Does African American Women’s Racial Identity Mediate Gendered Racism on Anticipated Relationship Threat?

Abstract

The present study examined the question of whether racial identity among African American women mediated the relationship between gendered racism and anticipated relationship threat. Using the Multicultural Assessment Intervention Process (MAIP) framework, we examined the relationship of gendered racism, racial identity, and anticipated relationship threat among a convenience sample of 411 African American women. A structural model was configured with gendered racism directly predicting anticipated relationship threat and racial identity serving as a mediator. Results indicated that greater levels of perceived gendered racism were associated with greater perceptions of anticipated relationship threat. Racial identity was found to not mediate the association with anticipated relationship threat. Individuals with less education experienced higher levels of concern regarding physical safety and controlling behaviors than those with more education. Implications for future relationship threat research with African American women are discussed.

Keywords

relationship threat African American women gendered racism racial identity Multicultural Assessment-Intervention Process model

The present study examined the predictive relationship between gendered racism (i.e., oppression due to race and gender) and African American women’s anticipated relationship threat (or relationship safety perceptions). In addition, we hypothesized that African American racial identity might mediate this relationship. Gendered racism is expected to be positively related to threat, but the possibility that this may only occur among African American women who possess higher levels of racial identity needs to be examined before the viability of such a conclusion may be put forward. Before offering a structural equation model that examines these variables, we briefly review the literature on relationship threat, gendered racism, and African American racial identity, within the framework of the Multicultural Assessment Intervention Process (MAIP) model.

Perceived Relationship Threat or Safety

Many domestic violence programs acknowledge that their primary aim is to help survivors achieve safety from intimate partner violence (IPV; Sullivan, 2011). IPV typically encompasses a variety of abusive behaviors that may include the following: emotional abuse (Shepard & Campbell, 1992), economic abuse (Smith, Whiting, Karakurt, Oka, & Servino, 2013), social isolation and control (Smith et al., 2013), physical abuse (Tjaden & Thoennes, 1998), sexual abuse (Mahoney & Williams, 1998), and stalking (Fremouw, Westrup, & Pennypacker, 1997); many of which may co-occur or interact with each other.

In comparison to White Americans, African American women have disproportionately higher rates of IPV throughout their lifetime (Kaslow et al., 2010; Taft, Bryant-Davis, Woodward, Tillman, & Torres, 2009). These disproportionate IPV rates have had devastating consequences for many African American women. For example, high lethal victimization rates among African American women have been observed among youth (West, 2004), limited education and low income individuals (Sharps, Campbell, Campbell, Gary, & Webster, 2003), disadvantaged community inhabitants (Benson, Wooldredge, Thistlethwaite, & Fox, 2004), and nonmarital cohabitating relationships (Mize, Shackelford, & Shackelford, 2009), or relationship separations (Williams, Oliver, & Pope, 2008). Thus, establishing and maintaining safety while intervening clinically with IPV populations (i.e., crisis services, shelter, support groups, individual counseling, legal and medical advocacy) is a top priority (Macy, Giattina, Montijo, & Ermentrout, 2010). Perceptions of relationship safety as well as actual safety-seeking behaviors among African American women are complex (K. A. Thomas, Goodman, & Putnins, 2015) and may inadvertently trigger negative consequences such as escalated violence, financial strain, residential instability, and custody battles (Bostock, Plumpton, & Pratt, 2009).

Due in part to the relative lack of published empirical work exploring the perceived relationship threat or safety construct (Novak, Smith, & Sandberg, 2015; K. A. Thomas et al., 2015), aspects of the MAIP framework (Dana, 1993; Gamst, Liang, & Der-Karabetian, 2011) suggest that cultural guidance in the selection of two constructs (gendered racism and African American racial identity) might predict relationship threat or safety. The MAIP offers a number of culturally responsive phases that encourage positive clinical outcomes during behavioral health service delivery. These phases include culturally appropriate intake contact (screening interview), client match (gender, language, culture) preference, cultural status assessment (ethnic identity, acculturation, discrimination, gender issues, etc.), provider self-perceived cultural competence, multicultural training, ethnic-specific or general interventions, and assessment of clinical outcomes and service satisfaction (Gamst et al., 2011).

MAIP-inspired empirical research has primarily examined client–practitioner ethnic/racial match (Gamst, Dana, Der-Karabetian, & Kramer, 2000, 2004), acculturation and ethnic identity (Gamst et al., 2002), practitioner self-reported cultural competence (Gamst, Dana, Der-Karabetian, et al., 2004; Keyser, Gamst, Meyers, Der-Karabetian, & Morrow, 2014), cultural shifting as a coping mechanism (Gamst et al., 2018; Johnson, Gamst, Meyers, Arellano-Morales, & Shorter-Gooden, 2016), Portuguese American quality of life predictors (Pinheiro Rocha, Gamst, Meyers, Der-Karabetian, & Magina, 2018), and Arab American mental health services attitudes (Balesh, Gamst, Meyers, Der-Karabetian, & Elias, 2018). Relationship threat or safety attitudes can also be conceptualized within the MAIP model, and hence the MAIP serves as a conceptual framework in which gendered racism acted as a predictor and African American racial identity operated as a mediator of gendered racism that predicts relationship threat among African American women. In the present study, the Multicultural Status Assessment phase of the MAIP rather than the entire model provided a more culturally refined set of predictors (gendered racism, African American racial identity) that is used to explore relationship threat or safety among African American women. The constructs of gendered racism and African American racial identity are reviewed in the following sections.

Gendered Racism

Oppression, particularly at the intersection of race and gender (gendered racism), is a major contributor to health and well-being challenges among women of color, particularly African American women (Essed, 1991; Franklin, Boyd-Franklin, & Kelly, 2006; Perry, Harp, & Oser, 2013; A. J. Thomas, Witherspoon, & Speight, 2008). African Americans are subjected on a daily basis to racial discrimination, microaggressions, stereotypes, and institutional discrimination (Gamst et al., 2011; Harrell, 2000; Stevens-Watkins, Perry, Pullen, Jewell, & Oser, 2014; Sue et al., 2008). For African American women, racial stressors are compounded by gender discrimination encountered at both the individual attitudinal level and a societal structural level (Klonoff, Landrine, & Campbell, 2000; Moradi & Subich, 2003).

