Revista da Associação Médica Brasileira (English Edition) Revista da Associação Médica Brasileira (English Edition)
Rev Assoc Med Bras. 2013;59:576-82 - Vol. 59 Num.06

Content validation of the 'Mosaic of Opinions on Abortion' (Mosai)

Denis Barbosa Cacique a, Renato Passini Junior b, Maria José Martins Duarte Osis c

a Center for Statistics and Information Hospital, Hospital da Mulher Professor Doutor José Aristodemo Pinotti, Center for Integral Attention to Women''s Health, Universidade Estadual de Campinas, Campinas, SP, Brazil
b Departament of Gynecology-Obstetrics, Faculty of Medical Sciences, Universidade Estadual de Campinas, Campinas, SP, Brazil
c Center for Research in Reproductive Health of Campinas, Campinas, SP, Brazil


Abortion. Validation studies. Healthcare providers. Health knowledge, attitudes and practice. Ethical analysis


Objective: This study is aimed to develop and validate the contents of the Mosaic of Opinions on Induced Abortion (Mosai), a structured questionnaire intended to be used as a tool to collect information about the views of health professionals about the morality of abortion.

Methods: The contents of the first version of the questionnaire was developed based on the technique of thematic content analysis of books, articles, films, websites and newspapers reporting cases of abortion and arguing about their practice. The Mosai was composed of 6  moral dilemmas (vignettes) related to induced abortion, of which outcomes should be chosen by the respondents and could be justified by the classification of 15  patterns of arguments about the morality of abortion. In order to validate its contents, the questionnaire was submitted to the scrutiny of a panel of 12 experts, an intentional sample consisted of doctors, lawyers, ethicists, sociologists, nurses and statisticians, who evaluated the criteria of clarity of writing, relevance, appropriateness to sample and suitability to the fields. These scores were analyzed by the method of concordance rate, while the free comments were analyzed using the analysis technique content.

Results: All the moral dilemmas and arguments were considered valid according to the rate of agreement, however, some comments led to the exclusion of a dilemma about emergency contraception, among other changes.

Conclusion: The content of Mosai was considered valid to serve as a tool to collect the opinions of healthcare professionals regarding the morality of abortion.



Considered natural advocates of women's health,1 physicians play a central role in the daily lives of induced abortion.2 They induce abortion in the cases provided by law (pregnancy resulting from rape or when there is a risk to the woman's life,3 as well as the situations of an anencephalic fetus4) and will be the most affected professional category by a possible easing of legal restrictions against induced abortion (IA). Even if such change does not occur, doctors are already inserted in a society where IA is widely practiced, often illegally and using unsafe methods,5 which end up involving them in one way or another, at least when they treat women with complications of unsafe IA.6

Not surprisingly, in recent years there have been a significant number of publications of opinion polls on abortion, especially studies carried out specifically with physicians and other health professionals, as they are the authorities on health issues and their attitudes can directly affect the access to healthcare.7 Among these studies, those of qualitative approach are the ones that have best produced in-depth knowledge on the theme.8 However, the comprehensiveness of such studies is known to be limited as, considering the in-depth analysis is performed, few individual cases are studied.9

The situation seems to be reversed in the case of quantitative researches, which, even when they include hundreds of professionals, reveal little of the reasons implicit in their attitudes.8 For instance, a survey verified that 40% of the consulted doctors would help a patient who asked for help to abort in case of an unwanted pregnancy, but less than 2% of them would do the procedure.10

These data point to a complex relationship between doctors and abortion, without, however, presenting justifications for such apparently contradictory attitudes. That is, why is it acceptable to advise a patient to terminate a pregnancy safely, but not do it personally? Is there an ethical and legal difference between the two actions? What do health professionals have to say about them?

