Acceptability of adolescents' abortion: effect of religious involvement and current legislation on people's views

Psicológica (2014), 35, 703-713. Acceptability of Adolescents’ Abortion: Effect of Religious Involvement and Current Legislation on People’s Views Félix Neto1 and Maria Teresa Munoz Sastre2 1University of Oporto, Portugal 2Jean-Jaures University, Toulouse, France Lay people’s views regarding the acceptability of adolescents’ abortion as a function of religious involvement, as a function of circumstances (e.g., fetus’ age), and as a function of current legislation in the country were examined. We compared data gathered in Portugal at a time when abortion on request was illegal, and data gathered in France where abortion on request has been legal since 30 years. Adult participants were presented with 64 vignettes of a few lines that were composed according to a five withinsubject factors design: the adolescent’s age (15 years vs. 17 ½ years) x the adolescent’s plans regarding schooling (leave school vs. go to college) x the number of months of the fetus (1, 2, 3, or 4 months) x the attitude of the adolescent’s family (agree vs. disagree) x the attitude of her boyfriend, 2 x 2 x 4 x 2 x 2. Through cluster analysis, three contrasting personal positions were found: Never acceptable, depending on circumstances, and always acceptable. The judgment rule of participants adopting the “depending” position was: Acceptability = Fetus’ age + Schooling plans + Partner’s agreement + (Parent’s agreement x Adolescent’s age). The percentages of participants endorsing the always unacceptable position were higher among regular attendees of religious service and among Portuguese participants than among other participants. The ways respect for religious tradition and respect for current laws impact of people’s views regarding adolescent’s abortion were shown to be independent the one from the other. The present study investigated lay people’s views regarding the acceptability of adolescents’ abortion as a function of circumstances (e.g., fetus’ age), as a function of their religious involvement, and as a function of the current legislation in their country. It compared data gathered in 1 Address correspondence to Dr. Félix Neto, e-mail: [email protected] F. Neto & M.T. Munoz 704 Portugal at a time when abortion on request was illegal, and data gathered in France where abortion on request has been legal since 30 years. During the second part of the 20th century, there has been a worldwide move towards liberalizing abortion laws. Many industrialized nations have issued compassionate abortion laws but even in a relatively homogeneous area as the European Union, legislation remains extremely disparate from one country to another: It ranges from extreme liberalism (e.g., Sweden) to extreme conservatism (e.g., Ireland). In Portugal, abortion on request was made legal in 2007, following a national referendum. The conditions for legal abortion are that pregnancy has not exceeded its tenth week, and that a three-day waiting period between request and abortion has been respected. Doctors have the right to refuse to perform abortions, and many of them use it. In France, abortion on request was made legal in 1975. The conditions are that pregnancy has not exceeded its twelfth week, and that a social worker has been contacted. A pregnant minor may ask for an abortion without her parents’ agreement if she is accompanied to the clinic by an adult (e.g., an aunt or a teacher). Many studies, using generic items, have been conducted on people’s attitude towards abortion in general (e.g., Strickler & Danigelis, 2002), and adolescents’ abortion (Adler, Ozer, & Tschann, 2003; Lohan, Cruise, O’Halloran, Aldernice, & Hyde, 2011) in particular. A majority of Western Europeans and North Americans support women’s right to get an abortion on request, and this support is even stronger regarding adolescents’ abortion. Although opinions about the right to abortion are reputed to be very polarized, what is usually found is a “normal” distribution of abortion attitudes, with most people’s views falling in between the two poles. This type of finding has sometimes been interpreted as reflecting people’s ambivalence towards abortion. Munoz Sastre, Pecarisi, Legrain, Mullet, and Sorum (2007) showed, however, that there was not necessarily ambivalence in people’s views because these views may in fact simply depend on concrete circumstances surrounding the abortion request. Using the methodological framework of Information Integration Theory (Anderson, 2008, 2012), they examined the situational factors that impact people’s acceptability judgments: the adolescent’s age, her future plans to continue schooling or not, the fetus’ age (1, 2, 3, or 4 months), and her parents’ and boyfriend’s agreement or not. They found three contrasted personal positions, which they called: Abortion is never acceptable (irrespective of circumstances, 8% of the participants), acceptability of abortion depends on the circumstances (63%), and abortion is always acceptable (in the conditions depicted in the scenarios, 23%). Among the Adolescents’ abortion 705 participants from the majority group, the fetus’ age, and the boyfriend and parents’ agreements were by far the most important determinants of acceptability. The present study complemented the previous study by Munoz Sastre et al. (2007), and used the same methodological framework (Anderson, 2008). The main question was: To what extent do people’s religious involvement and the current legislation in the country determine their adoption of one or another of these three positions? Do these factors have independent effects on people’s attitudes to adolescents’ abortion? Alternatively, do these effects interact? Previous studies have shown that religious involvement was, unsurprisingly, a strong predictor of attitude towards abortion (Ogland & Verona, 2011; Strickler & Danigelis, 2002). These studies have, however, considered abortion globally; that is, without distinguishing abortion on request from therapeutic abortion, and without specifying the context in which abortion is requested (e.g., the women’s age). As far as we know, no study has directly examined the effect of current legislation on attitude towards abortion: International comparisons have been conducted but they have relied on results obtained from studies using different methodologies. We replicated Munoz Sastre et al.’s study on two samples of participants, one that was composed of people living in Portugal, a country in which, as indicated above, abortion on request became legal only in 2007, and one that comprised French people. The French sample was composed in order to match, as closely as possible, the Portuguese sample on three criteria: age, gender and religious involvement. Our hypotheses were that (a) the three personal positions reported by Munoz Sastre et al. (2007) would be found, (b) these positions would be differently endorsed depending on the religious involvement of the participants and (c) these positions would be differently endorsed depending on the participants’ nationality. METHOD Participants. The participants were unpaid volunteers. We first contacted 250 people walking along sidewalks of the city of Oporto, Portugal, explained the study, asked them to participate, and, if they agreed, arranged where and when to administer the experiment. The researchers attempted to enroll participants evenly across a large age spectrum from 18 to 60 years of age. Of these 250, 117 (47%) participated: 57 were males and 60 were females, 42 were nonbelievers, 55 were believers in God and 20 F. Neto & M.T. Munoz 706 were regular attendees of religious service. Their mean age was 33.20. Secondly, we contacted people walking along sidewalks of the city of Toulouse, France. If their age, gender and level of religious involvement matched one of the participants in the Portuguese sample, they were explained the study and asked to participate. Out of the 117 French participants, 56 were males and 61 were females, 38 were nonbelievers, 58 were believers in God and 21 were regular attendees to religious services. Their mean age was 33.26. Material. The material consisted of 64 cards containing a vignette of a few lines, a question, and a response scale. The vignettes were composed according to a five within-subject factor design: the adolescent’s age (15 or 17½ years old) x the adolescent’s plans regarding schooling (“She has always wanted to leave school as early as possible” or “She has always wanted to go to college”) x the number of months of the fetus (1, 2, 3, or 4 months) x the attitude of the adolescent’s family (parents agree vs. disagree with the decision) x the attitude of the adolescent’s boyfriend (agrees vs. disagrees with the decision), 2 x 2 x 4 x 2 x 2. The boyfriend was understood to be the baby’s father. An example of scenario is the following: “Pauline is 15 years old. She has always wanted to leave school as early as possible. She is four months pregnant and has told her doctor she wants an abortion. Her parents do not consent, nor does her boyfriend.” Under each vignette were a question and a response scale. The question was, “To what extent do you believe that X’s abortion would be an acceptable medical/surgical procedure in this case?” The response scale was a 15-cm scale with a left-hand anchor of "Not at all acceptable” and a righthand anchor of "Completely acceptable." The cards were arranged by chance and in a different order for each participant. The participants answered additional questions asking for age, gender, education, opinion about the role of medical care, religious belief, and religious background. Strict anonymity was respected. Procedure. The Portuguese data were gathered in 2006; that is, before the referendum on abortion and before the radical change in the law. The French data were gathered in 2007. In both cases, the site was either a vacant classroom in the local university or the participant’s private home. Each person was tested individually. The original French version of the scenarios was used with the French participants and the Portuguese version was used with the Portuguese participants. (Authors are fluent in both languages.) Adolescents’ abortion 707 The session had two phases (Anderson, 2002). In the familiarization phase, the experimenter explained what was expected and