Research Design and Data Collection


The purpose of this study was to examine the risk factors predicting asthma among adult foreign-born African Americans in California. I used a quantitative correlation approach to undertake the research. The associated variables included the dependent variable (asthma status), independent variables (tobacco smoking, alcohol use, education level, income level, and health insurance), and confounders (age, gender, and marital status). The findings of this research could help to fill a research gap, which exists because previous researchers have not extensively explored the relationship between the aforementioned variables among foreign-born African Americans in California.

Reliability and Validity of CHIS Data

Validity and reliability are important considerations when testing or formulating research instruments. The Cronbach alpha technique was used to assess the validity of the CHIS data. This technique is among the most common methods used to test for reliability because it measures the internal consistency of a set of items in a group. According to Statistics Solutions (2017), the Cronbach’s alpha should typically give values from “0” to “1.”. These scores highlight the degree of internal consistency. Values, which are more or equal to 0.9, are perceived to have an “excellent” internal consistency, while those that are between 0.8 and 0.9 are deemed to have a “good” internal consistency.

Cronbach alpha values that are between 0.7-0.8 could be viewed to have an “acceptable” internal consistency, while those that are between 0.6 and 0.7 have “questionable” internal consistency. Lastly, values that are between 0.5 and 0.6 (or lower) are perceived to have a “poor” internal consistency, while those that have a negative value are viewed to have no internal consistency.

The internal consistency of the scales used to measure the variables in the CHIS database was found to be generally “excellent” because the Cronbach’s alpha was 0.89 for most of the research scales used, based on a sample of 100 households (California Health Interview Survey, 2014). This score of internal consistency is regarded as higher than most national data. For example, the Medical Expenditure Panel Survey has been used for medical research and has a validity of 0.7, which is deemed “acceptable” (Lindly, Zuckerman, & Mistry, 2016). Comparatively, a score of 0.89 for the CHIS data is deemed “excellent” (based on the evaluation metrics highlighted above).

According to a study by Wang, Ponce, Wang, Opsomer, and Yu (2015), which used CHIS data to generate health estimates by zip code found that the test-retest Pearson Correlation for CHIS data was “satisfactory” because the total score was 0.86. Although these statistics support the credibility and validity of the information obtained in the CHIS data, it is important to recognize that the responses obtained from the database were self-reported and could be vulnerable to selection bias.

However, as supported by the findings of Gonzalez, Sanders-Jackson, and Emory (2016), which also used CHIS data, this limitation has been accepted for many large-scale studies. It may also be true that symptomatic signs of asthma may be misclassified as the condition, but it is not projected that such errors could skew estimates (Gonzalez et al., 2016; Silverberg, Simpson, Durkin, & Joks, 2013; Rosser, Forno, Cooper, & Celedón, 2014).

A pilot study of interviewing adults with telephones was also conducted when developing the CHIS data. By doing so, the researchers evaluated the feasibility of sampling and interviewing the respondents using telephones. Using this technique, important information relating to the feasibility of conducting a lengthy health interview survey (via a telephone) was reviewed because many past surveys had a limited time to engage the respondents (California Health Interview Survey, 2014).

The second piece of information that was obtained from the pilot study was the feasibility of using random sampling methods in a household. To get the above information, a survey plan was formulated where 100 interviews involving households that had a telephone and an adult present were completed (California Health Interview Survey, 2014). The overall results of the pilot study showed that the first 100 interviews did not show any signs of poor validity or high rates of false negative or false positive responses, especially compared to prior responses provided in other national quantitative studies.

In fact, the findings of the study correlated with the Medical Expenditure Panel Survey, which had a Cronbach alpha of 0.70, as demonstrated by a study prepared by Elwood et al., 2017, which used CHIS data to measure the influence of legally recognized partnerships on the health and well-being of same-sex couples. Relative to this assertion, the California Health Interview Survey (2016) and Gonzalez et al. (2016) say that the CHIS database is a reliable measure for conducting population-based studies and adapting CHIS questions in health research. Therefore, the findings of the validity test showed that the questions correctly measured the constructs.

Although the findings of the pilot testing supported the internal validity of the research instruments, another measure of validity for the findings that were generated from the CHIS data is the inclusion of confounders (age, gender, and marital status). In other words, confounders were not integrated into the study to create a new purpose of the investigation, but rather to add to the internal validity of the relationship that was deduced from assessing the dependent and independent variables.

All data relating to variances, such as the actual number of foreign-born African Americans in California, were also weighted as a measure of validity. Replicate weights were used in this regard to address the complex sampling design used to collect the CHIS data and adjusted for non-response bias as well (UCLA Center for Health Policy and Research, 2018).

