Individual and School Correlates of DIT-2 Scores Using a Multilevel Modeling and Data Mining Analysis

Abstract

Moral reasoning was investigated with respect to individual characteristics (i.e., education level, political orientation and sex) and school-related (i.e., university/college) factors using multilevel modeling and data mining analysis. We used the multilevel modeling to detect school effects on moral reasoning as well as individual effects for 16,334 students representing 79 different higher education institutions across the U.S. The school-related factors, such as the racial composition, student–faculty ratio, average SAT score, institution type, institutions’ geographical region, frequencies of morally relevant words in college course catalog, college mission and value statements were collected through website searches. Data mining analysis was utilized to extract and calculate the frequencies of morally relevant words from the website content. There were significant effects for the individual characteristic of political orientation. Additionally, all school-related factors were significant. Only main effects were observed for some school-related factors (i.e., average SAT score, institution type, frequency of morally relevant words in mission statements, value statements and course catalogs). For other school-related factors (i.e., the region, student–faculty ratio and racial composition), main effects were also observed; however, these effects were particularly illuminating given their interactions with political orientation. Implications for educational communities are discussed.

1. Introduction

Higher education is the focus of much moral judgment development research, highlighting the link between more advanced moral judgment development and college attendance [1,2,3,4,5]. Before discussing moral judgment development in relation to higher education, it is prudent to discuss the nature of this construct and how it is measured.

1.1. Defining and Measuring Moral Judgment Development

Ref. [6] proposed a developmental stage view of moral judgment. According to [6], moral judgment develops in discrete stages, wherein individuals complete one stage before moving on to the next (much like a staircase). Ref. [6] created the Moral Judgment Interview (MJI) to measure moral judgment development. In the MJI, participants are prompted to articulate their moral reasoning or the process by which they decide which course of action the protagonist in the moral dilemma ought to take.

An extensive coding procedure is used to determine the developmental stage at which the interviewee is operating. Though revolutionary in his approach to studying moral development, Ref. [6]’s theory and methods are not without their limitations.

Ref. [4] and colleagues [7,8,9] advanced a Neo-Kohlbergian theory of moral judgment development that has its roots in [6]’s work; however, this Neo-Kohlbergian view differs in important ways from [6]’s view. For example, Refs. [7,8,9] rejected the staircase view of moral judgment development, instead endorsing a schemata view, which posits that an individual experiences shifts in the frequency with which they use moral judgment developmental schemata. Thus, on the schemata view, individuals can use more than one way of thinking about moral issues at a time, which contradicts the [6] staircase view.

According to [7,8,9], there are three primary moral judgment developmental schemata: personal interests, maintaining norms and postconventional reasoning. The personal interests schema is the least developmentally advanced of the three schemata, and reasoning at this level focuses on the self and those close to the self. The maintaining norms schema is more developmentally advanced compared to the personal interests schema, taking a society-wide perspective; reasoning at this level focuses on authority, rules, roles and the generally established way of life for a community in which the individual is a member. The postconventional schema is the most developmentally advanced, with reasoning at this level focusing on shared ideals that are open to scrutiny and deliberation, as opposed to simply relying on established way.

Moral judgment development, as defined in the Neo-Kohlbergian perspective [7,8,9], can be measured using the Defining Issues Test (DIT and the subsequent DIT-2; [7,8,9]). Unlike the MJI, the DIT does not require respondents to articulate their moral reasoning; rather, respondents are presented with moral dilemmas for which they must evaluate various statements’ importance regarding the course of action that the protagonist in the scenario should follow. These statements that respondents evaluate are representative of the three moral judgment developmental schemata described above. Measuring moral judgment development using recognition rather than production/articulation, as [6] did, allows researchers to detect a wider range of developmental reasoning levels and not simply the less developmentally advanced ones. (The Materials and Methods section provides additional information about the DIT).

1.2. Moral Judgment Development and Higher Education

The unique role that higher education plays in moral reasoning development is underscored by the findings of longitudinal studies [5]. For example, Ref. [5] conducted a longitudinal study measuring moral reasoning in participants during their senior of high school and then two, four and six years thereafter. The results indicated greater gains in moral reasoning for participants with three or more years of college experience compared to those with two or less years.

Nevertheless, Ref. [10] note that although research supports the advancement of moral judgment during the college years, it is interesting how little is known about the specific mechanisms responsible for this growth. A significant body of research has accumulated, however, pointing to relationships between moral reasoning development and various school-related factors [1,2,3,10,11,12,13,14,15,16,17,18]. Two factors that researchers have explicitly linked to moral judgment development are the institution type and the geographical region of institution.

