A Mixed-Methods Approach to Translation Challenges in Multi-Language Research: Assessing How Spanish-Speaking Adults in the U.S. Interpret the Terms Pro-Vida (Pro-Life) and Pro-Elección (Pro-Choice)

Abstract

In the US, abortion attitudes are categorized using the terms pro-life/pro-vida and pro-choice/pro-elección. In light of demographic changes in the U.S. population, it is important to explore whether these U.S.-centric abortion-related socio-political terms are interpreted similarly by English and Spanish speakers. We analyzed survey data from 1,025 adults and 15 cognitive-style interviews. We (1) quantitatively examined responses for abortion-related terms across English and Spanish, and (2) qualitatively assessed whether Spanish speakers are familiar with pro-vida and pro-elección, how they define these terms, their perceptions of people who self-identify with the terms, and their suggestions for alternative terms. Results suggest that Spanish speakers are less familiar with the contextual background of these terms. As such, they define them literally and conceptualize them from a broad perspective, which entails a degree of ambiguity for public opinion research. This study offers recommendations for abortion attitudes survey research and provides insights for survey methodologists and implications for multilingual survey research in culturally diverse settings. With Roe v. Wade being overturned in the U.S. in 2022, and thus abortion being on the ballot in many states hereafter, understanding how English and Spanish speakers interpret these terms is of utmost importance for polling and elections.

Introduction

In the US—where the Latinx population represents the largest and fastest-growing ethnic group, and Spanish is the second most spoken language—conducting survey research concurrently in English and Spanish has become a common survey design practice. Thus, it is paramount that researchers make efforts to ensure comparable cross-cultural understandings of survey items through cognitive interviewing and other evaluations (Behr, & Sha, 2018; Miller, 2018; Willis, 2015). Conducting cognitive interviews for survey translations with non-English speakers is important given that speakers of other languages and from other cultures might use different communication styles and norms that can result in different interpretations of items (Park, & Goerman, 2018). Cross-cultural cognitive interviewing can inform decisions about question-wording and translations adapted to a given cultural group (Miller, 2018). By employing a mixed-methods approach combining survey data and cognitive interviews conducted in 2021 with a sample of English- and Spanish-speaking U.S. adults, we explored how the terms pro-vida (pro-life) and pro-elección (pro-choice) were conceptualized and understood by a sample of U.S. Spanish-speaking participants. Understanding English- and Spanish-speakers’ interpretations of these terms is of utmost importance for polling research to document accurate sentiment toward abortion. In addition, understanding Spanish speakers’ interpretations is crucial in light of the rapid growth of Spanish-speaking Hispanic and Latinx populations in the US.

This research is particularly timely given the 2022 Supreme Court decision in Dobbs v. Jackson Women’s Health Organization that overturned Roe v. Wade (i.e., the Supreme Court decision establishing constitutional protection for abortion up to viability) resulting in states being able to restrict abortion beyond the bounds they were able to previously. Now, more than ever, public opinion on abortion is a topic that requires precision in measurement, given that abortion has been and will be on the ballot in many states (Felix et al., 2024). Given that Latinx-eligible voters represent the largest group of minority voters in federal elections (Krogstad et al., 2024), it is imperative to accurately measure Latinx public opinion on abortion.

The terms pro-life and pro-choice have been used in polls and public opinion research in the U.S. as proxy measures for abortion attitudes for decades (Rudy, 1996). Pro-life is broadly conceptualized as opposing abortion, while pro-choice is broadly conceptualized as advocating for the right to choose to have an abortion (Hess & Rueb, 2005), though people conceptualize and define these terms in varying ways (LaRoche, Jozkowski, Crawford, & Jackson, 2024; Solon et al., 2022b). These terms are often presented as dichotomous and in opposition to one another. Researchers often conflate people’s attitudes toward abortion with these labels—describing people who endorse legal abortion as holding “pro-choice attitudes” and those opposing legal abortion as holding “pro-life attitudes” (LaRoche et al., 2024). However, these terms should not be conflated with abortion attitudes as people who identify as pro-life indicate supporting abortion in some circumstances, while those who identify as pro-choice report opposing abortion in certain circumstances (Jozkowski, Crawford, & Hunt, 2018). Furthermore, such dichotomous and oppositional framing is not consistent with how people engage with the terms, as U.S. adults report simultaneously identifying as both pro-life and pro-choice (LaRoche et al., 2024; PerryUndem, 2017; Public Religion Research Institute, 2018). Given the inherent complexity of these terms among English speakers, diverse interpretations of the terms in survey research by different subpopulations, including Latinx and Spanish speakers might result in misleading or biased interpretations of public opinion on abortion.

Translation Issues in Abortion Attitudes Research

Translating pro-life and pro-choice across different languages adds a layer of complexity to survey research. These terms are largely recognizable and understood by most U.S. adults as pertaining to abortion, yet this may not be the case for the commonly used Spanish translations—pro-vida and pro-elección (Solon et al., 2022b; Valdez, Jozkowski, Montenegro, Crawford, & Jackson, 2022). Pro-vida and pro-elección are technically accurate respective translations, but they are not the only terms used as Spanish proxies for pro-life and pro-choice. For example, other alternative pairs, such as pro-aborto/anti-aborto (pro-abortion/anti-abortion) or derecho a decidir/derecho a vivir (right to choose/right to life), have been used when referring to abortion as a socio-political movement, demonstrating the use of other terminology (Valdez et al., 2022). However, more people interpret pro-elección as being related to issues other than abortion (e.g., voting or presidential elections) than pro-vida (Valdez et al., 2022). Furthermore, Spanish-speaking respondents are less familiar with these terms than English speakers, with Spanish-language survey-takers expressing more uncertainty in defining and more frequently misinterpreting the terms (Solon et al., 2022b). Such findings demonstrate that these U.S.-centric socio-political terms and translations—pro-vida and pro-elección—may not be similarly understood by English and Spanish speakers in the US. It is necessary to explore how and to what extent Spanish-speaking respondents understand and define pro-vida and pro-elección.

