CULTURAL norms about weight may influence beliefs African American women hold about their weight.1 These beliefs may influence their weight management behaviors.1 , 2 Many African American women choose a heavier body size as ideal3 , 4 and perceive body weight as smaller than their actual body mass index (BMI).5–8 These cultural norms reflect beliefs and values that are learned.9 Understanding the relationship between cultural weight beliefs, weight management behaviors, and actual weight in African American women has been limited by the availability of a culturally related questionnaire. Although many weight belief questionnaires exist, the majority measure general rather than personal weight beliefs and focus specifically on obesity10–13 or body appearance.14 , 15 These measures examine important aspects of weight beliefs, yet they do not consider the cultural-related beliefs about personal weight across the weight spectrum.1
This gap in the literature was addressed with the original development and testing of the Beliefs About Personal Weight Survey (BPWS).1 The original BPWS is a 65-item self-report instrument that contained items that measured 3 dimensions (descriptive characteristics, causal attributions, and consequences) of one's personal weight based on Fishbein's seminal work regarding belief formation.1 , 16 , 17 The initial items underwent principal component analysis to form 16 multi-item scales with 57 retained items. An example of a one of the multi-item scales is overweight descriptors consisting of 4 items (heavy, very overweight, chunky, and overweight). We originally hypothesized that these multi-item scales would fit a 3-factor model corresponding to the 3 discrete belief dimensions (descriptors, causes, and consequences)1 Confirmatory factor analysis did not support a 3-factor model. Exploratory confirmatory factor analysis was used to empirically determine alternative solutions that also made clinical sense. The obtained solution consisted of 2 body image factors (overweight acceptance and overweight concern) and 2 weight regulation factors (conventional and circumstantial)1 The overweight acceptance factor refers to overweight without negative attributions about appearance, social, emotional, or health consequences.1 The overweight concern factor refers to negative attributions of appearance with adverse social, emotional, and health consequences.1 Conventional weight regulation factor refers to physical activity and healthy eating, resulting in good health consequences.1 Circumstantial weight regulation factor refers to weight due to life circumstances thought to be outside of one's personal control or external factors.1 These factors were significantly correlated with eating behaviors, physical activity, and BMI.1 Analysis of the original BPWS revealed a promising instrument, with 4 factors composed of multi-item scales.1 In order for the survey to be clinically practical, more examination of the newly found factors was needed along with item reduction, refinement and analysis must be accomplished.
Statements of Significance
What is known or assumed to be true about this topic:
Culturally related beliefs about personal weight are thought to contribute to the high prevalence of overweight and obesity among African American women. However, no tools existed that measured beliefs about personal weight among young African American women. This gap in the literature was addressed with the initial development and testing of the BPWS.
What this article adds:
This article is significant for nursing and extends nursing knowledge by providing a refined tool to measure beliefs about personal weight that may be used in clinical and research environments to identify personal weight beliefs that may support or hinder personal weight management behaviors among young African American women.
The purpose of the study was to revise the BPWS by (a) developing additional items that explicitly tapped the identified 4 factors, and (b) examining evidence of internal structure and relationship with other variables with a larger sample of African American women.
The study used a quantitative, cross-sectional design to test the revised BPWS-2.
A nonprobability sampling plan was used to recruit a community-based sample of 200 premenopausal women, 18 to 40 years of age, who self-identified as African American. Inclusion criteria included not pregnant or breastfeeding, not having a medical diagnosis of diseases known to cause weight gain (eg, hypothyroidism, Cushing's syndrome, and polycystic ovarian syndrome), or taking medications known to influence weight (steroids, antiseizure medication, and antipsychotic) and no intellectual impairment as measured by the Short Portable Mental Status Questionnaire.18 The premenopausal age range was chosen to limit confounders of weight changes associated with menopause and increased risk of illness with age. Women who were pregnant and breastfeeding were excluded due to weight instability during pregnancy and breastfeeding. Participants with medical conditions or currently taking medications that are known to influence weight were excluded to control for a potential bias in weight beliefs due to the medical condition or medication-related weight changes.
