Breast cancer is the leading cause of cancer deaths among Hispanic women in the United States. Unfortunately Hispanic women exhibit poor mammography screening participation, are diagnosed at later stages of the disease, and have lower survival rates than non-Hispanic white women. Several cultural and psycho-social factors have been found to influence mammography screening participation among Hispanic women. We will begin by presenting the theoretical framework that grounded this research program to develop an instrument to assess factors contributing to poor mammography participation among Hispanic women. We will also summarize the early stages in the development of the English and Spanish Mammography Beliefs and Attitudes Questionnaire (MBAQ and SMBAQ) for use with low-health-literacy Mexican-American women. Next we will describe the initial psychometric testing of the MBAQ/SMBAQ, after which we will present the psychometric testing of the SMBAQ with low-health-literacy women. This will be followed by a discussion of the modification of the MBAQ and SMBAQ subscales. We’ll conclude with a discussion of the instruments and share our assessment regarding the limitations of this research program, where the program stands to date, and the implications for practice and future research.
Keywords: breast cancer, instrument development for low-health-literacy individuals, theory of planned behavior applied to mammography screening, Spanish translation, cultural adaptation of health-related instruments
The [MBAQ] and the [SMBAQ] are culturally appropriate and reliable tools...in assessing factors that most strongly impact the intention to seek mammography screening among low-health-literacy Mexican-American women. Breast cancer is the leading cause of cancer deaths among Hispanic women in the United States (US). However, only 58% of Hispanic women over the age of 40 in the US participate in regular mammography screening (American Cancer Society [ACS], 2007; Centers for Disease Control and Prevention [CDC], 2007; Intercultural Cancer Council [ICC], 2005). The percentage of low-income, low-health-literacy Hispanic women over the age of 40 who participate in regular mammography screening is even lower, with only 38% obtaining regular mammograms. This is in spite of federally subsidized programs that finance mammograms for this population. Hispanic women are also more likely to be diagnosed at a later stage of the disease and have lower survival rates than non-Hispanic white women (ACS, 2007; Fernandez, Palmer, & Leong-Wu, 2005; Li, Malone, & Daling, 2003; O’Brien et al., 2003)
Several cultural and psycho-social factors that impact the participation of Hispanic women in mammography screening have been identified in the literature. Some of these factors include low socio-economic levels, low educational levels, low English proficiency, attitudes of fatalism, religious and spiritual beliefs, embarrassment, and partner or family disapproval (ACS, 2007; ICC, 2005; O'Brien et al., 2003).
The purpose of this article is to describe the validation of a bilingual instrument (English and Spanish) that measures factors impacting the intentions of low income, low-health-literacy Mexican-American women to seek mammography screening (Lopez-McKee, 2010). We will begin by presenting the theoretical framework that grounded this research program to develop an instrument to assess factors contributing to poor mammography participation among Hispanic women. We will also summarize the early stages in the development of the English and Spanish Mammography Beliefs and Attitudes Questionnaire (MBAQ and SMBAQ) for use with low-health-literacy Mexican-American women. Next we will describe the initial psychometric testing of the MBAQ/SMBAQ, after which we will present the psychometric testing of the SMBAQ with low-health-literacy women. This will be followed by a discussion of the modification of the MBAQ and SMBAQ subscales. We’ll conclude with a discussion of the instruments and share our assessment regarding the limitations of this research program, where the program stands to date, and the implications for practice and future research. The focus of this article will be on describing the quantitative methods we utilized to validate the psychometric properties of the English and Spanish MBAQ.
Theoretical Framework
The opinions of family members, friends, and physicians significantly impact Hispanic women’s participation in mammography screening. The theory of planned behavior (TPB) (Ajzen,1985) was utilized to guide item development for the MBAQ because it is a theory that links attitudes and behavior. The TPB is a persuasive ‘prediction’ theory, which has been applied in various disciplines to predict a specific behavior. The TPB is an appropriate framework for the development of the MBAQ because it provides an item-development format that includes psychosocial factors as predictors of the intention to participate in health-seeking behaviors.
The TPB is based on psycho-social factors that are referred to as ‘behavioral determinants,’ such as ‘attitudes,’ ‘social norms,’ and ‘perceived behavioral control.’ These behavioral determinants are considered to be powerful predictors of the intention to perform a specific behavior. The cultural and psycho-social factors found to impact mammography screening among low-health-literacy Mexican-American women were identified by a review of the literature (see below) and were used as the basis for the development of items related to ‘attitudes,’ ‘social norms,’ and ‘perceived behavioral control’ for the MBAQ.
One cultural factor that contributes to poor participation in mammography screening by Hispanic women is the fatalistic attitude held by many Hispanic women toward cancer and cancer screening. One cultural factor that contributes to poor participation in mammography screening by Hispanic women is the fatalistic attitude held by many Hispanic women toward cancer and cancer screening. A fatalistic attitude is a feeling that one has no control over preventing cancer, along with a pessimistic attitude about the value of seeking screening. Several studies have reported that fatalistic attitudes held by many Hispanic women, especially those who are low income and have a low educational level, have a negative impact on their mammography screening participation (Deterborn, DuHamel, Butts, Thompson, & Jandorf, 2005; Fernandez et al., 2005; Lopez-McKee, McNeill, & Bader, 2008).