In Brazil, there are no publications of empirical research with in-depth answers and at the same time, comprehensive, for this type of question.8 The approach required to fill such gaps should employ a survey instrument that can be used in a comprehensive way, being self-applicable and feasible to be applied electronically or sent by post. Such a tool should allow the estimate of proportions, rates and percentages, without, however, neglecting the investigation of experiences and meanings implied by such elements as there is little point in saying that abortion - or any other controversial topic - is morally correct or incorrect if, in doing so, we do not present the criteria on which bases these perspectives are advocated.11

The aim of this article is to describe the process of development and validation of content Mosaic of Opinions on Induced Abortion (Mosai), an instrument intended to be used to meet comprehensively and in-depth the views of health professionals on induced abortion.


The present study was a methodological research12-15 aimed at developing the Mosai questionnaire, as well as the validation of its content by a panel of experts. The Mosai consists of six vignettes about individuals considering the possibility to abort. Vignettes are compact narratives of real or fictitious situation, written to elicit information on the perceptions, opinions and knowledge of the respondents about a phenomenon to be studied.16

In Mosai, all vignettes consist of moral dilemmas, i.e., circumstances in which moral obligations require a person to perform two or more mutually exclusive actions.17 These dilemmas were adapted from real cases reported in scientific publications or news about individuals who interrupted a pregnancy or considered doing so. Originally, the questionnaire had seven vignettes (not six), that focused on the key demands for abortion according to the official medical classification: eugenic, therapeutic, selective and voluntary.18  The original titles, type of demand according to official medical nomenclature, themes and sources used in each of these vignettes are summarized in Table 1.

Table 1 - Themes addressed by vignette.

After the outcome of each vignette, we inserted a series of statements that must be classified by a Likert scale with five levels, ranging from "strongly agree" to "strongly disagree." These phrases represent some of the main arguments used for the debate on the theme.

They were identified by the technique of thematic content analysis of articles, books, websites, magazines, newspapers and films on the subject.26 To introduce these arguments, we used the "fragment strategy" technique, writing each sentence (hereafter called the "fragment") as if it were the condensed form of an argument, rather than its complete logical structure.27-30 In all, we identified 14 patterns of arguments: seven supporting the right to abort and seven against it (Table 2). In addition to these 14 patterns of arguments, Mosai also uses a number of fragments with irrelevant content for the theme and these will only be scored for those individuals who answer the questions at random, thus allowing the elimination of poorly answered questionnaires.

Table 2 - Arguments for and against abortion and their central ideas.

After the development of the first version of the questionnaire, we performed its content validity, carried out by the analysis of experts on the subject, as recommended in the literature.13,31 Thus, we invited by e-mail or by telephone 24 researchers that were active in at least one of the following subject areas: ethics, bioethics, Law and abortion. It was an intentional sample,9,32 defined according to the logic of information-rich cases, consisting of specialists recognized for their expertise in the aforementioned areas and, therefore, able to contribute critically to the content validation of Mosai.

Of the 24 researchers invited, 12 expressed consent through the Term of Free and Informed Consent Form (FICF) and actively participated in the survey. For this, they were sent by mail an envelope containing: the original version of Mosai, a pre-stamped envelope, the FICF and a guide for validation. The vignettes had to be evaluated in two ways: through scores 0-5 for the criteria of "adequacy to sample", "relevance to theme" and "clarity of writing," and with freely written comments, criticisms and suggestions. The fragments should be evaluated with scores 0-5 for the criteria of "relevance to the domain", "adequacy to sample", "relevance to theme" and "clarity of writing," and with freely written comments, criticisms and suggestions.

These scores were evaluated according to the agreement rate, a score that describes the mean score received by each content.33 Generally, the agreement rate uses a Likert scale with 4 possible answers (1-4), in a way that only scores "3" and "4" indicate that the content is valid. In our research, we used a scale with five levels (1 for "very poor", 2 for "poor", 3 for "regular", 4 for "good" and 5 for "excellent") and considered satisfactory contents with mean scores between "4" and "5".

In percentages, contents were considered valid with means between 60% and 100%.

