Creation and Validation of the Self-esteem/Self-image Female Sexuality (SESIFS) Questionnaire.

INTRODUCTION: Self-esteem and self-image are psychological aspects that affect sexual function. AIMS: To validate a new measurement tool that correlates the concepts of self-esteem, self-image, and sexuality.
METHODS: A 20-question test (the self-esteem/self-image female sexuality [SESIFS] questionnaire) was created and tested on 208 women. Participants answered: Rosenberg's self-esteem scale, the female sexual quotient (FSQ), and the SESIFS questionnaire. Pearson's correlation coefficient was used to test concurrent validity of the SESIFS against Rosenberg's self-esteem scale and the FSQ. Reliability was tested using the Cronbach's alpha coefficient. RESULT: The new questionnaire had a good overall reliability (Cronbach's alpha r = 0.862, p < 0.001), but the sexual domain scored lower than expected (r = 0.65). The validity was good: overall score r = 0.38, p < 0.001, self-esteem domain r = 0.32, p < 0.001, self-image domain r = 0.31, p < 0.001, sexual domain r = 0.29, p < 0.001.
CONCLUSIONS: The SESIFS questionnaire has limitations in measuring the correlation among self-esteem, self-image, and sexuality domains. A new, revised version is being tested and will be presented in an upcoming publication.

Introduction

Self-esteem and self-image are psychological aspects that affect sexual function.1–3 Self-esteem is defined as how a person sees herself and it expresses attitudes of self-approval or self-rejection and self-judgment regarding competency and value.4,5 It is a personal judgment, revealed through attitudes that a person has regarding herself and personal beliefs about her abilities, capacities, and social relationships.6,7

Although self-esteem is a subjective experience, it can be assessed through observable behaviors or verbal reports.8 Self-esteem implies a commitment to becoming aware and responsible for one’s choice and to feel free to exercise one’s power of decision. Self-esteem is based on personal characteristics that, under specific circumstances, lead a person to exercise her capabilities of consciousness, seeking a meaning for herself and for her environment.3

There are theoretical controversies regarding the concept of self-esteem. Some investigators regard self-esteem as a stable personality trait based on the assumption that it is built slowly over many years through successful personal experiences being continually valued by significant people in an individual’s life. Others support the theory that self-esteem is a variable characteristic because it can be affected or manipulated by events that occur in a person’s life. A third theory proposes that self-esteem can be considered both as a condition and as a personality trait.3

When facing a challenge, success depends in part on a person’s emotional state, which is directly related to her self-esteem.8 Therefore, when a woman is facing the challenges of sexuality, the higher her self-esteem, the better are her chances of having a fulfilling sexual life.1–3 By being confident about herself, a woman can create a solid subjective reality related to sexuality that is free from cultural and religious influences, which often regulate sexual practices.4

Self-image also includes aspects of self-esteem but gives more emphasis to physical and bodily aspects. A person’s self-image is influenced by her status within socially and culturally established norms.9 Self-image is how a person perceives and feels about her own body and how she believes that others see her.10 The subjective perception that a person has of her own body can be more important than the objective reality of her appearance. Any socially imposed stereotyped beauty pattern can produce intense emotional suffering.11 According to some investigators, individuality is undervalued by modern societies, which tend to overemphasize the importance of image and endorse a narcissistic culture where the perfect body is regarded as ideal.9,12 It is more important to have perfect body measures than to be an individual that feels, thinks, or creates. Conscious or unconscious feelings about our bodies are important psychological forces.2,3 Persons who do not accept their own self-image will often have less satisfying sexual relationships because sexuality is directly related to accepting our own bodies.1

Problems in self-esteem and self-image are often linked with sexual performance. Difficulties in valuing oneself, whether physically or psychologically, can affect a woman’s sexual experiences because she can feel ashamed of her own body or believe that she does not have the right to seek sexual pleasure. Similarly, a woman with sexual problems can develop low self-esteem. The longer a woman delays seeking appropriate care for sexual dysfunctions, the higher the likelihood that this will affect her self-esteem.

