The U.S. Supreme Court overturned the long-time landmark ruling of Roe v. Wade in 2022 summer. Debates and legal challenges regarding legal abortion in the U.S. have been heated (Felix et al., 2023). Furthermore, residents in several states have or will cast their vote on a ballot measure to determine abortion rights at the state level. A Gallup poll released in 2023 summer found that about one third of voters indicated that they would only vote for a candidate who shared their views on abortion (Saad, 2023). Therefore, it is imperative to understand people’s attitudes toward abortion. Past research on such attitudes have mainly focused on the role of political ideology and religious belief (e.g., Hess and Rueb, 2005); however, to our knowledge, relatively few studies have been done to examine the psychological underpinnings. Here we propose that examining the correlations between psychological factors and attitudes toward abortion has the potential to make contributions from the perspectives of both research and practice.

First, compared to attitudes in everyday life such as attitudes toward a product or brand, attitudes toward abortion are unique because it often elicits strong emotional response and conflict experience (Foster et al., 2012; Scott, 1989). Moreover, such an attitude goes beyond individual preference as it is deeply intertwined with one’s moral and religious beliefs, cultural background, and societal norms. Debate on abortion is not merely about a personal choice; it is about the definitions of life, rights, and autonomy (Osborne et al., 2022; Scott, 1989). For abortion, the contrasting views may lead to polarized opinions. In contrast, disagreements about a product or brand preference are typically less emotionally charged and do not carry the same societal weight. Therefore, given the unique nature of attitudes toward abortion as described above, it remains unclear whether psychological factors that correlate with attitudes in other areas still apply and, if so, in what capacity they do so. Additionally, as introduced below, several studies in this area employed a qualitative approach (interview). While the qualitative approach offered valuable insights into individuals’ perspectives on abortion, we aim to expand upon these findings by employing a quantitative approach. Especially, the quantitative approach allows us to explore the unique relationship between psychology and abortion attitudes after statistically controlling for other powerful factors like religious belief and political ideology. Together, a major goal of the present study is to provide initial empirical evidence for the correlations between attitudes toward abortion and certain psychological factors. We will further detail how our study might fill research gaps when introducing specific psychological factors as described below.

Second, examining the correlations between psychological factors and attitudes toward abortion may also offer practical insights. Consider the role of thinking style, for instance. The decision to pursue an abortion is imperative and often a prominently salient one, impacting not just the pregnant woman but also her family and extensive social network. Such a decision is complex and challenging due to intense feelings (e.g., conflict) and the balance between a woman’s bodily autonomy and fetal rights. From this viewpoint, there might be a correlation between attitudes toward abortion and one’s thinking style, especially their willingness to address complex and difficult issues. Past research has highlighted the connection between rational decision-making and the availability of relevant information (Shafir and LeBoeuf, 2002). Hence, to facilitate informed decisions, comprehensive knowledge about abortion is both essential and beneficial. The present study will examine the relationship between thinking style and abortion attitudes. Should a correlation be identified, our study would suggest individuals engage more deeply in critical thinking about the issues of abortion to enhance abortion-related education and informed decision-making.

Together, the present study aims to shed more light on the unique role of psychology in abortion attitudes, particularly in the presence of political ideology and religious belief. Specifically, we choose to examine the factors of empathy, locus of control, and thinking style (need for cognition) based on three considerations. Firstly, from a face validity perspective, the psychological constructs are predicted to exhibit a relationship with abortion attitudes. For example, the internal locus of control aligns well with the pro-choice mantra, ‘my body, my choice. Secondly, as detailed below, although these constructs have been explored in previous studies, they have only received limited attention and their relations with abortion attitudes remain inconclusive. Hence, our study aims to fill the gaps from past research by further clarifying their roles in attitudes toward abortion. Thirdly, research has indicated significant intersections between elements like cognitive style, empathy, and locus of control with various decisions, especially in health contexts (Marton et al., 2021; Pfattheicher et al., 2020; Xu and Cheng, 2021). These elements are tied to motivation, information analysis, and make trade-offs (Fischhoff and Broomell, 2020). Building on this, our study seeks to explore the applicability of these factors to the deeply sensitive and polarizing decision of abortion. On the other hand, it is worth noting that the psychological factors examined in our study are not exhaustive or driven by theoretical considerations. However, as mentioned in recent publications (Osborne et al., 2022; Valdez et al., 2022), past research on abortion attitudes with a psychological perspective is still limited. Therefore, our hope is that the present study could provide initial yet meaningful empirical evidence to exhibit the sophisticated role of psychology in attitudes toward abortion. We detail our rationales for each factor below.