The resultant data were products of sequential series of logistic regression models, which focused on understanding asthma incidences, subject to adjustments of the independent and confounding variables. Weights were also used to adjust point estimates, variance estimates, and standard errors (UCLA Center for Health Policy and Research, 2018). All the weights used in the study had the same name – ‘RAKEDW0” (UCLA Center for Health Policy and Research, 2018).

The weighting technique accounted for sample selection probabilities and minimized the possibility of selection bias. They were also used in the study to account for sampling biases because some of the respondents included in the CHIS data could have had unequal chances of being included in the study if they were not used (UCLA Center for Health Policy and Research, 2018). Therefore, the weighting method was important in the collection of data because it provided a formula for analyzing the findings in a manner that is generalizable to the entire state of California. It also allowed the researchers to generate accurate data relating to standard errors and confidence intervals (Lee, Reed, & Berg, 2014). Tests of significance for population estimates were also generated using the same technique.

Threats to Validity

Threats to validity are issues that could affect the credibility of the research findings. Two types of threats to validity exist – internal and external validity threats (Guerrero et al., 2015). Internal validity threats are those, which undermine the cause and effect relationship between variables in a research, while external validity threats are those that affect the ability to generalize a study’s findings in other research settings (Nieswiadomy, 2012). The effects of these two types of threats to validity on the current research are explained in the subsections below.

Internal Validity

One of the threats to internal validity was the self-reporting nature of CHIS data. The accuracy of these self-reports is questionable because variations in language, race, or immigration status may cause differences in the accuracy of the responses given by the informants (Lee et al., 2014). The self-reporting nature of the CHIS also data made it difficult to verify asthma diagnoses independently. This threat to validity was partially addressed by undertaking the research using multiple languages (English, Chinese, Spanish, Korean, and Vietnamese) because it increased the probability that the respondents understood the questions asked, based on the convenience of choosing a language they were most comfortable with.

The specificity of variables was also another threat to the internal validity of the study because different criteria for analyzing the variables may have caused distortions in the findings. For example, a person’s tobacco smoking status may quickly change within the time a study is undertaken. This change may significantly affect the applicability of the study’s findings. To solve the problem of instrumentation, it was essential to include confounding variables in the study when analyzing the data to understand the effects of the use of different instrumentation techniques when developing the findings (Lang & Altman, 2014).

External Validity

The threats to validity in this study stem from the nature of the CHIS data. For example, the confined nature of the data to California alone is a threat because the health environment in the state is different from others in America because of the large presence of the health maintenance organization (HMO) (Wilson, 2016). The HMO is a type of health insurance that focuses on prevention and the alignment of financial goals with wellness (Wilson, 2016).

The state of California was among its early adopters and its health landscape has been predominantly defined by it (Wilson, 2016). Therefore, it may be difficult to extrapolate health data from the state to others which are not largely defined by the HMO or other aspects of California’s healthcare system. At the same time, California is a relatively heterogeneous state, in terms of racial hegemony. Therefore, the experience of being a foreign-born African-American in the state may be different from others in America. This is a limitation of the study.

Another threat to external validity was the failure to include institutionalized populations in the CHIS data (Becerra, Mshigeni, & Becerra, 2018). This makes it difficult to extrapolate the data to institutionalized populations such as those living in rehabilitation centers, prisons, correctional facilities, nursing homes, and assisted living facilities. The last threat to external validity stems from the fact that the sample population was predominantly Californian. This sampling procedure means that there may be a need to undertake future researcher to compare the findings of the study with a nationwide sample.

Application to Professional Practice and Implications for Social Change


The purpose of this study was to examine risk factors predicting asthma among adult foreign-born African Americans in California. The study was undertaken because of the lack of research on the health status of adult foreign-born African Americans in California. Study participants identifying themselves as African Americans in California that were born outside the USA made up the study sample. The findings of the study and an interpretation of the same information are undertaken in the sections below.

Interpretation of the Findings

Interpretation of the Findings in Relation to the Literature

Through my analysis of the data sampled, I found that there were no statistically significant associations between asthma status and tobacco smoking (p= 0.454), alcohol use (p=0.959), health insurance (p=0.700), and income level (p=0.966). However, I found that there was a statistically significant association between asthma status and education level (p= 0.016). These findings were achieved after accounting for the effects of confounders (age, gender, and marital status).

