motivation Questionnaire: Concept and Validation Rudolf Stark, PhD,*† Sabine Kagerer, Dipl.-Psych,† Bertram Walter, PhD,* Dieter Vaitl, PhD,* Tim Klucken, PhD,† and Sina Wehrum-Osinsky, PhD† *Bender Institute of Neuroimaging, Justus-Liebig-University, Giessen, Germany; †Department of Psychotherapy and Systems Neuroscience, Justus-Liebig-University, Giessen, Germany DOI: 10.1111/jsm.12843
ABSTRACT
Introduction. Trait sexual motivation defines a psychological construct that reflects the long-lasting degree of motivation for sexual activities, which is assumed to be the result of biological and sociocultural influences. With this definition, it shares commonalities with other sexuality-related constructs like sexual desire, sexual drive, sexual needs, and sexual compulsivity. Aim. The Trait Sexual Motivation Questionnaire (TSMQ) was developed in order to measure trait sexual motivation with its different facets. Methods. Several steps were conducted: First, items were composed assessing sexual desire, the effort made to gain sex, as well as specific sexual behaviors. Factor analysis of the data of a first sample (n = 256) was conducted. Second, the factor solution was verified by a confirmatory factor analysis in a second sample (n = 498) and construct validity was demonstrated. Third, the temporal stability of the TSMQ was tested in a third study (n = 59). Main Outcome Measure. Questionnaire data. Results. The exploratory and confirmatory factor analyses revealed that trait sexual motivation is best characterized by four subscales: Solitary Sexuality, Importance of Sex, Seeking Sexual Encounters, and Comparison with Others. It could be shown that the test quality of the questionnaire is high. Most importantly for the trait concept, the retest reliability after 1 year was r = 0.87. Conclusion. Our results indicate that the TSMQ is indeed a suitable tool for measuring long-lasting sexual motivation with high test quality and high construct validity. A future differentiation between trait and state sexual motivation might be helpful for clinical as well as forensic research. Stark R, Kagerer S, Walter B, Vaitl D, Klucken T, and Wehrum-Osinsky S. Trait sexual motivation questionnaire: Concept and validation. J Sex Med **;**:**–**. Key Words. Trait Sexual Motivation; Sexual Motivation; Sexual Desire; Retest Reliability
Introduction
S
exual motivation is a psychological construct that describes the driving force for sexual activity. It is related to the importance of sexual activity for the well-being of a subject. People’s extent of sexual activity varies over time due to different factors, for instance, the availability of a sexual counterpart, fitness, experienced stress, or age. Beside these fluctuating, mainly external influences, a more general motivation for sexual activity differs between individuals. Some people are eager © 2015 International Society for Sexual Medicine
for sexual experiences, while others seem to have almost no interest in sexual encounters. The sexual motivation of an individual might be reflected by the desire for sexual activities, the importance of sexual activities for a person, the effort to gain sexual activity, and the actual frequency of sexual activities. With this definition, sexual motivation shares commonalities with other constructs such as sexual interest, sexual desire, or maybe to a lesser extent with sexual drive. The term sexual drive links sexual activity closely to a biological need. J Sex Med **;**:**–**
2 Modern sexual motivation theories question the extent of the influence of a biological drive [1]. The differentiation of sexual motivation from constructs like sexual interest or sexual desire is difficult and thus the above-mentioned terms are often used synonymously. The history of the conceptualization of sexual motivation and related concepts started with Freud, who abstractly defined sexual drive as libido [2]. Later, Kinsey, Pomery, and Martin [3] proposed more specifically that the number of orgasms in a given time period (called outlets) best reflects sexual motivation. In contemporary models of sexual behavior (e.g., [1,4–6]), sexual motivation is conceptualized as the driving force behind sexual engagement accompanied by subjective sexual feelings (i.e., experienced sexual arousal), physiological sexual responses (e.g., genital responses), and sexual behaviors. The extent of sexual motivation depends on sexual incentives, cognitions, memories, and sex-related biological states. These incentive sexual motivation models are in line with up-to-date general motivation theories, which emphasize the importance of conditioned cues for motivated behavior [7,8]. Most of the sexual motivation models stress the importance of present incentives and the current biological state of the organism for the current sexual motivation, but they usually also propose more stable, long-lasting influences from biological (e.g., genes) or sociocultural factors. Therefore, it is reasonable to differentiate between state and trait sexual motivation. Trait sexual motivation can be conceptualized as stable over time, mainly reflecting genetic as well as long-lasting sociocultural influences (aspects of these