We were motivated to create this new questionnaire because we observed the frequent association of self-image and self-esteem issues with sexual complaints in women managed at our clinic. The objective of this new tool was not only to get an overall assessment of a woman’s self-esteem and self-image, but also to measure these aspects in relation to her sexuality. The use of a single questionnaire that assesses the three topics simultaneously could help us to better understand the interaction among them. Moreover, it would be easier for health professionals to use a single questionnaire to assess these three aspects at once. The female sexual clinic of the Department of Gynecology of São Paulo Federal University, where this study was conducted, has several projects to increase feminine self-esteem. Since self-esteem is essential for a healthy expression of sexuality, it is important to measure the correlation between self-esteem and sexuality.

This tool was created to assess new patients who present at our clinic with sexual desire, orgasm, or pain disorders. We think that a satisfying sexual life is closely linked to a positive self-image, as well as to feelings of competence and value associated with self-esteem. How a woman relates to her own body can affect her self-esteem and her sexuality, and these fundamental aspects are taken into account by our team of health professionals.

Motivated by the lack of an instrument that simultaneously assessed self-esteem and self-image and that correlated these two aspects with sexuality, we decided to create and validate a new questionnaire, the self-esteem/self-image female sexuality (SESIFS) questionnaire. In this study, we present and analyze the psychometric properties of this new questionnaire, a mandatory step in the validation of psychological instruments.13

Aims

To create and validate a new psychological instrument, the SESIFS questionnaire.

Methods

The SESIFS consists of 20 multiple-choice questions with four alternatives each, divided into three domains: self-esteem (questions 2, 5, 10–15, and 19), self-image (questions 1, 3, 4, 6, 16, and 20), and sexuality (questions 7–9, 17, and 18). This last domain assesses the influence of self-esteem and self-image on sexual satisfaction with the partner.

Process of creation of the SESIFS

We used several strategies to arrive at the pool of items used in this questionnaire. We started with a list of items that were frequently mentioned by our patients as being things that affected their sexual performance. Issues related to self-esteem and self-image, such as feeling unsatisfied or ashamed of their bodies or difficulties in communicating with their partners, were mentioned by many women. We then categorized issues related to “body dissatisfaction” into two groups: (1) related to the partner and (2) related to society in general. The issue of shame was also categorized into: (1) shame of sexuality and (2) shame of herself because of low self-esteem. This led to the creation of the three domains of the SESIFS questionnaire: self-esteem, self-image, and sexuality. We then selected items from the most popular scales used to measure self-esteem, self-image, and sexuality. These items were adapted so that they would measure not only self-esteem and self-image in general, but also self-esteem and self-image related to sexuality.

Pretest

After its creation, the questionnaire was pretested (semantic analysis) by 10 specialists in the field (judge test) to assess possible errors in the scale. These persons were asked to perform a series of tasks such as analyze the instructions and the questions, describe what the test was apparently measuring, point out items that could be confusing, and possible disagreements on answers.14 The panel of judges included five psychologists specialized in female sexual dysfunctions (one with experience in psychometric tests), three gynecologists, and two physiotherapists specialized in sexology. After this step, the reliability and validity of the final version of the questionnaire was tested on 208 women.

To assess the validity of the SESIFS, we used the technique of simultaneous criterion validation, which consists of correlating the new questionnaire with existing validated tools.15 The SESIFS was correlated with Rosenberg’s self-esteem scale and the female sexual quotient (FSQ) using Pearson’s correlation coefficient.

Sample

These volunteers were being managed during 2012 at the female sexual clinic of Department of Gynecology of São Paulo Federal University. Inclusion criteria were: age 18–60 years, being sexually active, and able to read. All participants had sexual complaints such as decrease in desire, problems in reaching orgasm, and pain during penetration.

Women who were pregnant or in the post-partum period, women with neurological diseases, or who could not, for any reason, understand instructions were ineligible. The volunteers received three self-responsive questionnaires: Rosenberg’s scale, the FSQ questionnaire, and the newly created SESIFS. It took less than one hour, on average, to answer all three questionnaires.

