Archives of Sexual Behavior, Vol. 15, No.

2, 1986

The GRISS: A Psychometric Instrument for the

Assessment of Sexual Dysfunction
John Rust, Ph.D. 1 and Susan Golombok, P h . D . 2"3

The Golombok Rust Inventory o f Sexual Satisfaction (GRISS) is a short

28-item questionnaire f o r assessing the existenee and severity o f sexual
problems. The design, eonstruction, and item analysis o f the GRISS are
described. The two separate male and female scales are shown to have high
split-half reliabilities o f 0.94 f o r women and 0.87for men. Validation o f
change scores in the GRISS on 30 elinical couples, before and after therapy,
showed correlations w#h therapists'blind ratings o f 0.54 (p < 0.001) f o r
men and 0.43 (p < 0.01) f o r women. Diseriminatory validity between
clinical (n = 69) and nonclinical (n = 59) groups was r = 0.63for women
and r = 0.37 f o r men. The 12 subscales o f impotente, premature
ejaculation, anorgasmia, vaginismus, noncommunication, infrequeney,
male and female avoidanee, male and female nonsensuality, and male and
female dissatisfaetion are also shown to have good reliability and validity.

KEY WORDS: sexual dysfunction; sex therapy; evaluation, test; outcome studies; psychornetric.

Evaluation of treatment efficacy in sexual dysfunction has been hindered by

the use of ad hoc measures. Although objective tests are available
(Derogatis and Melisaratos, 1979; Harbison et al., 1974; LoPiccolo and Steger,
1974; Thorne, 1966) these tend to be rather long and difficult to score, or they fall
to cover the fange o f behaviors and attitudes that are of interest to sexuai
dysfunction clinics. These tests are reviewed by Conte (1983), who points
out the need for two types o f instrument: reliable and valid single overall
scales for use in research, and subscale batteries that specify particular

1University of London Institute of Education, London, WC1, England.

2Urdversity of London, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London SE5,
England.
3To whom correspondence should be addressed.

157
0004-002/86/0400B157505.00/0 1986PlenumPublishingCorporation
158 Rust and Golombok

subsets of behavior and attitude relevant to more specific ~nterpretations

and diagnosis.
The Golombok Rust Inventory of Sexual Satisfaction (GRISS) (Rust
and Golombok, 1983, 1985a, 1985b) is a short measure of sexual
dysfunction which may be administered to heterosexual couples or
individuals who have a current heterosexual relafionship. It provides overall
scores, for men and women separately, of the quality of sexual functioning
within a relationship. In addition subscale scores of impotence, premature
ejaculation, anorgasmia, vaginismus, infrequency, noncommunication,
male dissatisfaction, female dissatisfacti0n, male nonsensuality, female
nonsensuality, male avoidance, and female avoidance can be obtained and
represented as a profile.

TEST SPECIFICATION

The development of the test specification raised the question of how

to define the quality of a sexual relationship. Purely objective parameters,
such as frequency of sexual intercourse or the incidence of unsuccessful
attempts, provide one possibility. But there is a 1arge variety in these
measures in the general popultion, and the majority of this variation is not
associated with any dysfunction or dissatisfaction by the partners
concerned. A bettet approach might be to define sexual adjustment in terms
of the partners, claimed satisfaction with their relationship, but this too is an
insufficient criterion when taken on its own. A couple could claim to be
satisfied by having no sexual contact whatever. Or one partner may claim
satisfaction and the other dissatisfaction. Fortunately there is now a body
of knowledge, initiated by Kinsey, Pomeroy, and Maron (1948) and Kinsey
et al. (1953) and more recently by Masters and Johnson (1970), about what
constitutes the norm for sexual behavior, and about those aspects of a
sexual relationship that can become more satisfying. This, as weil as
experience in the clinic, enables the sex therapist to identify the areas in a
couple's sexual relationship that could be improved. It is this body of
knowledge that was drawn upon in specifying the blueprint for the GRISS.
The test specification was drawn up by a "think tank" of sex therapists at
the Sexual Dysfunction Clinic of The Maudsley Hospital, London. It
specified seven major areas of interest: frequency, satisfaction, interest,
dysfunctions, anxiety, communication, and touching.