influences can be assessed, e.g., by the Sociosexual Orientation Inventory [9]). In contrast, state sexual motivation is the result of an interaction of trait sexual motivation and several fluctuating influences, e.g., the presence of adequate sexual cues, current biological state. This clear distinction between state and trait sexual motivation has—to our knowledge—not been made so far. Established concepts like sexual interest, sexual desire, or sexual drive reflect aspects of trait sexual motivation, but are sometimes also used for aspects of state sexual motivation. Therefore, a clarification by using a more precise terminology seems necessary. There is a long-lasting controversy about the relative contribution of biological and cultural factors in determining human sexuality [10]. Corresponding to this, the opposing positions are J Sex Med **;**:**–**
Stark et al. reflected by essentialism, which assumes that sexual phenomena are mainly determined by internal biological factors as, e.g., hormones and genes, and social constructionism, which emphasizes external, cultural influences [11]. The balance between nature and culture might be different between the sexes: Baumeister [10] presented evidence that the cultural influences are greater on female sexuality than on male sexuality, resulting also in greater erotic plasticity in women than in men. Is it possible that cultural influences, which can change over time, contribute to a trait, which is a longlasting property by definition? The answer is yes and no, depending on whether these influences are stable or unstable over time. The concept of trait sexual motivation is justified by findings from animal and human studies, which were able to show genetic influences on sexual behavior. From rat studies, it is known that dopaminergic transmission in the nucleus accumbens is involved in sexual behavior [12–14] and this transmission is partly under the control of genes (see review by Le Foll et al. [15]). A direct genetic contribution to sexual motivation has recently been demonstrated: Rats selectively bred for high novelty exploration were more sexually motivated than rats with low novelty exploration [16]. In humans, there is evidence that sexual motivation is at least partly determined by genetic influences. However, in these studies, aspects related to sexual motivation were assessed rather than the construct per se: Infidelity and the number of sexual partners were found to be under a moderate genetic influence [17], risky sexual behavior was found in a twin study to be partly determined by genetic factors [18], and sexual promiscuity was related to the dopamine D4 receptor gene variation [19]. Polymorphisms in the dopamine D4 receptor gene were also found to contribute to individual differences in human sexual desire [20]. Besides genetic influences, long-lasting sociocultural environments contribute to trait sexual motivation. For example, in a large international study, Schmitt [21] studied sociosexuality from Argentina to Zimbabwe in 48 nations. He reported a strong influence of sociocultural environments on sociosexual orientation, which reflects human mating strategies (e.g., monogamous vs. polygamous) that are probably also related to trait sexual motivation. Another example for long-lasting sociocultural influences is religiosity: For example, Ahrold et al. [22] could demonstrate an influence of religiosity on different sexual attitudes in a Canadian sample.
3
Trait Sexual Motivation Questionnaire Table 1
Questionnaires related to trait sexual motivation
Questionnaire
Authors
Focus
Missing aspects
Hurlbert Index of Sexual Desire
Apt & Hurlbert [30]
Sexual desire
Sexual Desire Inventory
Spector, Carey, & Steinberg [31] Toledano & Pfaus [32]
Sexual desire mainly focused on the last month Sexual arousal and desire with regard of a recent experience Motivational aspects and experiences related to different sexual activities Dispositional motives for sexual activities
Activities to gain sexual encounters Activities to gain sexual encounters Long-lasting perspective
Sexual Arousal and Desire Inventory Age, Gender, and Sexual Motivation Inventory
Quadragno [33]
Affective and Motivational Orientation Related to Erotic Arousal Questionnaire Multidimensional Sexual Self-Concept Questionnaire Sexual Sensation Seeking Scale
Hill & Preston [34]
Kalichman & Rompa [36]
Sexual Compulsivity Scale
Kalichman & Rompa [36]
Different aspects of the sexual self-concept Interest in exiting, novel sexual encounters Sexual urges
Sexual Drive Questionnaire
Ostovich & Sabini [37]
Sexual outlet
Snell [35]
Activities to gain sexual encounters Activities to gain sexual encounters Activities to gain sexual encounters Sexual desire activities to gain sexual encounters Sexual desire activities to gain sexual encounters Activities to gain sexual encounters
The table summarizes the focus of the questionnaire and describes the missing aspects to be suitable for measuring trait sexual motivation