The free comments written by the specialists were evaluated according to the content analysis technique.32 This resulted in the categorization of three types of comments on the vignettes: "clarity of writing", "adequacy of the text to the sample" and "contextual corrections". In the case of fragments, categorization resulted in five patterns of comments: "clarity of writing", "adequacy of text to the sample", "contextual corrections", "adjustments to the domain" and "ideological adjustment".


The sample consisting of 12  subjects was predominantly female, with only two men. All participants were residents of the Southeast region (seven from Campinas, four from São Paulo and one from Rio de Janeiro). Four individuals classified their occupation as "teacher", two as "researcher" and the others as "lawyer", "nurse", "statistician", "administrative professional" and "psychologist"; one participant did not provide this information. The abortion issue was not part of the professional activity of three study subjects. As for the remaining, four had worked professionally with the subject for at least 20 years, three between 10 and 19 years, and two between 5 and 9 years.

The development of research instruments had been part of the professional activity of three subjects for a period of time between 21 and 30 years; for five subjects, between 11 and 20 years, and for two subjects, between 5 and 10 years. Only one participant did not develop survey instruments. The mean age of the sample was 52 years.

No vignette received a mean score less than or equal to 85 for any of the evaluated criteria and thus, all were considered valid (Table 3).

Table 3 - Mean scores received by vignettes (score 0-100).

Regarding the specialists' free comments, in the "clarity of writing" category, corrections, suggestions and questions related to the formal qualities of the text (such as spelling, syntax and clarity) were indicated. As for the "adequacy of text to the sample", certain jargon recurrence was criticized, especially of the medical and bioethical context.

Finally, in the third category, "contextual corrections", the most commented vignette was "Angela", on emergency contraception (EC) use.

By including a vignette about EC in a questionnaire on abortion, we wanted to investigate how future study subjects would evaluate the morality of using this method, as some studies have shown technically wrong knowledge by professionals regarding their use and mechanisms of action.34 However, most of the specialists expressed concern about the relevance (and impact) of this objective. For one specialist, "the theme of EC is always wrongly associated to pregnancy termination, and that is why I do not recommend including this case, as it enables its correlation [with the abortive effect] instead of clarifying that its effects are similar to some contraception methods."

In view of this and other considerations, we decided to exclude the vignette "Angela" from the questionnaire. We also excluded the vignette "Mara", of which theme, sexual violence, was already addressed in the dilemma "Marina". At the end of this process of incorporating the specialists' criticisms and suggestions, the Mosai was reduced to five vignettes ("Jussara", "Tourmaline", "Graziane", "Marina" and "Rita"). Subsequently, we created and added a sixth vignette, "Paula", which deals with an unwanted pregnancy resulting from contraceptive failure.

As for the fragments, none of them received a mean score less than or equal to 80 for any of the evaluated criteria, so that all were considered valid. However, the fragments were also assessed according to the free comments written by the specialists, who warned against the use of misstatements on the theme. For instance, one of the fragments stated that "Terminating the pregnancy in a case like this might be the best way to protect the unborn child from a future full of suffering" (vignette "Jussara" about a case of anencephalic fetus). One of the specialists questioned whether the suffering actually occurred. Regarding the fragments of the "emotional appeal" argument, there were recurring statements such as "I think the sentence is too emotional". A specialist disagreed with the "disrespectful manner given to the embryo or fetus", which did not seem like an "emotional appeal", but the adjectives disqualify the contents of the argument without appealing to a relevant discussion."

It so happens that avoiding discussion and rational argumentation is the essential characteristic of the emotional appeal. Therefore, we chose to use fragments with aggressive contents, not because we agreed with them or wanted to endorse them, but because they are effectively employed among abortion debaters. Regarding the fragments of the "political argument against abortion," two specialists suggested mentioning the Pact of San José of Costa Rica, of November 22, 1969, an international treaty that protects the right to life from the moment of conception, as well as the recognition of the legal personality from the same moment on, so that, according to one specialist, "by legalizing abortion, the Brazilian government would be violating the Pact of San José."