Institution type. Institution type is a commonly studied school-related factor in investigations of moral reasoning. Based on a large and comprehensive meta-analysis, Ref. [1] concluded that liberal arts institutions were the only type reliably demonstrating a large effect size concerning undergraduates’ moral reasoning. Corroborating [1]’s findings, Ref. [13] observed higher moral reasoning scores in first-year liberal arts students compared to their regional or community college counterparts; the moral reasoning of first-year students at research institutions was also significantly higher than that of first-year regional or community college students.

Mayhew suggested that students who attend liberal arts and research institutions could have higher moral reasoning from the outset and thus choose colleges that they feel express a commitment to holistic education. Additionally, Ref. [11] observed significant gains in moral reasoning for freshmen and seniors attending a state university and a Christian liberal arts university, however, not for those attending a Bible university. Regarding Bible university students, Ref. [11] suggest that although Evangelical students may be capable of more advanced moral reasoning, “they deliberately set aside their own notions of fairness and endorse lower stage level DIT items out of conformity to their religious beliefs” (p. 275).

Region. Institutions’ geographical region also relates to students’ moral judgment development [12,19]. Ref. [12] found that an institution’s region relates to the average moral judgment development at that institution. Moreover, individual moral reasoning scores fluctuate based on region, as well as the institution’s political climate.

There are other factors that have yet to be explicitly linked to moral judgment development; however, literature suggests such relationships nonetheless. These factors include institutions’ racial composition and student–faculty ratio, as well as the institutions’ emphasis on moral and ethical issues in its mission and value statements and its curriculum.

Racial composition. According to [1], characteristics of the collegiate environment, such as diverse peer groups, are particularly illuminating in understanding moral judgment development. Moreover, Ref. [1] highlight the relatively higher importance of the unique social climate typically associated with the college years compared to academic involvement. Similarly, Ref. [10] note the significance of low-density friendship networks, or networks comprising multiple independent friend groups. Low-density networks are associated with moral judgment development presumably because being surrounded by different types of friends provides exposure to different ideas and viewpoints.

Interestingly, Ref. [16] found that students who experienced more negative interactions with diverse peers reported less advanced moral reasoning development. Perhaps students who experience such negative interactions avoid working through the resultant cognitive disequilibrium and instead rely on thinking patterns to which they have grown accustomed [16]. We acknowledge that institutions’ racial diversity does not guarantee positive or negative interactions among students. However, we assume that racial diversity can greatly increase the possibility of such interactions.

Student–faculty ratio. Class size, and therefore student–faculty ratio, is important to the extent that it affects limitations imposed and opportunities afforded by the classroom environment. For example, Ref. [15] found that the pedagogical practices of (1) discussions about diversity and (2) opportunity for reflection both were significantly related to social justice outcomes; therefore, it seems plausible that such practices would be employed more effectively with smaller class sizes. Additionally, Ref. [18] note the significance of high-quality faculty interactions. A lower student–faculty ratio may help facilitate such interactions.

Frequency of morally relevant words in college mission and value statements. According to [20], mission statements are essential to college communities because they outline major tenets of colleges’ structure and culture. Furthermore, mission statements describe colleges’ plans to set and achieve particular goals. Value statements provide similar guidelines for colleges’ directions; therefore, this study considers value statements as extensions of mission statements.

Skeptics may argue that mission statements are simply a rhetorical strategy to make institutions seem particularly impressive [21,22]; however, others uphold mission statements’ legitimacy [23,24]. For example, Ref. [24] asserted that mission statements are a reliable means of differentiating among schools, as they found schools to vary in predictable ways based on mission statement content. Additionally, Ref. [23] concluded from their analysis of college mission statements that they are more than a formality.

First, Ref. [23] found a lack of aspirational elements in mission statements, such as indicating a desire to be the best. Second, similar to [24]’s findings, Ref. [23] found mission statements of a particular institution type (public vs. private) to share certain elements; as such, rather than simply being a talking point or a formality, mission statements address the distinct challenges encountered by different types of institutions. Accordingly, it is plausible that the extent to which an institution’s mission statement and values emphasize morality and ethics could positively affect moral reasoning at that institution.

1.3. Frequency of Morally Relevant Words in College Course Catalog

The college experience promotes moral judgment development [1,4,5]. Individuals need not undergo formal moral education programs or interventions to develop their moral reasoning [1,4]; however, evidence suggests that these are beneficial [25,26,27]. For example, Ref. [25] found that exposure to Kohlberg’s developmental stage theory of moral reasoning promoted students’ moral reasoning development. Additionally, Ref. [27] asserts that a curriculum specifically designed to include ethics training is beneficial for professional students. For example, Ref. [26] examined the effects of an intensive six-week ethics course for physical therapy students and found significant increases in postconventional reasoning. Given these findings, it seems reasonable to suggest that as institutions place greater priority on ethics and morally relevant content, this will positively impact students’ moral judgment development.