In contrast to trait sexual motivation, state sexual motivation might best be described as the interaction of trait sexual motivation with several different factors. Beside the presence of sexual cues, sex steroids (e.g., [23,24]) and pharmacological influences, e.g., of Selective Serotonin Reuptake Inhibitors (overview by Serretti & Chiesa [25]), play a major role. Further, psychological factors affect state sexual motivation: Low sexual motivation is a frequent symptom of depression [26], stress [27], and negative mood [28]. Already Barlow [29] pointed out the influence of cognitive factors on sexual functioning and sexual motivation. The aim of the present study was to develop a self-report questionnaire to quantify trait sexual motivation reflected by the desire for sexual activities, the importance of sexual activities, the effort to gain sexual activity, and the actual frequency of sexual activities. There are several well-established questionnaires assessing partial aspects of this construct, but as shown in Table 1, all of them have a slightly different focus and miss some essential aspects of trait sexual motivation. This overview demonstrates that most of the existing questionnaires lack either action aspects of sexual motivation or a long-lasting time perspective. For this reason, we have developed the Trait Sexual Motivation Questionnaire (TSMQ) and evaluated it in several studies. In Study I, we first collected items from existing questionnaires and added further items in order to have a well sized item pool reflecting the different aspects of trait
sexual motivation. Then, we analyzed the factor structure of these items by principal component analysis. In Study II, we conducted a confirmatory factor analysis to verify the factor structure of our questionnaire. Further, we were interested in sex differences with regard to trait sexual motivation. Finally, we also investigated the relationship between the TSMQ with other sex-related questionnaires. In Study III, we investigated the stability of the responses in the TSMQ in order to test the hypothesis that the questionnaire assesses a trait. Subjects filled in the questionnaire twice with 1–1.5 years between the two examinations. Method and Results
Study I: Construction of the Raw Version of the TSMQ Aim Based on the concept of trait sexual motivation as described above, a questionnaire was developed: In a first step, items for the categories sexual interest/ sexual desire, activities to gain sexual encounters, and the occurrence of various sexual behaviors were compiled. In a second step, the analysis of the factor structure of these items was to answer the question, whether trait sexual motivation is a one- or a multidimensional construct. Method Subjects. Two hundred fifty-six subjects (133 women) participated in the study. The mean age was 24.21 years (standard deviation [SD]: 3.65, J Sex Med **;**:**–**
4 ranging from 20 to 39 years). Most of the participants were psychology students of the local university (n = 222) who were addressed either by personal contact or by advertisements. They received study credits if desired and had the chance to win 100 euros. Twelve subjects were married, 134 were engaged in a stable partnership, and 109 were single. The study was approved by the ethics committee of the German Society for Psychology and was in the accordance with the Declaration of Helsinki. Materials. The items of the raw questionnaire were obtained in several steps. Based on the considerations that sexual interest/sexual desire, activities to gain sexual encounters, and the occurrence of various sexual behaviors are the core aspects of trait sexual motivation, items from existing questionnaires were screened whether they covered one of the three aspects of trait sexual motivation. A typical item for category sexual interest/sexual desire is “I’m often interested in having sex” but also “Most of the people are less interested in sex than me”. An example for the category activities to gain sexual encounters is “I often go out to find a sexual adventure”. “I often watch pornographic material” would be an example of the category occurrence of various sexual behaviors. The existing questionnaires that were screened for constructrelated items were the Age, Gender, and Sexual Motivation Inventory [33], the Sexuality Scale [38], the Multidimensional Sexuality Questionnaire (MSQ [39]), the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ [35]), the Sexual Desire Inventory (SDI [31]), the Hurlbert Index of Sexual Desire (HISD [30]), the Hurlbert Index of Sexual Fantasy [40], the Sexual Arousal and Desire Inventory (SADI [32]), the Sexual Daydreaming Scale of the Imaginal Processes Inventory (SDS [41]), the Sex Fantasy Questionnaire (SFQ [42]), the Sex Drive Questionnaire (SDQ [37]), Sexual Inhibition Scale and Sexual Excitation Scale (SIS, SES [43]), the Questionnaire Measure of Sexual Interest [44], the Cowart-Steckler Scale of Sexual Experience [45], and the Human Sexuality Questionnaire [46]. All items assigned to one of the three content categories of sexual motivation were preselected by three female research assistants. The final selection was carried out by four of the authors (Bertram Walter, Sabine Kagerer, Sina WehrumOsinsky, and Rudolf Stark). The selected items were translated into German and if necessary rephrased. It was the goal of the first step to J Sex Med **;**:**–**
