Moral Foundations Theory (MFT) was developed with the intention of understanding variations in moral judgments across cultures, proposing five universal moral concerns, including “care/harm, fairness/cheating, loyalty/betrayal, authority/subversion, and purity/degradation.” The Moral Foundations Questionnaire (MFQ) was developed to measure these five foundations. Despite the extensive reference to MFT, and widespread use of the MFQ, critics have requested theoretical refinement and improvement of its psychometric properties. In a mega-study published in Journal of Personality and Social Psychology, Mohammad Atari and colleagues responded to this call.
“My main goal was to go beyond WEIRD [Western, Educated, Industrialized, Rich, and Democratic] morality and develop a measurement tool based on a genuinely diverse sample,” said study author Mohammad Atari (@MohammadAtari90), an assistant professor in the Department of Psychological and Brain Sciences at the University of Massachusetts Amherst.
MFT originally proposed five psychological foundations upon which cultures construct their moralities, undergirded by research in evolutionary and cultural psychology, and anthropology. These foundations were categorized into individualizing foundations (care and fairness) focusing on the protection of individuals, and binding foundations (loyalty, authority, and purity) centered around group cohesion and societal structures. The initial version of the MFQ (i.e., MFQ-1) was designed to measure the degree to which individuals endorse these five areas of morality.
MFQ-1’s extensive use has revealed that various moral foundations correlate with specific attitudes and behaviors, such as empathy, social justice, national security, tradition, and religious attendance. These foundations have further been linked to political ideology, emotional reactions, and other societal attitudes and behaviors. However, the application of MFT and MFQ-1 have also exposed gaps in the theory and measurement. For example, critics have pointed out the omission of moral concerns such as equality,and the inadequacy of MFT in addressing societal inequalities. The MFQ-2 was developed to address these gaps.
The goal of Study 1a was to develop a preliminary item pool for the MFQ-2, based on a top-down structure including six foundations: care, equality, proportionality, loyalty, authority, and purity. The team ensured cultural diversity and avoided Eurocentrism in item creation. A total of 840 participants from the United States and India were included. Participants first completed the MFQ-2 item pool, followed by the MFQ-1, and provided demographic details. The MFQ-2 item pool initially had over 100 items in a declarative format. The item pool was narrowed down for further studies.
Studies 1b and 1c further refined the MFQ-2 item pool. In Study 1b, 90 items were administered to participants from India, United States, and Iran, with the aim of fine-tuning the pool to a more focused set of items. A total of 71 items were retained. Study 1c aimed to finalize the item pool, administering 71 items to participants from the United States, Ecuador, and China. The final MFQ-2 had 50 candidate items.
The final 50-item pool from Study 1 was administered in Study 2. A total of 3902 participants from 19 different nations were recruited. This study had a goal of finalizing the MFQ-2 into a 36-item questionnaire. Exploratory Structural Equation Models were used to finalize the items, leading to the exclusion of 14 items. Reliability across nations was high, with omega coefficients ranging between .73 and .95. The study further revealed that the MFQ-2 had good structural validity across populations.
The researchers found significant cross-cultural differences in moral foundations. Purity was most strongly related to cultural variance, with participants from less WEIRD societies endorsing it more. Gender differences were also observed, with women scoring higher on equality and purity, and men scoring higher on proportionality, loyalty, and authority. Religious affiliation and religiosity showed significant relationships with different moral foundations.
Study 3 aimed to establish the convergence of MFQ-2 scores with MFQ-1 and to examine its capacity to predict criterion variables (e.g., psychopathy, social dominance orientation, disgust sensitivity, among others). This study involved 1410 participants from the United States, India, and Canada. The results showed strong correlations between MFQ-2 foundation scores and their counterparts in MFQ-1. Further, MFQ-2 demonstrated substantive associations with criterion variables and had greater predictive power than MFQ-1 in predicting related psychological variables. The addition of MFQ-2 scores improved the variance explained in outcome measures by an average of 13.7%.
There are several takeaways from this work. First, fairness in MFT was refined into two distinct foundations, including equality and proportionality. This allows for a better understanding of fairness-related norms and behaviors across cultures. The MFQ-2 demonstrated superior psychometric properties compared to MFQ-1, proving more effective in capturing morality cross-culturally. It also outperformed MFQ-1 in predicting various psychological outcomes.
The present findings support a pluralistic view of morality, showing that moral values and their importance greatly vary across cultures. The researchers discovered that moral foundations were not universal, but culture-dependent. There were notable differences in moral values based on gender, religion and political ideology. For example, women placed more importance on equality and purity compared to men, religious individuals scored higher on loyalty, authority, purity and equality, and conservatives scored higher on loyalty, authority, purity, and proportionality, while liberals valued care and equality to a greater extent.
A limitation the researchers note is that despite efforts to collect data from 25 different populations and seven languages, “the present results are still based on a subset of these populations who were educated enough to complete the surveys online.”
Overall, Atari shared that “[This work] suggests a lot of diversity in moral judgments around the globe. People’s endorsements of moral values vary across cultures and the nature of morality might vary from one culture to another.”
The study, “Morality Beyond the WEIRD: How the Nomological Network of Morality Varies Across Cultures”, was authored by Mohammad Atari, Jonathan Haidt, Jesse Graham, Sena Koleva, Sean T. Stevens, and Morteza Dehghani.