Item Analysis

The pilot version contained 96 items (48 for the man and 48 for the
woman) covering the area of the specification. Piloting was carried out on
G o i o m b o k Rust Inventory of Sexual Satisfaction 159

51 client couples at the Maudsley Sexual Dysfunction Clinic and 36

nonclinical couples from the Institute of Education, London, where orte of
the partners was taking a part-time course as a mature student.
From the psychometric point of view a particular problem arises in
looking at the sexual adjustment of a couple because we are measuring not
from one person but from two people and their interaction. Intuitively orte
might feel that the obvious solution would be to treat the couple as a single
unit and to elicit a one-dimensional test score, such that a couple with a high
score had a problematic relationship and vice versa. However, factor
analysis showed that this model was not appropriate for the present test.
The factor structure for the GRISS is fairly stable across versions, stan-
dardizations, and samples, and is illustrated by that for the standardiza-
tion sample given in Fig. 1. Two factors, orte for men and one for women,
are required to adequately describe the data.
The first stage of item analysis eliminated items with extreme scores or
with a large amount of response refusal. The second stage involved the
identification of stable subscale scores. Promax oblique factor analysis was
used to identify subscales in the first instance. The subscale items indicated
were then factor analyzed separately for each subscale, using orthogonl
rotations. The original subscale length varied between five and nine items.
The final subscales had four items each, which were selected to the
following criteria: (1) stability of the factor structure, with a common factor
accounting for more than 5070 of the variance, (2) an equal number of
items (four), two with positive and two with negative loadings, (3) content
continuity along the full length of the indicated dimension, (4) factorial
consistency between the clinical and student samples, and (5) face validity.
More details of the item analysis appear in the test handbook (Rust and
Golombok, 1985b). Four of the subscales thus generated were about the
specific problems of anorgasmia, vaginismus, impotence, and premature
ejaculation. All of these diagnostic categories proved to be continuous with
minor degrees of disturbance in the normal population. Six other subscales
gave separate male and female scores for avoidance, dissatisfaction, and
nonsensuality. The remaining two subscales measured infrequency and non-
communication about sex within the couple.
In the third stage, overall scales were sought to describe the state of
the couple's sexual relationship. This involved an orthogonal factor analysis
of the scored subscales, together with all the remaining items from the pilot
inventory which had been carried forward into Stage 2 of the item analysis.
This factor analysis yielded an orthogonal two-factor solution, and on the
basis of this two main scales were constructed. Details of this appear in the
GRISS Handbook (Rust and Golombok, 1985b). Eight of the items
retained in the questionnaire contribute toward these two main scales but
are not included in the subscales. Item dealing with interest in sex generally fall
into this category. They did not form a stable subscale but did have high
160 Rust and Golombok

loadings on the main scale factors. Items that disappeared that the third
stage of the analysis, because of low communality, included those dealing
with fantasy. These items additionally failed to yield a consistent subscale.
Following item analysis the GRISS now contains 56 items (28 for men and
28 for women).

Standardization

The GRISS was standardized on a sample of 88 sex therapy clients

from clinics throughout the United Kingdom. A combination of norm
referencing and criterion referencing yielded transformed scales that give a
good indication of the existence and severity of any problems. Transforma-
tions are to a pseudo-stannine scale (area based, from I to 9) with a score of
5 or above indicating a problem. Distributions of these transformed scales
are approximately normal for the clinical sample but skewed toward the
lower end of the scale to facilitate measurement in nonclinical populations.
As the pilot study involved more than one level of item selection, the
structure of the main scale and the subscales was replicated for the selected