Empathy

Empathy refers to a variety of cognitive and affective responses, including sharing and understanding, toward others’ experiences (Pfattheicher et al., 2020). Previous studies have demonstrated a positive association between empathy and prosocial behaviors, such as caring for others (Moudatsou et al., 2020; Klimecki et al., 2016), as well as a reduction in conflict and stigma (Batson et al., 1997; Klimecki, 2019). Recently, Pfattheicher et al. (2020) also demonstrated that inducing empathy for the vulnerable people could promote taking preventative measures during the Covid-19 pandemic. While researchers advocated for incorporating empathy into abortion-related mental health intervention (Brown et al., 2022), the role of empathy in attitudes toward abortion remains understudied. Hunt (2019) investigated the impact of empathy toward pregnant women by presenting testimonial videos in which a pregnant woman described the challenges she faced due to legal abortion restrictions in Arkansas. However, this manipulation did not significantly reduce participants’ support for the abortion restrictions. Research has found that people’s views on abortion tends to be stable over time (Jelen and Wilcox, 2003; Pew Research Center, 2022). Hence, a short video used in Hunt (2019) might not be able to change people’s long-held views on abortion. Instead, we here hypothesize that the pre-existing but not temporality induced empathy play a role in abortion attitudes.

Furthermore, in addition to the empathy toward pregnant woman, it is also reasonable to assume that (some) people may feel empathy toward the unborn. For instance, interviews with Protestant religious leaders exhibited empathy toward both pregnant women and unborn (Dozier et al., 2020). Embree (1998) asked participants to indicate their opinions when responding to different scenarios of abortion. As a result, the study found that 64% and 17% of participants showed a moderate and strong level of empathy for the unborn, respectively. Despite the informative findings, the relationship between attitudes toward abortion and empathy toward the unborn remains unclear, particularly when taking empathy toward pregnant woman and other factors (e.g., political ideology) into account.

Together, we raise three hypotheses regarding the role of empathy as shown below.

H1a: Empathy toward pregnant woman and unborn can coexist.

H1b: People’s empathy toward pregnant woman are positively related to the support toward abortion.

H1c: People’s empathy toward unborn are negatively related to the support toward abortion.

As empathy has been highlighted in the intervention process when dealing with abortion-related mental health issues (Brown et al., 2022; Whitaker et al., 2015), we hope our findings could generate implications for future research and practice.

Locus of control

Locus of control (LOC) refers to people’s beliefs regarding whether their life outcomes are controlled and determined by their own (internal LOC) or external resources (fate, chance and/or powerful people, external LOC) (Levenson, 1981). Before delving into details, it is important to note that the internal and external LOC refer to different dimensions and are not mutually exclusive (Levenson, 1981; Reknes et al., 2019). For example, a person’s success may be determined by both hardworking and support from others. Regarding abortion attitudes, Sundstrom et al. (2018) analyzed interview contents and found that some women’s thoughts on pregnancy and abortion aligned with an internal locus of control (e.g., “As women, we need to take control as much as possible of our reproductive health”), while others aligned with an external locus of control (e.g., “leave it in God’s hands…we’ll just play it by ear and if I get pregnant, I get pregnant”).

The findings from Sundstrom et al. (2018) were informative and consistent with common sense. For example, at face value level, the slogan of “my body my choice” well aligns with the concept of internal LOC. However, the role of internal LOC in abortion attitudes may be more complicated. That is, religious belief may complicate the association between internal LOC and abortion attitudes. Past studies, including a meta-analysis and a study with over 20,000 participants, found a positive relationship between internal LOC and religious belief (Coursey et al., 2013; Falkowski, 2000; Iles-Caven et al., 2020). As noted in these articles, there are similarities between internal LOC and religious belief. For instance, religious beliefs often provide individuals with a sense of meaning, purpose, and guidance in life. Meanwhile, people higher in internal LOC are more likely to report higher levels of existential well-being and purpose in life, which can be associated with religious belief and engagement (Kim-Prieto et al., 2005; Krause and Hayward, 2013). Thus, the relationship between internal LOC and religious belief may complicate how internal LOC is involved in the abortion attitudes. Sundstrom et al. (2018) used interviews to explore the role of LOC in thoughts about abortion. However, this method might not sufficiently differentiate the influence of religious beliefs. In this study, we adopt a quantitative approach, using a classical scale to measure LOC. We aim to empirically assess the relationship between internal LOC and attitudes toward abortion, especially when accounting for religious belief. Furthermore, considering that the relationship between internal LOC and abortion attitudes might be intertwined with religious beliefs, we refrain from positing a specific hypothesis at this point.

External LOC, on the other hand, does not appear to have a significant relationship with religious belief. Additionally, a few studies found that people higher in external LOC tended to attribute outcomes to external reasons (Falkowski, 2000; Reknes et al., 2019). Building on this concept, individuals with a higher external locus of control (LOC) may be more inclined to attribute pregnancy to external factors and place less emphasis on personal responsibility. Accordingly, we predict the hypothesis below.

H2: External LOC will be positively related to the support toward abortion.