Stark et al. collect a maximum of 80 items, which should be fairly equally distributed over the three categories. All items were evaluated with regard to redundancy, clarity, and comprehensibility by the above-mentioned authors. Because the screening of the existing questionnaires did not provide enough items in the category activities to gain sexual encounters, the authors added suitable items to this category. From the 70 items used for the factor analysis in Study I, 23 items composed the category sexual interest/desire, 20 items the category activities to gain sexual encounters, and 27 items the category occurrences of various sex behaviors. The numerical prevalence of the latter category was accepted because it was assumed that behavior-related items are more reliable than items from the other categories. All items were formulated as statements to which subjects could agree with “0” = “not at all” to “5” = “very much” (six-point Likert scale). Further items were added asking for background information of the subjects: Age, sex, religion, education, profession, relationship status (single, in a partnership, married), duration of an ongoing partnership, sexual orientation (homosexual, heterosexual, and bisexual), and number of children. The TSMQ was introduced with the following instruction: This questionnaire is about sexual motivation. Sexual motivation means interest in, the wish for, and the desire for sexual activity. If the terms “sex” and “sexual activity” are used in the questionnaire, then solitary sexual activity as well as partnered sexual activity is meant (e.g., masturbation, intercourse, petting, oral sex). You will find a number of sex-related statements in this questionnaire. Please relate these statements to the last five years. Read the statements carefully and decide to which degree the statements fit on a scale from“0” = “not at all” to “5” = “very much.” Although the trait concept would suggest not to use a fixed observation period of five years, we decided to use this time frame because the sample was quite young and we wanted to exclude experiences from puberty, which are more difficult to evaluate in a coherent manner. Beside the background items, the questionnaire consisted of 79 items. Sixty-six of the 79 items were not related to a current partnership and could thus be answered independent of the actual relationship status. Four of these items were infrequency items for the assessment of report accuracy (e.g., “I have never had an erotic thought”). Eight
Trait Sexual Motivation Questionnaire items addressed subjects living in a partnership. If a participant stated not to have had a sexual relationship within the last five years, these questions were marked as missing. Additionally, four items asked about the frequencies of different sexual activities (e. g. “how often were you sexually active in an average month?”). Finally, one question asked, “how long could you refrain from any sexual activity?” The latter five items and the four infrequency items were excluded from the factor analysis of the 70 remaining items. Procedure. Most of the questionnaires were distributed in courses of the psychology department. A research assistant involved in this study informed the audience about the aim of the study, the voluntariness of participation, and anonymity procedures. The students could—if wished—take additional copies for friends and acquaintances. Filled-in questionnaires were thrown into a letter box at the institute. If participants needed credit points or wanted to join the lottery, they could throw a separate note stating their name and email address into the box. Using this procedure, anonymity was ensured throughout the study. Statistical Analysis. If not explicitly noted, all statistical analyses were performed with SPSS 22 (IBM SPSS Statistics, Armonk, NY, USA). Because only items sharing a common variance with other variables should be included in a factor analysis, measurement of sampling adequacy (MSA) was calculated. MSA reflects the shared variance of one item with other items (Kaiser [47]). To ensure the applicability of factor analysis, all variables with an MSA < 0.8 were removed from the questionnaire. To determine the number of extracted factors, parallel analyses of the principal component analysis with 1,000 random repetitions were conducted. Only factors with eigenvalues above chance were considered appropriate factors [48]. An oblique rotation was applied for a better interpretation of the scales. For the construction of questionnaire scales, items have to be clearly allocated to the factors. Therefore, items with low maximum factor loadings were discarded. Further, items were removed when the difference between the highest and the next highest factor loading was lower than 0.2. This criterion ensures that there is a clear assignment of an item to a factor. Corrected item-total correlations as well as Cronbach’s alphas were calculated for the scales of the questionnaire. Further, the correlations
5 of the subscales with the total score and the intercorrelations of the subscales were calculated. The means of the respective items were used as scores of the subscales and the whole questionnaire.