These translation challenges are exemplified by the way some public opinion organizations approach measuring abortion identification. For instance, since 1995, Gallup has asked in English, “With respect to the abortion issue, would you consider yourself to be pro-choice or pro-life?” while in Spanish, they ask, “With respect to the abortion issue, do you consider yourself in favor of or opposed to abortion?” (Jones, & Saad, 2020);¹ importantly, these items are listed as equivalent. We argue that both questions hold value but assess two different constructs—the English version asks about identification with socio-political terms, and the Spanish version assesses broad support or opposition to abortion. That Gallup asks such different questions for items deemed equivalent in the different language versions may denote the difficulty of providing an accurate Spanish translation for pro-life and pro-choice. It may also indicate a presumption that Spanish-speaking respondents would have a different understanding of the corresponding Spanish translations (pro-vida and pro-elección) or an assumption that identifying as pro-choice is universally equivalent to being in favor of abortion and identifying as pro-life is equivalent to being opposed to abortion, which is not the case (Jozkowski et al., 2018; LaRoche et al., 2024; Solon et al., 2022b).

Assessing Abortion Attitudes Among Foreign-Born Populations

Language of survey administration matters because it can signify migration status, length of residence, and language dominance, which can, in turn, influence abortion attitudes. Indeed, a significant proportion of survey respondents in the U.S. who choose to answer a survey in Spanish are likely migrants (Goerman, Meyers, Sha, Park, & Schoua-Glusberg, 2018). U.S. Latinxs tend to endorse legal abortion less than U.S. non-Latinxs (Bruce, 2020; Thomas, Norris, & Gallo, 2017), English-dominant Latinxs endorse legal abortion more than Spanish-dominant Latinxs (Branton, Franco, Wenzel, & Wrinkle, 2014), and English-speaking respondents are more inclined to support upholding Roe v. Wade than Spanish-speaking respondents (Solon et al., 2022a). After migrating to the U.S., foreign-born Latinxs might carry specific cultural knowledge, values, norms, and language, which may influence how they conceptualize and understand abortion-related terms. For instance, common terms associated with the abortion movement in Argentina are pro-/anti-aborto (pro-/anti-abortion), while in Mexico, they are pro-elección/pro-vida (pro-choice/pro-life) (Tozzi, Marcilese, & Gonzáles, 2015; Zaremberg, & Lucero, 2018). Thus, when surveys are translated from English into Spanish, it is important to consider regional language variations, cultural nuances, and historical language development. Failure to adequately do so may harm survey validity and data quality and introduce measurement error, especially among foreign-born respondents (Valdez et al., 2022).

Multi-language survey methodologists have articulated the importance of measuring English-language proficiency when surveying foreign-born populations (Goerman et al., 2018; Pasick, Stewart, Bird, & D’Onofrio, 2001; Sha & Gabel, 2020). Language acquisition, fluency, and use are key determinants in explaining migrants’ integration processes and cultural adjustments in destination countries (Marin, & Gamba, 1996), which may, in turn, influence their views on various issues or at least responses on surveys. In addition, exposure to mainstream U.S. culture and cultural references may determine the extent that survey items might be cognitively or culturally challenging for foreign-born respondents (Park, Sha, & Pan, 2014) and the extent that such exposure may also influence opinions and attitudes. Indeed, Goerman et al. (2018) highlight the importance of including monolingual foreign-born respondents in cognitive interviewing since they might be better at recognizing translation issues in their native language, which could lead to inaccurate results compared with bilingual respondents. Second and subsequent generations of Latinxs tend to respond with greater endorsement of legal abortion than foreign-born Latinxs (Bartkowski, Ramos-Wada, Ellison, & Acevedo, 2012; Bueno et al., 2024; Jones & Cox, 2012), which could reflect variation in attitudes or interpretation of items. Thus, respondents’ preexisting familiarity with the English terms pro-life and pro-choice, or the extent that they report being bilingual in English, may explain differences in their familiarity, identification, and attitudes toward abortion compared with monolingual Spanish speakers without exposure to these English terms and their associated socio-political and cultural meanings. As such, previous research that has relied on using the terms pro-life/pro-vida and pro-choice/pro-elección to measure Spanish-speaking respondents’ abortion attitudes is questionable.

Current Study

To address this gap, we pursued the following research questions: To what extent do Spanish speakers self-identify with and interpret pro-vida and pro-elección differently than English speakers self-identify and interpret pro-life and pro-choice, respectively? How do Spanish speakers interpret these terms?

To answer these questions, we used (1) survey data collected from English- and Spanish-speaking respondents in the U.S. and (2) cognitive interviews from a subsample of Spanish-speaking survey respondents. This mixed-method approach allows us to first quantitatively explore patterns in respondents’ self-identification with abortion identity terms across English and Spanish; and secondly, to qualitatively examine (1) how Spanish-speaking US adults identify with, define, and interpret the terms pro-vida and pro-elección, (2) how they perceive others who self-identify with the terms, and (3) how familiar they are with the terms and their recommendations for other alternative terms in Spanish.

We selected the Spanish terms pro-vida and pro-elección because, a priori, they seem to represent close and accurate translations of the terms pro-life and pro-choice according to the survey (Solon et al., 2022b) and social media data (Valdez et al., 2022).