Need for cognition

Based on face validity, thinking style might pertain to one’s perception of abortion. For instance, individuals who prioritize comprehensive and empirical data might arrive at a different conclusion than those who lean on personal stories and emotional narratives. A few studies have tapped into the relationship between thinking style and attitudes toward abortion. Valdez et al. (2022) conducted qualitative interviews on abortion and employed natural language processing techniques to analyze the interviews. The study identified analytical thinking, which involved considering abortion from multiple perspectives, had a negative relationship with the number of cognitive distortions (such as polarized and rigid thinking about abortion). However, such a finding conflicted with another study by Hill (2004) where the concept of cognitive complexity (thinking beyond surface-level observations) did not correlate with attitudes toward abortion. The inconsistency might be due to methodological issues. For example, the correlations described above in Valdez et al. (2022) were derived from a small sample consisting of 16 participants. A low reliability of the cognitive complexity scale used in Hill (2004) might (partly) address the non-significant relationship. Thus, the present study will utilize the Need for Cognition scale, a widely recognized and validated instrument that measures thinking style, to examine its correlation with attitudes toward abortion in a larger sample.

Need for cognition (NFC) pertains to the inclination to derive satisfaction from and actively participate in effortful thinking (Cacioppo et al., 1984). Consistent with its concept, past research demonstrated that NFC was positively correlated with information seeking (Verplanken et al., 1992), academic achievement (Richardson et al., 2012), and logical reasoning performance (Ding et al., 2020). As for attitudes toward abortion, we hypothesize the following.

H3: There will be a positive correlation between NFC and attitudes toward abortion.

Our prediction is based on two reasons. First, NFC drives individuals to actively seek and update information and knowledge. It was discovered that acquiring a deeper understanding of abortion correlated with increased support for it (Hunt, 2019; Mollen et al., 2018). Second and relatedly, NFC was found to be negatively associated with various stereotype memories and positively related to non-prejudicial social judgments (Crawford and Skowronski, 1998; Curşeu and de Jong, 2017).

Overview

In sum, the present study aims to provide empirical evidence for the association between attitudes toward abortion and psychology by examining and clarifying the role of empathy, locus of control, and need for cognition. Past research has repeatedly found the involvement of political ideology and religious belief in abortion attitudes (e.g., Hess and Rueb, 2005; Holman et al., 2020; Jelen, 2017; Osborne et al., 2022; Prusaczyk and Hodson, 2018). Given their powerful and robust effect, it is crucial to gather additional empirical evidence to elucidate the distinct contribution of psychology to attitudes toward abortion, while considering the influence of political ideology and religious beliefs. Additionally, when describing attitudes toward abortion, the dichotomization of “pro-choice” and “pro-life” have been widely used for decades. However, some studies have criticized that the dichotomization oversimplified attitudes toward abortion (Hunt, 2019; Osborne et al., 2022; Rye and Underhill, 2020). That is, people’s views on abortion vary across different scenarios and reasons. For instance, people showed less support toward abortion with elective reasons than with traumatic reasons (Hoffmann and Johnson, 2005). With confirmatory analysis, Osborne et al. (2022) derived two types of abortion: traumatic (e.g., pregnancy due to rape) vs. elective (e.g., the woman does not want the child anymore). Building on prior research, the current study aims exploring potential variations in attitudes across different abortion reasons. Furthermore, we also intend to examine whether the psychological factors described above have varying associations with different types of abortion.

Methods

Participants

The study was approved by IRB before data collection. Participants were recruited from Amazon Mechanical Turk (mTurk) on October 20th, 2022. To be eligible for the study, participants must be an adult, a U.S. citizen, and have an approval rating greater than 98% in mTurk. A total of 300 participants were enrolled into the study. Each participant received $3 for compensation. Six participants did not complete at least 80% of the items and were removed from the study. Thus, the effective sample size was 294. Demographics are presented in the Results section.

Materials and procedures

Participants took an online survey developed by Qualtrics. Our study did not set a specific time restriction. Across 294 participants, the average survey completion time was 682.8 s (SD = 286.6 s). The median completion time was 595.0 s (IQR = 344.8 s). The following questionnaires were completed.

Attitudes toward abortion

Hoffmann and Johnson (2005) and Osborne et al. (2022) analyzed attitudes toward abortion with six different scenarios (scenarios a-f below) that were measured by the U.S. General Social Survey. We further added an additional item regarding underage pregnancy for two reasons. First, compared to other Western industrialized nations, the U.S. has historically had a higher rate of underage pregnancies. Additionally, underage pregnant individuals tended to have a higher likelihood of seeking abortions compared to their older counterparts (Lantos et al., 2022; Kearney and Levine, 2012; Sedgh et al., 2015). Second, underage pregnancy is linked to various adverse outcomes, such as increased risk during childbirth, heightened stress and depression, disruptions in education, and financial challenges (Eliner et al., 2022; Hodgkinson et al., 2014; Kearney and Levine, 2012). Given the significance and prevalence of underage pregnancy, we chose to include it as a scenario to understand the public’s perception. Additionally, we understood that people might feel conflict or uncertain toward one or more scenarios. Hence, instead of using binary response (yes/no format) adopted in the U.S. General Social Survey, we employed a 1 to 7 Likert scale for each scenario, with a higher score indicating stronger support for a pregnant woman to obtain legal abortion.

