Chapter 2
The Premarital Sexual Permissiveness Scales
Premarital Sexual StandardsPremarital Sexual Permissiveness
Universal Permissive Scales
Summary and Conclusions
Premarital Sexual Standards
The key focus of this study is upon sexual permissiveness as measured by Guttman scales. However, it is suitable to start this discussion by a brief glance at studies of sexual standards. After this discussion the Guttman scales will be presented and analyzed as an important way of measuring premarital sexual permissiveness.
A few studies in recent decades have probed the area of sexual attitudes rather than taking the more usual approach of focusing on sexual behavior. Bromley and Britten, two female journalists, studied attitudes among college students during the 1930s.1 Although their samples were not probability samples, it is interesting to note that they found 25 percent of the females and 52 percent of the males to be nonvirginal-rates that have frequently been reported in more recent studies up to the 1960s.2 Bromley and Britten divided their virgins and nonvirgins into subcategories depending on their attitudes, but the classification was not mutually exclusive and included married and homosexual categories. They called many of their male nonvirgins "hotbloods" and many others "pragmatists" (the exact meaning of such terms is hard to determine). Interestingly, they reported that half of the virginal female population approved of intercourse outside marriage even though they had not yet experienced it. The Bromley and Britten study, however, was not an attempt to develop a scientific classification of sexual standards but rather an attempt to afford a popular understanding of college sex life in the 1930s. Thus, it is only of peripheral interest in the search for a classification of sexual attitudes.
A more scholarly attempt at classifying sexual attitudes was published by Rockwood and Ford in 1945.3 They used two classification forms; the first had four categories of attitudes, one of which the respondent was to check as being in agreement with: (1) equal freedom for both, (2) complete chastity for both, (3) complete freedom for men, and (4) complete chastity for women. Clearly, these categories are not mutually exclusive. An individual could believe in the double standard and check either "complete chastity for women" or "greater freedom for men." He could believe in abstinence for both and agree with either "equal freedom for both" or "complete chastity for both." Rockwood and Ford recognized some of these shortcomings and revised this form. Table 2.1 presents the revised set of categories and the results obtained on the basis of 173 Cornell students.
Table 2.1
SEXUAL STANDARDS OF 173 CORNELL STUDENTS
Percentage
SEXUAL STANDARD MEN (Na = 73) WOMEN(N = 100) BOTH(N = 173) Sex relations for both 15 6 9 No sex relations for either 49 76 65 Sex relations for men only 23 11 16 Sex relations for engaged 11 6 8 Question left blank 1 2 2
Source: L. Rockwood and M. Ford, Youth, Marriage, and Parenthood, New York: John Wiley & Sons, copyright © 1945, p. 40.
a In this and all subsequent tables N stands for the number of respondents.
However, this set of categories is still not completely adequate. The first three standards cover the three logically possible attitudes that any culture may take toward premarital coitus. A culture may accept coitus for both sexes, for neither sex or predominantly for one sex.4 Thus, it follows that all additional standards are subtypes of these three standards. Such subtypes must be spelled out, for they give the particular flavor to a society's values in this area. The only subtype in this table is "sex relations for engaged." This is logically a subtype of "sex relations for both." The lack of subtypes means sexual beliefs that are quite different are being lumped together. "No sex relations for either" includes people who will pet to orgasm and those who will not even accept kissing. "Sex relations for both" includes people who will have coitus only when in love and those who accept coitus more casually. Also, there are important sexual standards that are left out entirely; for instance, in what category could an individual be placed who believed that men had more sexual rights than women but that women were allowed to have coitus when in love or engaged? If such a person checked "sex relations for both," then the equalitarian aspect is overestimated; and if he checked "sex relations for men only," the equalitarian aspect is understated. This is a popular standard in our culture, and yet there does not seem to be a clear-cut category for it in the Rockwood and Ford classification schema.
Shortly after the Rockwood and Ford study, Judson Landis questioned five thousand college students from twelve different colleges, using the same fourfold breakdown of standards.5 The results of his research are presented in Table 2.2, together with a summary of the original
Table 2.2
PERCENTAGE OF STUDENTS CHECKING EACH OF FOUR STATEMENTS
REPRESENTING ATTITUDES ON PREMARITAL SEX STANDARDS
APPROVED STANDARD 11 MICHICAN CORNELL
COLLEGES STATE U
1952-1955 1947 1940
(N = 3000) (N = 2000) (N = 1073)
SEXUAL RELATIONS
Males
For both 20 16 15
None for either 52 59 49
For men only 12 10 23
Between engaged only 16 15 11
Females
For both 5 2 6
None for either 65 76 76
For men only 23 15 11
Between engaged only 7 7 6
Source: Judson T. Landis and Mary G. Landis, Building a Successful Marriage, Third Edition, © 1958, P. 215. Reprinted by permission of Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Table has been rearranged.
results obtained at Cornell by Rockwood and Ford. The results were relatively similar, but due to the aforementioned shortcomings, it is difficult to establish the validity of these measures.