Average SAT score. SAT scores are positively correlate with moral reasoning; however, this relationship has been established at the student level not at the institutional level [13,14,18].

1.4. Individual Factors Associated with Moral Judgment Development

Ref. [28] note that, in addition to considering experiences (e.g., college education) when investigating moral judgment development, it is also important to consider individual factors. Two students may experience the same event; however, it is unlikely to carry the same meaning for both. For this reason, we now turn to the body of literature that supports a relationship between moral reasoning development and individual factors [1,5,7,8,12,29,30,31,32,33,34].

Education level. A primary indicator of validity for the Defining Issues Test (DIT-1 and subsequent DIT-2; [7,8]), a prominent test of moral judgment development, is differences in scores based on education level [7,8,29]. Overall, the research supports that formal education positively impacts moral judgment development [1,4,5,29].

Political orientation. Much research supports the notion that political orientation and moral reasoning are distinct yet related constructs [31,32].

Sex. According to the results of two meta-analyses, female undergraduate students demonstrate higher moral reasoning levels compared to their male counterparts [33,34].

As previously stated, research has long established the role of higher education when it comes to moral judgment development [1,2,3,5,10,11,12,13,14,15,16,17,18]. However, few studies have employed multilevel modeling techniques to explore moral reasoning scores across academic institutions while accounting for school and individual factors [12,13]. Accounting for both individual factors and wider contextual factors related to institutions of higher education, clarifies variability in moral judgment development scores [12,13].

The present study adds to the existing literature by investigating the aforementioned factor types using multilevel modeling techniques. Additionally, the present study uniquely contributes to the literature by expanding the consideration of school factors to include racial composition, student–faculty ratio, average SAT score and morally relevant words in college mission and value statements as well as morally relevant words in the course catalog.

2. Materials and Methods

2.1. Participants

Participants included 16,334 students representing 79 different higher education institutions across the U.S. The sample comprised 6478 freshmen, 1304 sophomores, 2018 juniors and 6520 seniors. The mean age was 21.23 years. There were 7923 male and 8904 female participants. We used the datasets provided by a third-party research center, the Center for the Study of Ethical Development at the University of Alabama.

The datasets were collected by more than 300 researchers who requested use of the Defining Issues Test (DIT-2) from the Center for the Study of Ethical Development in 2010–2014. The researchers were not able to score and to interpret the test results and use of the DIT-2 is provided only with the Center’s scoring service. The Center for the Study of Ethical Development retains the right to use the datasets, which the Center makes available upon request. IRB approval was not obtained for the present study because of the secondary nature of the data.

2.2. Materials

Defining Issues Test (DIT-2). The DIT-2 [7,8] measured moral judgment development. Respondents read about five separate moral dilemmas. The five dilemmas or stories were labeled “Famine”, “Reporter”, “School Board”, “Cancer” and “Demonstration”. For example, the story about Famine is that a father contemplates stealing food for his starving family from the warehouse of a rich man hoarding food. The Cancer story is about a doctor who must decide whether to give an overdose of pain-killer to a suffering but frail patient (see https://ethicaldevelopment.ua.edu/about-the-dit.html, accessed on 15 March 2022).

An example question for the Cancer story is “Isn’t the doctor obligated by the same laws as everybody else if giving an overdose would be the same as killing her?” For each dilemma, respondents indicated the choice that they believed the protagonist of the dilemma should make. Next, respondents rated the importance of 12 issue statements related to the action previously presented. Finally, respondents ranked the four most important issue statements.

According to [7,8], the issue statements and the moral dilemmas activate moral judgment developmental schemata. If the respondent reads an issue statement that makes sense and activates a preferred schema, then they assign a high rating/ranking; in contrast, if the respondent reads an issue statement that does not make sense and does not activate a preferred schema, then they assign a low rating, the issue statement is less likely to be ranked among the top four most important statements.

Data from the ratings and rankings are used to calculate the N2 score, which is intended to capture an individual’s emphasis on postconventional reasoning and deemphasis of reasoning based on personal interests [29]. Higher N2 scores thus indicate more advanced moral reasoning. N2 scores were analyzed in the present study. Cronbach’s alphas for the DIT-2 have ranged between the upper 0.70 s and the lower 0.80 s [7,8]. One DIT-2 subscale was used to index individuals’ political orientation. Scores on this index subscale from 1 (“very liberal”) to 5 (“very conservative”). In the present study, however, scores were converted to a scale ranging from 0 (“very liberal”) to 4 (“very conservative”).