The above-mentioned findings differ from those of other researchers because past researchers have pointed out that asthma shares a relationship with the independent variables (tobacco smoking, alcohol use, income level, and health insurance). For example, a study by Coogan et al. (2015), which investigated the relationship between asthma and smoking among black women over a 16-year period, found that asthma incidences increased with an increase in active smoking behaviors.

Perret, Bonevski, McDonald, and Abramson (2016) who investigated smoking cessation strategies for patients with asthma also found that the presence of smoking behaviors among young adolescents caused an increase in asthma incidences. They also pointed out that both smoking behaviors and asthma incidences interacted to diminish lung functioning (Perret et al., 2016). In a different study, authored by Kim et al. (2015), it was established that the prevalence of asthma increased with a similar increase in smoking behaviors because of airflow obstruction. These findings were developed after the researchers examined risk factors contributing to asthma incidences among 12,631 participants (Kim et al., 2015).

Researchers such as Ejebe, Jacobs, and Wisk (2014) also established a positive relationship between asthma incidences and income levels after demonstrating that low incomes are associated with increased asthma incidences. Their views were informed by the fact that low-income levels are associated with poor housing conditions, poor diets, and low education levels, all of which contribute to an increase in asthma incidences (Li et al., 2016).

Past studies have also shown a positive relationship between asthma incidences and low access to health insurance (Shin et al., 2018). Most of these studies have pointed out that minorities living in the U.S. are more prone to asthma because of poor access to health insurance – a phenomenon which loosely translates to a poor access to quality healthcare services (Shin et al., 2018). Lastly, studies have also shown a positive association between increased asthma incidences and alcohol use (Guidot & Mehta, 2013; Linneberg & Gonzalez-Quintela, 2016).

It was interesting to find out that education levels had a significant impact on asthma status while income levels did not share the same positive relationship because several studies such as those authored by Miller et al. (2017), Chung, Lim, Lee, Kim, and Kim (2017) have affirmed a significant correlation between income levels and education. Somanna (2016) explains that such findings have been reported before because the aforementioned relationship between family income and education levels is weak.

This view was developed after the researcher examined the relationship between income and educational levels using a General Social Survey, which analyzed cumulative data from surveys conducted between 1972 and 2012 (Somanna, 2016).

His findings showed that the weak relationship between education and income levels largely stems from the fact that family income could be a construct of other social, political, or economic variables affecting a community. Therefore, education levels may not directly correlate with this variable. If this analogy is extrapolated to the findings of the current study, it is imperative to be cautious about assuming a positive correlation between education and income levels. Thus, a positive association between asthma and education levels could emerge, independent of income levels.

Comprehensively, the above-mentioned studies show that asthma incidences have been associated with low levels of health insurance, active smoking behaviors, alcohol use, and low-income levels. However, the current study affirms no such relationship. The disparity in findings could stem from racial differences between the samples used. In the current study, African Americans were the only sample, while the above-mentioned studies were not keen on age or race. Comprehensively, these views imply that racial differences could explain disparities in the findings.

Interpretation of the Findings in the Context of the Theoretical Framework

As highlighted in this paper, the socioecological model was the main theoretical framework for the current study. This model describes the complex interplay between people’s health, the environment, and their health outcomes, through an assessment of societal, community and relationship factors. The socio-ecological model has five nested levels of interlocking behavioral and anthropological factors: individual, interpersonal, organizational, community, and public policy. The rationale for using this theoretical framework in this study stems from its ability to show how different levels of personal and environmental factors affect human behaviors and health outcomes.

The multifaceted nature of the model was appropriate for the study because it helped in the exploration of the influence of several health risk factors such as tobacco smoking, alcohol use, education level, income level, and health insurance in predicting asthma among adult foreign-born African Americans in California. This framework also helped to explain why there were significant differences between the findings of this study and those of other researchers, who also investigated the effects of various risk factors on asthma incidence among immigrant populations.

As highlighted above, the findings of this paper showed no significant relationship between asthma status, tobacco smoking, health insurance, and income levels. If this finding is contextualized within the wider prism of the socioecological model, which regards, social, community, relationship, and individual forces as being the greatest determinants of health, evidently environmental factors emerge as having an insignificant effect on the incidence of asthma among the sample population.

Comparatively, individual factors have a significant effect on asthma because the findings showed that education levels were directly correlated with asthma incidences Therefore, the positive correlation between asthma and education levels imply that individual factors have a greater role to play in influencing the occurrence of asthma among the sample population.