Results Due to the exclusion criterion of MSA
Trait Sexual Motivation Questionnaire: Concept and Validation Rudolf Stark, PhD,*† Sabine Kagerer, Dipl.-Psych,† Bertram Walter, PhD,* Dieter Vaitl, PhD,* Tim Klucken, PhD,† and Sina Wehrum-Osinsky, PhD† *Bender Institute of Neuroimaging, Justus-Liebig-University, Giessen, Germany; †Department of Psychotherapy and Systems Neuroscience, Justus-Liebig-University, Giessen, Germany DOI: 10.1111/jsm.12843
ABSTRACT
Introduction. Trait sexual motivation defines a psychological construct that reflects the long-lasting degree of motivation for sexual activities, which is assumed to be the result of biological and sociocultural influences. With this definition, it shares commonalities with other sexuality-related constructs like sexual desire, sexual drive, sexual needs, and sexual compulsivity. Aim. The Trait Sexual Motivation Questionnaire (TSMQ) was developed in order to measure trait sexual motivation with its different facets. Methods. Several steps were conducted: First, items were composed assessing sexual desire, the effort made to gain sex, as well as specific sexual behaviors. Factor analysis of the data of a first sample (n = 256) was conducted. Second, the factor solution was verified by a confirmatory factor analysis in a second sample (n = 498) and construct validity was demonstrated. Third, the temporal stability of the TSMQ was tested in a third study (n = 59). Main Outcome Measure. Questionnaire data. Results. The exploratory and confirmatory factor analyses revealed that trait sexual motivation is best characterized by four subscales: Solitary Sexuality, Importance of Sex, Seeking Sexual Encounters, and Comparison with Others. It could be shown that the test quality of the questionnaire is high. Most importantly for the trait concept, the retest reliability after 1 year was r = 0.87. Conclusion. Our results indicate that the TSMQ is indeed a suitable tool for measuring long-lasting sexual motivation with high test quality and high construct validity. A future differentiation between trait and state sexual motivation might be helpful for clinical as well as forensic research. Stark R, Kagerer S, Walter B, Vaitl D, Klucken T, and Wehrum-Osinsky S. Trait sexual motivation questionnaire: Concept and validation. J Sex Med **;**:**–**. Key Words. Trait Sexual Motivation; Sexual Motivation; Sexual Desire; Retest Reliability
Introduction
S
exual motivation is a psychological construct that describes the driving force for sexual activity. It is related to the importance of sexual activity for the well-being of a subject. People’s extent of sexual activity varies over time due to different factors, for instance, the availability of a sexual counterpart, fitness, experienced stress, or age. Beside these fluctuating, mainly external influences, a more general motivation for sexual activity differs between individuals. Some people are eager © 2015 International Society for Sexual Medicine
for sexual experiences, while others seem to have almost no interest in sexual encounters. The sexual motivation of an individual might be reflected by the desire for sexual activities, the importance of sexual activities for a person, the effort to gain sexual activity, and the actual frequency of sexual activities. With this definition, sexual motivation shares commonalities with other constructs such as sexual interest, sexual desire, or maybe to a lesser extent with sexual drive. The term sexual drive links sexual activity closely to a biological need. J Sex Med **;**:**–**
2 Modern sexual motivation theories question the extent of the influence of a biological drive [1]. The differentiation of sexual motivation from constructs like sexual interest or sexual desire is difficult and thus the above-mentioned terms are often used synonymously. The history of the conceptualization of sexual motivation and related concepts started with Freud, who abstractly defined sexual drive as libido [2]. Later, Kinsey, Pomery, and Martin [3] proposed more specifically that the number of orgasms in a given time period (called outlets) best reflects sexual motivation. In contemporary models of sexual behavior (e.g., [1,4–6]), sexual motivation is conceptualized as the driving force behind sexual engagement accompanied by subjective sexual feelings (i.e., experienced sexual arousal), physiological sexual responses (e.g., genital responses), and sexual behaviors. The extent of sexual motivation depends on sexual incentives, cognitions, memories, and sex-related biological states. These incentive sexual motivation models are in line with up-to-date general motivation theories, which emphasize the importance of conditioned cues for motivated behavior [7,8]. Most of the sexual motivation models stress the importance of present incentives and the current biological state of the organism for the current sexual motivation, but they usually also propose more stable, long-lasting influences from biological (e.g., genes) or sociocultural factors. Therefore, it is reasonable to differentiate between state and trait sexual motivation. Trait sexual motivation can be conceptualized as stable over time, mainly reflecting genetic as well as long-lasting sociocultural