More recently, Winston Ehrmann analyzed the sex codes of almost a hundred of his University of Florida students.6 As can be seen in Table 2.3,
Table 2.3
PERCENTAGE OF SINGLE AND DOUBLE STANDARDS IN THE PERSONAL
CODE OF THOSE SUBJECTS WHO HAD BEEN IN LOVE
STANDARDS[a] MALES (N = 45) FEMALES (N = 42) Double 33 0 Conservative single 20 86 General liberal single 42 7 Lover liberal single 5 7
Source: Winston W. Ehrmann, Premarital Dating Bei avior, Ncw York: Holt, Rinehart and Wittstott, copyright © 1959, p. 189.
a. "Conservative single" standard is thc same as rejecting coitus for both male and female. "General liberal single" is thc acccptance of coitus for male and female when in love or when not in love. "Lover liberal single" refers to the acceptance of coitus for both male and female but only when love is present.
the results seem to stress the male's emphasis on the double standard and to deemphasize the female's support of that standard. However, this difference is partially due to the method of classification used by Ehrmann. He counted all women who wanted to remain chaste as belonging in the "conservative single" standard, even though some of them would allow men to have coitus with other women and thus would not really hold a "single" standard,7 in the sense of a standard that applies equally to both sexes. The categories used by Ehrmann are only slightly different from those used in the other studies, and thus they are also subject to the same criticisms. Also, there is no clear place in this schema for the person who accepts coitus for the male at any time but for the female only when in love. This particular variation is quite popular, as evidence will presently show. In addition, there is a need for subdivisions within all standards.
It is not the purpose of this critical analysis to take away from the value of the above studies, but to point out that there are several important premarital sexual standards that are not classifiable by these approaches. It has not been attempted here to cover all previous work, but instead to select the studies most widely known and most typical of the work in this area.8
An attempt was made to overcome some of the limitations of previous studies by composing a set of standards that better fit the existing major premarital sexual standards in American society. As can be seen in Exhibit 2.1, four standards were decided upon: abstinence, double standard, permissiveness without affection, and permissiveness with affection.9 The abstinence standard is divided into four subtypes, depending on whether an
Exhibit 2.1
PREMARITAL SEXUAL STANDARDS IN AMERICA
1. Abstinence (premarital intercourse is considered wrong for both sexes)
(a) Petting without affection (petting is acceptable even when affection negligible)
(b) Petting with affection (petting is acceptable only in a stable, affectionate relationship)
(c) Kissing without affection (only kissing is acceptable, but no affection is req uired)
(d) I issing with affection (only kissing is acceptable, and only in affectionatc relationship)
2. Double standard (males are considered to have greater rights to premarital intercourse)
(a) Orthodox (males may have intercourse, but females who do so are condemnecl)
(b) Transitional (males have greater access to coitus, but females who are in love or engaged
are allowed to have intercourse)
3. Permissiveness without affection (premarital intercourse is right for both sexes
regardless of the amount of affection present)
(a) Orgiastic (pleasure is of such importance that precautions are not stressed)
(b) Sophisticated (pleasure is stressed, but precautions to avoid VD and pregnancy are of first importance)
4. Permissiveness with affection (premarital intercourse is acceptable for both
sexes if part of a stable, affectionate relationship)
(a) Love (love or engagement is a prerequisite for coitus)
(b) Strong affection (strong affection is a sufficient prerequisite for coitus)
Source: Adapted from Ira L. Reiss, Premarital Sexual Standards in America, New York; The Free Press, copyright © 1960, p. 251
individual accepts petting or kissing and whether or not he requires affection to be present. The double standard is divided into the "orthodox" subtype, which does not, under any conditions, accept coitus for women, and the newer, "transitional" subtype, which allows women to have coitus, but only when they are involved in a love relationship. The permissiveness-with-affection standard is divided into those who accept coitus for both men and women but require love and/or engagement to be present and those who require only strong affection to be present. The permissiveness-without-affection standard is composed of those who contend that physical attraction and pleasure are sufficient justification for coitus. The "orgiastic" subtype is less likely to take precautions against consequences such as venereal disease and pregnancy than is the "sophisticated" subtype.
This more refined breakdown has an advantage over the traditional fourfold basic breakdown of standards; it allows for more precise definitions of the four basic standards and their subtypes, so as to avoid overlapping classifications of individuals, and thereby aids in more accurate classification. Of course, even finer subtypes could be composed and finer distinctions made-for example, between mammary and genital petting and between those abstinent individuals who are equalitarian and those who are not. Part of this breakdown was actually used in the analysis presented in Chapter 6.