The seven scenarios in the present study included: (a) there is a strong chance of serious defect in the baby; (b) the woman’s own health is seriously endangered by the pregnancy; (c) the woman became pregnant as a result of rape; (d) the woman is married and does not want any more children; (e) the family has a very low income and cannot afford any more children; (f) the woman is not married and does not want to marry the man; and (g) the woman is underage.

Empathy

Following the wording used to measure empathy in Pfattheicher et al. (2020), we developed six items to measure the empathy toward the pregnant woman and unborn or fetus, respectively. The scale of empathy toward pregnant woman included: (a) I am very concerned about the pregnant woman who may lose access to legal abortion; (b) I feel compassion for the pregnant women who may lose access to legal abortion; and (c) I am quite moved by the pregnant women who may lose access to legal abortion. The scale of empathy toward unborn included: (a) I am very concerned about the fetus or unborn child; (b) I feel compassion for the fetus or unborn child; and (c) I am quite moved by the fetus or unborn child. Participants rated each item on a five-point Likert scale, with 1 being strongly disagree and 5 being strongly agree. Thus, a higher score demonstrated stronger empathy toward the target. The Cronbach’s α for the scale of toward pregnant woman was 0.90 in the present study. The Cronbach’s α for the scale of toward unborn was 0.92.

Need for cognition

The need for cognition scale (NFC, Cacioppo et al., 1984) intends to measure the tendency to engage into deep thinking. It has 18 items, such as “I only think as hard as I have to” and “I find satisfaction in deliberating hard and for long hours”. Participants rated each item on a five-point Likert scale, with a higher score indicating a greater tendency to enjoy deep thinking. In the present study, the reliability of this scale was 0.93.

Locus of control

The present study adopted Levenson multidimensional locus of control scale (Levenson, 1981). Across 24 items, this scale measures three dimensions of locus of control: internality (sample item: Whether or not I get to be a leader depends mostly on my ability); powerful others (sample item: I feel like what happens in my life is mostly determined by powerful people); and chance (sample item: To a great extent my life is controlled by accidental happenings). In the present study, participants rated each item on a 1 to 6 Likert scale, with a higher score indicating a stronger belief that fate was controlled by self, powerful others, or chance. The Cronbach’s α for the subscales of internality, powerful others, and chance was 0.84, 0.91, and 0.93, respectively. As shown below, there was a high agreement between powerful others and chance subscales (r = 0.87, p < 0.001). Hence, we combined these two subscales to form an external locus of control composite.

Demographics

After completing the scales described above, participants were asked to report their demographic information including race, age, gender, education, annual household income, current relationship status, abortion experience, religious belief, and political ideology. Gender was coded with 1 = male, 2 = female, and 3 = other. Race was coded with 1 = White or Caucasian, 2 = Hispanic or Latinx, 3 = Black or African American, 4 = Asian or Asian American, and 5 = Other. Education was coded with six levels: 1 = Less than high school graduate, 2 = High school graduate or equivalent, 3 = Some college or associate degree, 4 = Bachelor’s degree, 5 = Master’s degree, 6 = Doctoral degree. Annual household income was categorized into 13 levels and ranged between under $9,999 and above $120,000 with increments of $9,999. Current relationship status was coded into six levels: 1 = single and not dating, 2 = single but in a relationship, 3 = married, 4 = divorced, 5 = widowed, 6 = other. For abortion experience participants were asked “For any reason, have you had an abortion?”. For this question, the answer was coded with 1 = yes and 2 = no.

Religious belief was measured with three items. The first item asked “How often do you attend religious services?” Participants selected one option out of the following: 1 = never, 2 = a few times per year, 3 = once a month, 4 = 2–3 times a month, 5 = once a week or more. The second item asked “How important is religion to you personally?” Participants rated this question on a five-point Likert point, with 5 being most important. The third question asked “How would you describe your religious denomination”. The options included 1 = Christian, 2 = Islam, 3 = Judaism, 4 = Buddhism, 5 = Hinduism, 6 = other or atheism. In the present study, the first two items were highly correlated (r = 0.77, p < 0.001). Following Hunt (2019), we combined the two items to form a general religiosity composite, with a higher score indicating a stronger religious belief.

Political ideology was measured with two items: (a) Generally, how would you describe your views on most social political issues (e.g., education, religious freedom, death penalty, gender issues, etc.)? and (b) Generally, how would you describe your views on most economic political issues (e.g., minimum wage, taxes, welfare programs, etc.)? Participants rated each item with a five-point Likert scale, with 1 = strongly conservative 2 = conservative 3 = moderate 4 = liberal 5 = strongly liberal. We found a strong correlation between the two political ideology items, r = 0.76, p < 0.001. Hence, we combined the two items to form a general political ideology composite.