Media portrayals further castigate African American women as nurturing and passive, aggressive and emasculating, or sexually promiscuous (Donovan & West, 2015; A. J. Thomas et al., 2008; West, 2018; Woods-Giscombé, 2010). These stereotypes consistently portray African American women in a negative light arguably making them more susceptible to sexist discrimination than their White American counterparts (Sue, 2010). These persistent stereotypes continue to plague African American women and also foster new emergent stereotypes that include “welfare queen,” “gold digger,” and “video vixen” (Adams-Bass, Bentley-Edwards, & Stevenson, 2014). Furthermore, these negative stereotypes have entered the music industry (through hip-hop and rap videos) by hyper-sexualizing African American women as exotic dancers (Reid-Brinkley, 2008). African American women are also bombarded with White American beauty standards concerning hair and body image that have caused many women to look for support at local beauty salons (Linnan, 2007). Reality television shows’ (Bad Girls Club, The Real Housewives of Atlanta) depiction of African American women also helps to reinforce historically negative stereotypes (Allison, 2016). While LGBTQ characters have increased representation in films and television, African American lesbians are often portrayed as either aggressive, erotic, or victims of hate crimes (DeClue, 2011). Overall, the effect of these stereotypical portrayals of African Americans within the mass media appears to be having a tremendous influence in shaping perceptions about African American women among the general population (Punyanunt-Carter, 2008).

As a consequence of the dual threat of racist and sexist stressors, some African American women engage in shifting (i.e., altering one’s self-presentation due to environmental context) as a response to perceptions or expectations of gendered racist discrimination (Johnson et al., 2016). Recent evidence indicates that shifting, as a coping strategy, can be stressful for some African American women (Gamst et al., under review), and may have important consequences and implications for African American women’s perceived need to shift in the workplace (Dickens, Womack, & Dimes, 2019).

Among African American women, gendered racism can manifest itself in psychological distress (e.g., depression, post-traumatic stress, lower self-esteem; Szymanski & Lewis, 2016) and can negatively impact both mental and physical health outcomes (e.g., quality of life, less perceived stress; Perry et al., 2013).

African American Racial Identity

For nearly half a century, African American racial identity theory has evolved providing greater insight about client–practitioner racial match, as well as the process of “healthy assimilation” to an often-oppressive mainstream culture (Baldwin, 1981; Cross, 1971; Helms, 1986, 1990; Jackson, 1975; Myers, 1988; C. W. Thomas, 1971).

Three major conceptual approaches to African American racial identity theory have been identified (Constantine, Richardson, Benjamin, & Wilson, 1998; Wilson, Sellers, Solomon, & Holsey-Hyman, 2017). Nigrescence (or mainstream) stage models of African American racial identity (Cross, 1978, 1995; Helms, 1990, 1995; White & Parham, 1990) posit a linear progression of stages or statuses that African Americans progress from negative thoughts and feelings about being Black in the United States to an internalized positive view of themselves and other African Americans.

Alternative Africentric approaches to racial identity do not necessarily posit that African Americans first internalize a negative self-perception or progress through Nigrescence-like stages. Instead, Africentrism attempts to restore and reflect traditional African acculturation processes (Dana, 2002) grounded in African cultural worldview and values such as group cohesiveness, collaborative experiences, spirituality, and interdependence (Baldwin, 1984; Kambon, 1992; Nobles, 1980). Several investigators have developed measures to explore aspects of the Africentric construct (Baldwin & Bell, 1985; Grills & Longshore, 1996).

Finally, multidimensional approaches to African American racial identity, such as the Multidimensional Model of Racial Identity (MMRI; Sellers & Shelton, 2003; Sellers, Smith, Shelton, Rowley, & Chavous, 1998) or the MAIP (Dana, 2002; Gamst et al., 2011), place emphasis on the factors that shape an individual’s understanding of what it means to be Black. The MMRI has garnered the most empirical research regarding Black racial identity. The MMRI posits four racial identity dimensions that include salience (how relevant race is to one’s self-concept), centrality (the extent to which one defines themselves by their race), regard (public regard which is affective and evaluative, and private regard which involves positive and negative attitudes regarding Black individuals), and ideology (beliefs about how Black people should behave).

The Present Study

Critical variables assumed to underlie and drive processes involved in African American women’s self-perceptions of relationship threat and safety are examined in the present study. Explicating some of the salient predictors and mediators of relationship threat would enhance interventions for African American women. We hypothesized that gendered racism will positively predict anticipated relationship threat with African American racial identity mediating this association.

Method

Participants

All participants (N = 411) were African American women who were recruited from churches in two southern CA counties and were paid $15.00 each for completing the questionnaire. The participants in this convenience sample ranged in age from 18 to 84 years (M = 40.54, SD = 16.32). Additional sample demographics can be found in Table 1.

Table 1.

Sample Characteristics of Participants (N = 411).