Measurement of school-related variables. Racial composition referred to the percentage of the most common race reported at a particular school. As such, higher percentages indicated less diverse schools. The number of students for every one faculty member constituted the student–faculty ratio. SAT score referred to the average SAT score reported for admitted freshmen at a particular institution and region, based on the U.S. Census Bureau’s delineations.

The Institution type referred to the following categories of colleges/universities: public, private, religious, religious liberal arts and private liberal arts. These school-related variables were collected through website searches. With respect to the course catalog, mission statement and value statement variables, we conducted internet searches to obtain these documents for the institutions represented in this study and generated a list of words related to ethics and morality (see

Table A1. List of Morally Relevant Words.

No.	Words	No.	Words	No.	Words
1.	care	19.	empathy	37.	moral
2.	caring	20.	equality	38.	morality
3.	character	21.	equity	39.	mutually
4.	citizens	22.	ethical	40.	needs
5.	citizenship	23.	ethics	41.	outreach
6.	civic	24.	honest	42.	partnership
7.	civility	25.	honesty	43.	perspective
8.	collaborative	26.	humane	44.	respect
9.	community	27.	humanities	45.	responsibility
10.	compassion	28.	ideals	46.	sensitivity
11.	conscience	29.	inclusiveness	47.	serve
12.	cooperation	30.	inclusivity	48.	service
13.	cooperative	31.	integrity	49.	serving
14.	courage	32.	interdependence	50.	sharing
15.	courageous	33.	just	51.	society
16.	dignity	34.	justice	52.	stewardship
17.	diverse	35.	kindness	53.	support
18.	diversity	36.	law	54.	teamwork

).

Next, we conducted data mining analysis using R 4.1.1 software to create effective school level predictors. All words in the course catalogs, mission and value statements from websites were saved, and only words that matched those on the predetermined list were counted. The frequency (total number of words) in the mission and value statements for each college were used as predictors (see

Table A2. Frequency of Morally Relevant Words in Mission Statements.

Number of Ethical and Moral-Related Words in Mission Statement	Frequency by School (%)
0	7 (9.6%)
1	18 (24.7%)
2	14 (19.2%)
3	11 (15.1%)
4	4 (5.5%)
5	4 (5.5%)
6	6 (8.2%)
7	0 (0.0%)
8	1 (1.4%)
9	1 (1.4%)
10	2 (2.7%)
11	1 (1.4%)
12	2 (2.7%)
13	1 (1.4%)
14	1 (1.4%)
Total	73 (100.0%)

,

Table A3. Frequency of Morally Relevant Words in value statements.

Number of Ethical and Moral-Related Words in Value Statement	Frequency by School (%)
1	1 (2.3%)
2	3 (6.8%)
3	3 (6.8%)
4	2 (4.5%)
5	4 (9.1%)
6	4 (9.1%)
7	4 (9.1%)
8	2 (4.5%)
9	1 (2.3%)
10	6 (13.6%)
11	3 (6.8%)
12	2 (4.5%)
13	3 (6.8%)
14	1 (2.3%)
15	1 (2.3%)
16	1 (2.3%)
17	0 (0.0%)
18	2 (4.5%)
19	0 (0.0%)
20	1 (2.3%)
Total	44

and

Table A4. Frequency of Morally Relevant Words in Course Catalogs.

Coding for Course Catalog	Number of Ethical and Moral-Related Words in Course Catalog	Frequency by School (%)
0	0–300	5 (11.4%)
1	301–600	3 (6.8%)
2	601–900	8 (18.2%)
3	901–1200	10 (22.7%)
4	1201–1500	6 (13.6%)
5	1501–1800	2 (4.5%)
6	1801–2100	3 (6.8%)
7	2101–2400	2 (4.5%)
8	2401–2700	2 (4.5%)
9	2701–3000	1 (2.3%)
10	3001–3200	2 (4.5%)
	Total	44 (100.0%)

). As the number of morally relevant words in the course catalogs ranged from 0 to 3159, the course catalog variable comprised 11 ordered categories (see Table A4).

Procedure. The analysis included paper/pencil DIT-2 questionnaires collected between 2010 and 2014. An inclusion criterion was that more than 85% of the dataset needed to pass several reliability checks. Only data from American citizens who reported English as their primary language were used. Additionally, only data from undergraduate students were considered for this study. In total, 231 datasets included observations that met these inclusion criteria.