Such an extensive typological approach to premarital sexual standards also has its limits. At times it is difficult for an individual to have analyzed his sexual beliefs well enough to be able to see himself clearly as holding one or another of the sexual standards presented. In short, a typology calls for a global judgment that is often difficult for a person to make. In addition, a typological approach involves a mixture of dimensions, and it is difficult to compare all standards along any one continuum. It was for these reasons and others that it was decided to try and develop Guttman scales for this study that would measure premarital sexual permissiveness and also afford a basis for judging what sexual standard an individual held. Such scales would be unidimensional and thus allow all respondents to be compared in terms of being more or less permissive. Furthermore, such scales would involve many specific questions, which might well be easier for a respondent to answer and which could ultimately be used to categorize a person into one of the sexual standards. Thus, Guttman scales would yield a unidimensional measure of premarital sexual permissiveness, would present questions easy for the respondent to answer, and ultimately would allow the researcher to judge the sexual standard of the respondent.10
The dimension to be measured was premarital sexual permissiveness,11 for the degree of permissiveness seemed to be of primary importance in any comparison of sexual belief. Another key dimension was equalitarianism. An indirect measure of equalitarianism was worked into the study of permissiveness by having two separate scales, one with questions only about males and one with questions only about females. The similarity or difference in response to these two scales could be used as a measure of equalitarianism without ever asking any direct questions on equalitarianism.
It was felt that in American culture, permissiveness in sex was not only a matter of physical activity but also of the conditions under which the individualwould accept such physical activity. Permissiveness depended, therefore, on both the intimacy of the physical act and the conditions under which it occurred. The key condition was judged to be the amount of affection present in the relationship. The bare question, Do you accept premarital coitus? would not be unidimensional in American society, since it would not have a singular meaning to all respondents. A clearer way of probing this area seemed to be to ask questions like, Do you accept premarital coitus when in love? when not in love? and so forth. Consequently, the Guttman scales are based on questions concerning the individual's acceptance of various physical acts under various conditions of affection.
For the purpose of this study physical acts were divided into three categories: kissing, petting, and coitus. Conditions of affection were divided into four categories: engagement, love, strong affection, and no affection. Each of the three physical conditions was qualified by each of the four affection-related states, making a total of twelve statements which the respondent was asked to agree or disagree with either strongly, moderately, or slightly.12
Exhibit 2.2
MALE AND FEMALE PREMARITAL SEXUAL PERMISSIVENESS SCALESa
The following questions concern some attitudes of yours regarding courtship behavior. We realize that you may be tolerant of what others do and think, but we are not interested now in that. We are interested in your own personal views about the questions we will ask. These questions do not concern what you do- they concern what you believe about courtship. On this sheet we would like you to circle the degree of agreement or disagreement you have with each statement. Just answer these statements on the basis of how you feel toward the view expressed. Your name will never be connected with these answers, so please be as honest as you can. We use the words to mean just what they do to most people, but some may need definition:
Love means the emotional state which is more intense than strong affection and which you would define as love.
Strong affection means affection which is stronger than physical attraction, average fondness, or "liking," but less strong than love.
Petting means sexually stimulating behavior more intimate than kissing and simple hugging, but not including full sexual relations.
Male Standards (Both Men and Women Check This Section)
1. I believe that kissing is acceptable for the male before marriage when he is engaged to be married.
Agree: (1) Strong, (2) Medium, (3) Slight
Disagree: (1) Strong, (2) Medium, (3) Slight
2. I believe that kissing is acceptable for the male before marriage when he is in love.
(The same six-way choice found in statement 1, follows every statement.)
3. I believe that kissing is acceptable for the male before marriage when he feels strong affection for his partner.
4. I believe that kissing is acceptable for the male before marriage even if he does not feel particularly affectionate toward his partner.
5. I believe that petting is acceptable for the male before marriage when he is engaged to be married.
6. I believe that petting is acceptable for the male before marriage when he is in love.
7. I believe that petting is acceptable for the male before marriage when he feels strong affection for his partner.
8. I believe that petting is acceptable for the male before marriage even if he does not feel particularly affectionate toward his partner.
9. I believe that full sexual relations are acceptable for the male before marriage when he is engaged to be married.
10. I believe that full sexual relations are acceptable for the male before marriage when he is in love.
11. I believe that full sexual relations are acceptable for the male before marriage when he feels strong affection for his partner.
12. I believe that full sexual realtions are acceptable for the male before marriage even if he does not feel particularly affectionate toward his partner.