influences (aspects of these influences can be assessed, e.g., by the Sociosexual Orientation Inventory [9]). In contrast, state sexual motivation is the result of an interaction of trait sexual motivation and several fluctuating influences, e.g., the presence of adequate sexual cues, current biological state. This clear distinction between state and trait sexual motivation has—to our knowledge—not been made so far. Established concepts like sexual interest, sexual desire, or sexual drive reflect aspects of trait sexual motivation, but are sometimes also used for aspects of state sexual motivation. Therefore, a clarification by using a more precise terminology seems necessary. There is a long-lasting controversy about the relative contribution of biological and cultural factors in determining human sexuality [10]. Corresponding to this, the opposing positions are J Sex Med **;**:**–**
Stark et al. reflected by essentialism, which assumes that sexual phenomena are mainly determined by internal biological factors as, e.g., hormones and genes, and social constructionism, which emphasizes external, cultural influences [11]. The balance between nature and culture might be different between the sexes: Baumeister [10] presented evidence that the cultural influences are greater on female sexuality than on male sexuality, resulting also in greater erotic plasticity in women than in men. Is it possible that cultural influences, which can change over time, contribute to a trait, which is a longlasting property by definition? The answer is yes and no, depending on whether these influences are stable or unstable over time. The concept of trait sexual motivation is justified by findings from animal and human studies, which were able to show genetic influences on sexual behavior. From rat studies, it is known that dopaminergic transmission in the nucleus accumbens is involved in sexual behavior [12–14] and this transmission is partly under the control of genes (see review by Le Foll et al. [15]). A direct genetic contribution to sexual motivation has recently been demonstrated: Rats selectively bred for high novelty exploration were more sexually motivated than rats with low novelty exploration [16]. In humans, there is evidence that sexual motivation is at least partly determined by genetic influences. However, in these studies, aspects related to sexual motivation were assessed rather than the construct per se: Infidelity and the number of sexual partners were found to be under a moderate genetic influence [17], risky sexual behavior was found in a twin study to be partly determined by genetic factors [18], and sexual promiscuity was related to the dopamine D4 receptor gene variation [19]. Polymorphisms in the dopamine D4 receptor gene were also found to contribute to individual differences in human sexual desire [20]. Besides genetic influences, long-lasting sociocultural environments contribute to trait sexual motivation. For example, in a large international study, Schmitt [21] studied sociosexuality from Argentina to Zimbabwe in 48 nations. He reported a strong influence of sociocultural environments on sociosexual orientation, which reflects human mating strategies (e.g., monogamous vs. polygamous) that are probably also related to trait sexual motivation. Another example for long-lasting sociocultural influences is religiosity: For example, Ahrold et al. [22] could demonstrate an influence of religiosity on different sexual attitudes in a Canadian sample.
3
Trait Sexual Motivation Questionnaire Table 1
Questionnaires related to trait sexual motivation
Questionnaire
Authors
Focus
Missing aspects
Hurlbert Index of Sexual Desire
Apt & Hurlbert [30]
Sexual desire
Sexual Desire Inventory
Spector, Carey, & Steinberg [31] Toledano & Pfaus [32]
Sexual desire mainly focused on the last month Sexual arousal and desire with regard of a recent experience Motivational aspects and experiences related to different sexual activities Dispositional motives for sexual activities
Activities to gain sexual encounters Activities to gain sexual encounters Long-lasting perspective
Sexual Arousal and Desire Inventory Age, Gender, and Sexual Motivation Inventory
Quadragno [33]
Affective and Motivational Orientation Related to Erotic Arousal Questionnaire Multidimensional Sexual Self-Concept Questionnaire Sexual Sensation Seeking Scale
Hill & Preston [34]
Kalichman & Rompa [36]
Sexual Compulsivity Scale
Kalichman & Rompa [36]
Different aspects of the sexual self-concept Interest in exiting, novel sexual encounters Sexual urges
Sexual Drive Questionnaire
Ostovich & Sabini [37]
Sexual outlet
Snell [35]
Activities to gain sexual encounters Activities to gain sexual encounters Activities to gain sexual encounters Sexual desire activities to gain sexual encounters Sexual desire activities to gain sexual encounters Activities to gain sexual encounters
The table summarizes the focus of the questionnaire and describes the missing aspects to be suitable for measuring trait sexual motivation