The sample participants (N = 200) were African American women, primarily younger than 30 years (mean = 27.3 years; standard deviation = 6.8 years), with half of the sample having some college education (50.5%, n = 101). The majority of participants were employed either full-time (31.5%, n = 63) or part-time (26%, n = 53). More than half of the sample (65%, n = 136) had incomes less than $15 000 annually. The sample included women across the full weight spectrum. Five women (2.5%) were underweight (BMI ≤ 18.0), 65 (31.5%) were normal weight (BMI 18-24.9), and 39 (19.5%) were overweight (BMI 25-29.9). Nearly half of the sample was obese: 41 (20.5%) were grade 1 (BMI 30-34.9), 26 (13%) were grade 2 (BMI 35-39.9), and 23 (11.5%) were grade 3 (BMI ≥ 40).
Instruments and measures
Beliefs About Personal Weight Survey-2
The BPWS-2 examined personal weight beliefs among African American women. The complete 53-item survey is shown in the Supplemental Digital Content, Appendix A, available at: http://links.lww.com/ANS/A11. The construction of these items and how they were separated into 3 item pools is explained. A careful review of the 57 items from the original BPWS resulted in 18 items being deleted due to weak component loadings and redundancy in content due to items that were nearly identical but reversely worded. Twenty-seven items were retained as originally tested and 12 original items were retained but reworded to reduce their reading level to the seventh grade. Fourteen new items were added. One was needed to further assess physical activity, while 13 were developed to tap the new, 4-factor, solution identified by analyses of the initial BPWS.
Thus, the BPWS-2 contained a total of 53 items that were separated into 3 pools to facilitate psychometric analysis. These pools integrated the original 3 dimensions within the 4 factors. Specifically, pool 1 included items related to overweight acceptance, reflecting the descriptive dimension. Pool 2 contained items for the 2 weight regulation factors, conventional and circumstantial, reflecting the causes dimension, while pool 3 items addressed consequences as captured within the overweight concern factor. These item pools are shown in Table 1. It was hypothesized that the 4 factors identified in the initial BPWS (Overweight acceptance, overweight concern, conventional weight regulation, and circumstantial weight regulation) wound reemerge in the BPWS-2 with the newly written items contributing to each of the 4 factors.
The BPWS-2 retained the original 5-point Likert-type scale format, with responses ranging from 1 (strongly disagree) to 5 (strongly agree). Survey instructions asked participants to indicate what they believed about their current personal weight by choosing how much they agreed or disagreed with the survey items. The survey format facilitated administration and reduced respondent burden because all items in each of the hypothesized 4 factors had a common stem. For example, the common stem for items in the overweight acceptance factor was descriptive in nature: “I believe my weight is:” with item examples of heavy, attractive, and just right for my age. The common stem for items in the 2 weight regulation factors was, “I believe my current weight is due to:” Examples of item with this stem include stress and worry, family traditions, and my emotions. The common stem for items in the overweight concern factor was, “I believe my current weight has led to or will lead to.” Item examples for this factor include being tired all the time, breathing problems, and problems finding stylish clothes. An example of new items that were hypothesized to contribute to each of the 4 factors includes comfortable with my weight, I am too busy to eat right, weight control is important to me, and my weight has always been a problem for me.
Four measures were used to provide evidence of validity of the BPWS-2 based on relationships with other variables. The measured variables were body image, BMI, eating behavior patterns, and psychoemotional factors (depression and perceived stress). Demographic data and height and weight were also measured.
Body image is defined as perception of one's body, which includes body weight. The Pulver's Figure Rating Scale (PFRS)19 was developed to measure body image perception in African American men and women. In the current study, the PFRS was used to examine convergent validity of the BPWS-2. The PFRS consists of 9 female and 9 male silhouette drawings. The female silhouettes were used for this study. The 9 figures were arranged in order of body size from the smallest (A) to the largest (I). The body sizes correspond to an estimated BMI of 16 to 40. Participants were asked to choose the figure that matched their body size.19
Body mass index
BMI is an objective calculation of one's wei-ght category (underweight normal weight, overweight, and obesity). Participants' body weight was measured with the participant wearing light clothing using a standard digital calibrated scale. BMI was calculated20 using weight measured to the nearest 0.5 lb and height measured to the nearest 0.5 using a portable stadiometer (Seca 217).