Another cultural factor found to influence the decision of Mexican-American women to obtain mammography screening was the influence of ‘significant others.’ The opinions of family members, friends, and physicians significantly impact Hispanic women’s participation in mammography screening. Hispanic women participate in breast and cervical cancer screening more often if their family or friends are supportive; they also tend to participate in mammography screening more often when their physician recommends and encourages them to do so (Otero-Sabogal, Stewart, Brown, & Perez-Stable, 2003).
Other psycho-social factors also impact Hispanic women’s participation in breast cancer screening. These factors include: attitudes about the value of obtaining a mammogram; their perceived risk of developing breast cancer in their lifetime; perceptions of control over their own breast health; fear of undesired findings from mammography screening; religious or spiritual beliefs toward cancer; procrastination; and/or embarrassment (Carter, Park, Moadel, Cleary, & Morgan, 2002).
Description of Early Stages of the Research Program
This research program has utilized a measurement research design, along with certain qualitative-design components. IRB approval to conduct all stages of this research program was obtained from The University of Texas at El Paso.The geographic area for this study was El Paso, Texas, a mid-size city located on the U.S.-Mexico border, whose Hispanic population of Mexican descent is approximately 81% of the total population.
Development of Items on the MBAQ
Items on the English MBAQ were developed by a committee consisting of three oncology nurses who were of Mexican-American descent, bilingual in English and Spanish, and had worked with low-health-literacy Mexican-American women in their clinical practice. This committee was named the ‘Expert Committee.’ Permission was obtained to utilize an existing instrument, the Beliefs and Attitudes Questionnaire (BAQ) by Young, Lierman, Powell-Cope, and Kasprzyk (1991) as a model to develop the items of the MBAQ. The BAQ was selected as a model for developing the MBAQ because this instrument also utilized the theory of planned behavior (TPB) as the theoretical framework for its development. The BAQ applied the theoretical concepts from the TPB to breast self-exam, (BSE) and identified important content domains for determining a woman’s perceptions about her own breast health. The 65-item BAQ was also selected because of its reliability coefficients (0.70-0.86) in previous studies. The BAQ included items dealing with BSE attitudes, perceived behavioral control, behavioral beliefs, and normative beliefs; it utilized a 7- point Likert-type scale for each item.
The Expert Committee utilized a five-step process in developing the items for the MBAQ (Francis, Eccles, Johnston, Walker, Grimshaw, & Foy, 2004) based on the theory of planned behavior and on guidelines for TPB item development outlined by Ajzen (1985). The MBAQ included 25 items, each having four response options. The response options on the MBAQ were also based on the recommendations by Francis et al. (2004) for the development of TPB questionnaires. Response options were structured in a semantic differential manner that measured people's reactions to stimulus words and concepts in terms of ratings on a bipolar scale. For example, adjective pairs such as valuable/worthless and harmful/beneficial, as well as good/bad were utilized. Four different response options were offered between every set of bipolar adjectives on the MBAQ, with one option being the response option of don’t know (Francis et al.).
Spanish Translation of the MBAQ
The Mammography Beliefs and Attitudes Questionnaire was translated from English into Spanish utilizing a translation model (Lopez-McKee, 2005) that included the forward translation of the MBAQ by three independent certified translators and back-translation by another three independent translators. The Expert Committee then selected the ‘consensus’ Spanish version of each item on the MBAQ that best represented the constructs being measured in the English MBAQ. The Spanish version of the MBAQ was named the Spanish Mammography Beliefs and Attitudes Questionnaire (SMBAQ).
Cultural Adaptation
The Mammography Beliefs and Attitudes Questionnaire was culturally adapted with the input from two focus groups composed of low-health-literacy Mexican-American women. The Mammography Beliefs and Attitudes Questionnaire was culturally adapted with the input from two focus groups composed of low-health-literacy Mexican-American women. A total of 14 Mexican-American women were recruited to participate in one of two different focus groups to culturally adapt the SMBAQ. The inclusion criteria for recruiting women into each of the two focus groups were: (a) women who self-identified as being of Mexican-American descent, (b) women who were over 40 years of age, (c) women who had an educational level less than 8th grade, and (d) women who stated that their preferred language when writing and speaking was Spanish.
The determination of low-health-literacy in potential participants in each focus group was completed using two criteria: an educational level below the 8th grade, along with a statement that the preferred language when writing and speaking was Spanish, a preference that established a low-English-proficiency level (LEP). LEP individuals, according to the U.S. Department of Health and Human Services’ (2001) standards for assessment of low-health-literacy, can be considered to have a low-health-literacy level. The focus group discussions for both focus groups were conducted by a trained focus group leader who was a bilingual, Mexican-American woman over the age of 40. Each focus group gave input relative to the appropriateness of the content of each of the items on the SMABQ, as well as the Spanish words utilized for each item and the appropriateness of the questionnaire for use with Mexican-American women.
