Table 2 Results of the psychometric properties of the instruments included and rated by the COSMIN checklist
From: Systematic review of the psychometric properties of instruments to measure sexual desire
Instrument | COSMIN Assessment | Validity/Reliability |
|---|---|---|
Questionnaire Measure of Sexual Interest (QMSI) [57] | Box A: Poor Box B: Poor Box C: NR Box D: Poor Box E: Poor Box F: NR Box G: NA Box H: NR Box I: NR | Content Validity: This is the standard pattern for the questionnaire; the scale positions within each pair were then randomized. As there are four bipolar adjectives and five levels of sexual behavior to be rated, the questionnaire comprises 140 items. These are set out in random order, in the form of a questionnaire. Construct Validity: PCA identified five components. Factor loadings between 0.83–0.98; p < 0.001. With variances ranging from 69.2 to 92.9% for all three groups. Internal Consistency: Kuder Richardson Coefficient between 0.83–0.95. Test Retest: 0.68 to 0.92 (p < 0,01 or p < 0.001). |
Box A: Fair Box B: NR Box C: NR Box D: Poor Box E: Poor Box G: Poor Box H: NR Box I: NR | Content Validity: A questionnaire was developed consisting of 40 recognized fantasy themes drawn from scientific and erotic literature [2]. Construct Validity: PCA and varimax rotation identified four components. With a correlation of r = 0.04 to 0.32. Total variance explained by 45.47%. Factorial loads ≥0.30 and communalities that vary between 0.14–0.71. Internal Consistency: Cronbach’s alpha between 0.69–0.89 | |
Box A: Fair Box B: NR Box C: NR Box D: NR Box E: Poor Box F: NR Box G: NR Box H: NR Box I: NR | Construct Validity: CFA by LISREL in two samples and identified four oblique factors. Factor loadings ≥0.32. In the first sample (n = 195): x2 = 401.64; d.f = 246; GFI = 0.94; NNFI = 0.95; CFI = 0.96; RMSEA = 0.057 In the second sample (n = 315): x2 = 714.76; d.f = 246; GFI = 0.94; NNFI = 0.93; CFI = 0.93; RMSEA = 0.078 Internal Consistency: Cronbach’s alpha between 0.66–0.79 | |
Sexual Desire Conflict Scale for Women (SDCSW) [18] | Box A: Poor Box B: NR Box C: NR Box D: Poor Box E: Poor Box F: Poor Box G: NA Box H: NR Box I: NR | Content Validity: To begin the process of quantitatively assessing the experience of sexual desire and its dysfunctional manifestations, a set of 43 items were written by the first author and also upon clinicians’ reports of working with such patients, most typically the survivors of childhood sexual abuse. Construct Validity: EFA and varimax rotation identified three factors. Total variance explained by 46.1%.Eigenvalues of ≥3,24. Factor loadings between 0.30 to 0.84 for the three factors. Internal Consistency: Cronbach’s alpha of 0.927. |
Sexual Desire Inventory (SDI-2) [2] | Box A: Good Box B: NR Box C: NR Box D: Good Box E: Good Box F: Good Box G: NA Box H: NR Box I: NR | Content Validity: Items were selected by considering extant theoretical models of desire, diagnostic criteria used in the DSM-III-R for HSDD, and clinical experience in assessing and treating sexual desire disorders. The items were presented initially to sexology researchers and clinicians, who rated the face validity and the clarity of the items. Construct Validity: EFA and MLFA Maximum Likehood Factor Analysis. Eigenvalues of ≥1.85 for both factors. With a correlation of r = 0.35–0.49, (p < 0,001). With total variance explained by 65.11% Hypothesis testing: Was hypothesized that the model containing two correlated dimensions would provide a better fit to the data r = 0.35 Internal Consistency: Cronbach’s alpha between 0.86–0.96 |
Sexual Desire Inventory (SDI-2) [25] | Box A: Good Box B: Good Box C: NR Box D: Good Box E: Good Box F: NR Box G: Fair Box H: NR Box I: NR | Construct Validity: PCA identified two factors. Eigenvalues of 4.66 for the first factor and 1.85 for the second factor. With total variance explained by 64% Internal Consistency: Cronbach’s alpha between 0.78–0.87. Reliability: Test-Retest: Correlation between r = 0,76–0,83 and r = 0,82 for the global test. Cross-cultural Validity: The questionnaire was translated into German by two independent persons (including a native English speaker with excellent German language skills). The two versions were reviewed and revised in several steps for deviations. Thus, taking into account the clarity of the content and the correct German grammar, which was checked by a third person independent of the study, the final German version of the SDI-2 was developed. In terms of content clarity, the synonyms for English-speaking terms used in Germany were taken into account. |