presented each participant with 32 vignettes taken from the complete set. For each vignette, the participant read it out loud, was reminded by the experimenter of the items of information in the story, and then made an acceptability rating by putting a mark on the response scale. After completing the 32 ratings, the participant was allowed to look back at his or her responses and to compare and change them. In the experimental phase, each participant gave ratings for the whole set of 64 vignettes, working at his or her own pace, but was no longer allowed to look back at and change previous responses. In both phases, the experimenter made certain that each subject, regardless of age or educational level, was able to grasp all the necessary information before making a rating. The participants took 25-50 minutes to complete both phases. RESULTS A cluster analysis was first performed on the raw data (Hofmans & Mullet, 2013), and a three-cluster solution was chosen on the basis of interpretability of findings. Figure 1 shows the main results. For 108 participants (46%), the mean responses varied as a function of the information in the vignettes. An ANOVA was conducted on the raw data from this cluster. The design was Age x Plans x Months x Family x Boyfriend, 2 x 2 x 4 x 2 x 2. Owing to the great number of comparisons, the threshold value has been set at .0016, using the Bonferoni technique (.05/31). Main results are shown in Table 1. Acceptability was judged higher (a) when the adolescent was 15-year old (M = 8.73) than when she was 17½-year old (M = 7.99), F(1, 103) = 12.75, p < .001, η²p = .11, (b) when she wanted to go to school (M = 8.69) than when she wanted to leave (M = 8.03), F(1, 103) = 18.35, p < .001, η²p = .15, (c) when her parents agreed with her decision (M = 8.74) than when they disagreed (M = 7.98), F(1, 103) = 39.55, p < .001, η²p = .28, and (d) when her male partner agreed with her decision (M = 8.97) than when he disagreed (M = 7.75). F(1, 103) = 58.95, p < .001, η²p = .36. The effect of the fetus’ age was also significant, F(3, 309) = 150.40, p < .001, η²p = .59. Post-hoc analyses using Tukey’s honestly significant difference test showed that acceptability was judged higher (a) when the fetus was one-month old (M = 11.06) than when the fetus was two-months old (M = 10.10), (b) when the fetus was two-months old than when the fetus was three-months old (M = 7.51), and (c) higher when the fetus was three-months old than when the fetus was four-month F. Neto & M.T. Munoz 708 old (M = 4.77). The Age x Family interaction was significant. F(1, 103) = 13.57, p < .001, η²p = .12. The effect of the parents’ attitude on acceptability was stronger when the adolescents was 15-year old than when she was 17 ½-year old. This cluster was called “Depending on Circumstances”. Table 1. Demographic composition of the three Clusters in terms of frequency (and percentage). For 49 participants (21%), the mean responses were always on the extreme left of the scale; that is, close to the “Not at all acceptable” end. This cluster was called “Always Unacceptable”. The effect of the fetus ‘age was the only one to be significant, F(3, 120) = 20.10, p < .001, η²p = .33. Post-hoc analyses showed that acceptability was judged higher when the fetus was one-month old (M = 4.32) than when the fetus was Four-months old (M = 2.72). Adolescents’ abortion 709 For the remaining 77 participants, the mean responses were always on the extreme right of the scale; that is, close to the “Completely acceptable” end. This cluster was called “Always Acceptable”. Acceptability was judged higher (a) when the adolescent wanted to go to school (M = 12.72) than when she wanted to leave (M = 12.48), F(1, 68) = 17.02, p < .001, η²p = .205, (b) when her parents agreed with her decision (M = 12.83) than when they disagreed (M = 12.37), F(1, 68) = 26.33, p < .001, η²p = .28, and (c) when her male partner agreed with her decision (M = 13.14) than when he disagreed (M = 12.06), F(1, 68) = 25.56, p < .001, η²p = .27. The effect of the fetus’ age was also significant, F(3, 204) = 27.46, p < .001, η²p=.29. Post-hoc analyses using Tukey’s honestly significant difference test showed that acceptability was judged lower when the fetus was one-month old (M = 11.68) than in any other cases M = 12.91. Figure 1. Results of the Cluster Analysis. The three clusters of attitudes toward adolescent abortion are labeled, from left to right, “Never Acceptable,” “Depending on Circumstances”, and “Always Acceptable”. The X-axis displays the age of the fetus, and the two curves the attitude of the adolescent’s boyfriend (the baby’s father). The mean levels of acceptability for the participants of each combination of age and boyfriend’s attitude are shown on the Y-axis. These mean levels have been computed across levels of adolescent’s age, plans regarding schooling and parent’s attitude. F. Neto & M.T. Munoz 710 Table 1 shows the composition of the clusters as a function of age, gender, country, and religious involvement. Country, Chi² (2) = 12.90, p < .002, and religious involvement, Chi² (4) = 28.71, p < .001, were significantly associated with the clusters. The Country x Religion interaction was not