Table of Contents

Sociosexual Orientation Inventory

JEFFRY A. SIMPSON,1 University of Minnesota

STEVEN W. GANGESTAD, University of New Mexico

In the 1940s and 1950s, comprehensive surveys of the sexual practices of North American men (Kinsey, Pomeroy, & Martin, 1948) and women (Kinsey, Pomeroy, Martin, & Gebhard, 1953) documented that people differ dramatically on several “sociosexual” attitudes and behaviors. Although men, as a group, displayed greater sexual permissiveness than women on most sociosexual attitudes and behaviors (e.g., men have more permissive attitudes toward casual sex, and they are more likely to have sexual affairs), one of the most striking features of the Kinsey data is that much more variability in sociosexual attitudes and behaviors exists within each sex than between men and women. Some women, for example, are more sexually permissive than most men, and some men are less permissive than most women.

The Sociosexual Orientation Inventory (SOI; Simpson & Gangestad, 1991) was developed to measure individual differences in willingness to engage in casual, uncommit- ted sexual relationships. The SOI assesses individuals’ past sexual behavior, anticipated (future) sexual behavior, the content of their sexual fantasies, and their attitudes toward engaging in casual sex without commitment and emotional investment. Individuals who score high on the SOI have an unrestricted sociosexual orientation. These individuals report having a larger number of different sexual partners in the past year, anticipate having more partners in the next 5 years, have had more one-night stands (“hook-ups”), fantasize more often about having sex with people other than their current (or most recent) romantic partner, and believe that sex without emotional ties is acceptable. Individuals who score low on the SOI have a restricted sociosexual orientation. These individuals report fewer sexual partners in the past year, anticipate fewer partners in the next 5 years, are less likely to engage in “one-night stands,” rarely fantasize about extra-pair sex, and do not believe in having sex without love and commitment.

Description

The SOI has seven items. Two items ask respondents to report on their past sexual behavior: Item 1 (the number of sexual partners in the past year) and Item 3 (the number of times they have had sex with someone on only one occasion). Item 2 assesses future sexual behavior (the number of partners anticipated in the next 5 years). Item 4, answered on a Likert-type scale, inquires about sexual fantasies (how often they fantasize about having sex with someone other than their current [or most recent] romantic partner). Items 5, 6, and 7, all answered on Likert-type scales, ask about respondents’ attitudes toward engaging in casual sex. These seven items load on a higher-order factor labeled sociosexuality.

Response Mode and Timing

Items 1–3 on the SOI (those that inquire about past and future sexual behavior) require respondents to write down specific numbers of sexual partners. Items 4–7 (those that inquire about fantasies and sexual attitudes) are answered on Likert-type scales. The SOI takes 1–2 minutes to complete.

Scoring

Item 7 of the SOI must be reverse keyed. Items 5, 6, and 7 are then aggregated (summed) to create the attitudinal com- ponent of the SOI. The following weighting scheme is used when aggregating the five components: SOI = 5X (Item 1) + 1X (Item 2) + 5X (Item 3) + 4X (Item 4) + 2X (aggregate of Items 5–7). To ensure that Item 2 does not have disproportionate influence on the total SOI score, the maximum value of Item 2 is limited to 30 partners. This weighting scheme approximates the scores that individuals would receive if the five SOI components were transformed to z scores, unit-weighted, and then summed. Scores based on the current weighting scheme correlate at or above .90 with a unit-weighting system (Simpson & Gangestad, 1991).

SOI scores can range from 10 (a maximally restricted orientation) to 1,000 (a maximally unrestricted orientation). The normal range in college samples is 10–250. Because men tend to score higher on the SOI than women (Simpson & Gangestad, 1991, 1992), respondents’ gender should be partialed before statistical analyses are conducted, or analyses should be performed separately on women and men.

Some respondents will occasionally report very high numbers for Items 1–3. In college samples, 30 is the maxi- mum value for Item 2. If respondents report more than 20 partners for Items 1 or 2, these individuals may be outliers who could have undue influence on the results. Thus, outlier detection should always be done prior to analyzing SOI scores.

Reliability

The SOI is internally consistent (average Cronbach alpha = .75; Simpson & Gangestad, 1991, 1992). Test-retest reli- ability over 2 months is high (r = .94; Simpson & Gangestad, 1991; see Simpson, Wilson, & Winterheld, 2004, for addi- tional information).

Validity

Predictions for individuals who have restricted or unrestricted sociosexual orientations can be derived from the theoretical construct of sociosexuality (see Gangestad & Simpson, 1990; Simpson et al., 2004). Predictive validity evidence for the SOI is reviewed in Simpson et al. (2004). Evidence for its convergent and discriminant validity properties also exists. With regard to convergent validity, for example, more unrestricted individuals (relative to more restricted ones): (a) engage in sex earlier in their roman- tic relationships, (b) are more likely to have sex with more than one partner during a given time period, and (c) tend to be involved in sexual relationships characterized by less investment, less commitment, less love, and weaker emotional ties (Simpson & Gangestad, 1991). More unrestricted individuals also score higher on other scales known to tap related constructs (e.g., sexual permissiveness, impersonal sex).

More unrestricted people also desire, choose, and acquire romantic partners who have different attributes compared to more restricted people (Simpson & Gangestad, 1992). For example, more unrestricted individuals prefer partners who are more physically attractive and have higher social status, and they place less emphasis on kindness, loyalty, and stability. More restricted persons prefer partners who are kinder and more affectionate, more faithful and loyal, and more responsible, and they place less weight on attractiveness and social status. In dating initiation studies (Simpson, Gangestad, & Biek, 1993), more unrestricted persons—especially men—display more nonverbal behaviors known to facilitate rapid relationship development (e.g., more smiling, laughing, maintaining direct eye con- tact, flirtatious glances; for further validity information, see Simpson et al., 2004).

In terms of discriminant validity, Simpson and Gangestad (1991) found that more restricted persons (a) do not have appreciably lower sex drives and (b) do not score higher on scales assessing sexuality-based constructs that should not correlate with the SOI (e.g., sexual satisfaction, sex guilt, sex-related anxiety).

1Address correspondence to Jeffry A. Simpson, Department of Psychology, University of Minnesota, Minneapolis, MN 55455; e-mail: [email protected] or to Steven W. Gangestad, Department of Psychology, University of New Mexico, Albuquerque, NM 87131.