Eating behaviors are associated with weight management. Eating behaviors were measured using the Eating Behavior Pattern Questionnaire (EBPQ).21 In the current study, the EBPQ was used to measure concurrent validity with the BPWS-2. The EBPQ contains 6 dimensions of eating behavior that includes (a) low-fat eating, (b) snacking on sweets, (c) cultural/lifestyle behaviors, (d) haphazard planning, (e) meal skipping, and (f) emotional eating. The EBPQ is a 51-item, self-administered survey using a 5-point Likert-type scale with responses ranging from 1 (strongly disagree) to 5 (strongly agree). The instrument was initially validated using 278 adult African American women. Subscale Cronbach αs range from 0.70 to 0.88. Construct validity of the EBPQ was supported by each subscale being correlated with total energy and total fat intake (r = 0.23-0.53).21 Internal consistencies measured with the Cronbach α for the subscales from our study sample range from 0.87 to 0.63.
Depression and perceived stress are associated with eating behaviors, thus potentially able to influence weight.1 , 22–24 Psychosocial factors including depression and perceived stress have been conceptualized as foundational to the ability to perform self-care behaviors and indirectly affect eating behaviors, weight, and overall health.1 , 25 , 26 Depression symptoms were measured using the Patient Health Questionnaire.27 In the present study, the Patient Health Questionnaire-9 (PHQ-9) was used to examine concurrent validity of the BPWS-2. The PHQ-9 was developed to measure depression symptoms over the previous 2 weeks based on the Diagnostic and Statistical Manual of Mental Disorder, Fourth Edition (DSM-IV). The PHQ-9 is a 9-item questionnaire with 4-point Likert-type responses ranging from 1 (not at all) to 4 (nearly every day). Scores range from 0 to 27, with higher scores indicating more depressive symptoms, where 0 to 4 is considered normal and 24 to 27 severely depressed. The PHQ-9 has been used with African American samples, with the Cronbach α reliability ranging from 0.86 to 0.89.28 , 29 The Cronbach α for the current sample was 0.87.
Perceived stress was measured using the Perceived Stress Scale (PSS).30 The PSS measures the degree to which situations in one's life are appraised as being stressful over the past month. The PSS was used to measure concurrent validity with the BPWS-2. Perceived stress, a psychosocial factor, conceptualized as being foundational to the agency for self-care behaviors. The PSS is a 10-item self-administered scale using Likert-type ratings with responses ranging from 0 (never) to 4 (Very often). A single total sum score is created by summing all items (after reverse scoring as appropriate). Scores range from 0 to 40, with higher scores indicating greater perceived stress. The Cronbach α was 0.81. The PSS has been used with African American samples.28 The Cronbach α for the current sample was 0.80.
A researcher-developed demographic form was used to collect participants' race, age, years of education, and annual income.
Following institutional review board approval, participants were recruited using recruitment flyers, and word of mouth. Data collection was done in private rooms at a daycare center, an apartment complex, and 1 university in North Carolina, as well as 1 university in Pennsylvania. Data collection was done in-person at each site. After trained data collectors obtained informed consent, participants completed 6 questionnaires, including a demographic form, and had their height and body weight measured. The participants received $15.00 as a “thank you” for their time and effort in completing the study.
A common factor measurement model31 was used to evaluate the psychometric properties of items in each of 3 item pools (see Table 1). The common factor measurement model is a confirmatory factor analysis (CFA) model, with each item contributing to a single factor and with item residuals independent across factors. Thus, the number of factors and fixed zero loadings are hypothesized in advance and the size and statistical significance of the factor loadings are determined by the data. Items within the pool were omitted from the factor if they did not load significantly or if the residuals were highly correlated. Pool 1 hypothesized a 1-factor model, but 2 factors were needed to achieve homogeneous scales that represented the items. Pool 2 hypothesized and fit a 2-factor CFA, and pool 3 hypothesized and fit a 1-factor CFA. The models were estimated using full-information maximum likelihood with AMOS software (IBM SPSS). Modification indices (Lagrange multipliers) were used to identify correlated error terms that would potentially affect scale homogeneity. Items with nonsignificant factor loadings and items with multiple correlated error components were omitted from the model. The one exception to this was a set of intercorrelated items on the overweight acceptance factor (pool 1), which was remodeled as a second factor scale. A total of 18 items were removed from the models. Model fit was evaluated using a 2-criterion rule.32 Hu and Bentler32 proposed using the standardized root mean squared residual (SRMR) and 1 additional measure of fit. Because the root mean square error of approximation (RMSEA) is generally regarded as one the most informative of fit indices, we elected to go with it in addition to the SRMR. The recommended cutoff values for this 2-index strategy were an SRMR of 0.09 or less, and an RMSEA or 0.06 or less.32 Other model fit indices were also performed including χ2 and comparative fit index (see model fit in Table 2).