Variable	n	%	M	SD
Age	411	100	40.54	16.32
Gender
Female	411	100
Are you a mother?
Yes	257	62.5
No	154	37.5
Are you a grandmother?
Yes	122	29.7
No	289	70.3
Number of grandchildren			1.22	0.92
Are you a child caretaker?
Yes	67	16.3
No	344	83.7
Number of dependents			0.32	0.98
Relationship status
Married	112	27.4
Living with partner	45	10.9
Divorced	38	9.2
Separated	21	5.1
Widowed	31	7.5
Single	164	39.9
Highest level of education			0.32	0.98
Some high school or less	39	9.5
High school	68	16.5
Some college	125	30.4
Associate degree	46	11.2
Bachelor’s degree	56	13.6
Some graduate	25	6.2
Master’s degree	40	9.7
Doctoral degree	12	2.9
Employment status
Employed full-time	208	50.6
Employed part-time	55	13.4
Self-employed	34	8.3
Homemaker	23	5.6
Unemployed	29	7.1
Retired	42	10.2
Student (currently enrolled)	20	4.8
Annual income
Less than $30,000	184	44.7
$30,001-$39,999	83	20.2
$40,000-$74,999	92	22.4
$75,000-$99,999	29	7.1
$100,000 or more	23	5.6

Procedure

All participants were recruited from churches in Los Angeles and San Bernardino, CA, counties through clergy and church staff, who provided facilitation. Participants completed a consent form and hard-copy version of the questionnaire at church facilities after attending church activities (Sunday service, bible study, choir rehearsal). Participants understood that their involvement in this study was voluntary and their information would be confidential. Presentation of the order of scales used on the questionnaire was randomized.

Measures

The self-administered questionnaire contained 158 items comprising nine demographic questions and five scales.

Demographic Questions

The demographic questions consisted of nine items that included the participants’ age, mother and grandmother status, caretaker status, relationship status, education, employment status, and income.

Predictor and Mediator Measures

General Ethnic Discrimination Scale (GED)

The GED (Landrine, Klonoff, Corral, Fernandez, & Roesch, 2006) measures participant’s perceptions of ethnic discrimination in the past year (Recent Discrimination subscale, 18 items), one’s lifetime (Lifetime Discrimination subscale, 18 items), and an appraisal of the impact of these perceptions (Appraised Discrimination subscale, 18 items). In this study, the Recent Discrimination subscale (in public places, health care, etc.) was used. Items were measured on a 6-point summative response scale that ranged from 1 (never) to 6 (almost all the time). Higher scores indicated greater experience with discrimination. Landrine et al. (2006) reported high levels of internal consistency for the scores of the Recent Discrimination subscale (α = .93), which were similarly found in this study (α = .95).

Schedule of Sexist Events (SSE)

The SSE (Landrine & Klonoff, 1995) consists of 20 items (with four subscales) that assess participant’s perceptions of recent (past year) or lifetime sexist discrimination. The Recent Sexist Discrimination subscales were used in the present study and include the following: Sexist Degradation and Its Consequences subscale (8 items), Sexist Discrimination in Distant Relationships subscale (5 items), Sexist Degradation in the Workplace subscale (4 items), and Sexism in Close Relationships subscale (3 items). Items were measured on a 6-point summative response scale that ranged from 1 (never) to 6 (almost all the time). Higher scores indicated greater experience with recent discrimination. In this study, we found slightly higher reliabilities than those reported by Landrine and Klonoff (1995) with Cronbach’s alpha coefficients for the scores of the Recent Discrimination subscales as follows: Sexist Degradation and Its Consequences (α = .93), Sexist Discrimination in Distant Relationships (α = .91), Sexist Discrimination in the Workplace (α = .90), and Sexism in Close Relationships (α = .78).

Multidimensional Inventory of Black Identity (MIBI)

The MIBI (Sellers et al., 1998) is a 51-item self-report measure that assesses African American racial identity as a multidimensional construct, and includes the following six subscales: Centrality (i.e., how significant racial identity is to the individual, 8 items), Private Regard (i.e., how an individual feels about their racial group, 7 items), Assimilation (i.e., views correspondence between African Americans and mainstream society, 9 items), Humanist (i.e., emphasizes commonality among all humans, 9 items), Minority (i.e., views correspondence between African Americans and other oppressed groups, 9 items), and Nationalist (i.e., emphasizes the uniqueness of being of African descent, 9 items). All items were rated on a 7-point summative response scale with anchors of 1 (strongly disagree) to 7 (strongly agree). Higher scores indicated greater racial identity. Sellers, Rowley, Chavous, Shelton, and Smith (1997) reported Cronbach’s alpha coefficients for the six subscales as follows: Centrality (α = .77), Private Regard (α = .60), Assimilation (α = .73), Humanist (α = .70), Minority (α = .76), and Nationalist (α = .79). In the present study, internal consistency values were as follows: Centrality (α = .52), Private Regard (α = .52), Assimilation (α = .80), Humanist (α = .80), Minority (α = .79), and Nationalist (α = .65). Due to the low Cronbach’s alpha coefficients (indicating unreliability) for the Centrality, Private Regard, and Nationalist subscales, they were eliminated from all subsequent analyses.

Measurement of Acculturation Strategies for People of African Descent (MASPAD)

The MASPAD (Obasi & Leong, 2010) is a 45-item self-report measure used to assess an individual’s primary acculturation strategy. Two 20-item orthogonal scale dimensions are produced: Dimension 1 (MASPAD1) reflects an individual’s relative preference for maintaining their heritage ethnocultural group and was used in the present study, and Dimension 2 (MASPAD2) reflects an individual’s relative preference for having contact and participation with a different ethnocultural group; only Dimension 1 was used in the present study. All items were rated on a 6-point summative response scale with anchors of 1 (strongly disagree) to 6 (strongly agree). Higher scores indicated greater preference for an individual’s ethnocultural heritage. Obasi and Leong (2010) reported an internal consistency value of α = .87 for MASPAD1. In the present study, Cronbach’s alpha for MASPAD1 was .90.

Outcome Measures

Anticipated Relationship Threat

Anticipated Relationship Threat, as measured by the Self Assessment of Future Events Scale (SAFE; Smith et al., 2013), is a 15-item self-report measure of an individual’s anticipation that their partner will engage in future physical violence, verbal abuse, or controlling behaviors. The three 5-item subscales (Physical Safety, Verbal/Psychological Safety, Control) were each measured on a 6-point summative response scale that ranged from 1 (extremely unlikely) to 6 (extremely likely). Higher scores indicated greater perception of risk. Smith et al. (2013) reported internal consistency reliabilities for the Physical Safety, Verbal/Psychological Safety, and Control subscales as .85, .89, and .76, respectively. In the present study, good internal consistency was found for the Physical Safety (α = .91), Verbal Psychological Safety (α = .85), and Control (α = .70) subscales, respectively.