Female Standards (Both Men and Women Check This Section)
(The same twelve items occur here except that the female is the sex referent.)
a There are some very slight differences in the wording of this scale in the student and adult samples. The wording above is the preferred form and was used in the adult sample.
These twelve statements are the basis of both the male and female premarital sexual premissiveness scales; the only difference in each is the sex referent. In order to avoid ambiguity, some key terms used in the scales, such as love, and petting, and strong affection, were defined. The graphic portrayal in Table 2.4 shows the logic of the scales. A simple comparison of the responses to the male and female scales affords a measure of equalitarianism. A respondent can be classified in several ways: (1) how he responds to the scale statement of his own sex, (2) how he responds to the scale statements of the opposite sex, and (3) how equalitarian his responses are to both scales. On the basis of these types of responses to the scale statements, each respondent may be placed into one of the sexual standards
Table 2.4
GRAPHIC PORTRAYAL OF QUESTIONS
IN THE MALE AND FEMALE SCALESa
Affectionate Condition Type of Physical Aclivity
KISSING PETTING COITUS
Engaged 1 5 9
Love 2 6 10
Strong affection 3 7 11
No affection 4 8 12
a The same twelve items occur in both male and female scales except that the sex referent in the former is the male and the sex referent in the latter is the female.
listed in Exhibit 2.1 without ever having been asked directly to make a "pigeonhole" global judgment of himself.13 For example, a respondent who only agreed with the first three "kissing" statements on both the male and female scales would be a kissing-with-affection subtype under the abstinence standard. If the respondent also agreed with the fourth "kissing" statement for both the male and female scale, he would be a kissing-without-affection subtype. If he agreed with the same number of affectionate petting statements (5, 6, 7) for both scales, he would be a petting-with-affection subtype and if he also checked statement 8 he would be a petting-without-affection subtype. The orthodox double-standard male would allow some degree of coitus for himself but not for the female, and thus he might check all twelve items on the male scale with agreement but agree only to the first five or six items on the female scale. The transitional double-standard (see Exhibit 2.1) subtype might agree with one or more of the "affectionate coitus" items on the female scale but with more of these and with item 12 on the male scale. Permissiveness-with-affection subtypes would be detected as those who agree with the affectionate-coitus items on both scales, and permissiveness-without-affection subtypes as those who agree with all twelve items on both scales.
The only subtypes that cannot be detected in terms of the standards in Exhibit 2.1 are the subtypes of the permissiveness-without-affection standard. These subtypes are unique in that they do not relate to affectionate states but rather to the amount of precautions taken. They involve a different dimension than that measured by these scales. Thus, all permissive-nesswithout-affection adherents are counted as one group.
The empirical data revealed some other standards that had not been incorporated into the original schema. One was the nonequalitarian subtype of the abstinence standard that typically involved the acceptance of petting on the male scale but only of kissing on the female scale. More than one out of every six abstinence adherents, particularly females, accepted this subtype. In addition, it was found that some respondents displayed a slight "reverse" double standard, that is, they gave greater coital sexual freedom to the female. This latter response is almost certainly an error. Assuming that people do not have their answers to all twenty-four questions perfectly worked out in their minds, it is not surprising that some should unwittingly agree with one more question on the female scale than on the male scale. However, the matter was checked further. Using several hundred students from an Iowa college (the Iowa College Sample; see Appendix B and Chapters 6 and 7 ), a checklist was composed of the standards in Exhibit 2.1 and of the reverse double standard. No respondent checked the reverse double standard as his choice, despite the fact that these same respondents produced response patterns that were reverse double standard in their replies to the male and female scales. This is further evidence of the inadvertent nature of this response pattern. These reverse doublestandard respondents seem to fit very well within the general category of double-standard respondents in many characteristic ways. The "reverse" image was largely accidental and due to lack of precisely worked-out views in the area. Similarly, a reverse nonequalitarian subtype under the abstinence standard was found, and was shown by the Iowa sample to be a subtype that no one would directly choose and thus consisted mainly of error responses due to lack of clarity on the issues raised in the questions.