3. Results

We used multilevel analysis to analyze the data. Analyses included individual- and institutional-level variables, as this study’s purpose was to gain a better understanding of factors related to moral judgment development. The individual-level variables of interest were sex, education level and political orientation. The institutional-level variables were racial composition, student–faculty ratio, average SAT score, institutional region and type, morally relevant words in mission statements, value statements and course catalogs.

Table 1. Descriptive Statistics for N2 Scores as Functions of Individual and Institutional Factors.

	N2 Score
Variables	M	SD
Individual Characteristics
Sex
Male	32.57	15.83
Female	36.74	15.72
Education Level
Freshmen	31.42	15.09
Sophomores	35.38	16.12
Juniors	35.12	15.85
Seniors	38.07	15.97
School-Related Factors
Institution Type
Public	34.66	15.29
Private	32.24	15.30
Religiously Affiliated	39.11	15.71
Liberal Arts	30.76	15.06
Liberal Arts-Religious	44.29	15.00
Region
Northeast	34.81	14.88
South	32.34	15.95
West	36.93	14.71
Midwest	36.56	16.29

presents descriptive statistics for the individual and institutional variables. Appendix B contains the equation for our multilevel model. The intraclass correlation for the unconditional N2 score model indicated that institution explained 16.8% of the variability in N2 scores. The significant relationship between institution and N2 score allowed us to proceed with hierarchical linear modeling.

3.1. Main Effects of Institutional Factors

Various institutional factors emerged as significant predictors of N2 scores (see

Table 2. Parameter estimates for target variables.

Parameters	N2 Score
Level 1
Sex (γ10)	2.11
Education Level (γ20)	1.08
Political Orientation (γ30)	−0.68
Level 2
Region (reference = South)
West (γ01)	29.72 ***
Midwest (γ02)	−29.13 ***
Northeast (γ03)	3.02
Institution Type
Public (γ04)	29.40 **
Liberal Arts (γ05)	11.58 **
Not Religious (γ06)	0.85
SAT (γ07)	−0.04 *
Student–faculty Ratio (γ08)	−3.76 ***
Racial Composition (γ09)	0.89 ***
Morally Relevant Words
Mission (γ010)	−2.36 ***
Values (γ011)	3.59 ***
Course Catalog (γ012)	4.44 ***
Intercept (γ00)	25.75 ***

NOTE: * p < 0.05; ** p < 0.01; *** p < 0.001; Sex (1 = Female, 0 = Male); Education Level (3 = senior, 2 = junior, 1 = sophomore, 0 = freshman); Public (1 = public, 0 = else); Liberal Arts (1 = liberal arts and liberal arts-religious, 0 = else); Not Religious (1 = not religious = public, private and liberal arts, 0 = religious = religiously affiliated and liberal arts-religious); Mission = total number of words in the school’s mission statement related to ethics and morality; Values = total number of words in the school’s value statement(s) related to ethics and morality; Course Catalog = 10 categories of total number of words in the school’s course catalog related to ethics and morality (see Table A4); SAT, Student–faculty Ratio, Racial Composition, Mission, Values and Course Catalog all are centered around the grand mean.

). First, after controlling for other institutional factors, students at schools located in the West exhibited N2 scores 29.72 points higher than did students at schools in the South.

Additionally, students at schools in the Midwest exhibited N2 scores 29.13 points lower than those of students at schools in the South. Second, in terms of institution type, public institution students exhibited N2 scores 29.40 points higher than those of private institution students. Additionally, students at liberal arts institutions exhibited N2 scores 11.58 points higher than those of students at non-liberal arts institutions. Third, a one-point increase in a school’s average SAT score corresponded with a 0.04-point decrease in N2 score.

Fourth, a one-unit increase in a school’s student–faculty ratio corresponded with a 3.76-point decrease in N2 score. Fifth, a one-unit increase in the racial majority of an institution corresponded with a 0.89-point increase in the N2 score. Sixth, a one-word increase in the number of morally relevant words in a school’s mission statement corresponded with a 2.36-point decrease in N2 score. Seventh, a one-word increase in the number of morally relevant words in a school’s value statement corresponded with a 3.59-point increase in the N2 score. Eighth, a one-word increase in the number of morally relevant words in a school’s course catalog corresponded with a 4.44-point increase in the N2 score.

In sum, moral judgment development competencies significantly related to the institutions’ region, type, average SAT score, student–faculty ratio, racial composition and emphasis on moral and ethical content. Counterintuitively, higher average SAT scores, more racial diversity, and a stronger ethical focus in mission statements were associated with decreased N2 scores (see Table 2). These counterintuitive findings are addressed further in the Discussion.