Results

Table 2 presents Pearson correlations, means, standard deviations, and Cronbach’s alpha for the scale and subscale scores used in the present study. All values were generally in the range of those reported in the literature. All scales and subscales scores reported reached a minimally acceptable level of internal consistency of at least .70.

Table 2.

Correlations, Means, Standard Deviations, and Reliabilities for Key Variables.

	1	2	3	4	5	6	7	8	9	10	11	12	M	SD	α
1. GED Recent	—	.62**	.57**	.57**	.56**	−.11*	−.04	.01	−.12*	.31**	.37**	.23**	2.84	1.03	.95
2. SEE Rec Degrad		—	.74**	.79**	.82**	−.23**	−.28**	−.10*	−.27**	.50**	.48**	.31**	2.52	1.23	.93
3. SSE Rec Dis Rel			—	.87**	.70**	−.24**	−.27**	−.12*	−.32**	.48**	.39**	.31**	2.25	1.16	.91
4. SSE Rec Dis Work				—	.73**	−.24**	−.31**	−.15**	−.35**	.53**	.40**	.30**	2.16	1.23	.90
5. SSE Rec Close					—	−.23**	−.26**	−.11*	−.28**	.51**	.40**	.30**	2.41	1.29	.78
6. MIBI Assim						—	.59**	.49**	.49**	−.11*	−.08*	−.13*	3.30	.75	.80
7. MIBI Hum							—	.57**	.51**	−.20**	−.07	−.17**	3.44	.80	.80
8. MIBI Min								—	.47**	−.04	.04	−.02	3.25	.80	.79
9. MASPAD 1									—	−.13**	−.11**	−.04	3.88	.84	.90
10. SAFE Psy										—	.76**	.66**	2.45	1.35	.91
11. SAFE Verb											—	.57**	2.90	1.24	.85
12. SAFE Con												—	3.05	1.04	.70

Note. GED Recent = GED Recent Discrimination subscale; SSE Rec Degrad = SSE Recent Sexist Degradation; SSE Rec Dis Rel = SSE Recent Sexist Discrimination in Distant Relationships; SSE Rec Dis Work = SSE Recent Sexist Discrimination at Work; SSE Rec Close = SSE Recent Sexism in Close Relationships; MIB Assim = MIBI Assimilation; MIBI Hum = MIBI Humanist; MIBI Min = MIBI Minority; MASPAD1 = MASPAD Dimension 1; SAFE Phy = Physical Safety; SAFE Verb = Safe Verbal/Psychological Safety; SAFE CON = Safe Control.

p < .05. **p < .01.

The three SAFE subscales had high positive correlation ranging from r = .57 to .76, p < .01. Likewise, the SSE subscales were also positively correlated with rs ranging from .57 to .87, p < .01. Moderate positive correlations were observed among the MIBI subscales, with rs ranging from .49 to .59, p < .01.

The Structural Model

A structural equation model with three latent variables was configured as shown in Figure 1. The latent variable of Anticipated Relationship Threat (so named because higher values signify greater levels of anticipated threat) was established as the outcome variable using as indicators the three subscales of the SAFE (Physical Safety, Verbal/Psychological Safety, and Control). Gendered Racism was proposed as an exogenous predictor in the model as indicated by the Sexist Degradation and Its Consequences, Sexist Discrimination in Distant Relationships, Sexist Discrimination in the Workplace, and Sexism in Close Relationships subscales of the SSE together with the Recent Discrimination subscale of the GED. Finally, African American Racial Identity as indicated by the Assimilationist, Humanist, and Minority subscales of the MIBI as well as the Dimension 1 or MASPAD1 (heritage ethnocultural group) scale of the MASPAD was used as a proposed mediator.

Figure 1.

Structural model (N = 411) for anticipated relationship threat. Latent constructs are shown in ellipses, and observed variables are shown in rectangles.

Although the model fit was less than adequate in the initial analysis, we had not specified any correlations between pairs of error terms at that time. Indicator variables representing latent variables may be related to each other for several reasons beyond what they have in common with the factor (e.g., they may relate to a construct different from the factor, they may be reverse worded, they may contain common words or wording). However, it is difficult to anticipate which relationships between indicators share enough variance in advance of the initial data analysis so that failing to take those relationships into account can detract from model fit. Thus, on a post hoc basis, adding such correlations may be useful in portraying a better or more veridical indication of the fit of the model to the data than not including them in the model (Bowen & Guo, 2012; Brown & Moore, 2012; Kline, 2015; Wang & Wang, 2012).

Because the indicators of Gendered Racism shared sufficient variance outside of their common factor variance to impact model fit, we added correlations of their residuals to the model. In this re-specification of the model, performed in IBM SPSS Amos version 25, the chi-square value 158.319, p < .001, was statistically significant (45, N = 411), but the goodness-of-fit index (GFI), normed fit index (NFI), incremental fit index (IFI), Tucker–Lewis index (TLI), and comparative fit index (CFI) were .940, .950, .964, .947, and .964, respectively, and the root mean square error of approximation (RMSEA) was .078 (90% confidence interval [CI] = [.065, .092]). These indexes taken together suggested that the model yielded a good fit to the data and explained approximately 13% and 36% of the variance of African American Racial Identity and Anticipated Relationship Threat, respectively.

Gendered Racism predicted both African American Racial Identity (standardized path coefficient = –.362, unstandardized path coefficient = –.185 with a standard error of .029, p < .001) and Anticipated Relationship Threat (standardized path coefficient = .616, unstandardized path coefficient = .734 with a standard error of .062, p < .001). However, African American Racial Identity did not significantly predict Anticipated Relationship Threat (standardized path coefficient = .049, unstandardized path coefficient = .114 with a standard error of .122, p = .348), thus eliminating the possibility of it mediating the influence of Gendered Racism on Anticipated Relationship Threat.