Tables 2.5 and 2.6 show the breakdown of the various standards within each of the five high schools and colleges comprising the Five-School Student Sample and show how this varies by sex. The vast differences among these five schools shows up clearly here.14 More will be said about this
Table 2.5
PERCENTAGE ACCEPTING EACH PREMARITAL SEXUAL STANDARD
IN FIVE SELECTED SCHOOLS IN STUDENT SAMPLE[a]
STANDARD WHITE WHITE WHITE NEGRO NEGRO TOTAL
VIRGIN. VIRGIN. N.Y. VIRGIN. VIRGIN. STUDENT
COLLEGE HIGH COLLEGE COLLEGE HIGH SAMPLE
SCHOOL SCHOOL
1. Abstinence
Kissing with affection 6 12 1 4 4 5
Kissing without affection 5 8 1 0 2 4
Petting with affection 36 22 12 15 5 21
Petting without affection 2 2 2 0 0 2
Nonequalitarian 10 15 1 4 7 7
Reverse nonequalitarian 3 6 2 2 2 3
Total 62 65 19 25 20 42
2. Double standard
Transitional 10 5 21 19 11 14
Orthodox 14 10 6 14 4 11
Total 24 15 27 33 15 25
3. Permissiveness with affection
Engaged and/or love 6 6 11 12 16 9
Strong affection 2 4 20 13 21 10
Total 8 10 31 25 37 19
4.Permissiveness without affection
Total 3 2 14 8 16 7
5. Reverse double standard
Total 3 8 10 9 14 7
Number of respondents (264) (147) (188) (165) (57) (821)
a. To obtain a column total of 100 percent use only the five "total" percents. The rounding-off procedure may cause this total to be off one or two points.
b The percents in this column are calculatcd on the entire student sample. They are not simply averages of the percent in each of the five schools because the schools rcprcsent subsamples of various sizes that must be properly weighted.
Table 2.6
later.15 For the present, some general observations will suffice. Thus the general popularity of some of these standards, such as the petting-with-affection subtype of the abstinence standard, and the wide range of acceptance of abstinence, from 65 percent in the white high school to 19 percent in the New York college, was noteworthy. (Thus it would seem that the hoped-for diverse type of sample on which to test the scales was achieved. ) Negroes seemed generally more permissive than whites, although the white New York college equaled the Negro schools. Table 2.6 shows some striking male-female differences, such as twice as many females as males accepting abstinence, male preference for the transitional double standard versus female preference for the orthodox double standard, and permissiveness without affection as very heavily a male position.
The National Adult Sample shows similar features, as can be seen in Table 2.7. To make the results comparable to the student sample race and sex were controlled. The same type of differences between Negroes and whites, and males and females, appeared here. However, the greater conservatism of the adult sample is plainly visible in this table.
Probably the most important point that these findings bring out is that there is a way of measuring premarital sexual standards by use of Guttman scales.
Table 2.7
It is now pertinent to deal with the Guttman scales themselves, which are the basic measure of the dependent variable (premarital sexual permissiveness) used in this study. Although standards will be discussed, the main emphasis throughout this study is on the dimension of permissiveness measured by the scales. The key question of the entire study is: What sociocultural factors help explain the difference between high- and low-permissive groups?
Premarital Sexual Permissiveness
The question of whether arriving at standards via the twenty-four scale questions is more effective than via a simple checklist of the standards was tested on about three hundred Iowa college students. Roughly eighty percent of the students felt the scale questions were easier to answer; twenty percent felt they were equally difficult; virtually no one thought the checklist was easier to answer. The Guttman scales directly measure one dimension-namely, the individual's amount of premarital sexual permissiveness; this is most frequently used as a measure of the dependent variable. Premarital sexual standards can be indirectly measured by these scales, as has been already shown, and they serve as another measure of the dependent variable. Measuring premarital sexual permissiveness directly by the scales allows all respondents to be placed on a continuum of more or less acceptance of permissiveness and allows the continuum to be cut wherever wished. These qualities make permissiveness more useful and simpler than sexual standards. Of course, as shall become apparent, there are problems for which we must look at sexual standards, and both of these dependent variables were checked in all the analyses performed.
The twelve statements that make up each scale were selected because it was felt that they covered the dimension of premarital sexual permissiveness from its low to its high points. However, this can only be certified by testing the scales and by observing the order in which the items scale. The checks made on these scales in the student and adult samples show that the items do cover almost the full range of this dimension, the first item in the scale (see Exhibit 2.2 for description of items) receiving about ninety-five to ninety-nine percent agreement and the last item only receiving about seven to twenty-one percent agreement (see Table 2.8).
Table 2.9 shows the rank order of acceptance of these items in the two major samples. Basically, items 4 and 8 scaled in other than the order in which they were asked. Items 1, 2, and 3 also showed some variation, but the difference was so slight, involving only one or two percentage points, that it was not significant. Clearly these first three "kissing" items are so close to one another that it is best to treat them as one item. In comparing the two samples it can be seen that the student and adult samples scale differently
Table 2.8
PERCENTAGE AGREEING WITH EACH ITEM
IN MALE AND FEMALE SCALES IN THE ADULT
AND STUDENT SAMPLES
QUEST. ADULT STUDENT
NUMBER SAMPLE SAMPLE
TOTAL TOTAL
PERCENT PERCENT
Male Scale
1 95,3 97,5
2 93,6 98,9
3 90,2 97,2
4 58,6 64,2
5 60,8 85,0
6 59,4 80,4
7 54,3 67,0
8 28,6 34,3
9 19,5 52,2
10 17,6 47,6
11 16,3 36,9
12 11,7 20,8
N (1390) (811)
Female Scale
1 95,0 98,5
2 93,3 99,1
3 88,1 97,8
4 50,1 55,2
5 56,1 81,8
6 52,6 75,2
7 45,6 56,7
8 20,3 18,0
9 16,9 44,0
10 14,2 38,7
11 12,5 27,2
12 7,4 10,8
N (1411) (806)
with regard to items 4 and 8. These two items ranked higher (obtained higher relative support) in the adult sample than in the student sample. This is particularly true for item 8.
