Psychological traits and public attitudes towards abortion: the role of empathy, locus of control, and need for cognition

The study investigates whether and how psychological traits uniquely relate to public attitudes toward abortion beyond well-established influences of religious belief and political ideology. The authors argue that abortion attitudes are distinct from everyday preferences due to their moral, cultural, and emotional complexity, often eliciting strong conflict. They hypothesize that individual differences in empathy (toward the pregnant woman and toward the unborn), locus of control (internal vs external), and thinking style (need for cognition) are associated with abortion attitudes. They also examine whether attitudes depend on specific abortion scenarios and replicate a two-factor structure of abortion attitudes—traumatic vs elective—assessing whether the psychological predictors differentially relate to these categories.

Prior work links abortion attitudes largely to religiosity and ideology, with mixed or limited evidence on psychological underpinnings. Empathy generally promotes prosociality and reduced stigma, and during COVID-19, empathy increased preventive behaviors. However, empathy-based media did not shift abortion attitudes in one study, potentially due to attitude stability. Qualitative research suggests empathy can be directed toward both pregnant women and the unborn. Locus of control has been qualitatively noted in abortion contexts, and is multidimensional (internality, powerful others, chance); internal locus may intersect with religiosity and moral beliefs. Need for cognition relates to deep thinking, reduced stereotyping, informed decision-making, and political behavior; its role in abortion attitudes has been unclear. Research also criticizes the oversimplified pro-choice/pro-life dichotomy, showing scenario-based variability and a two-type structure (traumatic vs elective).

Design: Cross-sectional online survey approved by IRB (IRB 23-0018). Participants: N = 294 U.S. adults recruited via Amazon Mechanical Turk on October 20, 2022; eligibility: age 18+, U.S. citizen, ≥98% MTurk approval. Compensation: $3. Exclusions: 6 of 300 for <80% completion. Mean age 40.4 (SD 12.4), demographics reported. Survey administration via Qualtrics; average completion time 682.8 s (SD 286.6), median 595.0 s. Measures: - Abortion attitudes: Seven scenarios rated on 1–7 Likert (higher = more support): (a) serious defect in baby; (b) woman’s health endangered; (c) pregnancy due to rape; (d) married, does not want more children; (e) very low income; (f) not married and does not want to marry; (g) underage pregnancy. The underage item was added to reflect prevalence and associated risks. - Empathy: Six items adapted from Pfattheicher et al. (2020), three each for empathy toward pregnant women (α = 0.90) and toward the unborn/fetus (α = 0.92), rated 1–5 (higher = more empathy). - Need for Cognition (NFC): 18-item scale (Cacioppo et al., 1984), 1–5 Likert, higher = greater tendency to enjoy effortful thinking; α = 0.93. - Locus of Control: Levenson’s multidimensional scale (24 items, 1–6 Likert): internality (α = 0.84), powerful others (α = 0.91), chance (α = 0.93). Due to high correlation between powerful others and chance (r = 0.87, p < 0.001), these were averaged to form an external locus composite. - Religiosity: Two-item composite (attendance; personal importance), r = 0.77, p < 0.001; higher = stronger religiosity. Denomination was also recorded. - Political ideology: Two items (social and economic), 1 = strongly conservative to 5 = strongly liberal; r = 0.76, p < 0.001; averaged as a composite. - Demographics: race, age, gender (1 male, 2 female, 3 other), education (6 levels), income (13 brackets), relationship status (6 levels), abortion experience (yes/no; collected among females), religious denomination. Analytic strategy: Principal Component Analysis (PCA) with Varimax rotation on seven abortion scenarios (eigenvalue ≥ 1 criterion) to identify components and derive composite scores: traumatic vs elective. Descriptives reported. Zero-order correlations among key variables. Hierarchical linear regressions predicting traumatic and elective abortion support separately: Block 1—age, gender, income, education; Block 2—religious belief, political ideology; Block 3—psychological factors (empathy toward woman and unborn, internal and external locus of control, need for cognition). Suppression effects examined for internal locus of control.