The BPWS-2 scales were computed by summing all items that loaded on their respective factors. These are shown in Table 3. Internal consistency reliability (Cronbach α) of each scale was assessed. Evidence for validity was examined by correlating the BPWS-2 scales with the PFRS, BMI, eating behaviors (EBPQ subscales), depression (PHQ), and perceived stress measures (PSS)
To establish the validity of the BPWS-2, we determined the degree to which the evidence supported the intended interpretation of survey results. Evidence based on internal structure, and relationship to other variables, was evaluated.33
Evidence of validity based on internal structure
Confirmatory factor analysis was conducted to assess the relationship of the items to the underlying factors being evaluated. A good fitting measurement model was identified for each pool of items (RMSEA 0.047 [confidence interval 0.022-0.068] to RMSEA 0.059 [confidence interval 0.036-0.080]); see Table 2. However, we were unable to verify the hypothesized overweight acceptance factor from items in pool 1. The items in this pool split into 2 factors; the factors were named weight acceptance and excess weight acknowledgment. As expected the weight regulation items (pool 2) supported a 2-factor model with items for circumstantial weight regulation and conventional weight regulation. Also, as expected, the items in pool 3 (weight concern) defined the overweight concern factor.
Based on model fitting indices, the BPWS-2 was reduced from 53 to 35 items on 5 factors: weight acceptance (8 items), excess weight acknowledgment (5 items), conventional weight regulation (3 items), circumstantial weight regulation (9 items), and weight concern (10 items). The identified factors with their items are viewed as subscales of the BPWS-2.
Subscale scores and reliability
The BPWS-2 subscales were computed by taking the simple sum of items that loaded on each factor. The items making up each scale are shown in Table 3. Correlations among the scales ranged from −0.11 to 0.61, with a median correlation of 0.44. Scale means, standard deviations, and reliabilities are shown in Table 4. Internal consistency for the BPWS-2 scales, measured as Cronbach α reliabilities, ranged from 0.67 to 0.89.
Evidence of validity based on relationships with other variables
Correlations between the BPWS-2 scales and study measures are shown in Table 4. BMI was significantly associated with each of the BPWS-2 scales with the pattern of positive and negative correlations that made good theoretical sense (Table 4). The PFRS was significantly correlated with each of the BPWS-2 scales in the same way as BMI. This was not surprising, given the high correlation between the BMI and the PFRS (r = 0.85, P < .001). These results show that BPWS-2 scales represent both favorable and unfavorable weight belief constellations.
The psychosocial measures (depression and perceived stress) were significantly correlated in the same way as the BMI and PFRS (Table 4), indicating that favorable belief constellations (weight acceptance and conventional weight regulation) are positively associated less depression and stress, and negative belief constellations (excess weight acknowledgment, circumstantial weight regulation, and weight concern) are associated with more stress and greater depression.
The correlations of the EBPQ with BPWS-2 scales (Table 5) fit 2 distinct patterns. As noted earlier, weight acceptance and conventional weight regulation may be regarded as positive belief constellations. Examination of these scales in correlation to the EBPQ scales revealed that they were correlated with the EBPQ in the same way. Both were significantly positively associated with low-fat eating and negatively associated with emotional eating and haphazard mean planning (eat out frequently). One small exception was the correlation with mean skipping; it was significant for weight acceptance (r = 0.16, P < .05) and not significant for conventional weight regulation (r = −11, P > .05).