Given that the issue of Anticipated Relationship Threat was of particular interest in this study, an exploratory follow-up multivariate analysis of covariance (MANCOVA) concerning this construct was performed to focus on whether the degree of anticipated threat was related to level of education. The sample was categorized into three groups that served as the independent variable in the analysis: those whose education level was high school or less (n = 106), those who had college experience including achieving a bachelor’s degree (n = 227), and those having education at the graduate level (n = 77). The three subscales of the SAFE (Smith et al., 2013) served as dependent variables, and we used the quantitative variable of age as a covariate.

The data for this analysis conformed to the assumption of linearity of regression; that is, linear relationships were observed between the covariate of age and the dependent variables of the SAFE. The values of these three linear correlations were quite modest, ranging from –.135 to –.219, but were statistically significant (ps ≤ .006) with our sample size, suggesting that younger respondents anticipated somewhat more threat than older respondents. The data also met the assumption of homogeneity of regression at both the multivariate and univariate levels. The multivariate Education × Age interaction effect was not statistically significant, Wilks’ lambda = .012, F(6, 806) = 0.816, p = .558, and this pattern was mirrored at the univariate level for all three of the interaction effects. We therefore proceeded to perform the covariance analysis.

Bartlett’s test of sphericity was statistically significant (approximate chi-square = 596.491, df = 5, p < .001), indicating that the correlations of the adjusted SAFE subscores were sufficient to support the MANCOVA. These correlations were robust, ranging from .572 to .764, but we did not statistically combine the subscales in that there was still a possibility of obtaining a somewhat different pattern of results among them. Box’s test of the equality of the variance–covariance matrices was statistically significant, Box’s M = 31.331, F(12, 264774.543) = 2.576, p < .002, suggesting that the matrices adjusted for the covariate were not quite homogeneous; hence, Pillai’s trace was used to evaluate the multivariate effects as it is more robust to such violations.

The multivariate effects of both the covariate of age, Wilks’ lambda = .043, F(3, 404) = 6.120, p < .001, and the independent variable of education, Wilks’ lambda = .042, F(6, 810) = 2.890, p < .001, were statistically significant. Levene’s tests of homogeneity of variance were not statistically significant for any of the SAFE subscales, and so the univariate effects were evaluated with a Bonferroni corrected alpha level of .017 (.05/3).

The age covariate was statistically significant for all three SAFE subscales (ps ≤ .014). Of greater interest were the education univariate main effects. Although the Verbal/Psychological Safety subscale was not statistically significant, F(2, 406) = 2.242, p = .108, the Physical Safety, F(2, 406) = 4.124, p = .017, eta squared =. 012, and Control, F(2, 406) = 6.350, p = .007, eta squared =. 029, SAFE subscales both yielded statistically significant main effects.

The results indicated the following group differences. For the Physical Safety subscale, those with high school or less education reported higher levels of anticipated threat (adjusted M = 2.737, SE = 0.128, 95% CI = [2.483, 2.987]) than those with a college education (adjusted M = 2.294, SE = 0.088, 95% CI = [2.122, 2.466]); those with graduate levels of education did not differ from either of the other two groups (adjusted M = 2.521, SE = 0.150, 95% CI = [2.226, 2.815]). For the Control subscale, those with a college education (adjusted M = 2.943, SE = 0.067, 95% CI = [2.811, 3.076]) did not differ from those with graduate levels of education (adjusted M = 2.919, SE = 0.115, 95% CI = [2.692, 3.146]); however, respondents who had a high school level of education or lower reported significantly higher levels of anticipated controlling behaviors than the other two groups (adjusted M = 3.345, SE = 0.099, 95% CI = [3.151, 3.539]).

Discussion

Using the MAIP model for conceptual scaffolding, the present study examined the relationship of African American women’s Gendered Racism and African American Racial Identity in predicting perceived Anticipated Relationship Threat. Structural equation modeling indicated that Gendered Racism predicted both African American Racial Identity and Anticipated Relationship Threat. Interestingly, African American Racial Identity was found to not mediate the relationship between Gendered Racism and Anticipated Relationship Threat.

The hypothesized predictive relationship between Gendered Racism and Anticipated Relationship Threat was confirmed; higher levels of Gendered Racism lead to greater levels of perceived Anticipated Relationship Threat. This is consistent with work over the last three decades (Essed, 1991; Lewis, Williams, Peppers, & Gadson, 2017) that clearly indicates a link between increases in perceived gendered racism and mental and physical health challenges among African American women. The present research is the first to situate relationship threat self-perceptions in the same vein. Increases in perceived microaggressions and daily racial discrimination can result in African American women feeling insecure, fearful, and threatened in their personal intimate relations.

The present study failed to support the hypothesized mediation relationship of African American Racial Identity between Gendered Racism and Anticipated Relationship Threat. The latent African American Racial Identity variable (composed of the three MIBI subscales: Assimilationist, Humanist, Minority, and the MASPAD1 subscale) was not predictive of Anticipated Relationship Threat. The fact that African American Racial Identity neither appears to buffer nor accentuate the impact of Gendered Racism may suggest that racial identity, as operationalized in the present study, is not a particularly salient construct with regard to relationship safety or threat (cf. Sellers & Shelton, 2003).

Additional MANCOVA results, with a trichotomized education independent variable and the three SAFE subscales serving as dependent variables and age serving as a covariate, suggested that education influences African American women’s perceptions of relationship safety. Specifically, African American women who were high school graduates or less experienced greater levels of physical safety and control insecurity than their counterparts who obtained higher levels of education. This finding of education effects with relationship threat certainly merits further scrutiny in future research.

The utility of examining racial and cultural factors that contribute to relationship threat and identified by the MAIP model was underscored in the present study. Since gendered racism appears to be predictive of African American women’s perceptions that their partner may engage in future physical/verbal/psychological violence or controlling behaviors, behavioral health practitioners should consider assessing African American women’s perceptions of gendered racism in their daily lives (Sabri et al., 2014). Determining the coping mechanism or strategies employed by Black women experiencing IPV would also be helpful in determining treatment intervention options (K. A. Thomas et al., 2015).