The effect of race in influencing the above-mentioned outcomes is supported by studies, which have shown that cultural factors influence people’s health outcomes. For example, researchers who have investigated the health status of immigrant populations from parts of Asia and Africa show that the health status of immigrant populations tends to be better in their countries of origin compared to when the same population resides in the U.S. (Iqbal, Oraka, Chew, & Flanders, 2014). They say that cultural factors associated with their countries of origin, such as the stigma associated with women who smoke, enabled them to have better health outcomes compared to their counterparts in the U.S. (Iqbal et al., 2014).

The influence of confounding factors (age, gender, and marital status) on the association between asthma status and the independent variables could be largely confined to the first level of the social-ecological model, which postulates that individual factors influence people’s health outcomes. Relative to the findings of this investigation, it has been established that these individual attributes did not have significant effects on the association between asthma status and the independent variables. Therefore, it was possible to generalize the findings across different demographic variables underpinning the investigation.

The interpersonal level of the social-ecological model postulates that people who are close to a patient have an effect on their health. This tenet of the social-ecological model could influence the likelihood of the target population taking health insurance, consuming more alcohol, and using more tobacco (among other health risk factors). Therefore, they have a strong likelihood of influencing the association between asthma status and the independent variables.

The fourth level of the social-ecological model relating to societal factors influencing health outcomes could also have influenced the findings of this study by affecting some of the confounding and independent variables, such as access to health insurance, and alcohol use. Economics, education, and societal policies influencing the health status of foreign-born African Americans could also have had an effect on the risk factors predicting asthma status among the target population.

Limitations of the Study

Limitations of a study generally refer to aspects of a research investigation that are out of the control of the researcher and that may affect the overall integrity of the findings. One of the main limitations of the current study is the small number of confirmed cases of asthma among adult foreign-born African American immigrants who were sampled in the study. As highlighted in this paper, the sample size was comprised of 626 participants.

Out of this number, there were only 66 confirmed cases of asthma. If this number of contextualized within the greater population of respondents (626 participants) who took part in the study, the number of confirmed asthma cases emerges as being only 10% of the total sample. This small percentage of participants could have contributed to the null findings observed between the dependent variable (asthma) and all the independent variables (except for education levels) because a sample of 10% of the target population, which had asthma could have been too small to draw statistically significant correlations between the dependent variable (asthma) and the independent variables.

The self-reported nature of the CHIS data, which was used to develop current findings, was also a limitation of the study because the number of confirmed asthma cases highlighted in the report was not confirmed by medical records abstraction. In this regard, the number of confirmed asthma cases could be a limitation on the accuracy of the dependent variable (asthma). This limitation is confirmed by research studies, which have used a larger sample of respondents to come up with statistically significant findings between dependent and independent variables.

For example, the study by Kim et al. (2015) sampled 12,631 respondents to examine risk factors that contributed to asthma incidences. A similar study conducted by Becerra, Scroggins, and Becerra (2014) also derived a large number of respondents (19,841) from CHIS data to find associations between asthma and obesity among Asian American immigrants. Therefore, the small number of African American immigrants (626 respondents), which characterized the current study was a limitation.

The CHIS data used in this research provided useful information relating to risk factors predicting asthma among adult foreign-born African Americans in California. However, the findings obtained in this investigation are only indicative of the association between asthma (dependent variable) and its risk factors (tobacco smoking, alcohol use, education level, income level, and health insurance).

Although some of these limitations largely reflect those of the CHIS data, it was important to consider the recommendations of Linneberg and Gonzalez-Quintela (2016), which suggest that no data is perfect. Therefore, it was up to the discretion of researchers to balance the pros and cons of each research data for the advancement of a study’s objective. The findings I presented in this study were developed with this consideration in mind.

Causality was another weakness of the cross-sectional study, which informed the CHIS data because the findings cannot be used to ascertain the cause and effect of asthma and risk factors. Furthermore, it was difficult to establish the cause of asthma using the findings of the study. In addition, I realized that the questionnaire provided an insufficient assessment of alcohol use for the participants. However, it was the only question related to alcohol use in the CHIS questionnaire. Hence, I had no choice but to use the only available variable relating to alcohol use. The asthma status variable was not further confirmed by medical records abstraction, thereby making it self-reported asthma and a limitation on the accuracy of the dependent variable.


In this study, I sought to examine the risk factors predicting asthma among foreign-born African Americans in California. The findings of this study did not corroborate previous research on the deleterious association between asthma status and tobacco smoking, alcohol use, or the beneficial association with health insurance, and income level. The overall significant positive relationship between asthma and education level found in this study did confirm previous research studies.