Similarly, the 3 scales associated with less favorable belief constellations—excess weight acknowledgment, circumstantial weight regulation, and weight concern—were significantly positively correlated with emotional eating, snacking on sweets, haphazard planning (eating out frequently), and meal skipping. One small difference in this pattern was noted; cultural/lifestyle eating behavior (large meals with meat) was more strongly associated with circumstantial weight regulation and weight concern than with excess weight acknowledgment.
In this study we reported the revisions and psychometric evaluation of the BPWS-2 with a sample of young African American women. The goal of this study was to strengthen the measurement of the 4 newly identified factors that emerged from the original BPWS and to reduce and refine the original items. The BPWS-2 was improved by addressing the limitations of the original survey. The BPWS-2 items were reduced to the 7th-grade reading level, which supports the recommended reading level of self-administered surveys.34 A physical activity item was added to the revised survey (not getting enough exercise) as the original survey contained 1 physical activity item (regular exercise). However, only 1 exercise item (regular exercise) was retained after analysis due to model fit indices. Overall, the BPWS-2 retained 28 of the original items and 7 new items, reducing the survey from 53 to 35 items. This shortened version increases the usefulness of the survey in clinical and research settings while maintaining the integrity of the beliefs about personal weight construct.
There were 3 key findings from the psychometric evaluation of the BPWS-2. First, the findings indicate that the beliefs about personal weight construct represented by BPWS-2 factors were similar to the original 4 factors (weight acceptance, weight concern, conventional weight regulation, and circumstantial weight regulation) that reemerged as hypothesized. This factor structure supports the validity of the BPWS-2, in that similar factors were found in 2 different samples representing 350 African American women. The excess weight acknowledgment factor emerged as a fifth factor from this current sample of African American women. This fifth factor represents a distinction between acknowledgment of excessive weight (beliefs about personal weight being overweight, very overweight, chucky, and heavy) and acceptance of weight (beliefs about personal weight being attractive, just write for my age, and normal).
Second, the evaluation of the 5 factors representing the BPWS-2 achieved good internal consistency and acceptable evidence of validity based on significant relationships with other variables measuring body image, BMI, eating behavior patterns, and psychosocial factors (depression and perceived stress). The validity evidence indicated that the 5 factors representing beliefs about personal weight were able to distinguish between certain eating behaviors as well as BMI. For example, conventional weight regulation beliefs (eg, items—regular exercise, living a healthy lifestyle) were associated with low-fat eating and lower BMIs, while circumstantial weight regulation beliefs (eg, items—stress and worry, emotions) were associated with emotional eating, haphazard meal planning, snacking on sweets, and higher BMIs. Consistent with findings from the original BPWS,1 the BPWS-2 contains 4 factors that support a personal weight belief profile that distinguishes between favorable and unfavorable eating behaviors and BMI.
The new factor on the BPWS-2, excess weight acknowledgment (eg, items—very overweight, heavy, chucky) was closely associated with body image and BMI but may contain components that may be unique to the factor. This novel finding may indicate that identifying one's body silhouette (body image) and/or BMI as overweight or obese may be somewhat different from acknowledging one's personal weight as being excessive. This finding is of note because the acknowledgment of excessive weight was not associated with eating behaviors (low-fat eating) leading to weight loss but was associated with eating behaviors leading to increased BMI. This finding may indicate that acknowledgment of excess weight may not be motivation for weight management behaviors among African American women. More research is needed to examine acknowledgment of excess weight and weight management behaviors in this population.
Third, results revealed that all the BPWS-2 subscales were significantly associated with emotional eating. Emotional eating is characterized as consumption of highly palatable foods (high-fat, high-sugar, high-caloric foods) in response to stress and worry and negative emotions.35 This definition is consistent with our study findings in which participants believed that stress and worry are causes of their personal weight as indicated by association with the circumstantial weight regulation subscale. The BPWS-2 subscales that were negatively associated with emotional eating were more conventional beliefs about weight management, which includes describing weight as normal and well-proportioned (weight acceptance) and believing cause of weight was regular exercise and eating low-fat foods (conventional weight regulation). The association of the BPWS-2 subscales with emotional eating is particularly important for African American women because emotional eating is noted as a factor that contributes to reduced weight loss in weight management interventions.