Several limitations of the present study can be identified. First, the convenience sample of church-going African American women may not be representative of all Black women living in the United States today. Second, the lack of a mediating effect of African American Racial Identity may have been due to the choice of the four measured variables used in the present study. Several MIBI subscales were excluded from the analyses due to score unreliability and may have affected the study outcome. Third, due to the location of the present data collection (i.e., churches) individuals who identified as LGBTQ may not have been free to accurately self-identify, which in turn may have had an impact on the data analysis in unpredicted ways. Finally, this study employed a cross-sectional design and was unable to examine dynamic change over time. Future research with this population would benefit from employing longitudinal designs to explore the effects of gendered racism on relationship threat over time.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Glenn Gamst

Author Biographies

Giana N. Young received her PsyD degree in clinical psychology from the University of La Verne in 2018. Her practice and research interests focus on multicultural assessment and interventions among diverse populations.

Glenn Gamst, PhD, is a professor and chair of the Psychology Department at the University of La Verne. His research focuses on the effects of multicultural variables on clinical outcome.

Lawrence S. Meyers, PhD, is a professor of psychology at California State University, Sacramento. His areas of expertise include test development and validation.

Aghop Der-Karabetian, PhD, is a Professor Emeritus at the University of La Verne. His research focuses on world mindedness and Armenian ethnic identity.

Cheryl T. Grills, PhD, is a professor of clinical psychology at Loyola Marymount University. Her research focuses on African psychology and African-centered models of treatment engagement with African Americans.

References

Adams-Bass

V. N.

Bentley-Edwards

K. L.

Stevenson

H. C.

(2014). That’s not me I see on TV . . .: African American youth interpret media images of Black females. Women, Gender, and Families of Color, 2, 79-100.

Allison

(2016). Black women’s portrayals on reality television: The new Sapphire. Lanham, MD: Lexington Books.

Baldwin

J. A.

(1981). Notes on an Africentric theory of Black personality. Western Journal of Black Studies, 5, 172-179.

Baldwin

J. A.

(1984). African self-consciousness and the mental health of African-Americans. Journal of Black Studies, 15, 177-194.

Baldwin

J. A.

Bell

(1985). The African Self-Consciousness Scale: An Africentric Personality Questionnaire. Western Journal of Black Studies, 9, 61-68.

Balesh

Gamst

Meyers

L. S.

Der-Karabetian

Elias

(2018). A structural model predicting Arab American’s attitudes towards mental health services from ethnic identity, acculturation, and spirituality. Journal of Mental Disorders and Treatment, 4(2), 1-9.

Benson

M. L.

Wooldredge

Thislethwaite

A. B.

Fox

G. L.

(2004). The correlation between race and domestic violence is confounded with community context. Social Problems, 51, 326-342.

Bostock

Plumpton

Pratt

(2009). Domestic violence against women: Understanding social processes and women’s experiences. Journal of Community & Applied Social Psychology, 19, 95-110.

Bowen

N. K.

Guo

(2012). Pocket Guides to Social Work Research Methods. Structural equation modeling. New York, NY: Oxford University Press.

10.

Brown

T. A.

Moore

M. T.

(2012). Confirmatory factor analysis. In Hoyle

R. H.

(Ed.), Handbook of structural equation modeling (pp. 361-379). New York, NY: Guilford.

11.

Constantine

M. G.

Richardson

T. Q.

Benjamin

E. M.

Wilson

J. W.

(1998). An overview of Black racial identity theories: Limitations and considerations for future theoretical conceptualizations. Applied & Preventive Psychology, 7, 95-99.

12.

Cross

W. E.

(1971). The Negro to Black conversion experience: Toward a psychology of Black liberation. Black World, 20, 13-27.

13.

Cross

W. E.

(1978). The Thomas and Cross models of psychological Nigrescence: A literature review. Journal of Black Psychology, 4, 13-31.

14.

Cross

W. E.

(1995). The psychology of Nigrescence: Revising the Cross model. In Ponteroto

J. G.

Casas

J. M.

Suzuki

L. A.

Alexander

C. M.

(Eds.), Handbook of multicultural counseling (pp. 93-122). Thousand Oaks, CA: SAGE.

15.

Dana

R. H.

(1993). Multicultural assessment perspectives for professional psychology. Needham Heights, MA: Allyn & Bacon.

16.

Dana

R. H.

(2002). Mental health services for African Americans: A cultural/racial perspective. Cultural Diversity and Ethnic Minority Psychology, 8, 3-18.

17.

DeClue

(2011). Lesbian cop, queer killer: Leveraging Black queer women’s sexuality on HBO’s The Wire. Race, Sexuality and Television, 31(2), 53-62.

18.

Dickens

D. D.

Womack

V. Y.

Dimes

(2019). Managing hypervisibility: An exploration of theory and research on identity shifting strategies in the workplace among Black women. Journal of Vocational Behavior, 113, 153-163. doi:10.1016/j.jvb.2018.10.008

19.

Donovan

R. A.

West

L. M.

(2015). Stress and mental health: Moderating role of the strong Black woman stereotype. Journal of Black Psychology, 41, 384-396.

20.

Essed

(1991). Understanding everyday racism: An interdisciplinary theory. Thousand Oaks, CA: SAGE.

21.

Franklin

A. J.

Boyd-Franklin

Kelly

(2006). Racism and invisibility: Race-related stress, emotional abuse and psychological trauma for people of color. Journal of Emotional Abuse, 6, 9-30.

22.

Fremouw

W. J.

Westrup

Pennypacker

(1997). Stalking on campus: The prevalence and strategies for coping with stalking. Journal of Forensic Science, 42, 666-669.

23.