1.()
FiletOT
AVF NSF

DISM
DISF e
INF
ANORG oNCO
VAG

-[ 0 1o Factor 11

NSM AVM

-1.0

F i g . 1.Factor analysis of the transformed subscales on the stan-

dardization sample (88 clinical couples). AVF, female avoidance;
N S F , female nonsensuality; DSM, male dissatisfaction; A N O R G ,
anorgasmia; DSF, female dissatisfaction; INF, infrequency; N C O ,
noncommunication; VAG, vaginismus; PE, premature ejacula-
tion; AVM, male avoidance; NSM, male nonsensuality; IMP, im-
potence.
Golombok Rust Inventory of Sexuai Satisfaction 161

CENTROIDSCORES

."/

j~

~iN SCALE~iNTS

.~../
"
"

-9

.......... Fema[e Scale

Male Scale
Fig. 2. Discriminant function analysis for the main scales.
Discrimination is between the nine transformed scale points.
Discriminating variables are the subscale scores.

items. The characteristics o f the factor analysis were stable across both the
pilot and the standardization samples. The factor analysis for the 88 clinical
couples appear in Fig. 1. Discriminant funcfion analyses were carred out to
test the linearity of the scales (Fig. 2). The ordering o f scale and subscale
points was found to be linear, showing only occasional nonlinear
discrepancies for some o f the subscales.

Reliability

For the main scales the split-half reliabilities were extremely high, 0.94
and 0.87 for the female and the male scales, respectively. The reliabilities o f
162 Rust and Golombok

Table I. Internal Consistencies of the Subscales and

Estimated Test-Retest Reliability
Internal Test-retest
Subscale consistencya reliability~
Impotence 0.78 0.79
Premature ejaculation 0.78 0.84
Male nonsensuality 0.69 0.57
Male avoidance 0.76 0.64
Male dissatisfaction 0.69 0.61
Infrequency 0.79 0.66
Noncommunication 0.61 0.52
Female dissatisfaction 0.64 0.47
Female avoidance 0.82 0.62
Female nonsensuality 0.78 0.61
Vaginismus 0.73 0.82
Anorgasmia 0.83 0.61
aCalculated as square root variance of factor 1 for
each subscale in the 88-subject standardization
sample.
Calculated from the pre-post treatment correlation
of 42 clinical subjects. This is an underestimate since
there was significant improvement overall during
therapy.

the subscales are given a minimum value by the internal consistencies, which
were obtained from the factor analysis of the items in the standardization
sample (square root of percentage variance for Factor 1). The values
obtained are high for scales with this number of items, averaging 0.74, and
ranging between 0.61 for noncommunication and 0.83 for anorgasmia (see
Table I). Test-retest reliabilities were calculated for pre- and posttherapy
data on 41 clinical couples, 20 of whom had marital therapy (Bennum,
Rust, and Golombok, 1985) and 21 had sex therapy. Both groups showed
significant changes with therapy, so that the figures obtained are under-
estimates. The values obtained were 0.76 for the male scale and 0.65 for the
female scale. Subscale test-retest reliabilities ranged from 0.47 for female
dissatisfaction to 0.84 for premature ejaculation, and averaged 0.65 (Table
I).

Validation

For 68 men and 63 women, of whom 62 were couples, from sexual

dysfunction clinics, therapists completed validation questionnaires in which
they were asked to define the severity and nature of any sexual problems for
men and women separately. Twenty-four men were diagnosed as impotent,
19 as having premature ejaculation, 15 as having low interest in sex, and 10
Golombok Rust Inventory of Sexual Satisfaction 163