The overall evaluation of the BPWS-2 subscales with other variables indicate that the subscales were sensitive enough to determine unique eating behaviors and psychosocial factors that may influence BMI among African American women. The development and analysis of BPWS-2 advances the existing knowledge regarding beliefs about personal weight and its relationship with weight management behaviors among African American women.
As in all studies, this study has limitations. The n/k (sample size/number of items) ratio was lower than usually expected, making comprehensive assessment of the hypothesized factors difficult. Therefore, the hypothesized factor models were assessed using item pools. Much of the study data were collected using self-report questionnaires, which may increase the chance of participants providing socially desirable responses.36 This bias, however, was minimized by informing participants that there were no right or wrong responses and having participants self-administer the survey. Also, nonprobability sampling was used, which prevents the findings from being generalized beyond the sample and may have introduced selection bias. Selection bias may have been minimized by using a community sample from a number of different sites.35
The BPWS-2 measures an important new dimension for weight control. Understanding the role of personal weight beliefs may be a critical component in developing effective weight management strategies. Further research is needed to test the BPWS-2 in larger samples of African American women, as well as to expand and test the instrument in men, and other racial and ethnic groups. Research is also needed to test the feasibility of using the BPWS-2 in clinical settings, as well as to determine whether targeting weight management strategies to personal weight belief results increased weight loss compared with standard interventions.
Results demonstrated the reliability and validity of the 5-factor BPWS-2. The identified subscales showed a pattern of association with eating behaviors, BMI, and body image that may have important clinical implications to be tested in future research. Given that the BPWS-2 is the only instrument designed to specifically measure beliefs about one's personal weight, the BPWS-2 makes an important contribution to the literature and to understanding weight management in African American women.
1. Pickett S, Peters RM, Templin T. Development and validation of the beliefs about personal weight survey among African American women
. Adv Nurs Sci. 2017;40(2):E13–E27. doi:10.1097/ANS.0000000000000155.
2. Altabe M, Thompson JK. Body image: a cognitive self-schema construct? Cogn Ther Res Cogn Ther Res. 1996;20(2):171–193.
3. Chithambo TP, Huey SJ. Black/White differences in perceived weight and attractiveness among overweight women. J Obes. 2013;2013:320326. doi:10.1155/2013/320326.
4. Kronenfeld LW, Reba-Harrelson L, Von Holle A, Reyes ML, Bulik CM. Ethnic and racial differences in body size perception and satisfaction. Body Image. 2010;7(2):131–136. doi:10.1016/j.bodyim.2009.11.002.
5. Lynch EB, Kane J. Body size perception among African American women
. J Nutr Educ Behav. 2014;46(5):412–417. doi:10.1016/j.jneb.2014.03.002.
6. Hendley Y, Zhao L, Coverson DL, et al Differences in weight perception among blacks and whites. J Womens Health (Larchmt). 2011;20(12):1805–1811. doi:10.1089/jwh.2010.2262.
7. Sutcliffe CG, Schultz K, Brannock JM, Giardiello FM, Platz EA. Do people know whether they are overweight? Concordance of self-reported, interviewer-observed, and measured body size. Cancer Causes Control. 2015;26(1):91–98. doi:10.1007/s10552-014-0487-y.
8. Ard JD, Zunker C, Qu H, et al Cultural perceptions of weight in African American and Caucasian women. Am J Health Behav. 2013;37(1):3–13. doi:10.5993/AJHB.37.1.1.
10. Crandall CS, Martinez R. Culture, ideology, and antifat attitudes. Personal Soc Psychol Bull. 1996;22(11):1165–1176. doi:10.1177/01461672962211007.
11. LeBow MD, Ness D, Makarenko P, Lam T. Attitudes, perceptions and practices of Canadian teenagers towards obesity
. J Obes Weight Regul. 1989;8:53–65.
12. Janssen I, Katzmarzyk PT, Ross R. Waist circumference and not body mass index explains obesity
-related health risk. Am J Clin Nutr. 2004;79(3):379–384. doi:10.1111/j.1365-2265.2012.04487.x.
13. Brown TA, Cash TF, Mikulka PJ. Attitudinal body-image assessment: factor analysis of the Body-Self Relations Questionnaire. J Pers Assess. 1990;55(1/2):135–144. doi:10.1080/00223891.1990.9674053.