Gamst

Arellano-Morales

Meyers

L. S.

Serpas

D. G.

Balla

Diaz

Aldape

(under review). Shifting can be stressful for African American women: A structural mediational model. Journal of Black Psychology.

24.

Gamst

Arellano-Morales

Meyers

L. S.

Tolstoy

Garcia

Sjobeck

. . . Casas

(2018). Development and validation of the Latina American Shifting Scale (LASS). Gender Issues. doi:10.1007/s12147-018-9225-1

25.

Gamst

Dana

R. H.

Der-Karabetian

Aragon

Arellano

Kramer

(2002). Effects of Latino acculturation and ethnic identity on mental health outcomes. Hispanic Journal of Behavioral Sciences, 24, 479-505.

26.

Gamst

Dana

R. H.

Der-Karabetian

Aragon

Arellano

Morrow

Martenson

(2004). Cultural competency revised: The California Brief Multicultural Competence Scale. Measurement and Evaluation in Counseling and Development, 37, 163-183.

27.

Gamst

Dana

R. H.

Der-Karabetian

Kramer

(2000). Ethnic match and client ethnicity effects on global assessment and visitation. Journal of Community Psychology, 28, 547-564.

28.

Gamst

Dana

R. H.

Der-Karabetian

Kramer

(2004). Ethnic match and treatment outcomes for child and adolescent mental health center clients. Journal of Counseling and Development, 82, 457-465.

29.

Gamst

G. C.

Liang

C. T. H.

Der-Karabetian

(2011). Handbook of multicultural measures. Thousand Oaks, CA: SAGE.

30.

Grills

Longshore

(1996). Africentrism: Psychometric analyses of a self-report measure. Journal of Black Psychology, 22, 86-106.

31.

Harrell

S. P. A.

(2000). Multidimensional conceptualization of racism-related stress: Implications for the well-being of people of color. American Journal of Orthopsychiatry, 70, 42-57.

32.

Helms

J. E.

(1986). Expanding racial identity theory to cover counseling process. Journal of Counseling Psychology, 33, 62-64.

33.

Helms

J. E.

(1990). Black and White racial identity: Theory, research, and practice. Westport, CT: Greenwood Press.

34.

Helms

J. E.

(1995). An update of Helm’s White and people of color racial identity models. In Ponteroto

J. G.

Casas

J. M.

Suzuki

L. A.

Alexander

C. M.

(Eds.), Handbook of multicultural counseling (pp. 181-198). Thousand Oaks, CA: SAGE.

35.

Jackson

(1975). Black identity development. Journal of Educational Diversity, 2, 19-25.

36.

Johnson

J. C.

Gamst

Meyers

L. S.

Arellano-Morales

Shorter-Gooden

(2016). Development and validation of the African American Women’s Shifting Scale (AAWSS). Cultural Diversity and Ethnic Minority Psychology, 22, 11-25.

37.

Kambon

(1992). The African personality in America: An African-centered framework. Tallahassee, FL: NUBIAN Nation.

38.

Kaslow

N. J.

Leiner

A. S.

Reviere

Jackson

Bethea

Bhaju

. . . Thompson

M. P.

(2010). Suicidal, abused African American women’s response to a culturally informed intervention. Journal of Consulting and Clinical Psychology, 78, 449-458.

39.

Keyser

Gamst

Meyers

L. S.

Der-Karabetian

Morrow

(2014). Predictors of self-perceived cultural competence among children’s mental health providers. Cultural Diversity and Ethnic Minority Psychology, 20, 324-335.

40.

Kline

R. B.

(2015). Principles and practice of structural equation modeling (4th ed.). New York, NY: Guilford Press.

41.

Klonoff

E. A.

Landrine

Campbell

(2000). Sexist discrimination may account for well-known gender differences in psychiatric symptoms. Psychology of Women Quarterly, 24, 93-99.

42.

Landrine

Klonoff

E. A.

(1995). The African American Acculturation Scale II: Cross-validation and short form. Journal of Black Psychology, 21, 124-152.

43.

Landrine

Klonoff

E. A.

Corral

Fernandez

Roesch

(2006). Conceptualizing and measuring ethnic discrimination in health research. Journal of Behavioral Medicine, 29, 79-94.

44.

Lewis

J. A.

Williams

M. G.

Peppers

E. J.

Gadson

C. A.

(2017). Applying intersectionality to explore the relations between gendered racism and health among Black women. Journal of Counseling Psychology, 64, 475-486.

45.

Linnan

(2007). Beauty salons. Health Education & Behavior, 34, 517-530.

46.

Macy

R. J.

Giattina

M. C.

Montijo

N. J.

Ermentrout

D. M.

(2010). Domestic violence and sexual assault agency directors’ perspectives on services that help survivors. Violence Against Women, 16, 1138-1161.

47.

Mahoney

Williams

L. M.

(1998). Sexual assault in marriage: Prevalence, consequences, and treatment of wife rape. In Jasinski

J. L.

Williams

L. M.

(Eds.), Partner violence: A comprehensive review of 20 years of research (pp. 113-157). Thousand Oaks, CA: SAGE.

48.

Mize

K. D.

Shackelford

T. K.

Shackelford

V. A.

(2009). Hands-on killing of intimate partners as a function of sex and relationship status/state. Journal of Family Violence, 24, 463-470.

49.

Moradi

Subich

L. M.

(2003). A concomitant examination of the relations of perceived racist and sexist events to psychological distress for African American women. The Counseling Psychologist, 31, 451-469.

50.

Myers

L. J.

(1988). Understanding an Afrocentric world view: Introduction to an optimal psychology. Dubuque, IA: Kendall/Hunt.

51.

Nobles

W. W.

(1980). Extended self: Rethinking the so-called Negro self-concept. In Jones

R. R.

(Ed.), Black psychology (2nd ed., pp. 99-105). New York, NY: Harper & Row.

52.

Novak

J. R.

Smith

D. B.

Sandberg

J. G.

(2015). The association between the report of abuse in childhood and perceptions of safety in current adult intimate relationships. The American Journal of Family Therapy, 43, 425-440.