Content Validity Assessment
Content validity of the MBAQ and SMBAQ was evaluated by the same bilingual Expert Committee, using the Lynn content validity index (Lynn, 1986). Each committee expert evaluated each of the items on the MBAQ and the SMBAQ to determine if the content of each item on each instrument was appropriate and applicable for use with English- and Spanish-speaking, low-health-literacy Mexican-American women.
Reading Level
The reading level of the MBAQ and the SMBAQ was evaluated for use with low-English-proficiency and low-health-literacy Mexican American women by the same Expert Committee using the Fry Graph (Fry, 1977) for the MBAQ and the Readability Adaptation for Spanish Evaluation (FRASE graph) (Vari-Cartier, 1981) for the SMBAQ. These are established measures for the determination of the reading level of English (Fry Graph) and Spanish (FRASE graph) written materials. The bilingual Expert Committee rated the reading level of both the MBAQ and the SMBAQ below the 6th grade level. A detailed description of the different steps utilized during the development of the MBAQ and SMBAQ was previously published in OJIN: The Online Journla of Issues in Nursing (Lopez-McKee, 2010).
Initial Psychometric Testing of the MBAQ abd SMBAQ
A total of 259 Mexican-American, bilingual women (English and Spanish), were recruited for the initial psychometric testing phase of the MBAQ and SMBAQ. The psychometric testing of the MBAQ and SMBAQ included the following steps: a) an estimate of the internal consistency reliability, utilizing Cronbach’s alpha reliability coefficient, b) factor analysis, using principal components analysis with Varimax rotation, and c) paired t-test comparison of the means of the MBAQ and SMBAQ to establish their equivalence.
The sample size of 259 women was determined according to the formula of calculating 10 participants per item on a questionnaire (Nunnally & Bernstein, 1994). Since there were 25 items on both the MBAQ and SMBAQ, 250 participants were recruited into the study plus an additional nine participants to allow for questionnaires that would need to be eliminated because of incomplete or missing data on questionnaire items. Inclusion criteria for participants in the initial psychometric testing of the MBAQ and SMBAQ included women who: (a) were 18 years of age or older, (b) self-identified as being Hispanic of Mexican-American descent, and (c) stated that they were bilingual in English and Spanish.
We recruited women under the age of 40 into this initial validation study because the main purpose of the study was to determine the psychometric properties and language equivalence of the MBAQ and SMBAQ. This required the inclusion of bilingual women who could answer the items in both English and Spanish, even if they were under the age of 40. We recognize that screening mammograms are not routinely recommended for women under the age of 40. However, it was not necessary to include women in this study who were representative of the population for which the instrument was intended. Their responses were analyzed only in relation to their internal consistency, factor structure, and equivalence on both the MBAQ and SMBAQ.
Recruitment of Participants
Participants were recruited from a university campus, a community college campus, and several community and senior centers located in the city of El Paso, Texas. Flyers were posted at these different sites specifying the required inclusion criteria. Flyers also specified the dates when the administration of the MBAQ and SMBAQ would take place at the different sites. Participants were offered a $10 gift card to participate in the study. Flyers specified that participants would be completing the MBAQ and SMBAQ during the same session.
Administration of the MBAQ and SMBAQ
Each of the 259 participants was asked to complete the MBAQ and SMBAQ during the same session at the sites at which they were recruited. Participants completed the MBAQ and SMBAQ in random order, with some participants completing the SMBAQ first, while other participants completed the MBAQ first. The research assistant read each of the 25 items verbatim on the MBAQ and the SMBAQ to each participant. Participants were asked, after each question was read, to circle their answer to each question on both of the questionnaires. No other information was offered to participants during the administration of either the MBAQ or the SMBAQ. Participants generally took about 15 minutes to complete both instruments.
Each completed questionnaire was coded with an identifying number used to enter both the information from the completed questionnaires and also the demographic information from each participant. Any identifying information about participants was removed before entering data for analyses. Computerized data were protected by a password accessible only to the principal investigator (PI); the original completed questionnaire forms were filed in a locked cabinet in the PI’s office.
Findings
Demographic characteristics. The mean age for participants in this initial psychometric testing of the MBAQ and SMBAQ was 38.9 years of age, with a mean educational level of 15.1 years. Most participants were Catholic (80%); 51% were single and 28% reported an income level of $10,000 to $20,000. The majority (59%) listed the US as their country of origin and 58% listed English as their preferred language when writing (See Table 1). This study included only participants who were bilingual in English and Spanish. This necessitated including Hispanic women who had a higher educational level than the low-health-literacy Mexican-American population for whom the MBAQ and SMBAQ was intended. Bilingual Mexican-American women tend to have higher levels of education than do the women for whom this instrument was designed. It was essential to include bilingual women in this initial testing phase so that each woman could complete both the MBAQ and the SMBAQ during the same session, in order to compare the psychometric equivalence of MBAQ to the SMBAQ.