Data and Methods

We fielded an online survey and selected a subsample of respondents to participate in cognitive interviews. All stages of this study were approved by the Institutional Review Board of Author(s)’s University and NORC. Respondents were offered compensation for their participation in each phase.

The multi-language design of the survey items and the interview protocol followed a parallel development approach with English as the source language. The translation process adhered to the Translate, Review, Adjudicate, Pretest, and Document framework (Harkness, Vijver, & Mohler, 2003). This approach entailed a team-based translation process with several iterations of review and reconciliation. The translators were two native Spanish-speaking researchers (one being the first author). They conducted two independent translations of the instruments that were later reviewed and reconciled into one final version. Two other bilingual team members reviewed the reconciled version before the adjudication stage, whereby modifications to both language versions were made iteratively to achieve closely corresponding instruments.

Given our prior awareness of translation issues with the terms pro-vida and pro-elección, in the Spanish version of the survey, some questions were followed by the English terms in parentheses (i.e., pro-vida [pro-life] and pro-elección [pro-choice]). This decision was made for clarification purposes and in an attempt to facilitate parallel meaning (Goerman et al., 2018).

Quantitative Data and Analysis

We launched a web-administered survey in 2021, using NORC’s AmeriSpeak® panel, a U.S. nationally representative probability-based panel. The sample consists of 1,025 adults; 20% responded in Spanish. The sampling strata included age, race/ethnicity, education, and gender, with a proportional sampling model, except for Spanish-preference panelists, who were oversampled.

We compared Spanish speakers’ (n = 204) respondents with English-speaking respondents (n = 821). First, we explored whether participants were familiar with the terms in their respective languages (see Supplementary Table A1, Item 1). Second, we compared Spanish-speaking respondents’ familiarity with pro-vida and pro-elección with their level of English-language proficiency. Disregarding survey language, all Latinx respondents (n = 361) were asked to self-report their English-language proficiency (see Supplementary Table A1, Item 2). Third, we examined participants’ self-identification with the terms across a nominal 9-point scale (see Supplementary Table A1, Item 3). Chi-square analyses were conducted to examine the associations of our variables in the three analyses above. Fourth, as an additional test, we compared the response distribution of English- and Spanish-speaking participants on two distinct abortion-related measures: (1) self-identification with the terms using two unipolar questions (see Supplementary Table A1, Item 4); and (2) participants’ general support or opposition to abortion in response to the question (see Supplementary Table A1, Item 5). See Supplementary Table A2 for the characteristics of the quantitative sample.

Qualitative Data and Analysis

Upon survey completion, respondents were invited to participate in a follow-up interview. Among those who volunteered, we selected a diverse sample based on demographic characteristics and responses to abortion attitude items. Our interview protocol (available from the authors upon request) drew on cognitive interviewing principles (e.g., Willis, 2005), assessing issues such as item comprehension and interpretability. Cognitive interviewing is a qualitative research technique aimed at assessing respondents’ cognitive processes to interpret and respond to survey items; an important purpose is to improve item comprehension and clarity (Willis, 1999). The protocol was designed based on retrospective probing (Beatty & Willis, 2007) with scripted and spontaneous probing. In this way, we obtained information from participants via their responses to the survey items and then by asking them to explain their responses during the interview. However, we did not adhere to strict recommendations for cognitive interviewing, such as completing the cognitive interview immediately after the survey. As such, during the interview, participants were shown their survey responses. Our interview protocol also included construct-generating information, such as asking for alternative terms of pro-vida and pro-elección.

We conducted 10 pilot interviews with a convenience sample (half in Spanish with respondents of diverse origins). During the pilot phase, the research team discussed and revised the protocol. In spring of 2022, we conducted 51 interviews, 15 in Spanish. The sample contains respondents from Mexico, Cuba, and several other Central American, South American, and Caribbean countries. See Supplementary Table A3 in the Appendix for details of the qualitative sample.

We conducted Zoom® interviews lasting an average of 60 min; they were professionally transcribed in Spanish and translated into English. Before each interview, interviewers prepared a visual aid document with some survey items and interviewees’ responses. The document was shown to participants through screen sharing, allowing the interviewer to engage participants in discussing their cognitive interpretation of the survey items. The interview protocol included two types of questions. First, interpretation questions assessed how participants understood and conceptualized the question and the response options offered. Second, participants could elaborate on the reasoning and thought process underlying their responses. We also asked participants (1) how they define and interpret the terms pro-vida and pro-elección, (2) how they perceive other people who self-identify as pro-vida and pro-elección, and (3) how familiar they are with pro-vida and pro-elección and whether they think there are better or alternative terms in Spanish. We provided descriptions of pro-vida and/or pro-elección to those participants who stated that they did not know one or both terms at the time of the interview.²

Data were analyzed using Dedoose® in Spanish to respect the cultural nuances expressed by participants in their native language; we provide in-text quotes in Spanish (English translations are provided in Supplementary Appendix B). The first two authors, both native Spanish speakers, undertook a multi-step approach to conduct content and thematic analyses of the data. After an iterative review of the transcript data, we grouped categories of information in a codebook, and identified recurring response patterns. Following recommendations that assert that reliability can be established via negotiation (Elliott, 2018; Saunders et al., 2023), the first two authors coded 50% of the interviews as first coders and the remaining 50% as debriefers of the interviews initially coded by the other author. We discussed each alteration of the codebook, which resulted in consensus among coders. We met regularly to consider the analytic plan and interpretations of findings and resolved rare disagreements through discussion.