Instruments

Rosenberg’s self-esteem scale16,17 is an international and widely used instrument that was validated to Brazilian Portuguese by Hutz.18,19 In 2006, it was adapted and transculturally validated by Avanci et al.4 It assesses global self-esteem through a set of feelings of self-esteem and self-acceptance. This uni-dimensional scale consists of 10 questions that use a four-point Likert response scale, with answers ranging from 1 (complete disagreement) to 4 (complete agreement). Higher total scores indicate higher self-esteem. The construct validity of the Rosenberg scale analyzed the main components using Varimax rotation and resulted in a unifactorial solution. This scale has a high internal consistency (Cronbach’s alpha 0.90).

The FSQ was created in 2006 by ProSex, a project coordinated by the Psychiatry Department of São Paulo Medical School, in São Paulo, Brazil.20,21 This instrument evaluates all phases of the female sexual cycle, as well as other aspects (desire, arousal, foreplay, excitement, harmonization with the partner, comfort, orgasm, and satisfaction). It consists of 10 questions with answers ranging from 0 (never) to 5 (always). Higher scores indicate a better performance/sexual satisfaction. Overall scores are classified as Excellent (82–100 points), Good (62–81 points), Regular (42–61 points), Poor (22–41 points), or Very Poor (0–21 points). This instrument was validated using the simultaneous validity strategy, comparing 30 women with sexual dysfunction to 30 women without sexual dysfunction, according to clinical criteria, selected randomly from 200 patients who participated in the ProSex project. Statistically significant differences were detected in the mean scores of the two groups (p < 0.001, Mann–Whitney test). This instrument has a high internal consistency (Cronbach’s alpha 0.98).

This study was approved by the ethics committee of São Paulo Federal University, and all participants gave written informed consent.

The main outcome measures were the reliability and validity of the new questionnaire. Cronbach’s alpha coefficient was used to measure the internal reliability (or consistency) of the new questionnaire. Cronbach’s alpha coefficient,15,22 the most widely used objective measure of reliability, produces final scores that range from 0 to 1. Higher scores indicate higher reliability, and 0.70 is considered the usual lower cutoff for good reliability.

Results

The age of the 208 participants ranged from 18 to 60 years (mean 36.6, standard deviation 11.5 years). Most of the participants had 12 years or more of schooling: 37.0% had high school and 45.2% had college degrees.

The total overall scores of each of the three questionnaires were calculated, as well as the individual scores for each of the three domains (self-image, self-esteem, and sexuality) of the SESIFS questionnaire. The scores of the SESIFS were correlated with Rosenberg’s self-esteem scale and with the FSQ. Cronbach’s alpha coefficient and Pearson’s correlation test were used to assess the reliability and validity of the new questionnaire, respectively. These statistical tests were used because the three variables of the SESIFS are categorical. An explanatory factorial analysis was also performed to evaluate the internal structure of the questionnaire and confirm that the three variables (self-esteem, self-image, and sexuality) were latent domains of the SESIFS.

The self-image and self-esteem domains of the SESIFS questionnaire had good reliability, with Cronbach’s alpha coefficients >0.70; the sexual domain scored slightly below this cut-off (Table 1). Using a 1% significance threshold, the null hypothesis could not be rejected for the self-image (r = 0.83, p < 0.001) and self-esteem (r = 0.75, p < 0.001) domains, but it could be rejected for the sexual domain (r = 0.65). Therefore, we concluded that the self-image and self-esteem domains are reliable, while the sexual domain does not fulfill this psychometric requirement. However, the overall reliability of the SESIFS questionnaire was good (r = 0.862, p < 0.001).

Table 1

Reliability of the SESIFS questionnaire.

DOMAINS	INTERNAL CONSISTENCY α-CRONBACH**
	VALUE	95% CI	α-CRONBACH TEST		STANDARDIZED
			p VALUE*	CONCLUSION AT 5%
Self-image	0.830	(0.793; 0.863)	<0.001	Reject H0	0.832
Self-esteem	0.752	(0.696; 0.800)	0.018	Reject H0	0.754
Sexual	0.655	(0.574; 0.725)	0.443	Accept H0(1)	0.653
Total	0.862	(0.833; 0.889)	<0.001	Reject H0	0.865

Notes:

Cronbach-α test.

Cronbach-α Coef ranges from 0 to 1. Bold p values—at 5% level, reject H0, the hypothesis that Cronbach-α >0.700.