Construct validity. The MBAQ and the SMBAQ each included 25 items divided into five subscales. These subscales included: (Subscale A) Attitudes about mammography screening; (Subscale B) Intentions to seek a mammogram; (Subscale C) Perceived control over getting screened for breast cancer; (Subscale D) Perceived risk of getting breast cancer; and (Subscale E) Subjective norm - the influence of family or physician recommendations in seeking mammography screening.
When factor analysis was conducted on the 25-item MBAQ and SMBAQ using principal components analysis with Varimax rotation, it revealed that only three distinct subscales could be identified on both the MBAQ and SMBAQ. Criteria utilized to determine the number of subscales on the MBAQ and SMBAQ included the use of Eigenvalues greater than 1.0, factor loadings greater than 0.50, and the use of the Scree plot. Only 19 out of the original 25 items on the MBAQ and SMBAQ exhibited eigenvalues greater than 1.0 and factor loadings greater than 0.50. The Scree plot also confirmed that there were only three distinct subscales on the MBAQ and SMBAQ. After review of the six items that exhibited eigenvalues less than 1.0 and factor loadings less than 0.5, these six items were removed. Removal of these six items improved the factor structure of the remaining 19 items, confirming that there were only three distinct subscales. Please see Table 2 for the factor loadings of the 19-item MBAQ and SMBAQ.
Internal consistency reliability. Estimates of internal consistency of the 25-item MBAQ and the SMBAQ were obtained using Cronbach’s alpha reliability coefficients. Results revealed reliability estimates of 0.80 for the MBAQ and 0.83 for the SMBAQ in this sample of 259 participants. Estimates of internal consistency, conducted on the 19-item MBAQ and SMBAQ after removal of the six items that exhibited poor psychometric properties and reliability, were 0.82 for the MBAQ and 0.83 for the SMBAQ, using the same sample of 259 participants.
Equivalence. A paired-samples t-test comparison was conducted between the MBAQ and SMBAQ, after the six items were removed, using the same sample of 259 bilingual Mexican-American women. No statistically significant difference was found between the MBAQ and SMBAQ (t = -1.54; p > 0.12).
Psychometric Testing of the SMBAQ with Low-Health Literacy Women
Since the first psychometric testing of the MBAQ and SMBAQ was not conducted with low-health-literacy Mexican-American women, it was necessary to test the psychometric properties...using a sample of low-health-literacy Mexican-American women. Since the first psychometric testing of the MBAQ and SMBAQ was not conducted with low-health-literacy Mexican-American women, it was necessary to test the psychometric properties of the SMBAQ using a sample of low-health-literacy Mexican-American women. The SMBAQ was tested next with a sample of 68 low-health-literacy Mexican-American women recruited from a database at the El Paso Cancer and Chronic Disease Consortium (EPCCDC) (Lopez-McKee et al., 2008). The main purpose of this 2008 study was to compare factors affecting repeat mammography screening participation among two groups of low-income Mexican-American women. It compared women who had infrequent patterns of mammography screening with women who had regular patterns of mammography screening. This research did not focus on the low-health-literacy level of these women or on the psychometrics of the instruments used in this study (which included the SMBAQ). However, this 2008 EPCCDC study did provide us with a second opportunity to test the psychometric properties of the SMBAQ, using a low-health-literacy sample of Mexican-American women.
In addition to the IRB approval obtained from The University of Texas at El Paso to conduct this study, permission was also obtained from the directors at the EPCCDC. The EPCCDC followed HIPPA regulations to protect the privacy of women who had been referred for mammography screening through the EPCCDC by requiring that the study be conducted at the EPCCDC office utilizing their computer system, and by requiring that one of their staff members be trained as the research assistant in the study. The EPCCDC employee who was trained as the research assistant in this study de-identified all information obtained during the study prior to delivering data to the PI.
Recruitment of Participants
Participants were recruited to participate in this study by telephone calls from the research assistant. Participants eligible to participate were identified from a database of their mammography screening participation history through the EPCCDC. Inclusion criteria for recruitment of participants into this study included women in the EPCCDC database who self-identified as being of Mexican-American descent, were 50-64 years of age, had never been diagnosed with breast cancer, and had received at least one mammogram during the previous five years. Only women who were 50-64 were selected because they qualified for free mammograms through the EPCCDC. All participants in this study met the criteria for low income according to the U.S. Department of Health and Human Services Annual Federal Poverty Income Guidelines (2002) as this was a requirement by the EPCCDC in order to qualify for free mammograms. Also, all the women recruited into this study stated that their preferred language when speaking and writing was Spanish.
Administration of the SMBAQ
Participants were given the option by the research assistant to complete all questionnaires in either English or Spanish. All participants chose to complete the demographic questionnaire, the SMBAQ, and the Powe Fatalism Inventory (PFI) (Powe 1995a) in Spanish (SPFI) (Lopez-McKee et al., 2007).