Sexual Desire Inventory (SDI-2) [26] | Box A: Good Box B: NR Box C: NR Box D: NR Box E: Good Box F: Fair Box G: Poor Box H: NR Box I: NR | Construct Validity: CFA and the oblique rotation identified four factors: x982 = 299.50, GFI = 0.98, TLI = 0.96, RNI = 0.96, RMSEA = 0.058 attested to two additional factors tested (fantasies and erotophilia). Hypothesis testing: Was hypothesized that the model containing two correlated dimensions would provide a better fit to the data r = 0.49 (p < .001). Internal Consistency: Cronbach’s alpha between 0.87–0.88 Cross-cultural Validity: Was translated via the parallel-blind technique. This approach requires the participation of at least two people translating an inventory from the source (English) to the target language (Spanish). Then, these translators compare their individual work and collaborate on the final version. |
Menopausal Sexual Interest Questionnaire (MSIQ) [19] | Box A: Good Box B: Good Box C: NA Box D: Good Box E: Good Box F: Excellent Box G: NA Box H: NR Box I: Good | Content Validity: It was designed to focus on sexual desire, responsiveness, and satisfaction in postmenopausal women. The MSIQ was the primary outcome measure of a double-blind efficacy trial. Construct Validity: PCA and varimax rotation identified three factors. Total variance explained by 75%. Eigenvalues of 6.44; 1.05; 0.67 respectively. Factor loadings between 0.30 to 0.95 Hypothesis testing: For the desire domain of the MSIQ, the correlation coefficient was r = 0.82 and for the total score of the MSIQ r = 0.81. In contrast, the MSIQ total score was inversely related to the domains in the KSQ and MENQOL: KSQ-Anxiety, − 0.002; KSQ-Depression, − 0.14; and MENQOL-Sexual, r = − 0.42. Reliability: Test-Retest: Correlation for all items (r = 0.52–0.76) and the total correlation scale (r = 0.79) Internal Consistency: Cronbach’s alpha of 0.87 |
The Sexual Interest and Desire Inventory-Female (SIDI-F) [20, 21, 67, 68] | Box A: Good Box B: Excellent Box C: NR Box D: Excellent Box E: Good Box F: Excellent Box G: NA Box H: NR Box I: NR | Content Validity: Was developed as a clinician administered assessment tool to quantify severity of HSDD, and to assess change in response to treatment. A preliminary version of the SIDI-F was presented to 21 volunteers, nine participants without sexual complaint and 12 participants diagnosed with HSDD. This pilot study was carried out in order to assess comprehension and ease of use of the tool, as well as to determine that relevant symptoms were being assessed. A panel of experts in scale development and sexual dysfunction in women from academia and the pharmaceutical industry was convened and met regularly via teleconferences and face-to-face meetings, with e-mail review of ideas and changes [17]. Construct Validity: Adequate discrimination in 13 items, however, 5 items were not sensitive to identify the severity of hypoactive sexual desire disorder. Scores ranging from 23 to 58 (average of 37). Hypothesis testing: The SIDI-F was highly correlated with the arousal, desire, and satisfaction domains of the FSFI (all correlations > 0.8) but not with the lubrication, orgasm, and pain domains. The total score on the SIDI-F was also highly correlated with the arousal, desire/frequency, desire/interest, and pleasure domains of the IVR-CSFQ (all correlations > 0.7) but not with the orgasm subscale. [57]. Internal Consistency: Cronbach’s alpha of 0,90 Reliability: Intraclass Correlation Coefficient (ICC) between 0.85–0.90 |
The Sexual Interest and Desire Inventory-Female (SIDI-F) [28] | Box A: Poor Box B: Fair Box C: NR Box D: Fair Box E: NR Box F: NR Box G: Poor Box H: NR Box I: NR | Content Validity: The data collected from face-to-face interviews were checked to determine the participants’ understanding and interpretation of the questionnaire. The qualitative content validity of the Farsi version of SIDI-F was determined by 14 experts in the field of sexual and reproductive health and psychiatry. CVR scores for all items were equal or more than ≥0.79. Internal Consistency: Cronbach’s alpha of 0.89 Reliability: Intraclass Correlation Coefficient (ICC) between 0.577–0.995 Cross-cultural Validity: The inventory was translated into Farsi by two Persian translators. The review and comparison of the two scripts were done by both translators and the head researcher. The final version of the questionnaire was agreed upon after selecting the most appropriate translated phrases, and then two English translators translated the script into English again to ensure the accuracy of the primary translation. |