Mosai also underwent a lot of changes due to discussions among the authors of the research. That was the case, for instance, of changing the name of the protagonist of the second vignette, i.e., "Tourmaline", to "Maria". A more structural change was the reformulation of the questions subsequent to the vignettes. Originally, after each vignette, the questionnaire asked the respondent if he was in favor of abortion in that case. It so happens that, as Faúndes warns, "virtually no one is pro-abortion 'per se' or against abortion, without exception6;" in fact, "The real dilemma is to condemn or not to condemn the woman who aborts", so that "some believe that the solution to the problem of abortion is to condemn, while others think that condemning the woman who aborts is no solution".35

Given these considerations, we are convinced of the need to rewrite that question, this time no longer seeking to know whether the respondents are in favor or opposed to abortion, but whether or not they agree with the legal configuration of abortion for each of the cases discussed in the vignettes. In addition to rephrasing the question of the right to abort, we added a question about what should be the role of the Brazilian Public Health System - SUS - in each of the circumstances shown by the questionnaire. For instance, in the dilemma "Jussara", which discusses the gestation of a fetus with Down syndrome, we asked: "If Jussara could legally terminate a pregnancy, would you be in favor of the procedure being performed by SUS?".

Another important change occurred in view of the ruling of the Brazilian Supreme Court (STF) on the possibility of terminating a pregnancy in cases of an anencephalic fetus.4 After the Supreme Court decision, the questionnaire had to be changed and started to refer to the possibility of the procedure in a lawful and safe manner, being no longer necessary to file a legal petition to obtain authorization to do so.


Mosai is an original questionnaire set in the Brazilian scenario and its creation and content validation were based on methodological parameters recommended for this type of study. Its development was aimed at contributing to the increased amplitude and depth of data on the opinions of health care professionals on the issue of abortion, including physicians, nurses, social workers and psychologists, among others. This type of information is essential for the making of laws that are more adequate to the perspectives of society, to improve the working conditions of health professionals and to improve the quality of care that these workers offer to women having an abortion.

It is expected that the Mosai can also contribute with a greater variety of professional categories included in opinion surveys on abortion. Although the technical standards of humanized care for abortion stipulate the need of multidisciplinary care for women,36 very few publications have shown data on social workers and psychologists, and even so, very modest ones.8 The Mosai can facilitate the inclusion of a greater variety of professionals, given the possibility that the tool can be answered electronically, objectively and at the same time, in depth. One possible merit of Mosai for the in-depth study of opinions on abortion is the use of vignettes. The gain of using this strategy is that the participants' answers tend to be closer to what they would do in real-life situations, whereas, when a simple hypothetical situation is used, the participants' answers do not usually reflect the decisions they would make in everyday life.37

It is also expected that the Mosai can contribute to the collection of data on the allocation of public resources to assist patients who wish to perform the abortion, a subject rarely discussed in the scientific literature. Among the few studies on the subject, there was a 95.3% rate of approval for the allocation of public resources to assist patients seeking abortion.38 Similar data were found in the United States39 and Argentina.40

We recognize that a possible limitation of the process of content validation of the tool is the lack of specialists from other Brazilian regions in addition to the southeast. It should be emphasized, however, that subjects were invited from other states (Bahia, Rio Grande do Sul and Distrito Federal), but could not participate. Another possible limitation is the unbalanced ratio between men and women in the composition of the expert panel. It should be stressed, however, that the composition of the panel was scheduled for a more even number of men and women. However, of the eight men invited to join the study, only two were able to participate.


The objective of this research was to develop and validate the contents of Mosai. We considered we have achieved this goal satisfactorily. The interventions of specialists allowed us to correct conceptual and contextual misconceptions, as well as to identify ambiguities in the writing of the questionnaire. Following the validation process, Mosai will continue. The next steps will consist of its pre-test on a sample of 10 specialists in gynecology and obstetrics in order to verify its applicability. The psychometric properties of the questionnaire also need to be evaluated in future studies, including the construct validity, internal consistency and reliability.