53.

Obasi

Leong

(2010). Construction and validation of the Measurement of Acculturation Strategies for People of African Descent (MASPAD). Cultural Diversity and Ethnic Minority Psychology, 16, 526-539.

54.

Perry

B. L.

Harp

K. L.

Oser

C. B.

(2013). Racial and gender discrimination in the stress process: Implications for African American women’s health and well-being. Sociological Perspectives, 56, 25-48.

55.

Pinheiro Rocha

Gamst

Meyers

Der-Karabetian

Magina

. (2018). Predictors of quality of life of Portuguese Americans: Identity, acculturation, and religiosity. Mental Health, Religion, & Culture, 21, 780-796.

56.

Punyanunt-Carter

N. M.

(2008). The perceived realism of African American portrayals on television. The Howard Journal of Communications, 19, 241-257.

57.

Reid-Brinkley

S. R.

(2008). The essence of res(ex)pectability: Black women’s negotiation of Black femininity in rap music and music video. Meridians, 8, 236-260.

58.

Sabri

Stockman

J. K.

Campbell

J. C.

O’Brien

Campbell

Callwood

G. B.

. . . Hart-Hyndman

(2014). Factors associated with increased risk for lethal violence in intimate partner relationships among ethnically diverse Black women. Violence and Victims, 29, 719-741.

59.

Sellers

R. M.

Rowley

S. A.

Chavous

T. M.

Shelton

J. N.

Smith

M. A.

(1997). Multidimensional Inventory of Black Identity: Preliminary investigation of reliability and construct validity. Journal of Personality and Social Psychology, 73, 805-815.

60.

Sellers

R. M.

Shelton

J. N.

(2003). The role of racial identity in perceived racial discrimination. Journal of Personality and Social Psychology, 84, 1079-1092.

61.

Sellers

R. M.

Smith

M. A.

Shelton

J. N.

Rowley

S. A. J.

Chavous

T. M.

(1998). Multidimensional Model of Racial Identity: A reconceptualization of African American racial identity. Personality and Social Psychology Review, 2, 18-39.

62.

Sharps

Campbell

J. C.

Campbell

Gary

Webster

(2003). Risky mix: Drinking, drug use, and homicide. National Institute of Justice Journal, 250, 8-13.

63.

Shepard

M. F.

Campbell

J. A.

(1992). The Abusive Behavior Inventory: A measure of psychological and physical abuse. Journal of Interpersonal Violence, 7, 291-305.

64.

Smith

D. B.

Whiting

J. B.

Karakurt

Oka

Servino

(2013). The Self Assessment of Future Events Scale (SAFE): Assessing perceptions of risk for future violence in intimate partner relationships. Journal of Marital and Family Therapy, 39, 314-329.

65.

Stevens-Watkins

Perry

Pullen

Jewell

Oser

C. B.

(2014). Examining the associations of racism, sexism, and stressful life events on psychological distress among African-American women. Cultural Diversity and Ethnic Minority Psychology, 20, 561-569.

66.

Sue

D. W.

(2010). Microaggressions in everyday life: Race, gender, and sexual orientation. Hoboken, NJ: Wiley.

67.

Sue

D. W.

Nadal

Capodilupo

Lin

Torino

G. C.

Rivera

D. P.

(2008). Racial microaggressions against Black Americans: Implications for counseling. Journal of Counseling and Development, 86, 330-338.

68.

Sullivan

C. M.

(2011). Evaluating domestic violence support service programs: Waste of time, necessary evil, or opportunity for growth? Aggression and Violent Behavior, 16, 354-360.

69.

Szymanski

D. M.

Lewis

J. A.

(2016). Gendered racism, coping, identity centrality, and African American college women’s psychological distress. Psychology of Women Quarterly, 40, 229-243.

70.

Taft

C. T.

Bryant-Davis

Woodward

H. E.

Tillman

Torres

S. E.

(2009). Intimate partner violence against African American women: An examination of the socio-cultural context. Aggression and Violent Behavior, 14, 50-58.

71.

Thomas

A. J.

Witherspoon

K. M.

Speight

S. L.

(2008). Gendered racism, psychological distress, and coping styles of African American women. Cultural Diversity and Ethnic Minority Psychology, 14, 307-314.

72.

Thomas

C. W.

(1971). Boys no more. Beverly Hills, CA: Glencoe Press.

73.

Thomas

K. A.

Goodman

Putnins

(2015). “I have lost everything”: Trade-offs of seeking safety from intimate partner violence. American Journal of Orthopsychiatry, 85(2), 170-180.

74.

Tjaden

Thoennes

(1998). Stalking in America: Findings from the National Violence Against Women Survey. Washington, DC: National Institute of Justice and Centers for Disease Control and Prevention.

75.

Wang

(2012). Structural equation modeling: Applications using Mplus. West Sussex, UK: Wiley.

76.

West

C. M.

(2004). Black women and intimate partner violence: New directions for research. Journal of Interpersonal Violence, 19, 1487-1493.

77.

West

C. M.

(2018). Mammy, Sapphire, Jezebel, and the bad girls of reality television: Media representation of Black women. In Chrisler

Golden

(Eds.), Lectures on the psychology of women (5th ed., pp. 139-158). Long Grove, IL: Waveland Press.

78.

White

J. L.

Parham

T. A.

(1990). The psychology of Blacks: An African-American perspective. Englewood Cliffs, NJ: Prentice-Hall.

79.

Williams

O. J.

Oliver

Pope

(2008). Domestic violence in the African American community. Journal of Aggression, Maltreatment and Trauma, 16, 229-237.

80.

Wilson

S. L.

Sellers

Solomon

Holsey-Hyman

(2017). Exploring the link between Black racial identity and mental health. Journal of Depression and Anxiety, 3, 1-4.

81.

Woods-Giscombé

C. L.

(2010). Superwoman schema: African American women’s views on stress, strength, and health. Qualitative Health Research, 20, 668-683.