Quantitative Results

English-speaking participants indicated a significantly higher level of familiarity with pro-life and pro-choice than Spanish participants (χ² (1) = 66.1082 and 83.5293, for pro-vida and pro-elección, respectively, p’s < .001). Figure 1 shows that more than one-third of Spanish-speaking respondents were unfamiliar with pro-vida (37.4%), and more than two-fifths were unfamiliar with pro-elección (41.1%). Comparatively, only 10% of English-speaking respondents indicated unfamiliarity with the two terms. We further assessed whether the lack of correspondence observed for Spanish-speaking respondents could be related to their English-language proficiency level. Figure 2 shows a positive relationship between English-language proficiency and familiarity with both terms (χ² (3) = 42.7223 and 39.3250, for pro-life and pro-choice, respectively, p’s < .001). More than 70% of people who reported not speaking English at all said they were not familiar with pro-vida and pro-elección.

Second, we found differences in participants’ endorsement of pro-life/pro-vida and pro-choice/pro-elección across the 9-point scale by language (χ²(8) = 29.5310, p < .001). According to Figure 3, the percentage of English-speaking respondents who identified as completely pro-choice/pro-elección (33.6%) was more than double that of Spanish-speaking respondents (14.0%). Alternatively, Spanish-speaking respondents who identified as completely pro-life/pro-vida (25.4%) surpassed English-speaking ones (17.0%) by eight percentage points. In addition, more Spanish-speaking respondents indicated they prefer not to answer (17.7%), in contrast with 5.7% of English-speaking participants. We also assessed self-identification with pro-life/pro-vida and pro-choice/pro-elección using unipolar scales from zero to six and compared findings with a question of overall support or opposition to abortion, similar to the Spanish version of the abortion labels questions asked by Gallup (see Figure 4). This comparison allowed us to identify correspondence among English-speaking respondents who identified as not at all pro-life (i.e., selected value “0”) and support abortion (95.4%) and those who identified as completely pro-choice (i.e., selected value “6”) and support abortion (94.1%). However, this pattern differs for Spanish-speaking respondents. Less than two-thirds of Spanish-speaking respondents (63.4%) who identified as not at all pro-vida indicated they support abortion, and only half (52.9%) of Spanish-speaking respondents who identified as completely pro-elección also indicated that they support abortion.

Qualitative Results

Cognitive interviews helped us further explore the patterns observed in our quantitative assessment of Spanish-speaking participants’ conceptualization of pro-vida and pro-elección.

Pro-Vida

Defining and interpreting pro-vida.

“Respetar” (respect), “defender” (defend), “preservar” (preserve), “dar” (give), “luchar por” (fight for), “en favor de la vida” (in favor of life) were some of the verbs used by respondents to define pro-vida.

Most participants referred to pro-vida as related to human life broadly across the life course (e.g., providing care and education, favoring gun control, opposing the death penalty), including abortion. Participants also more narrowly restricted their conceptualization of pro-vida to fetuses and babies. Alicia,³ a 53-year-old woman from Central America,⁴ described these two views:

…si hubiera más educación, más acceso a las mujeres a tener una prevención, eso es parte de ser provida, pero por supuesto que se interpreta de otra manera. Se interpreta “no al aborto,” pero cuando no es esa [interpretación], esa palabra tiene mucho significado, porque después que nacen los niños, incluso hay que seguirles dando educación, eso es parte de ser provida. (Supplementary Appendix B: Quote 1)

Half of the respondents defined pro-vida as synonymous with “anti-abortion,” like Paula, a 38-year-old woman from Central America, said, “yo me considero provida porque me opongo al aborto.” (Supplementary Appendix B: Quote 2). Sometimes, pro-vida was associated with notions of “killing” or “taking life away.” This was the case of Rosa, a 34-year-old woman from an unspecified Hispanic/Latino country:

Pues no, pienso que quitarle la vida a alguien no sería lo correcto y no somos quienes para hacerlo, …la gente no lo quiera ver, así seriamos asesinos, prácticamente estaríamos practicando el asesinato. (Supplementary Appendix B: Quote 3)

Generally, pro-vida was conceptualized as protecting babies and children; for example, Paula referred to the fetus/unborn child as “personitas indefensas que no pueden ni siquiera levantar la mano.” (Supplementary Appendix B: Quote 4). Of note, only one participant defined pro-vida as being against “legal” abortion.

Alternatively, some people conceptualized pro-vida from a more holistic perspective that extended beyond human life to the life of animals and plants as well. Brenda, a 44-year-old woman from South America, referred to this broad perspective:

Hay seres que son muy primitivos y eso no quiere decir que no tengan vida, la tienen. Para mí todo lo que tiene que ser vivo debe ser respetado. Todo. Yo no quiero ni que mi hijo arranque las flores del jardín, ni que coja la mariposa. Yo no quiero… Entonces, ¡imagínate si no soy provida! (Supplementary Appendix B: Quote 5)

Perceptions about people who identify as pro-vida.