The research assistant, who was located inside the EPCCDC office, completed all participant recruitment efforts and the administration of questionnaires over the telephone from the EPCCDC office. Participants were asked if they could participate during the initial telephone call, or if a more convenient time could be set up. The 25 questions on the SMBAQ were read to each participant over the telephone by the research assistant who emphasized that there were no right or wrong answers to any of the questions. No additional information was offered to participants; their responses were not questioned. The research assistant recorded each participant’s answer on the SMBAQ. Each participant was mailed a $20 gift card by the research assistant for participating in the study.
Findings
Demographic Characteristics. All 68 women recruited into this study self-identified as being of Mexican-American descent. The mean age for participants was 58.4 years (SD=4.6). Most participants were married (n=42, 62%). Participants who were not married were able, when completing Question 3 on Subscale C (See Figure), to select the option of ‘not at all.’ Mean years of living in the US was 27.7 (SD=13.6). The majority of the women reported being Catholic (n=54, 79%). Mean educational level was 7.1 years. All women recruited into this study (100%), reported that they preferred Spanish when speaking and writing.
Internal consistency reliability. Estimates of internal consistency reliability of the SMBAQ were obtained using Cronbach’s alpha reliability coefficient. Results revealed a reliability estimate of 0.69 utilizing 25 items and 0.70 with 19 items (when the same six items removed from the MBAQ and SMBAQ in the original testing were excluded).
Modification of MBAQ and SMBAQ Subscales
After reviewing the reliability estimates and factor structure from the initial psychometric testing of the MBAQ and SMBAQ, which included 259 bilingual Mexican-American women, and the results of the additional psychometric testing of the SMBAQ with 68 low-health-literacy Mexican-American women, the total number of items on the MBAQ and SMBAQ was decreased from 25 to 19 for future studies utilizing the MBAQ and SMBAQ.
The factor structure significantly improved when six items were deleted from both the MBAQ and SMBAQ in the original psychometric testing completed by 259 participants. However, the factor structure was not conclusive when factor analysis was conducted with the sample of 68 women from the Cancer Consortium database, possibly because of the small sample size.
Examples of items that caused confusion to participants included questions relating to assessing their risk of getting cancer in their lifetime and questions related to feeling in control over their own breast health. Several of the six items that were removed from the MBAQ and the SMBAQ had also been identified as being confusing to answer by participants in each of the focus groups that were originally used to culturally adapt the MBAQ and SMBAQ. While the wording was modified for these items prior to administering them to participants, these items continued to produce confusion among women who completed the MBAQ and SMAQ in the two separate studies. Examples of items that caused confusion to participants included questions relating to assessing their risk of getting cancer in their lifetime and questions related to feeling in control over their own breast health (Lopez-Mckee, 2010).
When the six items exhibiting poor psychometric properties were removed from the MBAQ and SMBAQ, the number of subscales was decreased to three subscales, which were renamed Subscale A (five items addressing ‘attitudes’), Subscale B (four items dealing with ‘perceived control’), and Subscale C (ten items considering ‘subjective norms’) (See Figure).
Discussion
Testing of the reliability estimates of the MBAQ and SMBAQ with two different samples provided evidence that these estimates were improved with the removal of six items. Factor analysis during the first psychometric testing of the 25-item MBAQ and SMBAQ extracted three distinct factors. After removal of the six items that exhibited poor psychometric properties, 48.6% of the variance in the MBAQ and 49.2% of the variance in the SMBAQ was explained.
While the mean number of years of education for participants recruited into the initial validation study for the MBAQ and SMBAQ was 15.1 years, the mean number of years of education for participants included when the SMBAQ was validated with low-health-literacy Mexican-American women was 7.1 years. The second validation of the SMBAQ provided adequate reliability estimates, which confirmed that the SMBAQ is an appropriate instrument for use with low-health-literacy Mexican-American women.
While the mean age of participants who completed the MBAQ and SMBAQ during the initial psychometric testing was 38.9 years of age, the mean age of the 68 low-health-literacy Mexican-American women who completed the SMBAQ was 58.4 years of age. The second validation of the SMBAQ provided evidence that the SMBAQ is appropriate for use with Mexican-American women over the age of 50, which is the population for which the instrument was developed.
Limitations
One limitation of the psychometric testing of the MBAQ and SMBAQ was that only the SMBAQ was validated with low-health-literacy Mexican-American women. The MBAQ has only been validated with bilingual, Mexican-American participants having a relatively high educational level. Additionally the SMBAQ was tested with only 68 low-health-literacy Mexican-American women. Validation of the SMBAQ with this small sample did not provide sufficient evidence of the SMBAQ’s psychometric properties with a low-health-literacy population. Further testing of the MBAQ and SMBAQ with low-health-literacy Mexican-American women will be necessary in order to confirm its psychometric properties with this population.