3.2. Interactions of Individual and Institutional Factors

Multiple interactions emerged between institutional factors and individuals’ political orientation, though the results demonstrated a nonsignificant main effect for political orientation (γ30) (see Table 2 and

Table 3. Parameter Estimates for Cross-Level Interactions and Random Effects.

Parameters	N2 Score
Cross-Level Interactions
Sex × Public (γ11)	1.00
Sex × Liberal Arts (γ12)	1.09
Sex × Not Religious (γ13)	4.07
Sex × SAT (γ14)	0.012
Sex × Student–faculty Ratio (γ15)	0.26
Sex × Racial Composition (γ16)	0.15
Sex × Mission (γ17)	−0.70
Sex × Values (γ18)	0.10
Sex × Course Catalog (γ19)	0.42
Education Level × Mission (γ21)	−0.28
Education Level × Values (γ22)	−0.23
Education Level × Course Catalog (γ23)	−0.69
Political Orientation × West (γ31)	−9.00 ***
Political Orientation × Midwest (γ32)	1.99 *
Political Orientation × Northeast (γ33)	−0.29
Political Orientation × Public (γ34)	1.04
Political Orientation × Liberal Arts (γ35)	−0.012
Political Orientation × Not Religious (γ36)	−1.16
Political Orientation × Student–faculty Ratio (γ37)	0.48 **
Political Orientation × Racial Composition (γ38)	−0.15 ***
Random Effects
Residual (var(rij))	203.70 ***
Education Level (var(u2j))	4.24

NOTE: * p < 0.05; ** p < 0.01; *** p < 0.001.

). Before we discuss the findings related to political orientation, recall that lower political orientation scores indicate a more liberal political orientation and higher political orientation scores indicate a more conservative political orientation. Additionally, interaction effects must be interpreted alongside significant main effects.

First, there was a significant interaction between political orientation and institutions’ region. Among those with very liberal to somewhat conservative political orientations, students at schools in the West exhibited moral reasoning scores higher than those of students at schools in the South. However, more conservative students in the West exhibited lower moral reasoning scores than did those in the South (γ₀₁West + γ₁₃Political orientation × West = 29.72 × 1 − 9 × 4 × 1 = −6.28 for West region (=1) and very conservative (=4)).

Regardless of political orientation, the results showed that moral reasoning scores for students at schools in the Midwest tend to be lower compared to those of students in the South (γ₀₂Midwest + γ₃₂Political orientation × Midwest = −29.13 (=−29.13 × 1 + 1.99 × 0 × 1) for very liberal to −21.17 (=−29.13 × 1 + 1.99 × 4 × 1) for very conservative). The magnitude of the difference in moral reasoning scores between students in the Midwest and the South depended on students’ political orientation. The difference in scores between the two regions among more liberal students was more pronounced than that among more conservative students (region difference = 29.13 for very liberal students and 21.17 for very conservative students).

Third, political orientation interacted with student–faculty ratio. Specifically, the results showed lower moral reasoning scores when the student–faculty ratio was greater, regardless of political orientation; however, moral reasoning scores among more liberal students were affected to a greater extent (by a larger student–faculty ratio) compared to moral reasoning scores among more conservative students.

For example, while very liberal students exhibited a 3.76-point decrease in N2 scores, very conservative students exhibited a 1.84-point decrease in N2 score when same student–faculty ratio was assumed (using γ₀₈Ratio + γ₃₇Political orientation × Ratio, −3.76 (=−3.76 + 0.48 × 0) × Ratio for very liberal students and −1.84 (=−3.76 + 0.48 × 4) × Ratio for very liberal students). Fourth, the analysis revealed an interaction between political orientation and racial composition.

Specifically, N2 scores were adversely affected when schools’ racial composition was more diverse, regardless of students’ political orientation; however, N2 scores for more liberal students were affected to a greater extent by a more diverse racial composition compared to N2 scores for more conservative students. For example, while very liberal students exhibited a 0.89-point decrease in N2 score, very conservative students exhibited a 0.29-point decrease (0.89 (=0.89 − 0.15 × 0) × Racial Composition for very liberal students and 0.29 (=0.89 − 0.15 × 4) × Racial Composition for very liberal students using γ₀₉Racial Composition + γ₃₈Political orientation × Racial Composition).

4. Discussion

Using an undergraduate student sample, we employed multilevel modeling techniques to explore the relationships of individual and institutional factors with moral reasoning development. Next, we discussed further the interaction of political orientation with institutional factors. Subsequently, institutional factors that did not interact with individual characteristics are addressed, which included the average SAT score, institution type, number of morally relevant words in mission statements, value statements and course catalogs.