Results

SPSS 24.0 was employed to perform all the analyses. Across 294 participants, age ranged from 21 to 79, with a mean of 40.4 and a standard deviation of 12.4. Table 1 displays the descriptive statistics for the variables of gender, race, education, annual household income, current relationship status, religious denomination, and abortion experience.

Table 1 Descriptive statistics for race, education, and income.
Full size table

Table 2 presents the descriptive statistics of attitudes toward abortion in different scenarios, religious belief, political ideology, and the scores of the psychological scales. Similar to the results obtained from the large-scale surveys in the U.S. and New Zealand (Osborne et al., 2022), the support toward abortion was strong (neutral = 4) across all scenarios.

Table 2 Descriptive statistics of attitudes toward abortion in different scenarios, religious belief, political ideology, and the scores of the psychological scales.
Full size table

To examine the structure of attitudes toward abortion in different scenarios, a Principal Component Analysis (PCA) with a Varimax orthogonal rotation was performed on all seven scenarios. With eigenvalue ≥ 1 as the threshold, two components were generated, accounting for 81.34% of the variability. Table 3 presents the PCA results. As shown, we obtained two distinct components. The first one included the scenarios of baby defection, pregnant woman’s health being endangered, pregnancy caused by rape, and underage pregnancy. The second component included the scenarios of not wanting the child, low income, and not wanting to marry. Such a differentiation between the two components was consistent with the notion in Osborne et al. (2022). Following this paper and the face validity of the scenarios, we labeled the two components traumatic abortion and elective abortion, respectively. Accordingly, we also computed a composite score for each component by averaging the corresponding items. In line with previous research (Hoffmann and Johnson, 2005), the support was significantly stronger toward the traumatic abortion (mean = 5.84, SD = 1.24) than the elective abortion (mean = 4.94, SD = 1.74), t(293) = 11.51, p < 0.001, Cohen’s d = 0.67.

Table 3 Principal component analysis for attitudes toward abortion across different scenarios.
Full size table

Table 4 presents the zero-order correlations between attitudes toward traumatic and elective abortions, demographics, and scores of the psychological factors. Consistent with the findings from past research (e.g., Hess and Rueb, 2005; Holman et al., 2020), a stronger religious belief was negatively related to the support toward both types of abortions. A stronger liberal ideology was positively related to the support toward both types of abortions. Additionally, empathy toward the pregnant woman was positively associated with the support toward both types of abortions whereas empathy toward unborn or fetus had an opposite effect. Based on the zero-order correlation, we did not find a significant relationship between internal locus of control and attitudes toward either type of abortion. The external locus of control (either powerful others or chance), on the other hand, was positively related to the support toward elective but not traumatic abortion. As there was a high agreement between the two external locus of control subscales (r = 0.87, p < 0.001), we formed a general external locus of control composite by averaging the two items in the following regressions. Finally, need for cognition was positively related to attitudes toward elective abortion but not traumatic abortion.

Table 4 Correlations between attitudes toward abortion, demographics, and psychological factors.
Full size table

While the zero-order correlations were informative, we were mindful that the Type I error might be greatly inflated due to a vast amount of repeated testing. Moreover, one goal of the study was to examine the role of psychological factors in the presence of religious belief and political ideology. Thus, we performed a hierarchical linear regression on each type of abortion, with age, gender, income, and education in the first block, religious belief and political ideology in the second block, and psychological factors in the third block. We separated the regression between the two types of abortion because the role of predictors might vary. This approach was also employed in Osborne et al. (2022). Table 5 exhibits the regression results.

Table 5 Hierarchical regressions on traumatic and elective abortions.
Full size table

As shown in Table 5, the demographic variables of age, gender, education, and income did not account for a significant portion of the variability in attitudes toward either type of abortion. The present study added to the literature that there might not necessarily be a difference in attitudes toward abortion between males and females (Bilewicz et al., 2017; Jelen and Wilcox, 1997). By contrast, in the second block, religious belief and political ideology collectively explained a sizable portion of the variability in attitudes toward both types of abortion. In block 3, in the presence of demographic variables including religious belief and political ideology, psychological factors could still account for a significant portion of the variability.

Looking at the individual psychological predictors (for more detailed interpretations please refer to the discussion part), consistent with our hypothesis, empathy toward the pregnant woman was positively associated with the support toward both types of abortion. By contrast, empathy toward the unborn or fetus was negatively associated the support toward abortion. For the factor of locus of control, the internal locus of control was not related to any type of abortion attitudes when zero-order correlation was used (Table 4); yet it was positively related to abortion attitudes after all other predictors were taken into account, indicating a suppressing effect. Upon further examination, we identified two suppressors: religious belief and empathy toward the unborn. After removing these two variables, internal locus of control was no longer significant. The observed pattern reflected our previous prediction, indicating that the role of internal locus of control could be complicated by religious beliefs. External locus of control, on the other hand, was positively correlated with the support toward elective abortion. Similarly, need for cognition (NFC) also had a positive relationship with the support toward elective abortion. Neither external locus of control nor NFC had a significant correlation with attituded toward traumatic abortion. Hence, our hypotheses regarding external locus of control and NFC were partially supported. We detailed out interpretation and discussion of the results below.