Item 8 refers to the acceptance of petting with someone for whom there is no particular affection. The lower support of this item, relative to other items, by the student sample, indicates a greater willingness on the part of students as compared to adults to accept coitus when affection is present (items 9, 10, and 11) than to accept petting when affection is absent (as in item 8). The adult sample reversed this priority and accepted petting without affection (item 8) before they would accept coitus with affection (items 9, 10, and 11). This is an important difference in ranking, for it
Table 2.9
SCALE ORDER OF QUESTIONS ON MALE AND FEMALE
SCALES IN THE ADULT AND STUDENT SAMPLES [a]
RANK ADULT STUDENT
SAMPLE SAMPLE
TOTAL TOTAL
QUEST. QUESTION
ORDER ORDER
1st (highest support) 1 2
2nd 2 1
3rd 3 3
4th 5 5
5th 6 6
6th 4 7
7th 7 4
8th 8 9
9th 9 10
10th 10 11
11th 11 8
12th(lowest support) 12 12
a. Same results were obtained for both male and female scales.
would seem to be a characteristic difference distinguishing high- and low-permissive individuals.
This high-permissive-low-permissive difference can be seen by comparing the white and Negro groups in the adult sample in Tables 2.10 and 2.11. Negroes show higher percentages accepting the more permissive items and also rank items 4 and 8 lower (less relative support) than do whites. This is a comparable situation to the relative ranking between the total student sample and the more conservative adult sample as seen in Table 2.9. The Negro segment of the adult sample has almost the exact same high-permissive scale order of items as the total student sample. (Compare Tables 2.9 and 2.11.)
Dividing the student sample into the five different schools also bears out this low- and high-permissive difference. Tables 2.12 and 2.13 show the percent accepting each item and the rank order of each item by school. The schools that seem to have a high level of permissiveness among their students give less relative support (low rank) to items 4 and 8. The Negro schools and the white New York college are the three high-permissive schools. In the male scale, item 4 ranks ninth and tenth in these schools but ranks sixth in the two low-permissive schools. Item 8 ranks eleventh in the high-permissive schools but ninth in the low-permissive schools. This means that the high-permissive schools give more support to having intercourse with someone you are in love with than to kissing someone you don't care for (item 4) and give more support to intercourse with someone you have strong affection for (item 11 ) than to petting with someone for whom
Table 2.10
you do not have affection (item 8). The low-permissive schools reverse these rankings and give the priority to affectionless kissing and petting. It is important to note that these high- and low-permissive differences hold regardless of whether whites and Negroes or students and adults are compared.
There seems to be some sort of compensatory behavior operating here. Those who are high on permissiveness differ from those who are low on permissiveness primarily by their greater willingness to choose a higher level of physical intimacy with affection as more desirable than a lower level of physical intimacy without affection.16 In place of greater physical
Table 2.11
Table 2.12
intimacy with affection the low permissives allow themselves greater access to affectionless behavior that is less intimate, while the high permissives prefer the more intimate affectionate behavior. For example, it seems that the low permissives feel the stronger taboo against coital behavior and thus will even resort to affectionless petting in order to stay away from coitus.17 The high permissives do not exhibit such a taboo on coitus and thus prefer coitus with affection to affectionless petting. This is not by any means art inevitable result, for the high permissives could have been much more acceptant of affectionless kissing and petting.
Some other features stand out in these scale results. The great importance of affection as a means of increasing acceptance of sexual behavior is quite visible. This can be seen most clearly by looking at the "affectionless" statements (items 4, 8, and 12). All groups rated the behavior indicated in item 4 as at least more permissive than some petting behavior, and all groups except the adult whites rated the behavior indicated in item 8 as at least more permissive than some form of coitus. In addition, the "strong affection" statements almost always rank as less-approved, or higher on permissiveness, than do the "engaged" or "love" statements. (See Tables 2.11 and 2.13.) To summarize-when affection decreases, the same sexual behavior is viewed as considerably more permissive. This is particularly true for high-permissive sexual groups.