Most participants had either positive or neutral perceptions of people who identify as pro-vida or of those involved in the pro-life movement. Often, perceptions about other people reiterated their descriptive definitions of pro-vida. For example, interviewees argued that pro-vida are “people who defend or respect life” or “people who are against abortion." However, other times, their perceptions denoted positive emotions toward those who associate with pro-vida. For example, some participants perceived pro-vida people as “generous,” “good people who have good values,” or “people who represent peace and love”; for example, Beatriz, a 44-year-old woman from an unspecified Hispanic/Latino country, stated:

Me imagino una persona muy, con valores muy buenos, con deseos de vivir, una persona con mucho amor, una persona también altruista, una persona como lo decimos nosotros, buena gente. (Supplementary Appendix B: Quote 6)

Negative perceptions emerged, though less frequently; for instance, one participant expressed that “pro-vida people condemn and demonized women.” Additionally, some participants perceived pro-vida people as “extreme,” “very conservative,” or “guided by fixed moral views tightly linked to religion.” Rodrigo, a 43-year-old man from Mexico, shared this perspective:

los grupos provida creo que son demasiado moralistas… … los veo como demasiado polarizados hacia el “¡no!, siempre tiene que ser la vida”… los grupos provida siento que son muy…, pues eso como muy conservadores, muy cercanos a la iglesia, a la religión, y esas cosas. (Supplementary Appendix B: Quote 7)

Noteworthy, Rodrigo who partly self-identified as pro-vida (as well as partly pro-elección) distanced himself from pro-vida people, exemplifying the complexities and nuances associated with the pro-vida term and individuals’ own self-identifications:

Yo me considero a favor de la vida, pero no en ese extremo, o sea, creo que estoy a favor de la vida, pero no me identifico con esos grupos provida. (Supplementary Appendix B: Quote 8)

Familiarity, adequacy, and alternative terminology.

Beyond the literal meaning interpreted from pro-vida, some participants were not entirely familiar with the term as related to abortion. For example, Claudia, a 33-year-old woman from an unspecified Hispanic/Latino country, stated, “Yo creo que [el] otro día vi un sticker en un carro.” (Supplementary Appendix B: Quote 9). As a sample of primarily foreign-born Spanish-speaking respondents, some emphasized that the term was new to them since it was not commonly used or heard of as related to abortion in their country of origin and pertained mostly to the U.S. context. In this regard, when asked about the term, Miguel, a 64-year-old man from Cuba, said, “Aquí sí, pero en Cuba eso no” (Supplementary Appendix B: Quote 10). Similarly, Alicia told us:

Bueno, ese concepto, la imagen de acá, que la aprendí acá, porque [el término provida] está en contra del aborto totalmente (Supplementary Appendix B: Quote 11)

Generally, most participants stated that pro-vida was an adequate term in Spanish as a translation for pro-life. Only one participant offered alternative terms in Spanish, such as pro-esperanza (pro-hope) or pro-pensar (pro-think), possibly implying their perspectives on abortion.

Pro-Elección

Defining and interpreting pro-elección.

Most participants mentioned a common definition of pro-elección: “el derecho a elegir” (the right to choose). Notably, some people focused their definitions on the concept of “elegir” (choose) or “decidir” (decide), referring for example, to a woman’s decision or to a couple’s decision to have or not to have an abortion. Others identified the term directly as synonymous with pro-abortion, defining it as “support for” or “being in favor of” abortion. Paula defined pro-elección as:

Para mí, pro-elección son para las personas que sí están de acuerdo con el aborto, aquellas que piensan de que sí están en su derecho de elegir porque es su cuerpo, es su vida, es su elección, tienen derecho de elegir eso que quieren. (Supplementary Appendix B: Quote 12)

Although less frequently, some participants defined pro-elección as supporting the killing of lives or “taking life away.” For instance, Rosa said, “si apoyo pro-elección es como que estoy de acuerdo, que …estaría apoyando que sigan asesinando vidas.” (Supplementary Appendix B: Quote 13). Additionally, many participants argued that being pro-choice entails support for choosing generally, in other words, it includes both support for choosing to continue with or terminate a pregnancy. Brenda argued, “Pro-elección serían las dos cosas. Porque si yo elijo la vida, también estoy eligiendo.” (Supplementary Appendix B: Quote 14).

In contrast with the definitions provided for pro-vida—which were associated almost exclusively with human life and abortion—participants defined pro-elección from a broad perspective that sometimes incurred interpretations of the term or part of the term beyond abortion. For example, some defined it as pre-elección (pre-choice), or making an anticipated choice, making plans before choosing something, or preparing to choose. Carmen, a 61-year-old woman from South America who was not familiar with the term, shared:

Pro-elección, pues me… tomar este, decisiones anticipadas, ¿Sería? Prepararse para tomar elecciones. No sé, podría... si va por ese lado, no sé. (Supplementary Appendix B: Quote 15)

Others interpreted pro-elección as supporting politicians or political elections or conceptualized the term as “being in favor of something” as if they were defining the prefix “pro-.” Julia, a 35-year-old woman from Mexico, conceptualized pro-elección as:

[Es] algo con lo que estás de acuerdo… Puede referirse a algo más, como a la elección de algún político, pro-elección de tal político, pro-elección de…Entonces está algo generalizado. No necesariamente tiene que ver con lo de pro-vida. (Supplementary Appendix B: Quote 16)

Perceptions about people who identify as pro-elección.

Participants expressed positive and negative perceptions of people who identify as pro-elección and the pro-elección movement. For example, positive perceptions that emerged characterized people as “smart,” “very sure and not afraid people,” “good decision-makers,” “careful thinkers before making choices,” and “those that care about other people making their own decisions.” While it was unclear whether some responses pertained to abortion, within the context of the interview, most participants referred to abortion. Others, like Yolanda, a 53-year-old woman from Mexico, unfamiliar with the term in relation to abortion, offered a broad perception of pro-elección people:

Una persona inteligente, calculadora, que sabe cómo vivir, que tiene buenas decisiones porque primero analiza lo que quiere y después lo elige, me imagino. (Supplementary Appendix B: Quote 17)

Other respondents offered perceptions that denoted negative emotions toward the term and, thus, undoubtedly to abortion. For example, some participants perceived pro-elección people as “nonspiritual,” “extreme,” “polarized towards choice,” “irresponsible,” “empty and immoral,” “selfish,” “superficial, materialistic, and cold” or people who “choose to kill someone.”