Test hypotheses: H0: Cronbach-α is <0.7. H1: Cronbach-α is >0.7. (1) rejects H0, Cronbach-α is <0.58 (p = 0.034).

Using Rosenberg’s self-esteem scale as reference, the validity of the SESIFS questionnaire was statistically significant both in the overall assessment of the instrument (r = 0.38, p < 0.001) and in the three individual domains: self-image (r = 0.31, p < 0.001), self-esteem (r = 0.32, p < 0.001), and sexual function (r = 0.29, p < 0.001). Similarly, when compared to the FSQ, the overall validity of the SESIFS questionnaire (r = −0.47, p < 0.001) and the scores of each of the three domains were also statistically significant (self-image r = −0.21, p < 0.001; self-esteem r = −0.42, p < 0.001; sexual function r =−0.60, p < 0.001).

The three variables of the SESIFS questionnaire accounted for 47.2% of the total variability of the data, thus confirming the statistical representativeness of these three domains. Therefore, the three latent domains of the SESIFS (self-esteem, self-image, and sexual function) were confirmed as variables that can be studied through this questionnaire.

The overall reliability of the new questionnaire was good, with a high Cronbach’s alpha coefficient. The sexual domain had the lowest reliability scores, out of the three domains of the SESIFS questionnaire. However, because it scored only slightly below the 0.70 cut-off, this domain should not be completely disregarded. The slightly lower internal consistency of the sexual domain of this new questionnaire may be because of the fact that two questions in this domain (8 and 18) offer limited answers to questions that are very broad and subjective. In fact, while all the other answers in the SESIFS related to frequency of events, these two specific questions offer answers that relate to feelings or sensations, which are more subjective and can therefore lead to confusion.

There are several possible explanations for low reliability of a domain or measurement tool, including the use of imprecise words.8 Some answers in this new questionnaire tried to define subjective feelings related to sensations experienced during sexual contact. Several different feelings could have been listed to define these sensations, and it seems that for some women, none of the alternatives presented reflected the exact feelings that they had experienced. Another possible reason for the lower reliability of the sexual domain was that the alternatives to two questions (8 and 18) do not clearly express a scale of values. Our idea was to offer first the most positive alternatives (pleasure) and end with the most negative alternatives (repulsion or disgust). However, perhaps the formulation of the answers does not make this scale clear, since they refer to feelings or attitudes that could appear to be disconnected from each other, giving the impression of free and independent alternatives. Because of this, it is possible that some women felt they could not select any of the alternatives. Although a pretest of the questionnaire was performed, these deficiencies were not detected or reported by the participating specialists. Therefore, this inadequacy was only detected after the 208 questionnaires were analyzed.

To assess the validity of the new questionnaire, it was tested against two other validated instruments (Rosenberg’s scale and the FSQ). Rosenberg’s scale was selected because it has been widely used worldwide to assess self-esteem.23,24 The FSQ is a simple, validated instrument that is widely used in Brazilian studies on sexual function.25,26 When compared to Rosenberg’s scale and the FSQ, the overall validity of the SESIFS questionnaire, as well as the validity of each of its two domains (self-esteem and self-image) was good. However, the sexuality domain should be reviewed and validated again in a future study.

Conclusions

The new SESIFS questionnaire was reliable and accurate in the assessment of self-esteem and self-image. However, the sexuality domain of this questionnaire had a slightly lower than expected Cronbach’s alpha coefficient, which indicates low reliability for this domain. Questions 8 and 18 will be reviewed and better adapted to the objectives of the questionnaire. The lower reliability of one of the domains compromised the main objective of this tool, which was to assess the interaction among self-esteem, self-image, and sexuality. This precludes the use of this questionnaire for clinical practice or research purposes, at present. The current version of the SESIFS is under development and has a limited capacity to measure the interaction among self-esteem, self-image, and sexuality in adult Brazilian women. A new, revised version of this questionnaire is being tested and will be presented in an upcoming publication.

Annex 1 SESIFS Questionnaire (Author’s Translation from Portuguese)

“This questionnaire refers to the self-esteem, the self-image and the sexual function of women. Please read each question carefully and choose the alternative that best describes your behavior or how you feel.”