Another limitation is that a power analysis was not done to calculate the sample size in either the original testing of the MBAQ and SMBAQ or in the second administration of the SMBAQ. Rather the ‘rule of thumb’ formula by Nunnally and Bernstein (1994), requiring 10 subjects per variable, and which has yielded reliable results in the past was used in the first testing. The authors recognize that utilizing a power analysis would have allowed a more accurate interpretation of the results of both studies. Furthermore the results of this study may be difficult to generalize because of important cultural differences that exist among different Hispanic subgroups.
Conclusions: Where the Program Stands To Date
Two instruments, the MBAQ and the SMBAQ, were developed to measure cultural and psychosocial variables that impact the intention to seek mammography among low-health-literacy Mexican-American women. No statistically significant differences were found between the MBAQ and SMBAQ when administered to a sample of 259 bilingual Mexican-American women. The MBAQ and SMBAQ can both be used interchangeably within the same study.
...this description of the process utilized to develop these instruments will assist others in developing culturally appropriate, health-related materials to facilitate health literacy among the Hispanic population.The Mammography Beliefs and Attitudes Questionnaire and the Spanish Mammography Beliefs and Attitudes Questionnaire are culturally appropriate and reliable tools for use by clinicians in assessing factors that most strongly impact the intention to seek mammography screening among low-health-literacy Mexican-American women. By assessing these factors, healthcare providers can develop interventions to influence attitudes, increase social support, and/or increase women’s perceived control over obtaining mammography screening. Also, this description of the process utilized to develop these instruments will assist others in developing culturally appropriate, health-related materials to facilitate health literacy among the Hispanic population.
Using the TPB as a framework for the item development of the MBAQ and SMBAQ provided a means for inclusion of psychosocial factors that impact mammography screening in low-health-literacy Mexican-American women. The MBAQ and SMBAQ will be useful tools for assessing the link between the attitudes held by low-health-literacy Mexican-American women about mammography screening and the actual behavior of obtaining a mammogram.
Implications for Future Research
...further testing of the MBAQ and SMBAQ with Hispanic women belonging to different subgroups...is needed to determine the generalizability of findings using these two instruments to low-health-literacy Hispanic women in different populations.Further testing of the 19-item MBAQ and SMBAQ with larger samples of low-health-literacy Mexican-American women will be needed to confirm the factor structure and reliability estimates of both instruments. In addition, further testing of the MBAQ and SMBAQ with Hispanic women belonging to different subgroups, such as Puerto-Rican and Cuban populations, is needed to determine the generalizability of findings using these two instruments to low-health-literacy Hispanic women in different populations.
Acknowledgements: This study was funded by the National Institutes of Health, National Center on Minority Health and Health Disparities (Grant No. P20 MD000548), through the Hispanic Health Disparities Research Center, El Paso, Texas and by the National Institutes of Health, National Center for Research Resources (Grant No. 5G12 RR008124).
Copies of the MBAQ and SMBAQ may be requested from Dr. Lopez-McKee.
Author
Gloria Lopez-McKee, PhD, RN
E-mail: gmckee@utep.edu
Dr. Lopez-McKee is currently an Assistant Professor at the University of Texas at El Paso School of Nursing. Her research focuses on the development, translation, and cultural adaptation of health-related instruments into Spanish. She has developed the Spanish version of the Powe Fatalism Inventory, which involved the cultural adaptation of the instrument to a Mexican-American population. Dr. Lopez-McKee also developed the English and Spanish Mammography Beliefs and Attitudes Questionnaire (MBAQ and SMBAQ) and validated these instruments with a bilingual population of Mexican-American women. Dr. Lopez-McKee received her BSN from the University of Texas System School of Nursing (SON), her MSN from the University of Texas at El Paso SON and Allied Health, and her PhD in Nursing from the University of Houston Health Sciences Center SON in Houston, TX.
Julia Bader, PhD
E-mail: jbader@utep.edu
Dr. Bader is a consultant to faculty at the University of Texas at El Paso. She has extensive experience in assisting faculty with the design of research studies, the data analysis, and the interpretation of the findings. Dr. Bader has collaborated with Dr. Lopez-McKee on several studies related to health-instrument development and psychometric validation of these instruments. Dr. Bader received her BS in Statistics from the University of the Philippines, Los Banos, Laguna, Philippines, her MS in Statistics from Iowa State University, Ames, IA, and her PhD in Statistics from Texas A & M University, College Station, TX.
Francis, J., Eccles, M., Johnston, M., Walker, A., Grimshaw, J., Foy, R.,…Bonetti, D. (2004). Constructing questionnaires based on the Theory of Planned Behavior: A manual for health services researchers. UK: Centers for Health Services Research.