as having other problems. There was some overlap between the categofies,
with three of the men having both premature ejaculation and impotence
and eight having both impotence and low interest in sex. Those with other
problems were subdivided into five with delayed ejaculation, one with lack
of sex education, one with fear of sex, one with difficulty in showing
affection, one with relationship problems resulting from his interest in
cross-dressing, and orte with impotence that was considered to be
organically based. The remaining men had no problem but accompanied a
dysfunctional partner. Of the women, 14 were diagnosed as anorgasmic, 26
as having low interest in sex, 5 as having vaginismus, and 6 as having other
problems. Nine women with anorgasmia also had low interest in sex. In the
" o t h e r problems" category were two women with lack o f sex education,
one with anxiety about sex, one with preoccupation about her husband's
cross-dressing, and orte who was unhappy about her husband's interest in
watching her make love to other men. The remaining women had no
problem but accompanied dysfuncfional partners.
Those subjects (n = 42 women, n = 57 men) in the clinical group who
had been diagnosed as having a problem were compared with a control
group of 59 subjects (29 men and 30 women) taken from a random sample
of general practitioner patients (Golombok, Rust, and Pickard, 1985). Both
the overall fernale scale (point biserial r = 0.63, p < 0.001) and the overall
male scale (point biserial r = 0.37, p < 0.005) were found to discriminate
between the clinical and nonclinical groups. Only fur female clinical
subjects scored lower than the mean for the control group. O f these, three
had specific difficulties coping with their partner's cross-dressing, delayed
ejaculation, or anger about premature ejaculation, respectively. Fourteen
men had scores less than the mean for the control group; o f these, five had
severely dysfunctional partners, three had quite severe premature
ejaculation (which is known from the factor analysis to have a relafively
small loading on the male scale), and three had delayed ejaculation.
The specific dysfunctional groups as diagnosed by the therapists
(impotence, premature ejaculation, vaginismus, and anorgasmia) were also
compared with the control group. All clinical groups differed from the
control group on their target subscale. For impotence (r = 7.55, p < 0.001)
none of the clinical group scored lower than the mean for the control group.
For premature ejaculation (t = 5.37, p < 0.001), only one clinical subject
scored lower than the mean o f the control group. In this case the subject
had a severely dysfunctional partner. For anorgasmia (t = 3.46, p < 0.005),
three clinical subjects scored lower than the mean of the control group.
None o f these three were having sexual intercourse with their partner. For
vaginismus, the five women so diagnosed all obtained higher scores on the
vaginismus subscale than any control subject.
164 Rust and Golombok

T-test comparisons were also carried out between the two groups for
the eight subscales that did not measure specific dysfunction. Infrequency,
male and female dissatisfaction, and female avoidance were all significant
at the 0.001 level, whereas female nonsensuality was significant at the 0.005
level. Noncommunication, male nonsensuality, and male avoidance were
not significantly different between the two groups. Male avoidance attained
the 0.025 level of significance, however, in a comparison between the 15
men diagnosed as having low interest in sex and the control group.
A further measure o f validity was obtained by correlating between the
therapists' rafings of severity of problems (ranging from 0 = no problem, 1 =
slight problem, 2 = m o d e r a t e problem and 3 = severe problem) with the
overall male and female scales. These were r = 0.56, (n = 63, p < 0.001)
for women and r = 0.53, (n = 68, p < 0.001) for men, good for an
instrument of this type.
Follow-up validation of the main scales against therapists' esfimates of
i m p r o v e m e m during therapy was carried out on 30 clinical couples after
their fifth sex therapy session. The therapists, blind to the GRISS results,
rated both the man and the w o m a n separately on a 5-point scale ranging
f f o m 0 = improved a great deal, through 1 = improved moderately, 2 =
slightly improved, 3 = not improved at all, to 4 = got worse. For the men,
the correlation between the therapists' ratings of improvement and the change
in the main male score was 0.54 (p < 0.005). For the women the equivalent
correlation was 0.43 (p < 0.05).

DISCUSSION

The GRISS is a reliable measure of sexual dysfunction. It

discriminates well between those with and without sexual problems and is a
good outcome measure o f change during therapy. It relates closely to
therapists' ratings of diagnosis and severity o f sexual problems and its
subscales are successful at identifying impotence, premature ejaculation,
anorgasmia, and vaginismus, as well as infrequency, male and female dis-
satisfaction, female avoidance, and female nonsensuality. The lack of a
significant difference in noncommunication between the clinical and
nonclinical groups m a y be an artifact of the clinical situation, in that
attending a clinic m a y select out a group of " c o m m u n i c a t o r s " f r o m those in
the population with sex problems. The lack of a difference between the
groups on male nonsensuality and male avoidance m a y reflect a difference
between the sexes in the nature of sexual dysfunction. Men m o r e often
present with specific problems, whereas women tend to present with
generalized lack of interest and enjoyment (Bancroft, Tyrer, and Warner,
Golombok Rust Inventory of Sexual Satisfact[on 165