Rodrigo was explicit in self-identifying as pro-elección but did not want to be associated with “radical pro-elección” groups. He noted the polarization between pro-vida and pro-elección groups:

Pues, como que los grupos pro-elección, no sé, es que también creo que están muy polarizados, están totalmente del otro extremo… (Supplementary Appendix B: Quote 18)

Besides Rodrigo, very few participants explicitly mentioned the polarization and politicization of the pro-life/pro-choice movements in the U.S.

Familiarity, adequacy, and alternative terminology.

Two-thirds of interviewees were not entirely familiar with the term pro-elección as related to abortion. For example, Alfredo, a 62-year-old man from South America, indicated not being familiar with the term:

Yo creo que alguna cosa he escuchado, no lo recuerdo. Ahorita, para serle honesto, le diría que no. (Supplementary Appendix B: Quote 19)

Similar to pro-vida, some participants emphasized that pro-elección was new to them since it is U.S.-centric and was not commonly used or heard of in their country of origin regarding abortion.

Most participants (including those unfamiliar with the term but that received an explanation and contextualization as pertaining to the U.S. abortion debate), stated that pro-elección was an adequate term in Spanish, and an adequate translation of pro-choice. Lorena, a 49-year-old woman from Cuba, represented this view:

Está bien pro-elección, sí pro-elección, el derecho de la persona de elegir obviamente… en español sí se entiende, por supuesto, qué cosa es pro-elección y pro-vida, sí, se entiende perfectamente (Supplementary Appendix B: Quote 20)

However, in contrast with pro-vida, more participants offered alternative terms for pro-elección in Spanish, such as derecho a decidir (right to decide), pre-elección (pre-choice), pro-acceso [a la salud] (pro-access [to healthcare]), pro-decisión (pro-decision), and pro-muerte (pro-death), the latter also potentially implying their views on abortion.

Discussion

We used insights from survey data exploring differences in how English- and Spanish-speaking U.S. adults identify with pro-life/pro-vida and pro-choice/pro-elección to then qualitatively explore how a subsample of Spanish speakers understand and conceptualize pro-vida and pro-elección using cognitive-style interviews. This study exemplifies challenges associated with survey translation in multi-language public opinion research (Goerman et al., 2018) regarding abortion that involves cultural-specific references and terminology.

Familiarity and Language Proficiency

Approximately 40% of Spanish speakers were unfamiliar with pro-vida and pro-elección (compared with 7% of English speakers). This pattern corresponds with Spanish-speaking respondents being foreign-born. Indeed, some of our interviewees were unfamiliar with at least one of the terms (i.e., commonly pro-elección) before they were provided with an explanation. Some foreign-born interviewees acknowledged having only heard of the terms once they migrated to the U.S. This may explain why Gallup asks whether people identify as pro-life or pro-choice in English but then asks whether people oppose or support abortion in Spanish. However, given that our findings demonstrate a disconnect between participants’ self-identification with the terms and their general support/opposition to abortion, coupled with other work indicating that identification with these terms should not be conflated with attitudes toward abortion (e.g., Jozkowski et al., 2018), Gallup’s wording in Spanish as a proxy for identification with the terms in English likely contributes to a misleading interpretation of their data. The support/opposition item used by Gallup has utility, but we recommend that it not be used as a proxy for people’s self-identification with these socio-political abortion labels, as doing so may contribute to misleading and/or inaccurate conclusions about these different aspects of abortion attitudes and identification with labels. Instead, we recommend that researchers temper their language when using these labels to describe attitudes and be more precise in describing the constructs they are measuring (e.g., endorsement of abortion being legal; self-identification with abortion labels).

Approximately 18% of Spanish-speaking respondents chose prefer not to answer, compared with only 6% of English-speaking respondents. This might demonstrate a lack of familiarity with the terms compared with English-speaking respondents. Lack of familiarity with pro-vida and pro-elección might result from these terms being idiosyncratic to the U.S. socio-political landscape. As such, our findings underscore the need to collect information about migration or length of stay in the U.S. when assessing abortion attitudes. Since migrants might not have sufficient exposure to all aspects of U.S. politics, nor to the same cultural referents about abortion as non-migrants, failing to account for this information may lead to inaccurate or imprecise ways of asking questions, which in turn may lead to misleading findings. Concurrently assessing migration generation or length of stay in the U.S. could help researchers contextualize Spanish speakers’ familiarity with these terms when presenting their findings.

We also found that among Spanish-speaking respondents unfamiliar with pro-vida and pro-elección, 70% reported not speaking English at all, potentially indicating that monolingual individuals are less familiar with the U.S. linguistic and cultural contexts (Goerman et al., 2018). Thus, variation in English-language proficiency may also explain why Latinx populations have different conceptualizations of abortion-related terms. As such, some variability in abortion attitude response patterns may be an artifact of the interpretation of terms and not of actual attitudinal differences. It may also be important for researchers to assess the degree of English fluency and consider whether assessments of abortion attitude labels are appropriate for all segments of a study sample.