Cues for Sexual Desire Scale CSDS [23] | Box A: Good Box B: NR Box C: NR Box D: Good Box E: Excellent Box F: Fair Box G: NA Box H: NR Box I: NR | Content Validity: Fifty women (age range 18–67 years) were involved in the item generation stage. The 125 generated items were listed using a conventional questionnaire format. Construct Validity: PCA and varimax rotation identified four factors. Factor loadings between 0.43 to 0.89 Hypothesis testing: Concurrent validity was assessed by calculating relations between the four factor scores and the total score of the CSDS with the FSFI desire domain scores for women with HSDD (N = 30). The different construct revealed r = 0.10–0.24 between the FSFI and CSDS. Internal Consistency: Cronbach’s alpha between 0.78–0.92 |
Cues for Sexual Desire Scale CSDS [27] | Box A: Good Box B: NR Box C: NR Box D: NR Box E: Excellent Box F: NR Box G: Fair Box H: NR Box I: NR | Cross-cultural Validity: The original version of the CSDS was translated into Portuguese by three independent persons fluent in Portuguese and English. The final version was back-translated by a native English speaker. The translation of the English version into Portuguese was semantically equivalent to the English original accordingly to the retro-translation. Construct Validity: CFA identified four factors: x2/d.f. = 24.5; CFI = 0.793; GFI =0.754; RMSEA = 0.08; P[RMSEA < 0.05] < 0.001 PCA and varimax rotation identified six factors, but the scree plot extracted five factors. Total explained variance of 58.3% with factorial loads between 0.492–0.854, eigenvalue > 1 and communalities between 0.408–0.750 Internal Consistency: Cronbach’s alpha between 0.87–0.90 |
Screener for Hypoactive Sexual Desire Disorder in Menopausal Women (SHSDD) [22] | Box A: Good Box B: Good Box C: NR Box D: Good Box E: Good Box F: NR Box G: NA Box H: NR Box I: NR | Content Validity: The tool was then optimized based on feedback from focus groups consisting of physicians and patients in the European Union (Italy, the Netherlands,United Kingdom) and the United States. Each of the four patient focus groups included seven participants over the age of 50 years, who were recruited through general advertising. The four primary care physician focus groups consisted of five participants each. Intrarater reliability with a kappa statistic of 0.97. Construct Validity: PCA identified one factor with eigenvalue of 2.98. Total variance explained by 62.9% Factorial loads and communalities were not reported. Internal Consistency: Cronbach’s alpha of 0.79 Reliability: Intraclass Correlation Coefficient (ICC) of 0.70 |
The Sexual Arousal and Desire Inventory (SADI) [69] | Box A: Good Box B: NR Box C: Good Box D: Fair Box E: Good Box F: Good Box G: NA Box H: NR Box I: NR | Content Validity: List of 86 English descriptors was compiled by interviewing approximately 500 men and women. Construct Validity: PCA and varimax rotation identified four factors. Total variance explained by 41.3%. Factor loadings ≥0.30 for the total of factors. Hypothesis testing: The SADI demonstrated a correlation with an instrument measuring the same construct, MISSA—Sexual Arousal with the SADI—Evaluative r = 0.578–0.805 (p < 0.01) two tailed. The FS—Positive Affect with the SADI—Motivational and the SADI—Physiological r = 0.531–0.554 (p < 0.01) two tailed. The SDI-2 with the SADI—Evaluative and SADI—Motivational r = 0.536–0.595 (p < 0.01) two tailed. Cronbach’s alpha: 0.905 Measurement error: SEM > 0.041 |
Female Sexual Desire Questionnaire (FSDQ) [24] | Box A: Good Box B: NR Box C: NR Box D: Good Box E: Good Box F: Good Box G: NA Box H: NR Box I: NR | Content Validity: Preliminary items for the FSDQ were determined through individual interviews with 40 heterosexual partnered women. Interview data were analyzed using the principles of interpretive phenomenological analysis, and questionnaire items were developed to reflect the themes extracted from these data. Questions assessing the DSM-IV-TR criteria for HSDD. Approximately 250 candidate items were peer reviewed by a group of three researchers/clinicians for item appropriateness, relevance, redundancy, and ease of understanding. Construct Validity: EFA and the oblimin rotation identified six factors. Total variance explained by 60%. Factor loadings > 0.40 Hypothesis testing: The FSDQ demonstrated a correlation with an instrument measuring the same construct, The sexual desire factor the FSDQ with the SDI-2 dyadic domain r = 0.78 (p < 0.01) two tailed. The solitary desire factor the FSDQ with the SDI-2 solitary domain r = 0.83 (p < 0.01) two tailed. Cronbach’s alpha: 0.80–0.92 |