Discussion

The present study aimed to provide empirical evidence for the correlations between psychological factors and attitudes toward abortion. As introduced earlier, while it is common to find the involvement of psychology in everyday life attitudes and preferences, attitudes toward abortion are unique and drastically different. Given its unique nature, it lacks empirical evidence regarding whether psychological factors that interplay with attitudes in other areas still apply and, if so, in what capacity they do so. Past research has primarily focused on the role of religious belief and political ideology. Our study demonstrated a substantial involvement (R2 change = 0.27 and 0.24 for traumatic and elective abortion, respectively) of the psychological factors, after controlling for religious belief and political ideology. More importantly, these effects were comparable to the variability accounted for by religious belief and political ideology combined, particularly in the elective abortion category. The results highlighted the influential role of psychological factors in shaping attitudes toward abortion.

Additionally, past research has shown the interconnection between psychology and the public’s attitudes toward major societal events. For example, during the Covid-19 pandemic, while the perception of mask-wearing and/or social distancing was highly politicized, studies found that attitudes toward these preventative measures to be related to thinking style, self-control, numeracy, and working memory capacity (Steffen and Cheng, 2023; Xie et al., 2020; Xu and Cheng, 2021). In line with this, our study further underscored the significant influence of psychology on another pressing societal topic: abortion. In the sections below, we detail our findings and relevant implications. We are fully aware that our study was preliminary and hope it could serve as a starting point for future research and practice. We also acknowledge the limitations of our study and address them at the end.

Empathy

Some past studies on empathy and abortion only considered the empathy toward the pregnant woman (e.g., Brown et al., 2022; Homaifar et al., 2017; Hunt, 2019; Whitaker et al., 2015). The present study identified two types of empathy when dealing abortion: empathy toward the pregnant woman and empathy toward the unborn. In the presence of each other, we found that greater empathy toward the pregnant woman was associated with more support toward abortion, whereas greater empathy toward the unborn or fetus was associated with less support toward abortion. Such a pattern suggested that empathy might be a source of conflict feeling. That is, when considering abortion, concerns and care toward pregnant woman and unborn could coexist, potentially leading to conflict and dilemma when people thought about abortion. While the present study examined the public’s attitudes toward abortion with a diverse sample, pregnant women might have a similar pattern of empathy and hence feel conflict and dilemma when thinking about abortion. To cope with such a conflict, it might be beneficial for a counselor to acknowledge conflicting emotions that arise from empathizing with both the unborn and the pregnant individual. Moreover, the counselor could guide the client through the process of reconciling these emotions to alleviate feelings of isolation or confusion the client may experience. Future research in the realms of mental health and counseling should consider integrating these dual empathy perspectives and empirically assess the efficacy of such therapeutic interventions.

Additionally, Hunt (2019) did not find a significant influence of empathy on abortion attitudes change when participants were exposed to testimonial videos featuring pregnant women discussing the legal obstacles they faced. The disparity between Hunt’s (2019) findings and our own could potentially be attributed to the inherent stability and longstanding nature of abortion attitudes. Research has found that people’s views on abortion tends to be stable over time (Jelen and Wilcox, 2003; Pew Research Center, 2022). As a result, it is possible that pre-existing empathy, rather than empathy induced temporarily, was the factor correlated with individuals’ perception and consideration of abortion. Our findings were consistent with this possibility. Together, our findings supported H1a to H1c. Moreover, our study shed more light on empathy by showing its association with distinct views on abortion. The results suggest that future research could investigate how different types of empathy are formed and how they influence the shaping and persuasion of abortion attitudes.

Locus of control

Through qualitative interviews, Sundstrom et al. (2018) unveiled individual differences in the locus of control when discussing opinions on abortion. However, these interviews might not have fully captured the interplay between internal and external locus of control and other factors involved attitudes toward abortion. To fill the gap, our study employed a quantitative approach to delve deeper into how locus of control correlated with abortion attitudes. Consistent with Levenson (1981) and Reknes et al. (2019), we found that the constructs internal locus of control and external locus of control were differentiated but not unidimensional. For internal locus of control, interestingly, we found a suppressing effect. As discussed earlier, the role of internal locus of control in abortion attitudes might be complicated. That is, on the one hand, by face validity, the internal locus of control well aligned with the concept of “my body, my choice” (Sundstrom et al., 2018). On the other hand, in line with past research (Coursey et al., 2013; Falkowski, 2000; Iles-Caven et al., 2020), our study found that internal locus of control was positively related to religious belief. Furthermore, as shown in Table 4, internal locus of control was also positively related to the empathy toward the unborn, and such a relationship was significantly mediated by religious belief (mediation effect = 0.21, SE = 0.5, 95% CI = [0.13, 0.31]). Therefore, when using zero-order correlation, the effect of internal locus of control might be neutralized by the two opposite parts (“my body, my choice” vs. religious belief) discussed above. By contrast, in regression, the “my body, my choice” part stood out because the religiosity part was partialled out by the variables of religious belief and empathy toward the unborn.