The tendency of the behavior indicated in items 4 and 8 to move into a "higher" type of sexual behavior (petting or coitus) is of research value.
Table 2.13
For example, if for some reason the researcher cannot ask questions about premarital coitus to a group of respondents, then item 8 can be used to obtain some information concerning coitus without ever directly asking about it. It can be predicted (for all groups except white adults) that over ninety percent of those who agree with item 8 will also accept coitus when affection is present. This is so because item 8 ranks higher in permissiveness than do some of the coitus-with-affection items (9, 10, or 11), and the cumulative quality of a Guttman scale guarantees that in about ninety percent of the cases agreement to a higher-ranking item involves agreement with all lower-ranking items. It should be added, however, that this rule does not work both ways-that there are people who will accept coitus when affection is present but who will not accept petting without affection; and thus item 8 does not give information on all who accept coitus. Nevertheless, it is a way of gaining some information on attitudes toward coitus without ever asking the question directly. In an analogous fashion, item 4 would allow one to gain information on petting without ever directly asking questions on petting. In this way a less controversial item may be used to predict a more controversial attitude.
The scale order (rank order) of items in both the male and female scales is virtually identical in all adult and student groups when the highly "mobile" items, 4 and 8, are left out. The only exceptions are those cases where the difference is due to a few respondents, such as in item 1 in the student sample. The universal order is 1, 2, 3, 5, 6, 7, 9, 10, 11, 12. Furthermore, the movement of items 4 and 8 have empirical value in that they allow us a basis for judging the level of permissiveness of the group being measured.
The universal order of items, which holds up in all samples, may be used for comparative purposes. However, it is simpler to combine some items. For example, since over ninety-five percent accept items 1, 2, and 3, these separate items might well be dropped, and those who only accept one or more of these three items might be called scale-type zero. Then, since items 5 and 6 are so close in percentage agreement, they can be combined into one scale type, and all those who also agree with one or more of them could be called scale-type one. Scale-type two would, in addition, accept item 7, and scale-type three would, in addition, accept either item 9 or 10 or both. Scale-type four would also accept item 11, and finally, scale-type five would accept also item 12. This short universal scale would then consist of five items, or combined items: (5, 6), 7, (9, 10), 11, 12. A scale-type-zero person would disagree with all of these, and a scale-type-five person would agree with all of these. (For a fuller discussion of these scale types see Appendices D and E. )
Table 2.14 presents the race and sex breakdown on this short universal form for the adult and student samples. For the sake of simplicity, this scale has been dichotomized to designate scale-types zero, one, and two as "low" and scale-types three, four, and five as "high." In effect, this means that those who accept coitus will be called high and all others low. In actuality, all tables were run with the scale open and with other cuts, and it was found that this made little difference. Most of the relations with permissiveness show a linear quality, and thus the aforementioned dichotomy is suitable. Where a curvilinear relation occurs, it is presented accordingly.
Table 2.14
In order to obtain a "self-permissiveness" score, the scale was used with men being scored by their answers to the male scale and women being scored by their answers to the female scale. This presents the self-permissiveness of the person rather than the permissiveness for the opposite sex or than an average general permissiveness. Throughout this study the universal scale is used to measure self-permissiveness, and thus the respondent is scored on the scale of his or her own sex.
Appendix D gives full details on this universal scale as well as on the basic twelve-item scale. It is unnecessary to say more here than that without dropping or adding any items, the original twelve items formed a Guttman scale that met all Guttman-scaling criteria to quite a high degree. This includes a check on item intercorrelation, coefficient of reproducibility, coefficient of scalability, minimal marginal reproducibility, percent pure scale types, and other checks. The same was true of the six-scale-type universal scale described previously. Appendix E explains a simple way to use this six-scale-type universal scale in empirical work.
The Guttman scales have worked successfully in every sample in which they have been used. This method of measuring permissiveness is believed to have advantages over more "global" approaches, for the items used in the scale are easier to respond to, and the scale has a unidimensional quality that makes comparison simple. Also, the scale may be used to classify people according to standards as well as according to degree of permissiveness. This can be done by comparing the respondent's score on both the male and female scales, thereby adding the dimension of equalitarianism to the dimension of permissiveness.
It was noted earlier that there was a compensatory behavior that distinguished high- and low-permissive groups; that is, the high-permissive groups gave a lower relative rank to kissing and petting without affection than did the low-permissive groups, who to various degrees ranked such behavior as more acceptable than intercourse with affection.