1982). This m a y also a c c o u n t f o r t h e r e l a t i v e l y lower d i s c r i m i n a t i o n o f t h e

male m a i n scale.
T h e m a j o r a p p l i c a t i o n o f t h e G R I S S is to assess i m p r o v e m e n t as a
result o f sexual o r m a r i t a l t h e r a p y a n d to c o m p a r e the e f f i c a c y o f d i f f e r e n t
t r e a t m e n t m e t h o d s . It c a n also b e used to investigate t h e r e l a t i o n s h i p
b e t w e e n sexual d y s f u n c t i o n a n d e x t r a n e o u s v a r i a b l e s . F o r m o s t r e s e a r c h
p u r p o s e s , scores on t h e m a i n scales are the m o s t i m p o r t a n t . Because these
have higher r e l i a b i l i t y a n d v a l i d i t y , t h e y c o n s e q u e n t l y are m o r e sensitive to
differences in sexual f u n c t i o n i n g . H o w e v e r , even for r e s e a r c h , the subscales
are o f use, once o v e r a l l effects are f o u n d , in i d e n t i f y i n g m o d e s o f
o p e r a t i o n . T h e subscales a r e also h e l p f u l in d i a g n o s i s . F o r ease o f
i n t e r p r e t a t i o n t h e y h a v e b e e n scaled to give a p r o f i l e o f the p a t t e r n o f sexual
f u n c t i o n i n g within t h e c o u p l e , which can b e o f g r e a t b e n e f i t in designing a
treatment program.

REFERENCES

Bancroft, J., Tyrer, J., and Warner, P. (1982). The classification of sexual problems in women.
Brit. J. Sex. Med. 9: 30-37.
Bennun, I., Rust, J., and Golombok, S. (1985). The effects of material therapy on sexual dysfunc-
tion. Scand. J. Behav. Ther, 14: 65-72.
Conte, H. R. (1983). Development and use of Self-Report Techniques for assessing sexual func-
tioning: A review and critique. Arch. Sex. Behav. 12: 555-576.
Derogatis, L. R., and Melisaratos, N. (1979). The DSFI: A multidimensional measure of sexual
functioning. J. Sex. Mar. Therap. 5: 244-281.
Golombok, S., Rust, J., and Pickard, C. (1985). The prevalence of sexual dysfunction in the
general population. Brit. J. Sex. Mod. 11:171-175.
Harbison, J. J. M., Graham, P., Quinn, J. T., MacAllister, H., and Woodward, R. (1974).
A questionnaire measure of sexual interest. Arch. Sex. Behav. 3: 357-366.
Kinsey, A. C., Pomeroy, W. B., and Martin, C. E. (1948). Sexual Behavior in the Human Male.
Saunders, Philadelphia.
Kinsey, A. C., Pomeroy, W. B., Martin, C. E., and Gebhard, P. (1953). Sexual Behavior in
the Human Female. Saunders, Philadelphia.
LoPiccolo, J., and Steger, J. (1974). The sexual interaction inventory: A new instrument for
assessment of sexual dysfunction Arch. Sex. Behav. 3: 585-595.
Masters, W. H., and Johnson, V. E. (1979). Human Sexual Inadequacy. Churchill, London.
Rust, J., and Golombok, S. (1983). The Handbook o f the Golombok RUst Inventory o f Sexual
Satisfaetion. City Press of Glasgow, Scotland.
Rust, J., and Golombok, S. (1985a). The Golombok Rust Inventory of Sexual Satisfaction
(GRISS). Brit. J. Clin. Psychol. 24: 63-64.
Rust, J., and Golombok, S. (1985b). The Handbook o f the Golombok Rust Inventory o f Sex-
ual Satisfaction (2nd ed.). N.F.E.R., Winsor, England.
Thorne, F. C. (1966). The Sex Inventory. J. Clin. Psychol. 22: 367-374.