14. Spangler DL, Stice E. Validation of the beliefs about appearance scale. Cognit Ther Res. 2001;25(6):813–827. doi:10.1023/A:1012931709434.
15. Durso LE, Latner JD. Understanding self-directed stigma: Development of the weight bias internalization scale. Obesity
(Silver Spring). 2008;16(suppl 2):S80–S86. doi:10.1038/oby.2008.448.
16. Fishbein M. An investigation of the relationships between beliefs about an object and the attitude toward that object. Hum Relations. 1963;16(3):233–239. doi:10.1177/001872676301600302.
17. Fishbein M, Ajzen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975:1–18. doi:10.1017/CBO9781107415324.004.
18. Roccaforte WH, Burke WJ, Bayer BL, Wengel SP. Reliability and validity of the Short Portable Mental Status Questionnaire administered by telephone. J Geriatr Psychiatry Neurol. 1994;7(1):33–38. doi:10.1177/089198879400700107.
19. Pulvers KM, Lee RE, Kaur H, et al Development of a culturally relevant body image instrument among urban African Americans. Obes Res. 2004;12(10):1641–1651. doi:10.1038/oby.2004.204.
20. Jensen MD, Ryan DH, Apovian CM, et al 2013 AHA/ACC/TOS Guideline for the Management of Overweight and Obesity
in Adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Obesity
Society. Circulation. 2014;129(25, suppl 2):S102–S138. doi:10.1161/01.cir.0000437739.71477.ee.
21. Schlundt DG, Hargreaves MK, Buchowski MS. The Eating Behavior Patterns Questionnaire predicts dietary fat intake in African American women
. J Am Diet Assoc. 2003;103(3):338–345. doi:10.1053/jada.2003.50049.
22. Sims R, Gordon S, Garcia W, et al Perceived stress and eating behaviors in a community-based sample of African Americans. Eat Behav. 2008;9(2):137–142. doi:10.1016/j.eatbeh.2007.06.006.
23. Stevenson RJ. Psychological correlates of habitual diet in healthy adults. Psychol Bull. 2017;143(1):53–90. doi:10.1037/bul0000065.
24. Whitaker KM, Sharpe PA, Wilcox S, Hutto BE. Depressive symptoms are associated with dietary intake but not physical activity among overweight and obese women from disadvantaged neighborhoods. Nutr Res. 2014;34(4):294–301. doi:10.1016/j.nutres.2014.01.007.
25. Orem DE, Dorothea E, Taylor SG, Renpenning KM. Nursing: Concepts of Practice. St Louis, MO: Mosby; 2001.
26. Pickett S, Peters RM, Jarosz PA. Toward a middle–range theory of weight management. Nurs Sci Q. 2014;27(3):242–247. doi:10.1177/0894318414534486.
27. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–613.
28. Boyle GJ, Saklofske DH, Matthews G. Criteria for selection and evaluation of scales and measures. In: Measures of Personality and Social Psychological Constructs. Amsterdam, the Netherlands: Elsevier; 2015:3–15. doi:10.1016/B978-0-12-386915-9.00001-2.
29. Huang FY, Chung H, Kroenke K, Delucchi KL, Spitzer RL. Using the Patient Health Questionnaire-9 to measure depression among racially and ethnically diverse primary care patients. J Gen Intern Med. 2006;21(6):547–552. doi:10.1111/j.1525-1497.2006.00409.x.
30. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24(4):385. doi:10.2307/2136404.
31. McDonald RP. Test Theory: A Unified Treatment. New York, NY: Taylor & Francis; 1999.
32. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model A Multidiscip J. 1999;6(1):1–55. doi:10.1080/10705519909540118.
34. Streiner DL, Norman GR, Cairney J. Health Measurement Scales: A Practical Guide to Their Development and Use. New York, NY: Oxford University Press; 2015.
35. Kukull WA, Ganguli M. Generalizability: the trees, the forest, and the low-hanging fruit. Neurology. 2012;78(23):1886–1891. doi:10.1212/WNL.0b013e318258f812.
36. Van de Mortel T. Faking it: social desirability response bias in self- report research Australian Journal of Advancing. Nursing. 2008;25(4):40–48.