Conflicts of interest

The authors declare no conflicts of interest.

Article history:
Received 8 May 2013
Accepted 30 June 2013

Study conducted at Universidade Estadual de Campinas, Campinas, SP, Brazil.

* Corresponding author.
E-mail:"> (D.B. Cacique).


1. Shaw D, Faundes A. What is the relevance of women''s sexual and reproductive rights to the practising obstetrician/gynaecologist? Best Pract Res Clin Obstet Gynaecol. 2006;20(3): 299-309.
2. Gonzalez D, Aguirre L. Los Médicos y el Aborto. Salud Publica Mexico. 1995;37(3):248-55.
3. Nomura RMY, Brizot MdL, Liao AW, Hernandez WR, Zugaib M. Gêmeos unidos e autorização judicial para o aborto. Rev Assoc Med Bras. 2011;57(2):205-10.
4. A ntônio Henrique da Mata C, Ana Cristina Viana C. Antecipação terapêutica do parto do feto anencéfalo: uma discussão necessária. Rev Bioética. 2012;20(3):417-24.
5. Marta GN, Marta TN. Aborto de fetos anencefálicos. Rev Assoc Med Bras. 2010;56(5):506-10.
6. Faúndes A, Barzelatto J. O drama do aborto: em busca de um consenso. Campinas: Editora Komedi;2004.
7. Yam EA, Dri es-Daffner I, García SG. Abortion opinion research in Latin America and the Caribbean: a review of the literature. Stud Fam Plann. 2006;37(4):225-40.
8. Cacique DB, Passini-Junior R, Osis MJMD. Opiniões, conhecimento e atitudes de profissionais da saúde sobre o aborto induzido: uma revisão das pesquisas brasileiras publicadas entre 2001 e 2011. Rev Saúde Sociedade. 2013. (no prelo).
9. Mayan MJ. Una in troducion a los métodos cualitativos: módulo de entrenamiento para estudiantes y profesionales. Mexico: Qual Institute Press;2001.
10. Faúndes A, Duarte GA, Andalaft Neto J, Olivatto AE, Simoneti RM. Conhecimento, opinião e conduta de ginecologistas e obstetras brasileiros sobre o aborto induzido. Rev Bras Ginecol Obstet. 2004;26(2):89-96.
11. Jones K, Chaloner C. Ethics of abortion: the arguments for and against. Nurs Stand. 2007;21(37):45-8.
12. Contrandiopoulos AP ea. Saber preparar uma pesquisa: definição, estrutura, financiamento. 2nd ed. São Paulo: Hucitec;1997.
13. Polit D, Hungler BP. F undamentos de pesquisa em enfermagem. 3rd ed. Porto Alegre: Artes Médicas;1995.
14. Demo P. Metodologia cie ntífica em ciências sociais. 3rd ed. São Paulo: Editora Atlas S.A.;1995.
15. Demo P. Pesquisa e constru ção de conhecimento: metodologia científica no caminho de habermas. 5th ed. Rio de Janeiro: Tempo Brasileiro;2002.
16. Galante AC, Aranha JA, Berald o L, Pelá NTR. A vinheta como estratégia de coleta de dados de pesquisa em enfermagem. Rev Latino-Amer Enferm. 2003;11(3):357-63.
17. Beauchamp TL, Childress JF. Pri nciples of biomedical ethics. 4th ed. New York: Oxford University Press;1994.
18. Diniz D, Almeida MD. Bioética e aborto In: Costa SIF, Garrafa V, Oselka G, editores. Iniciação à bioética. Brasília (DF): Conselho Federal de Medicina;1998.
19. Gallo C. O aborto dos outros. São Pau lo;2008.
20. Sousa JIGdS, Ribeiro GTF, Melo APC. Sí ndrome de Down: sentimentos vivenciados pelos pais frente ao diagnóstico. Pediatria. 2009;31(2):100-8.