U.S. Department of Health and Human Services. (2002-2006). U.S. Poverty guidelines. Retrieved March 8, 2010 from: http://aspe.hhs.gov/POVERTY/index.shtml
| Characteristic | Mean | SD |
| Age (years) | 38.9 | 18.5 |
| Educational Level (years), n=256 | 15.1 | 3.6 |
| Years living in U.S.A. | 29.6 | 20.6 |
| Frequency | Percent | |
| Age |
|
|
| 18-39 years | 154 | 59.5 |
| 40-88 years | 105 | 40.5 |
| Educational Level, n=256 |
|
|
| 0-12 years | 62 | 24.2 |
| 13-21 years | 194 | 75.8 |
| Marital Status |
|
|
| Married | 80 | 30.9 |
| Single | 131 | 50.6 |
| Divorced | 29 | 11.2 |
| Widow | 19 | 7.3 |
| Religion |
|
|
| Catholic | 207 | 79.9 |
| Other | 52 | 20.1 |
| Country of Origin, n=257 |
|
|
| United States | 152 | 59.1 |
| Mexico | 105 | 40.9 |
| Household Income, n=255 |
|
|
| $10,000 and less | 55 | 21.6 |
| <$10,000-$20,000 | 72 | 28.2 |
| <$20,000-$30,000 | 50 | 19.6 |
| <$30,000-$40,000 | 25 | 9.8 |
| <$40,000-$50,000 | 25 | 9.8 |
| Over $50,000 | 28 | 11.0 |
| Preferred Language Spoken |
|
|
| English | 122 | 47.1 |
| Spanish | 90 | 34.8 |
| Both | 47 | 18.1 |
| Preferred Language Written |
|
|
| English | 150 | 57.9 |
| Spanish | 69 | 26.7 |
| Both | 40 | 15.4 |
| MBAQ | SMBAQ | |||||
| Factor 1 | Factor 2 | Factor 3 | Factor 1 | Factor 2 | Factor 3 | |
|
| Attitudes | Perceived control | Subjective Norm | Attitudes | Perceived control | Subjective Norm |
| Part 1, Question 1 | 0.72 | 0.00 | 0.09 | 0.74 | -0.01 | 0.12 |
| Part 1, Question 2 | 0.82 | -0.08 | 0.08 | 0.84 | -0.01 | 0.09 |
| Part 1, Question 3 | 0.80 | 0.05 | 0.23 | 0.79 | 0.06 | 0.25 |
| Part 1, Question 4 | 0.87 | 0.06 | 0.14 | 0.87 | 0.02 | 0.16 |
| Part 1, Question 5 | 0.84 | 0.09 | 0.19 | 0.84 | 0.04 | 0.19 |
| Part 2, Question 1 | 0.04 | 0.56 | 0.12 | 0.05 | 0.41 | 0.06 |
| Part 2, Question 2 | -0.05 | 0.72 | -0.07 | -0.02 | 0.58 | -0.13 |
| Part 2, Question 3 | 0.10 | 0.66 | -0.01 | 0.04 | 0.60 | 0.06 |
| Part 2, Question 4 | 0.00 | 0.53 | 0.04 | 0.01 | 0.67 | 0.03 |
| Part 3, Question 1 | 0.32 | 0.03 | 0.45 | 0.27 | 0.07 | 0.51 |
| Part 3, Question 2 | 0.25 | -0.30 | 0.56 | 0.19 | -0.38 | 0.53 |
| Part 3, Question 3 | 0.02 | -0.20 | 0.51 | 0.06 | -0.23 | 0.51 |
| Part 3, Question 4 | 0.14 | 0.09 | 0.77 | 0.07 | 0.15 | 0.81 |
| Part 3, Question 5 | 0.09 | 0.04 | 0.76 | 0.06 | 0.07 | 0.79 |
| Part 3, Question 6 | 0.15 | -0.04 | 0.68 | 0.13 | 0.12 | 0.78 |
| Part 3, Question 7 | 0.15 | 0.15 | 0.66 | 0.13 | 0.08 | 0.68 |
| Part 3, Question 8 | 0.06 | -0.05 | 0.57 | 0.08 | -0.16 | 0.56 |
| Part 3, Question 9 | 0.10 | 0.14 | 0.56 | 0.14 | 0.02 | 0.51 |
| Part 3, Question 10 | 0.04 | 0.13 | 0.54 | 0.11 | -0.03 | 0.56 |
| Eigenvalue | 2.33 | 1.76 | 5.15 | 2.43 | 1.59 | 5.33 |
| Percent of Variance Explained | 12.2 | 9.3 | 27.1 | 12.8 | 8.4 | 28.0 |
| Figure. Illustration of MBAQ questions |
| Subscale A: (Attitudes about mammography screening)
|
| Subscale B: (Perceived Control over seeking mammography screening) (Scale end points: Disagree……………………Strongly agree)
|
Subscale C: (Subjective Norm: Influence of Family and Friends in seeking mammography screening)
|
| The Spanish version of this instrument can be obtained by e-mailing the author and asking for a copy of the SMBAQ. |
References
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckman (Eds.), Action-control: From cognition to behavior (11-39). Heidelberg: Springer.
American Cancer Society. (2007). Cancer facts and figures for Hispanics and Latinos 2006. Atlanta, GA: Retrieved from http://www.cancer.org/.
Calderon, J., Smith, S., & Baker, R. (2007). A simple method for enhancing readability of Patient information. Annals of Behavioral Science and Medical Education, 13(1), 20-24.