4.1. Political Orientation and Institutional Region

The finding of an interaction between political orientation and institutional region is consistent with previous research that found political beliefs and region significantly relate to moral reasoning [12]. However, whereas this study investigated individuals’ political orientation, Ref. [12] investigated institutions’ political climate, finding that individuals attending schools with more politically conservative climates in the South were likely to have lower moral reasoning scores.

Although our analysis uncovered lower moral reasoning scores among students at schools in the South, when political orientation was considered in conjunction with institutional region, more nuanced results emerged. For example, students with very conservative political orientations attending schools in the West exhibited lower moral reasoning scores compared to students with very conservative political orientations attending schools in the South.

However, individuals with very liberal to somewhat conservative political orientations attending schools in the West exhibited higher moral reasoning scores compared to students with very liberal to somewhat conservative political orientations attending schools in the South. It could be that if there is a large discrepancy between individual political orientation and the surrounding political climate, then the individual may be less inclined to open up and engage in peer discussions, and research shows that discussions with diverse peers can promote moral judgment development [1,10]. Very conservative political views may be more uncommon in the Western region of the U.S. compared to more liberal views.

As such, very conservative students attending institutions in the West may feel intimidated by engaging in discussions with peers who think differently than them, ultimately missing opportunities for growth. The results also demonstrated that students attending schools in the Midwest exhibited lower moral reasoning scores compared to students attending schools in the South, regardless of their political orientation; however, more liberal political views corresponded with a greater decrease in moral reasoning scores compared to more conservative views. Following the same reasoning as above, perhaps this pattern of results concerns the alignment between students’ political orientation and the surrounding political climate.

That is, more conservative political views may correspond to smaller decreases in moral reasoning scores because such views align more strongly with the wider political climate (of the Midwest, for example). Students who hold more conservative views may generally feel more supported and, in turn, be more likely to engage in growth-inducing experiences, such as exploring viewpoints that diverge from their own, a practice associated with the development of moral judgment [1,10].

4.2. Political Orientation and Racial Composition

With respect to the interaction between political orientation and racial composition, the results demonstrated that less racial diversity at a school corresponded with higher N2 scores, regardless of students’ political orientation. As mentioned previously, research demonstrates that interactions with diverse peers can promote moral judgment development [1,10], which contradicts the present findings. However, Ref. [14] caution that negative interactions with diverse peers can stunt growth where moral reasoning is concerned.

Ref. [14] was a cross-sectional study, and they did not conclude that negative interactions with diverse peers result in decreases decrease in moral reasoning development. It is possible, however, that enough negative interactions with diverse peers could motivate someone to take refuge in familiar perspectives and thought processes rather than risking an uncomfortable and potentially negative encounter with a diverse peer; ultimately, then, moral judgment development could be affected adversely. It is important that higher education institutions provide effective support for diverse student populations in terms of fostering positive relations.

Interestingly, more liberal students’ N2 scores were affected to a greater degree by their school’s racial composition compared those of more conservative students. For example, more liberally oriented students exhibited larger increases in N2 scores compared to more conservatively oriented students when their institutions had the same level of racial diversity. Perhaps students with more conservative political orientations require the cognitive disequilibrium that accompanies interacting with racially diverse peers to experience growth in moral reasoning, whereas more liberally oriented students can experience growth without such salient encounters with diversity.

4.3. Political Orientation and Student–Faculty Ratio

Regarding the interaction between political orientation and student–faculty ratio, the results showed that a larger student–faculty ratio corresponded with lower moral reasoning scores, regardless of political orientation. However, moral reasoning scores decreased more for liberal students compared to conservative students. Fewer opportunities for interaction with faculty may affect students with liberal political orientations more because they are more likely to interact with faculty members in the first place and find value in such exchanges, which could be due to faculty’s sharing similar political views. To the extent that students with more conservative political orientations are apprehensive about interactions with faculty members with whom they feel they cannot relate, larger student–faculty ratios may not affect them as much.

4.4. Average SAT Score

The results demonstrated that higher average SAT scores corresponded with lower moral reasoning scores. As mentioned previously, previous research has demonstrated a positive relationship between individuals’ SAT scores and moral judgment development [13,14,18]. Recall, however, that the SAT score was measured at the institutional (rather than individual) level in this study. Accordingly, we do not know whether the individuals comprising our sample had SAT scores higher or lower than their school’s average. In this study, SAT score was indicative of a school’s selectivity.