In contrast to trait sexual motivation, state sexual motivation might best be described as the interaction of trait sexual motivation with several different factors. Beside the presence of sexual cues, sex steroids (e.g., [23,24]) and pharmacological influences, e.g., of Selective Serotonin Reuptake Inhibitors (overview by Serretti & Chiesa [25]), play a major role. Further, psychological factors affect state sexual motivation: Low sexual motivation is a frequent symptom of depression [26], stress [27], and negative mood [28]. Already Barlow [29] pointed out the influence of cognitive factors on sexual functioning and sexual motivation. The aim of the present study was to develop a self-report questionnaire to quantify trait sexual motivation reflected by the desire for sexual activities, the importance of sexual activities, the effort to gain sexual activity, and the actual frequency of sexual activities. There are several well-established questionnaires assessing partial aspects of this construct, but as shown in Table 1, all of them have a slightly different focus and miss some essential aspects of trait sexual motivation. This overview demonstrates that most of the existing questionnaires lack either action aspects of sexual motivation or a long-lasting time perspective. For this reason, we have developed the Trait Sexual Motivation Questionnaire (TSMQ) and evaluated it in several studies. In Study I, we first collected items from existing questionnaires and added further items in order to have a well sized item pool reflecting the different aspects of trait
sexual motivation. Then, we analyzed the factor structure of these items by principal component analysis. In Study II, we conducted a confirmatory factor analysis to verify the factor structure of our questionnaire. Further, we were interested in sex differences with regard to trait sexual motivation. Finally, we also investigated the relationship between the TSMQ with other sex-related questionnaires. In Study III, we investigated the stability of the responses in the TSMQ in order to test the hypothesis that the questionnaire assesses a trait. Subjects filled in the questionnaire twice with 1–1.5 years between the two examinations. Method and Results
Study I: Construction of the Raw Version of the TSMQ Aim Based on the concept of trait sexual motivation as described above, a questionnaire was developed: In a first step, items for the categories sexual interest/ sexual desire, activities to gain sexual encounters, and the occurrence of various sexual behaviors were compiled. In a second step, the analysis of the factor structure of these items was to answer the question, whether trait sexual motivation is a one- or a multidimensional construct. Method Subjects. Two hundred fifty-six subjects (133 women) participated in the study. The mean age was 24.21 years (standard deviation [SD]: 3.65, J Sex Med **;**:**–**
4 ranging from 20 to 39 years). Most of the participants were psychology students of the local university (n = 222) who were addressed either by personal contact or by advertisements. They received study credits if desired and had the chance to win 100 euros. Twelve subjects were married, 134 were engaged in a stable partnership, and 109 were single. The study was approved by the ethics committee of the German Society for Psychology and was in the accordance with the Declaration of Helsinki. Materials. The items of the raw questionnaire were obtained in several steps. Based on the considerations that sexual interest/sexual desire, activities to gain sexual encounters, and the occurrence of various sexual behaviors are the core aspects of trait sexual motivation, items from existing questionnaires were screened whether they covered one of the three aspects of trait sexual motivation. A typical item for category sexual interest/sexual desire is “I’m often interested in having sex” but also “Most of the people are less interested in sex than me”. An example for the category activities to gain sexual encounters is “I often go out to find a sexual adventure”. “I often watch pornographic material” would be an example of the category occurrence of various sexual behaviors. The existing questionnaires that were screened for constructrelated items were the Age, Gender, and Sexual Motivation Inventory [33], the Sexuality Scale [38], the Multidimensional Sexuality Questionnaire (MSQ [39]), the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ [35]), the Sexual Desire Inventory (SDI [31]), the Hurlbert Index of Sexual Desire (HISD [30]), the Hurlbert Index of Sexual Fantasy [40], the Sexual Arousal and Desire Inventory (SADI [32]), the Sexual Daydreaming Scale of the Imaginal Processes Inventory (SDS [41]), the Sex Fantasy Questionnaire (SFQ [42]), the Sex Drive Questionnaire (SDQ [37]), Sexual Inhibition Scale and Sexual Excitation Scale (SIS, SES [43]), the Questionnaire Measure of Sexual Interest [44], the Cowart-Steckler Scale of Sexual Experience [45], and the Human Sexuality Questionnaire [46]. All items assigned to one of the three content categories of sexual motivation were preselected by three female research assistants. The final selection was carried out by four of the authors (Bertram Walter, Sabine Kagerer, Sina WehrumOsinsky, and Rudolf Stark). The selected items were translated into German and if necessary rephrased. It was the goal of the first step to J Sex Med **;**:**–**