Literal Interpretations of the Terms

We also observed that among Spanish-speaking respondents who identified as completely pro-elección, approximately 47% opposed abortion, compared with only 6% of English-speaking respondents. Likewise, 37% of those who answered they were not at all pro-vida also opposed abortion, compared with 5% of English-speaking respondents. These results are not entirely surprising given the previously mentioned higher unfamiliarity with the terms among Spanish-speaking respondents. According to our qualitative data, some participants stated that pro-elección is an ambiguous term because it might entail either making a choice for “life” or making a choice for “abortion.” This conceptualization, while directly associated with abortion, suggests a more literal interpretation, potentially free of preconceived notions pro-choice has within the politicized abortion debate in the U.S. It is reasonable to think that for some Spanish speakers—particularly first-generation migrants—pro-elección is not as “charged” with contextual meanings and cultural connotations as pro-choice is in the U.S. abortion debate. In fact, Spanish speakers interpreted pro-elección in broad, flexible, or even literal ways. This is partly because the word “elección” (choice) is a polysemic noun that might take two different meanings: choice/decision or (political) elections.

Alternatively, “vida” uniquely translates to “life” in English, and participants’ definitions of pro-vida were often straightforward, disregarding awareness of the abortion term. However, some of our interviewees also revealed a broad and literal interpretation of pro-vida by envisioning a holistic conceptualization, which went beyond abortion (i.e., protecting human/animal life beyond birth and through the entire life course). This conceptualization is consistent with some pro-life organizations in the U.S., which ascribe to a “consistent life-ethic” philosophy regarding the protection of human and sometimes non-human life from “womb to tomb” (e.g., Rehumanize International). However, most mainstream pro-life organizations seem to focus predominantly on abortion. In our study, respondents who conceptualized pro-vida from a broad perspective were not sufficiently familiar with the term in the context of abortion or were familiar but consciously challenged its narrow focus. Notably, these interpretations of pro-vida mainly were limited to the notion of life versus death, in contrast with the interpretations that participants provided for the term pro-elección—(e.g., abortion vs. elections).

Finally, consistent with Solon et al. (2022b), we also found that some participants misinterpreted the prefix “pro-” as “pre-” to indicate “before [making a choice or decision]/[life]”; this occurred for both pro-vida and pro-elección. Collectively, these findings suggest that people unfamiliar in advance with the terms might be interpreting them literally when asked to define them, which may result in potential validity issues when using these terms in multi-language survey research.

Recommendations for Public Opinion Research on Abortion

When studying Spanish-speaking populations in the U.S., samples will comprise monolingual or predominantly Spanish-speaking first-generation migrants who may not be as familiar with U.S. socio-political culture as English speakers or English-dominant Spanish speakers (Goerman et al., 2018). Survey methodologists argue that public opinion research aiming to represent Spanish-speaking communities—and other non-English speakers—in the U.S. should purposely consider item wording to ensure comprehension by all respondents, disregarding their cultural origin and socioeconomic background (Warnecke et al., 1997). We provide several recommendations for survey researchers, polling organizations, and policymakers designing abortion-related surveys for Spanish-speaking respondents and discuss the benefits and limitations of each recommendation. Likewise, survey researchers may consider our results when examining other culturally contextual issues.

We first recommend that researchers deeply contemplate whether there is value in assessing whether people identify with these terms. Given the complexity associated with their interpretation, it is important to recognize the potential for misinterpretation. We argue that these terms should not be used as proxies for assessments of abortion attitudes, such as measuring support/opposition to abortion and whether abortion should be legal or illegal, because they represent a different facet of abortion attitudes and beliefs (LaRoche et al., 2024). However, there is value in assessing identification with these terms among English and Spanish speakers. Indeed, abortion attitudes—whether measured via endorsement of abortion legality, support, or opposition to upholding Roe v. Wade, or in association with specific gestational time points or fetal markers—are often differentially influenced by the abortion label people identify with; this is true for both English- and Spanish-speaking respondents (Crawford, Jozkowski, Turner, & Lo, 2021, 2022, 2024; Solon et al., 2022a). For instance, when assessing people’s attitudes toward upholding or overturning Roe v. Wade, those who identify as neither pro-life/pro-vida nor pro-choice/pro-elección responded in ways that were similar to those who identified as pro-choice/pro-elección (Crawford et al., 2022; Solon et al., 2022a). Alternatively, when assessing people’s attitudes toward fetal development-based bans on abortion, those who identify as neither pro-life/pro-vida nor pro-choice/pro-elección responded in ways that were similar to those who identified as pro-life/pro-vida (Crawford et al., 2021). As such, it is important to improve ways of measuring people’s identification with these labels, such as assessing people’s identification with these labels as separate from measures of abortion attitudes and by measuring identification with these labels via two different items as we did in the present study. Specifically, given that some people—both English and Spanish speakers—interpreted the terms as inclusive of both identities and identified with aspects of both terms, we recommend including separate items to assess the degree that people identify with each term separately over a single bi-directional survey item that asks identification within a continuum from pro-life/pro-vida to pro-choice/pro-elección. By asking separately, a more granular and nuanced response that seems to align with some people’s conceptualizations can be obtained.

Second, given the variability in interpretations of pro-vida and pro-elección, researchers could reference abortion in the item stem to signal to participants that the question pertains to abortion. We employed this approach in the present study and still experienced some interpretation issues. However, we maintain this approach is preferable to providing definitions of the terms, for example, because of the diverse conceptualizations people have of these terms in both English and Spanish that may introduce measurement error (Solon et al., 2022b).