The findings seem to be fundamentally different from those of previous researchers who have highlighted a strong relationship between the risk factors examined in this study and asthma. This departure in conventional findings accentuates the importance of understanding the community health outcomes of different immigrant groups. Based on this understanding, environmental and social issues concerning immigrant populations seem to affect their health outcomes and they need to be explored separately as well (this was the basis for the recommendations).

The need to explore the specifics of community health behavior is an approach that should be further entrenched in healthcare practice because different populations have unique health characteristics that should be exclusively explored. Increased support for genetic and genomic services in the healthcare practice is one approach that appreciates individual and societal differences in health care outcomes. The same should be encouraged in the management of asthma as a health issue.


Becerra, B. J., Scroggins, C. M., & Becerra, M. B. (2014). Association between asthma and obesity among immigrant Asian Americans, California health interview survey, 2001-2011. Preventing Chronic Disease, 11(1), 105-111. Web.

Becerra, M. B., Mshigeni, S. K., & Becerra, B. J. (2018). The overlooked burden of food insecurity among Asian Americans: Results from the California health interview survey. International. Journal of Environmental Research, 15(8), 1684-87.

Chung, W., Lim, S. J., Lee, S., Kim, R., & Kim, J. (2017). Gender-specific interactions between education and income in relation to obesity: A cross-sectional analysis of the fifth Korea national health and nutrition examination survey. BMJ Open, 7(12), 142-56. Web.

Coogan, P. F., Castro-Webb, N., Yu, J., O’Connor, G. T., Palmer, J. R., & Rosenberg, L. (2015). Active and passive smoking and the incidence of asthma in black women’s health study. American Journal of Respiratory and Critical Care Medicine, 191(2), 168-76.

Ejebe, I. H., Jacobs, E. A., & Wisk, L. E. (2014). Persistent differences in asthma self-efficacy by race, ethnicity, and income in adults with asthma. The Journal of Asthma: Official Journal of the Association for the Care of Asthma, 52(1), 105-13.

Guidot, D. M., & Mehta, A. J. (2013). Alcohol use disorders and the lung: A clinical and pathophysiological approach. New York, NY: Springer Science & Business Media.

Kim, H. J., Baek, S., Kim, H. J., Lee, J. S., Oh, Y. M., Lee, S. D., & Lee, S. W. (2015). The impact of smoking on airflow limitation in subjects with a history of asthma and inactive tuberculosis. PloS One, 10(4), 538-543. Web.

Lee, J., Reed, P. L., & Berg, J. P. (2014). Asthma characteristics among older adults: Using the California health interview survey to examine asthma incidence, morbidity, and ethnic differences. Journal of Asthma, Early Online, 51(4), 1-6.

Lindly, O. J., Zuckerman, K. E., & Mistry, K. B. (2016). Clarifying the predictive value of family-centered care and shared decision making for pediatric healthcare outcomes using the medical expenditure panel survey. Health Services Research, 52(1), 313-345.

Linneberg, A., & Gonzalez-Quintela, A. (2016). The unsolved relationship between alcohol and asthma. International Archives of Allergy and Immunology, 171(3), 155-157.

Li, Z., Leite, W. L., Thompson, L. A., Gross, H. E., Shenkman, E. A., Reeve, B. B., DeWalt, D. A., … Huang, I. C. (2016). Determinants of longitudinal health-related quality-of-life change in children with asthma from low-income families: A report from the PROMIS® pediatric asthma study. Journal of the British Society for Allergy and Clinical Immunology, 47(3), 383-394.

Miller, L. C., Joshi, N., Lohani, M., Rogers, B., Mahato, S., Ghosh, S., & Webb, P. (2017). Women’s education level amplifies the effects of a livelihoods-based intervention on household wealth, child diet, and child growth in rural Nepal. International Journal for Equity in Health, 16(1), 183.

Perret, J. L., Bonevski, B., McDonald, C. F., & Abramson, M. J. (2016). Smoking cessation strategies for patients with asthma: Improving patient outcomes. Journal of Asthma and Allergy, 9(1), 117-28. Web.

Shin, J. Y., Sohn, K. H., Shin, J. E., Park, M., Lim, J., Lee, J. Y., & Yang, M. S. (2018). Changing patterns of adult asthma incidence: Results from the national health insurance service-national sample cohort database in Korea. Scientific Reports, 8(1), 15052. Web.

Somanna, M. (2016). Is there a relationship between highest education attained by US residents and their family income? Web.

UCLA Center for Health Policy and Research. (2018). Frequently asked questions. Web.

Wilson, K. (2016). HMO enrollment in California: The dynamics of decline, 2004-2015. Web.