When you look at yourself in the mirror, you feel:

() very satisfied

() not very comfortable

() uncomfortable

() very uncomfortable

When someone praises you, you feel:

() very satisfied

() not very comfortable

() uncomfortable

() very uncomfortable

You enjoy taking care of yourself (putting on make-up, selecting clothes, taking care of your hair and nails, etc…):

() always

() only when there is a party

() almost never

() never

When you see yourself in a picture, you feel:

() very satisfied

() not very comfortable

() uncomfortable

() very uncomfortable

How much do you like yourself:

() a lot

() enough, reasonably

() not much

() very little

You like your body:

() a lot

() enough, reasonably

() not much

() very little

Do you feel uncomfortable with your body during a sexual relationship?

() never

() sometimes

() almost always

() always

When your partner caresses you, what do you feel?

() pleased

() apprehensive

() uncomfortable

() disgusted or repulsed

Do you avoid having sex with your partner because you do not feel comfortable with your body?

() all the time

() frequently

() sometimes

() never

Do you avoid going to social activities (the beach, clubs or parties) because you feel uncomfortable with your body or with your appearance?

() all the time

() frequently

() sometimes

() never

Do you feel comfortable to express your opinions?

() always

() most of the time

() sometimes

() never

Do you take the initiative to start new activities or new projects in life?

() always

() frequently

() sometimes

() never

Do you think that you deserve good things in life?

() always

() most of the time

() sometimes

() never

Do you feel inferior to persons around you (at work, in the family or friends)?

() never

() sometimes

() almost always

() always

Do you feel comfortable when you go out to shop for clothes?

() always

() most of the time

() sometimes

() never

Do you avoid going to a shop to buy clothes because you think that nothing will look good on you?

() never

() sometimes

() almost always

() always

During sexual intercourse, do you pretend you are having pleasure when you are not, to please your partner?

() never

() sometimes

() almost always

() always

After you have sexual intercourse, you feel:

() pleased

() apprehensive

() uncomfortable

() repulsed or disgusted

Do you feel that you deserve the love of your partner?

() always

() most of the time

() sometimes

() never

Do you feel fatter or thinner than people around you?

() never

() sometimes

() most of the time

() always

Table 2

Validity of the SESIFS questionnaire.

VARIABLE	EXTERNAL VALIDITY
	SESIFS SELF-IMAGE	SESIFS SELF-ESTEEM	SESIFS SEXUAL	SESIFS TOTAL
Rosenberg self-esteem scale (1)
Pearson’s correlation	0.310	0.326	0.294	0.383
p	<0.001	<0.001	<0.001	<0.001
Conclusion at 5%	Reject H0	Reject H0	Reject H0	Reject H0
FSQ
Pearson’s correlation	−0.215	−0.423	−0.596	−0.473
p	<0.001	<0.001	<0.001	<0.001
Conclusion at 5%	Reject H0	Reject H0	Reject H0	Reject H0

Table 3

Explanatory factorial analysis of the SESIFS questionnaire.

QUESTIONS	FACTOR 1	FACTOR 2	FACTOR 3
SESIFS 20	0.740	−0.104	0.036
SESIFS 16	0.705	0.164	0.167
SESIFS 6	0.697	0.251	0.275
SESIFS 1	0.693	0.230	0.110
SESIFS 10 inv	0.639	−0.023	0.221
SESIFS 7	0.582	0.169	0.329
SESIFS 4	0.576	0.388	0.005
SESIFS 15	0.560	0.430	0.066
SESIFS 11	0.034	0.733	−0.036
SESIFS 12	0.180	0.660	0.011
SESIFS 13	0.029	0.625	0.354
SESIFS 14	0.162	0.554	0.276
SESIFS 18	0.100	0.081	0.757
SESIFS 8	0.047	0.269	0.682
SESIFS 9 inv	0.280	−0.021	0.587
SESIFS 19	−0.030	0.425	0.487
SESIFS 17	0.211	0.005	0.426
SESIFS 2	0.278	0.456	0.169
SESIFS 5	0.451	0.452	0.261
SESIFS 3	0.388	0.369	−0.163

Notes: Extraction method: Analysis of the main component. Rotational method: Varimax with Kaiser normalization.