Stark et al. collect a maximum of 80 items, which should be fairly equally distributed over the three categories. All items were evaluated with regard to redundancy, clarity, and comprehensibility by the above-mentioned authors. Because the screening of the existing questionnaires did not provide enough items in the category activities to gain sexual encounters, the authors added suitable items to this category. From the 70 items used for the factor analysis in Study I, 23 items composed the category sexual interest/desire, 20 items the category activities to gain sexual encounters, and 27 items the category occurrences of various sex behaviors. The numerical prevalence of the latter category was accepted because it was assumed that behavior-related items are more reliable than items from the other categories. All items were formulated as statements to which subjects could agree with “0” = “not at all” to “5” = “very much” (six-point Likert scale). Further items were added asking for background information of the subjects: Age, sex, religion, education, profession, relationship status (single, in a partnership, married), duration of an ongoing partnership, sexual orientation (homosexual, heterosexual, and bisexual), and number of children. The TSMQ was introduced with the following instruction: This questionnaire is about sexual motivation. Sexual motivation means interest in, the wish for, and the desire for sexual activity. If the terms “sex” and “sexual activity” are used in the questionnaire, then solitary sexual activity as well as partnered sexual activity is meant (e.g., masturbation, intercourse, petting, oral sex). You will find a number of sex-related statements in this questionnaire. Please relate these statements to the last five years. Read the statements carefully and decide to which degree the statements fit on a scale from“0” = “not at all” to “5” = “very much.” Although the trait concept would suggest not to use a fixed observation period of five years, we decided to use this time frame because the sample was quite young and we wanted to exclude experiences from puberty, which are more difficult to evaluate in a coherent manner. Beside the background items, the questionnaire consisted of 79 items. Sixty-six of the 79 items were not related to a current partnership and could thus be answered independent of the actual relationship status. Four of these items were infrequency items for the assessment of report accuracy (e.g., “I have never had an erotic thought”). Eight
Trait Sexual Motivation Questionnaire items addressed subjects living in a partnership. If a participant stated not to have had a sexual relationship within the last five years, these questions were marked as missing. Additionally, four items asked about the frequencies of different sexual activities (e. g. “how often were you sexually active in an average month?”). Finally, one question asked, “how long could you refrain from any sexual activity?” The latter five items and the four infrequency items were excluded from the factor analysis of the 70 remaining items. Procedure. Most of the questionnaires were distributed in courses of the psychology department. A research assistant involved in this study informed the audience about the aim of the study, the voluntariness of participation, and anonymity procedures. The students could—if wished—take additional copies for friends and acquaintances. Filled-in questionnaires were thrown into a letter box at the institute. If participants needed credit points or wanted to join the lottery, they could throw a separate note stating their name and email address into the box. Using this procedure, anonymity was ensured throughout the study. Statistical Analysis. If not explicitly noted, all statistical analyses were performed with SPSS 22 (IBM SPSS Statistics, Armonk, NY, USA). Because only items sharing a common variance with other variables should be included in a factor analysis, measurement of sampling adequacy (MSA) was calculated. MSA reflects the shared variance of one item with other items (Kaiser [47]). To ensure the applicability of factor analysis, all variables with an MSA < 0.8 were removed from the questionnaire. To determine the number of extracted factors, parallel analyses of the principal component analysis with 1,000 random repetitions were conducted. Only factors with eigenvalues above chance were considered appropriate factors [48]. An oblique rotation was applied for a better interpretation of the scales. For the construction of questionnaire scales, items have to be clearly allocated to the factors. Therefore, items with low maximum factor loadings were discarded. Further, items were removed when the difference between the highest and the next highest factor loading was lower than 0.2. This criterion ensures that there is a clear assignment of an item to a factor. Corrected item-total correlations as well as Cronbach’s alphas were calculated for the scales of the questionnaire. Further, the correlations
5 of the subscales with the total score and the intercorrelations of the subscales were calculated. The means of the respective items were used as scores of the subscales and the whole questionnaire.
Results Due to the exclusion criterion of MSA