- PCA identified two components accounting for 81.34% variance (eigenvalues 4.64 and 1.06): traumatic abortion (defect, woman’s health endangered, rape, underage) and elective abortion (not wanting a child, low income, not marrying). Support was higher for traumatic (M = 5.84, SD = 1.24) than elective (M = 4.94, SD = 1.74); t(293) = 11.51, p < 0.001, d = 0.67. - Descriptives (Table 2): Mean support per scenario ranged from 4.84 (not married) to 6.02 (rape). Religiosity M = 2.85 (SD 1.40); political ideology M = 3.53 (SD 1.18); empathy toward women M = 3.96 (SD 1.05); empathy toward unborn M = 3.44 (SD 1.20); internal LOC M = 4.59 (SD 1.11); external LOC components: powerful others M = 3.85 (SD 1.21), chance M = 3.76 (SD 0.87); NFC M = 3.21 (SD 0.87). - Zero-order correlations (Table 4): Religiosity negatively related to support for both traumatic and elective abortions; liberal ideology positively related to both. Empathy toward pregnant women positively, empathy toward unborn negatively, related to both. External locus (powerful others, chance) and NFC positively related to elective, but not traumatic, support. Internal locus showed no zero-order relation with either type. - Hierarchical regressions (Table 5): • Block 1 (demographics) explained little variance (R² change = 0.02 traumatic; 0.01 elective). • Block 2 (religiosity, ideology) explained substantial variance (R² change = 0.33 traumatic, p < 0.01; 0.25 elective, p < 0.05). Coefficients: religiosity negative (B = −0.32 traumatic**, −0.22 elective*), ideology positive (B = 0.45 traumatic**, 0.66 elective**). • Block 3 (psychological factors) added significant variance (R² change = 0.27 traumatic**, 0.24 elective*). Key predictors: empathy toward women positive (B = 0.66 traumatic**, 0.67 elective**); empathy toward unborn negative (B = −0.13 traumatic*, −0.37 elective**); internal LOC positive (B = 0.24 traumatic**, 0.32 elective**), with evidence of suppression by religiosity and empathy toward unborn; external LOC positive for elective only (B = 0.34*, ns for traumatic); NFC positive for elective (B = 0.22*), ns for traumatic. In Block 3, religiosity remained negative for traumatic (B = −0.20*), ideology remained positive for both (B = 0.15 traumatic**, 0.26 elective**).

Findings show that psychological traits contribute uniquely and substantially to abortion attitudes beyond religiosity and political ideology. Empathy operates bidirectionally depending on its target: toward pregnant women increases support, whereas toward the unborn decreases support, revealing a potential source of internal conflict. Internal locus of control relates positively to support once confounds are controlled, consistent with autonomy framing, while its zero-order null relation reflects suppression by religiosity and empathy toward the unborn. External locus of control and need for cognition are selectively associated with support for elective abortions, suggesting that attributional tendencies and deliberative thinking matter more in less extreme, less emergent scenarios, whereas traumatic scenarios elicit broad support with less variability for these traits to explain. The comparable variance explained by psychological factors to ideology/religiosity, especially for elective abortion, underscores their relevance for understanding nuanced public opinion and informs educational and counseling practices that acknowledge dual empathies, autonomy, and contextual attributions.

The study demonstrates that empathy (toward both pregnant women and the unborn), locus of control (internal and external), and need for cognition make distinct contributions to public abortion attitudes and vary by scenario type (traumatic vs elective). Psychological traits explain meaningful variance beyond religiosity and political ideology, indicating that attitudes toward complex moral issues are shaped by individual differences in cognition and emotion as well as sociocultural factors. Future research should employ theory-driven frameworks (e.g., dual-process models), test proposed mechanisms (e.g., knowledge as mediator for need for cognition; attributional pathways for external locus), and evaluate applications in education and counseling that address conflicting empathies and autonomy-externality dynamics.

The study is exploratory rather than theory-driven and does not exhaustively assess psychological factors. Proposed mechanisms (e.g., attribution processes for external locus; abortion knowledge mediating need for cognition effects) are speculative and require empirical testing. Although counseling implications are discussed, the study is not designed for counseling outcomes, limiting direct application. No attention check items were included; despite high internal reliabilities and replication of known associations with religiosity and ideology, future online surveys should include attention checks.