Carter J., Park, E., Moadel, A., Cleary S., & Morgan, C. (2002). Cancer knowledge, attitudes, beliefs, and practices of disadvantaged women in the South Bronx. Journal of Cancer Education, 17(3), 142-149.
Centers for Disease Control and Prevention. (2007). Mammography percentages by race and ethnicity: 2007 fact sheet. Retrieved March 8, 2010, from: www.cdc.gov/omhd/AMH/factsheets/cancer.htm.
Detterborn, L., DuHamel, K., Butts, G., Thompson, H., & Jandorf. (2005). Cancer fatalism and its demographic correlates among African American and Hispanic women effects on adherence to cancer screening. Journal of Psychosocial Oncology. 22(4), 47-60.
Fernandez, M., Palmer, R., & Leong-Wu, C. (2005). Repeat Mammography screening among low-income and minority women: a qualitative study. Cancer Control, November issue.
Francis, J., Eccles, M., Johnston, M., Walker, A., Grimshaw, J., Foy, R.,…Bonetti, D. (2004). Constructing questionnaires based on the Theory of Planned Behavior: A manual for health services researchers. UK: Centers for Health Services Research.
Fry, R. (1977). Fry’s readability graph: Clarifications, validity and extension to level 17. Journal of Reading, 21, 241-252.
Intercultural Cancer Council. (2005). Hispanics/Latinos & cancer fact sheet. Online facts sheet. Retrieved February 21, 2011 from http://iccnetwork.org/cancerfacts/cfs4.htm.
Li, C., Malone, K., & Daling J. (2003) Differences in breast cancer stage, treatment, and survival by race and ethnicity. Archives of Internal Medicine, 163(1), 49-56.
Lopez-McKee, G. (2005). Spanish Translation, cultural adaptation and validation of power fatalism inventory among a Mexican-American population. Published doctoral dissertation, University of Houston Health Sciences Center School of Nursing.
Lopez-McKee, G. (2010) Development of the mammography beliefs and attitudes questionnaire for low-health-literacy Mexican-American women. OJIN: The Online Journal of Issues in Nursing, 16(1). Available: www.nursingworld.org/MainMenuCategories/ANAMarketplace/ANAPeriodicals/OJIN/TableofContents/Vol-16-2011/No1-Jan-2011/Articles-Previous-Topics/Mammography-Beliefs-and-Mexican-American-Women.aspx DOI: 10.3912/OJIN.Vol16No01PPT04.
Lopez-McKee, G., McNeill, J., Eriksen, L., & Ortiz, M. (2007). The Spanish translation and cultural adaptation of the powe fatalism inventory. Journal of Nursing Scholarship, 39(1), pp. 68-70.
Lopez-McKee, G., McNeill, J., Bader, J., & Morales, P. (2008). Comparison of factors affecting repeat mammography screening of low-income Mexican-American women. Oncology Nursing Forum, 35(6), 941-946.
Lynn, M.R. (1986). Determination and quantification of Content Validity. Nursing Research, 35, 382-385.
Nunnally, J.C., & Bernstein, I. (1994). Psychometric theory. McGraw-Hill, 3rd ed.
O’Brien, K., Cokkinides, V., Jemal, A., Cardinez, C.J., Murray, T., Samuels, A., Thun, M.J. (2003). Cancer statistics for Hispanics. CA: A Cancer Journal for Clinicians, 53(4), 208-226.
Otero-Sabogal R., Stewart S., Sabogal F., Brown B., & Perez-Stable E. (2003). Access and attitudinal factors related to breast and cervical cancer rescreening: why are Latinas still underscreened? Health Education and Behavior, 30(3); 337-359.
Powe, B. (1995a). Cancer fatalism among elderly Caucasians and African Americans. Oncology Nursing Forum, 22(9), 1355-9.
U.S. Department of Health and Human Services, Office of Minority Health. (2001). National standards for culturally and linguistically appropriate services in health care. Washington, DC: National Academic Press.
U.S. Department of Health and Human Services. (2002-2006). U.S. Poverty guidelines. Retrieved March 8, 2010 from: http://aspe.hhs.gov/POVERTY/index.shtml
Vari-Cartier, P. (1981). Development and validation of a new instrument to assess readability of Spanish prose. The Modern Language Journal, 65(2), 141-148.
Weiner, J., Aguirre, A., Ravenell, K., Kovath, K., McDevit, L., Murphy, J., Shea, J.A. (2004). Designing an illustrated patient satisfaction instrument for low-literacy populations. The American Journal of Managed Care, 10(11), 853-860.
Young, H., Lierman, L., Powell-Cope, G., Kasprzyk, D., & Benolict, J. (1991). Operationalizing the theory of planned behavior. Research in Nursing and Health, 14, 137-144.
Table 1. Demographic Characteristics of 259 Participants
Table 2. Factor analysis loadings and Eigenvalues with 19 Questions (out of 25) of the MBAQ and SMBAQ, using the Varimax Rotation with 3 Factors, and 3 subscales