In addition to internal locus of control, we also discovered that external locus of control was involved in abortion attitudes. Specifically, we found a positive relationship between external locus of control and support toward elective abortion (H2 was partially supported). Past research has found that locus of control is related to attribution (Falkowski, 2000; Reknes et al., 2019). Thus, our finding was in line with the notion that those with a greater level of external locus of control might be more likely to attribute unwanted pregnancy to external reasons (not personal responsibility), and hence showed more support toward abortion.

Our findings regarding locus of control suggest that individuals might simultaneously believe in personal autonomy (“my body, my choice”) while also feeling that certain life events, like unwanted pregnancies, are influenced by external factors beyond their control. This is particularly true when thinking about elective abortion. Education and counseling practices might be designed to reflect this duality. For example, materials and discussions could simultaneously emphasize the importance of personal choices and responsibilities, while also exploring societal, cultural, or circumstantial factors that might influence abortion decision. Incorporating both perspectives would allow to create a supportive environment where individuals feel seen and acknowledged in their complexities.

Need for cognition

As introduced earlier, past research on the relationship between thinking style and abortion attitudes was inconclusive. To clarify the relationship, the present study adopted the validated need for cognition scale. Need for cognition has demonstrated its involvement in consequential events, such as political elections and the adoption of preventive measures during the Covid-19 pandemic (Sohlberg, 2019; Xu and Cheng, 2021). In the present study, we discovered that need for cognition was positively related to the support toward elective abortion. Such a finding was consistent with the notion that need for cognition was negatively related to stereotypes (Crawford and Skowronski, 1998; Curşeu and de Jong, 2017). Additionally, as need for cognition drives individuals to seek and update knowledge, our result was also in line with the finding that gaining knowledge about abortion led to more positive view on abortion (Hunt, 2019; Mollen et al., 2018). Our study implied that future research could empirically evaluate if indeed abortion knowledge mediates the relationship between need for cognition and abortion attitudes.

It is worth noting that the present study also clarified the role of need for cognition in attitudes toward abortion by examining a potential artifact. Specifically, the observed positive relationship between need for cognition and support for abortion might be an artifact, given that liberal ideology is positively correlated with both abortion attitudes and need for cognition (Young et al., 2019). However, as shown in our regression, the relationship between need for cognition and elective abortion remained significant in the presence of other variables, including political ideology. Thus, the finding suggested that at least part of the relationship between need for cognition and attitude toward abortion was unique and not driven by political ideology.

Our findings related to need for cognition had an implication on abortion-related education. As discussed earlier, having adequate knowledge about abortion could facilitate the support for making informed decisions. As need for cognition was found to be related to openness and motivation to seek and update information (Russo et al., 2022), our finding suggested that cultivating willingness to engage into critical thinking might be beneficial for education on abortion and reproductive rights. While we are fully aware that correlation does not equate to causation, our study still offers a starting point for future research and practice on abortion-related education.

Traumatic abortion vs. elective abortion

While some researchers argued that the dichotomization of “pro-choice” and “pro-life” was oversimplified, to date, only two studies have empirically examined attitude variation between different abortion scenarios (Hoffmann and Johnson, 2005; Osborne et al., 2022). Both studies demonstrated that public views on abortion can be grouped into two categories: traumatic and elective. Our research not only replicated these findings but also introduced two significant advancements. First, we incorporated a scenario addressing underage pregnancy, given its high prevalence and significance. Secondly, instead of a binary response, we employed a 7-point Likert scale, allowing us to more accurately capture potential conflicting attitudes among participants.

Furthermore, our findings revealed that the roles of external locus of control and need for cognition varied in relation to attitudes toward the two types of abortion. Interestingly, we observed that neither of these variables significantly related to attitudes toward traumatic abortion, as indicated by both zero-order correlation and regression analyses. Conceptually, the scenarios of traumatic abortion (e.g., pregnancy caused by rape; mother life endangered) tend to be more extreme and emergent than the scenarios of elective abortion. Hence, there might be less room for psychological factors, such as thinking or attribution, to function in traumatic abortion than in elective abortion. Our interpretation was also consistent with the statistical pattern between the two abortions. That is, compared to elective abortion, the standard deviation of traumatic abortion was smaller. Additionally, there were more participants rated seven on the Likert scale in the scenarios of traumatic abortion (29.6%) than in the scenarios of elective abortion (18%). Despite the difference between the two types of abortion, it is essential to acknowledge that elective abortion does not imply a stress-free experience. Both traumatic and elective abortions involve significant levels of stress and emotional challenges. While traumatic abortion scenarios can be considered more extreme, it is crucial to recognize that individuals undergoing elective abortion may also experience considerable emotional distress.