It is possible that more selective schools (i.e., higher average SAT scores) may differ in ways that negatively impact moral judgment development. For example, perhaps more selective schools’ students tend to be more conservative, a political orientation that aligns with the less developmentally advanced form of moral reasoning known as maintaining norms [7,8,9]. Maintaining norms reasoning justifies action based on established practices rather than on shared ideals that are open to scrutiny and debate, with the latter being related to more developmentally advanced postconventional reasoning [7,8,9].

4.5. Institution Type

The institution type was significantly related to moral reasoning development, and this relationship was particularly salient when comparing public institutions to private institutions and liberal arts institutions to non-liberal arts institutions. First, the results demonstrated that public institution students exhibited more developmentally advanced moral reasoning scores compared to those of private institution students. Ref. [2] also found public institution students to have more advanced levels of moral judgment development compared to private institution students.

However, this study did not delineate types of public institutions, whereas [2] indicated that it was students at public research institutions specifically who were more developmentally advanced in moral reasoning compared to students at private institutions. Second, compared to students at non-liberal arts institutions, students at liberal arts institutions exhibited significantly higher N2 scores. This is an intuitive finding given the results reported by [1], which indicated that liberal arts was the only institution type that reliably demonstrated a large effect size for comparisons in which liberal arts students exhibited higher moral reasoning scores compared to other students at other types of institutions.

4.6. Morally Relevant Words in Mission and Value Statements and Course Catalogs

For value statements and course catalogs, a greater number of morally relevant words corresponded with more developmentally advanced moral judgment in students. However, for mission statements, a greater number of morally relevant words corresponded with less developmentally advanced moral judgment. It could be that the content of value statements and course catalogs is more central to the inner workings of higher education communities and in turn transfers more practically to campus life. Mission statements, on the other hand, may be more political in nature and less connected to the academic climate.

Additionally, Ref. [35] notes that the cynicism that is created when leadership contradicts itself, undermining its integrity, may lead to a lack of commitment to the organization and its designated objectives. Thus, it is possible that the greater extent to which the student body views a mission statement as unrepresentative of their school’s leadership, the less committed that student body may become to the mission; if the mission emphasizes ethical and moral reasoning and behavior, a lack of commitment may manifest in lower moral reasoning scores.

4.7. Future Directions

The present study provides fertile ground for future research. First, the interactions between political orientation and (1) institutional region, (2) racial composition and (3) student–faculty ratio are interesting findings requiring clarification. For example, regarding political orientation and institutional region, future research could use interview methods to explore goodness-of-fit as an explanation for why more developmentally advanced moral reasoning is associated with students with a stronger liberal orientation in the West compared to students with a stronger liberal orientation in the South. Second, less racial diversity corresponded with more developmentally advanced moral reasoning.

Future research should investigate this interesting finding; focusing on how the degree of racial diversity in a college community is associated with the nature of student interactions may be fruitful, given [14]’s claim that negative interactions with diverse peers are not conducive to moral judgment development. Third, higher average SAT scores corresponded with less developmentally advanced moral reasoning. Future research might focus on characteristics of more selective schools (e.g., higher average SAT scores) that negatively relate to moral reasoning development. For example, studies could explore whether selective schools’ student bodies are more politically conservative.

Fourth, more research is necessary to explore the link between moral judgment development and mission statements, value statements and course catalogs. For example, it may be fruitful to account for the frequency with which mission and value statements are updated and revised to match the constantly changing campus culture. Perhaps mission and value statements that are updated and revised more often relate differently to prosocial outcomes, including moral judgment development, compared to mission and value statements that are not modified periodically to align with campus culture.

5. Conclusions

This study’s findings have implications for higher education communities. First, the prominent role of political orientation highlights the necessity for educators to recognize and respond appropriately to the various types of students with whom they interact. For example, our findings demonstrate that the student–faculty ratio affects liberally oriented students more so than conservatively oriented students, perhaps because the former are more likely to initiate interactions with faculty members compared to the latter. If this is the case, faculty should be intentional in their efforts to facilitate interactions with all students, as such interactions can be effective growth experiences for students.

Second, with greater diversity relating to decreases in moral reasoning, it is imperative that schools make diversity initiatives a priority, helping to cultivate positive interactions among diverse students and, in turn, fostering growth in moral reasoning. Third, the positive relationship between moral judgment development and morally relevant content in an institution’s value statements indicates the significant effect of institutions’ principles on student development; as such, making the values of an institution explicit is beneficial, particularly for the student body.

Similarly, the positive relationship between moral judgment development and morally relevant content in an institution’s course catalog indicates the curriculum’s impact on student development and, therefore, highlights the importance of making these types of courses available for all students (e.g., offering the courses at various times/days each semester and to students at all levels).