The specific standards are discussed later at greater length, particularly in Chapters 6 and 7. It is enough to say here that the exact percent of adherents in each standard should not be taken as more than a general indication of the standard's relative popularity. The attempt was to get at informal, operational standards instead of at formal ones. The extent to which this was accomplished is also discussed later. Finally, it should be noted that the self-permissiveness score that is dichotomized into high and low is generally used as the key measure throughout the study. The basic attempt is to explain some of the sociocultural factors related to being high or low on permissiveness. To aid in this process, the empirical findings of the next seven chapters have been integrated into seven propositions, and in the final chapter an attempt has been made to integrate these propositions into one theory. The basic task of this study, then, is to utilize the scales developed to present empirical findings and to integrate those findings into propositions and a theory.
----------------------
NOTES:
1 Dorothy D. Bromley and Florence Britten, Youth and Sex. New York: Harper & Row, Publishers, 1938.
2 For a recent statement see Mervin B. Freedman, "The Sexual Behavior of American Colege Women: An Empirical Study and a Historical Survey," Merrill-Palmer Quarterly of Behavior and Development, 11 (January 1965), pp. 33-39.
3 Lemo D. Rockwood and Mary E. Ford, Youth, Marriage, and Parenthood. New York: John Wiley & Sons, 1945.
4 Ira L. Reiss, Premarital Sexual Standards in America. New York: The Free Press, 1960. Chap. 3 contains a full discussion of this area.
5 Judson T. Landis and Mary G. Landis, Building a Successful Marriage, 3d ed. Englewood Cliffs, New Jersey: Prentice-Hall, 1958, chap. 11.
6 Winston Ehrmann, Premarital Dating Behavior. New York: Holt, Rinehart and Winston, 1959, chap. 5.
7 Ehrmann, p. 304, fn. 13.
8 For another recent report see chap. 4 of Rose K. Goldsen et al, eds., What College Students Think. Princeton, New Jersey: D. Van Nostrand Co., 1960.
9 It should be clear that the concern here is not with homosexual relationships or masturbation nor with marital or postmarital sexual relationships. Rather it is exclusively with heterosexual premarital sexual standards, attitudes, and behaviors toward kissing, petting, and coitus. These other important topics require separate careful treatment. Only what our culture defines as our basic sexual standards before marriage have been selected here. Names for the four basic standards were chosen that would convey the meaning of the standard and would not be purely a deseription of behavior. This was done because attitudes and not behavior were being dealt with, and thus "permissiveness with affection" seemed a better name than "coitus with affection."
10 For an approach showing ranking regularities regarding sexual customs among various cultures see Julia S. Brown, "A Comparative Study of Deviation from Sexual Mores," American Sociological Review 7 (April 1952), pp. 135-146.
11 For an early presentation of findings see Ira L. Reiss, "The Scaling of Premarital Sexual Permissiveness," Journal of Marriage and the Family, 26 (May 1964), pp. 188-198.
12 Engagement is considered to be an affection-related state involving a higher degree of affection than is present in other love relationships that do not also involve engagement, and thus it ranks at the "top" of the affection-related conditions.
13 A respondent's sexual standard is arrived at by first examining his responses to the four "coital" statements in both the male and female scales. If he has agreed with one or more of these statements then a check is made to see if the response is the same on both the male and female scales. If it is, then the respondent accepts coitus equally for both men and women and would be either an adherent of permissiveness with affection or permissiveness without affection. If the respondent gave a different response to the "coital" statements on the male and female scales, then he would be classified in one of the double-standard categories. Those who disagreed with all coital statements on both the male and female scales would then be examined on the four "petting" statements in a fashion similar to that on the four coital statements. Finally, those who also disagreed with all petting statements on both the male and female scales would be similarly examined on the four "kissing" statements. In such a basic fashion respondents can be logically classified into the sexual standards and their subtypes.
14 Two other all-white college student samples have been used to test these scales. One was an Iowa sample, and the other was from the state of Washington. In both cases the scales met all Guttman-scaling criteria. The responses fitted generally with those of the other student populations, with the Iowa school showing more liberality than the white Virginia college but less than the New York college, and with the Washington school showing about the same type of response achieved from the white Virginia college. The Washington school was known for being a conservative college, and this result adds to the validity of the scales. The scales have been translated into Swedish under the direction of Professor Floyd Martinson of Gustavus Adolphus College, and they were administered in Sweden in the summer of 1966. Results have not yet been fully analyzed.
15 Only those respondents who answered all scale questions were used in the study. Fortunately, over ninety percent of those in all the samples did answer fully.
16 This does not mean that high permissives will have a lower percent supporting, say, item 8. Rather, it means that in a high-permissive group item 8 will have a lower percent than item 9, and that this will be much less the case in a low-permissive group.
17 Harold Christensen found that the low-permissive group (Mormon) was most likely to have "terminal petting." See his "Scandinavian and American Sex Norms," Journal of Social Issues, 22 (April 1966), pp. 60-75.