Taken together, with concrete evidence, our study demonstrated that the public’s attitude toward abortion depended on abortion reasons. Our study also implied that future research should focus on attitudes toward specific abortion scenarios rather than a holistic concept of abortion. Furthermore, the differentiation between the traumatic and elective abortions suggested the limitation and potential ineffectiveness of one-size-fits-all legislative solutions. Given the varying and often conflicting attitudes that people harbor, it would be reasonable for legislative frameworks to be flexible, adaptive, and cognizant of the different circumstances surrounding abortion. This will not only be more reflective of public opinions but also more supportive of individuals who undergo different types of abortion experiences, each of which carries its own set of emotional and psychological challenges.

Expanding findings with a quantitative approach

Some past studies employed a qualitive approach when dealing with attitudes toward abortion (e.g., Dozier et al., 2020; Sundstrom et al., 2018; Valdez et al., 2022; Woodruff et al., 2018). These investigations have provided insights and served as inspirations for our own research. However, the relationship between abortion attitudes and pertinent factors may remain somewhat opaque. This is particularly true when considering the intricate interconnectedness among these factors. The present study demonstrated that findings from qualitative studies could be extended and enriched with a quantitative approach. For instance, we utilized quantitative scales to measure empathy toward the unborn —a variable that was previously identified through interviews in the study by Dozier et al. (2020). Moreover, we further exhibited the role of empathy toward the unborn when statistically controlled other variables, including empathy toward the pregnant. Similarly, the role of internal locus of control was revealed in interviews in Sundstrom et al. (2018). With validated scales, we exhibited the correlation with internal locus of control in both types of abortion. Furthermore, by detecting and interpreting a suppressing effect, we showed the interplay between internal locus of control, religious belief, and attitude toward abortion. Thus, our study implied that using quantitative scales and analyses was a viable approach to examine attitude toward abortion and could deepen the understanding of relevant factors.

Limitations and future directions

Despite the contributions, limitations should be acknowledged as well. First and foremost, we believe our study was still in the explorative stage. The specific psychological factors tested in the present study were not exhaustive and not theoretically driven. We hope the present study could provide initial empirical evidence to show the sophisticated role of psychology in attitudes toward abortion. Future studies could use a more theoretical driven approach to examine the specific psychological involvement in abortion attitudes. For example, given the correlation between need for cognition and attitudes toward abortion, future research could further elucidate the role of thinking style in attitudes toward abortion by incorporating the Dual-Process Theory (Evans, 2008). The Dual-Process Theory posits that humans have two distinct systems of information processing: System 1, which is intuitive, automatic, and fast; and System 2, which is deliberate, analytical, and slower. By examining the interplay between these two systems, researchers might gain insights into how intuitive emotional responses versus more deliberate cognitive analyses influence individuals’ attitudes toward abortion. For instance, are individuals who predominantly rely on System 1 more swayed by emotive narratives or imagery related to abortion?

Second, when analyzing and discussing the results, we proposed several possible underlying mechanisms that might elucidate the relationships observed. To illustrate, we employed the concept of attribution to shed light on the role of an external locus of control, positing that individuals with a strong external locus might attribute abortion decisions to external factors or circumstances rather than personal choices. Furthermore, we suggested that the observed positive relationship between the need for cognition and abortion attitudes might be mediated through abortion knowledge. This implies that individuals with a higher need for cognition could potentially seek out more information on abortion, leading to more informed attitudes. However, while these interpretations offer potential insights, we recognize their speculative nature. It’s crucial to emphasize that our proposed mechanisms require rigorous empirical testing for validation. For example, it would be of interest to test whether indeed, gaining various types of abortion knowledge improves views of abortion.

Third, as described above, we strived to show how our findings could be potentially used in abortion-related counseling. However, we acknowledge that our study is explorative but not counseling focused. Therefore, while we believe our findings offer meaningful implications, we caution against over-extrapolating their direct applicability to counseling contexts. Future research could delve into empirically investigating how psychological factors, such as varying empathy types and loci of control, could be utilized to alleviate negative feelings associated with abortion decisions. Additionally, understanding how various psychological factors interact with cultural and social norms could further help tailor counseling approaches.

Fourth, the present study did not include an attention check item. We believe the quality of our survey could have been improved had we included one or more attention check items. However, the reliabilities of our scales were relatively high (ranged from 0.84 to 0.93). Additionally, we also replicated some major findings from previous research (e.g., the associations between attitudes toward abortion and religious belief and political ideology). Thus, we believe that overall, inattention did not affect the quality of our data. Future online surveys could consider using attention check items for quality control.

In conclusion, the present study demonstrates the unique contribution of empathy, locus of control, and need for cognition to how people perceived abortion in different scenarios. The findings suggests that attitudes toward complex moral issues like abortion are shaped by individual psychological traits and cognitive needs, in addition to societal, religious, and cultural norms. Future research could use our study as a starting point to expand on these findings, exploring other psychological traits and cognitive processes that may similarly affect perceptions of abortion and other controversial subjects.