Third, in considering the polysemic meaning of “elección,” researchers could also consider alternative Spanish terms that may culturally resonate better with different Spanish-speaking countries. This is challenging given that the Spanish language has multiple regional variations, and the abortion debate in each country has been evolving in reaction to complex social and political public debates generally linked to legislative action. As a recent example, in Argentina, activists in support of abortion changed the discourse by moving away from the moral/ethical dimensions in favor of the legal dimensions of abortion. They did so by systematically bringing the topic of legal abortion to the forefront of public opinion and media attention (Rosenberg, 2021). Unfortunately, given the diversity of migrants from Spanish-speaking countries to the U.S. and the cultural and linguistic nuances in Spanish-language usage, there are no apparent alternative terms that unequivocally mean the same for all foreign-born Spanish-speaking respondents (Valdez et al., 2022). Two of the most common alternatives to pro-elección recommended by our participants were “derecho a decidir” (right to decide) and “derecho a elegir” (right to choose). While “to choose” or “to decide” are not ideal, since they can still be interpreted broadly (e.g., applied to mask-wearing during COVID-19), similar to “pro-choice,” these alternatives remove the polysemic meaning of “elección” as related to political elections. In this vein, we underscore our recommendation to add the word ‘abortion’ to question stems to direct participants to think about how they identify with regard to abortion to potentially mitigate the divergent interpretations of “elección” or “vida.”

Fourth, and more generally, given the present results, when translating heavily cultural terms, translation teams should rely on recommended support procedures and systems (e.g., particularly translation notes) to ensure that the intended and most linguistically and culturally salient meaning is conveyed in the translated terms (Behr, & Scholz, 2011; Harkness, Pennell, & Schoua-Glusberg, 2004).

Fifth, beyond specific suggestions that pertain to abortion identity labels, when working with Spanish-speaking respondents, many of whom are foreign-born, we recommend collecting data on other individual characteristics, such as English-language proficiency, migrant generation, and length of stay in the U.S., since these characteristics may be associated with respondents’ familiarity and understanding of these terms, which can help researchers contextualize their findings.

Conclusion

Given that some of our participants were unfamiliar with the terms pro-vida and pro-elección, our findings suggest that the terms may be culturally ingrained and strongly linked to the social and political context of the abortion debate in the U.S. Although most Spanish speakers also interpreted the terms correctly, their use seemed less common, thus creating greater potential for different interpretations and misunderstandings. Different cultural backgrounds, values, norms, and even variations in the language of people from different Spanish-speaking countries have resulted in various approaches to understanding or identifying with these terms. We encourage public opinion researchers to carefully consider the rationale for specific survey questions and the corresponding translations used across multi-language studies. In the case of assessing Spanish speakers’ abortion attitudes in the U.S., we argue that researchers and pollsters should also account for migration status, length of residence, and language dominance. Such findings apply to cross-cultural research in a particular setting and translational research in multiple settings. In the first case, a common interpretation of a specific issue in a particular context for native speakers or native populations might not be interpreted the same way for non-native speakers and foreign-born populations. Some constructs may lead to different interpretations across national contexts, while others may not be directly applicable.

We must also acknowledge the contentious debate regarding abortion in different socio-political contexts, potentially exposing foreign-born populations in the U.S. to changing cultural references regarding abortion when they draw on experiences from their countries of origin. While our survey did not ask about years of residency in the U.S. or country of birth, which is a limitation of our study, given the potential influence of different socio-political contexts, future research could further explore the role of length of residence in the U.S. and country of origin on abortion attitudes and terminologies. The politicization of the abortion debate and its related terminology entails even further nuances. For example, the reproductive justice movement uses “anti-abortion” instead of “pro-life,” and there are some preferences for “reproductive justice” over “pro-choice.” In contrast, the pro-life movement attempts to use “pro-abortion” instead of “pro-choice.” This terminology is politicized in other countries as well. Therefore, there might be different cultural understandings of these terms, but also nuance when implementing cross-country survey research on the topic of abortion.

This research provides timely contributions to cross-cultural public opinion research as abortion will be on the ballot in many states in the future, and Latinx-eligible voters represent the largest group of minority voters in the U.S. Nevertheless, although we focused on abortion specifically for this study, we provide additional empirical evidence for the implications and importance of carefully conducting multi-language research in multicultural settings that might serve survey methodologists in multiple substantive research areas.

Funding

This work was supported by a grant from a confidential foundation. The funder played no part in the study design, the collection, analysis, or interpretation of data, the writing of the article, or in the decision to submit it for publication.

References

Biographical Notes

Bingbing Zhang (Ph.D., Pennsylvania State University) is an assistant professor at the School of Journalism and Mass Communication at the University of Iowa in the United States. Her research interests focus on political communication and media effects on individuals’ political beliefs, attitudes, and behaviors. She has published in JCR peer-reviewed journal articles such as The International Journal of Press/Politics, Information, Communication, and Society, and Journalism and Mass Communication Quarterly.

Araceli Mateos Díaz is a Full Professor at the University of Salamanca. She has a doctorate in Political Science and a Master’s degree in Latin American Studies and a degree in Sociology from the same institution. Her lines of research revolve around the study of parliamentary elites, political participation, culture and public opinion, and research methodology.

Homero Gil de Zúñiga holds a Ph.D. in Politics at Universidad Europea de Madrid and a Ph.D. in Mass Communication at the University of Wisconsin—Madison, serves as a Distinguished Research Professor at the University of Salamanca where he directs the Democracy Research Unit (DRU), as Professor at Pennsylvania State University, and as Senior Research Fellow at Universidad Diego Portales, Chile. His research addresses the influence of new technologies and digital media over people’s daily lives, as well as the effect of such use on the overall democratic process. He has published nearly a dozen books/volumes and over 100 JCR peer-reviewed journal articles (i.e., Journal of Communication, Journal of Computer-Mediated Communication, Political Communication, Human Communication Research, New Media & Society, Communication Research, etc).

Footnotes