21. Frazão M. Gravidez de risco coloca mãe e m dilema. Jornal de Londrina;2008 [accessed 16 Nov 2008]. Available at: id=828257.
22. Borges VTM, Magalhães CG, Martins AMVC, Ma tsubara BB. Síndrome de Eisenmenger na gravidez. Arq Bras Cardiol. 2008;90(5):e40-e1.
23. Hardy E, Duarte GA, Osis MJD, Arce XE, Possa n M. Anticoncepção de emergência no Brasil: facilitadores e barreiras. Cad Saúde Pública. 2001;17(4):1031-5.
24. Pedroso D, Drezett J. Histórias de mulheres em si tuação de violência e aborto previsto em lei. 2008 [accessed 6 Oct 2010]. Available at: pdf.
25. Briozzo L, Bedone JA. Aborto Inseguro: prevenção e re dução de riscos e danos. Campinas: Komedi;2009.
26. Minayo MCdS. O desafio do conhecimento: pesquisa qualitat iva em saúde. 2nd ed. São Paulo: Hucitec-Abrasco;1993.
27. Rest JR, Narvaez D, Thoma SJ, Bebeau MJ. DIT2: Devising and testing a revised instrument of moral judgment. J Educ Psychol. 1999;91(4):644-59.
28. Rest J, Narvaez D, Bebeau MJ, Thoma SJ. Postconventional mo ral thinking: a neo-Kohlbergian aproach. London: Lawrence Erlbaum associates, Inc.;1999.
29. Rest J, Thoma SJ, Narvaez D, Bebeau MJ. Alchemy and Beyond: Indexing the Defining Issues Test. J Educ Psychol. 1997;89(3): 498-507.
30. Rest J, Thoma S, Edwards L. Designing and validating a meas ure of moral judgment: Stage preference and stage consistency approaches. J Educ Psychol. 1997;89(1):5-28.
31. Coolican H. Research methods and statistics in psychology. 2nd ed. London: Hodder & Stoughton;1999.
32. Turato ER. Tratado da metodologia da pesquisa clínicoquali tativa: construção teórico-epistemológica, discussão e aplicação nas áreas da saúde e humanas. 2nd ed. Petrópolis: Editora Vozes Ltda.;2003.
33. Wynd C, Schmidt B, Schaefer M. Two quantitative approaches for estimat ing content validity. West J Nurs Res. 2003;25(5): 508-18.
34. Galvao L, Diaz J, Diaz M, Osis MJ, Clark S, Ellertson C. Emergency con traception: knowledge, attitudes and practices among brazilian obstetrician-gynecologists. Int Fam Plann Perspect. 1999;25(4):5.
35. Faúndes A. El falso dilema de estar a favor o en contra del aborto. 20 04 [accessed 19 Set 2011]. Available at: http://www.
36. Ministério da Saúde. Secretaria de Atencão à Saúde, Depar tamento de Açõ es Programáticas Estratégicas, Atencão Humanizada ao Abortamento: Norma Técnica. Brasília (DF);2005.
37. FeldmanHall O, Mobbs D, Evans D, Hiscox L, Navrady L, Dalgleish T. What we say and what we do: The relationship between real and hypothetical moral choices. Cognition. 2012;123(3):434-41.
38. Goldman LA, García SG, Díaz J, Yam EA. Brazilian obstetriciangynecologists and abo rtion: a survey of knowledge, opinions and practices. Reprod Health. 2005;2:10.
39. Chuang CH, Martenis ME, Parisi SM, Delano RE, Sobota M, Nothnagle M, et al. Contracept ion and abortion coverage: what do primary care physicians think? Contraception. 2012;86(2): 153-6.
40. Vasquez DN, Das Neves AV, Golubicki JL, Di Marco I, Loudet CI, Roberti JE, et al. Criti cal care providers'' opinion on unsafe abortion in Argentina. Int J Gynecol Obstet. 2012;116(3):249-52.