significant. Finally, an ANOVA was conducted on the participants’ mean ratings using a Country x Religious Involvement, 2 x 3, design. French participants’ mean ratings (M = 9.25) were higher than Portuguese participants’ mean ratings (M = 7.68), F(1, 228) = 10.84, p < .002. The religious involvement factor was also significant, F(2, 228) = 10.43, p < .001. Post-hoc analysis showed that non-believers mean rating were higher (M = 9.85) than regular attendees’ mean ratings (M = 6.94), p < .001. The interaction was not significant, p = .15. DISCUSSION As hypothesized, the three contrasting personal positions reported by Munoz Sastre et al. (2007) were retrieved, and their meaning was similar. As hypothesized, the percentages of participants endorsing the always unacceptable or the always acceptable positions were, respectively, higher or lower among regular attendees of religious service than among others. These results are consistent with findings reported in previous studies (Ogland & Verona, 2011; Munoz Sastre et al., 2007; Strickler & Danigelis, 2002). Interestingly, however, the majority position among regular attendees of religious service was not the always unacceptable one: Their attitude clearly depended on circumstances of the request for abortion. In other words, a majority of Portuguese and French regular attendees of religious service considered that in some circumstances (e.g., fetus aged 3 months and male partner unfavorable) abortion on request was not acceptable whereas in other circumstances it could be. If generic items had been used (e.g. pro-con items), it is not sure that similarly nuanced judgments would have been obtained. The depending-on-circumstances position was also the majority position among believers in God who were not regular attendees. Among non-believers, the majority position was always acceptable. As hypothesized, the percentage of participants endorsing the alwaysunacceptable position (as opposed to always-acceptable) was higher (lower) among Portuguese than among French participants. In both cases, however, the majority position was the depends-on-circumstances position. This result supports the idea that current legislation in a country naturally impacts on people’s views about societal issues. The effect is probably bi- Adolescents’ abortion 711 directional; that is, (a) changes in people’s opinions may lead politicians to adjust current laws to people’s opinions, and (b) changes in laws can make at least some people to realize that their views must change. In the case of Portugal, it was the progressive change in public opinion (illustrated by the results of a national referendum on abortion held in 2007) that prompted the change in legislation (National Commission of Elections, 2007). Interestingly, the effects of religious involvement and current legislation on participants’ views did not significantly interact. These factors had additive effects. In other words, the ways in which respect for religious tradition and respect for current laws impact of people’s views regarding adolescent’s abortion are largely independent the one from the other. As a result, a change in current laws is not likely to affect people’s attachment to religious values for what concerns adolescent’s abortion. In addition, neither gender nor age were related to personal position, a finding that is consistent with Munoz Sastre et al.’ (2007) results. The current findings, and the ones reported by Munoz Sastre et al. (2007, see also Mullet et al., 2012), have also methodological implications. As indicated early, many previous studies have assessed people’s attitudes regarding abortion by asking participants to indicate whether they were favorable or unfavorable to it (Carlson, Nelson & Coleman, 2000). At first glance, such question seems natural and reasonable. Unfortunately, this type of question already contains part of the response. It suggests that the only way to regard abortion is to be either con or pro. Not being clearly either pro or con is assimilated to be “ambivalent”. Adopting such a methodological stance amounts to ignoring that ethics is, according to Aristotle (1955, p. 41), an issue of practical reasoning, concrete deliberation, and applied wisdom. Already more than 2500 years ago, this philosopher knew that ethics was fundamentally a question of circumstances. Ignoring circumstances at the time of judging amounts to stop applying one’s reasoning abilities, to stop taking into account the pros and cons of each concrete situation, and to stop relying on what Aristotle called “one’s human virtues” when deciding on the more acceptable course of action; that is, largely acting like an automat (see also Rachels, 2010). The present study shows that methodologies that are consistent with Aristotle’s views about the nature of ethics exist and can be implemented in the field of empirical bioethics (Anderson, 2012; Theuns & Hofmans, in press). It shows that, for a majority of people, circumstances matter. These people’ judgment process can, according to Anderson (2013), be formalized in the following way: Acceptability = Fetus’ age + Schooling plans + Partner’s agreement + (Parent’s agreement x Adolescent’s age). This