Abstract
The purpose of this study is to examine the risk factors predicting asthma among adult African immigrants in California. The research problem is founded on the failure of many health studies to include African immigrants as a minority group in health studies that investigate the relationship between risk factors and asthma. From this gap in the literature, this study seeks to answer one key research question, which pivots on examining the risk factors predicting asthma among adult African immigrants in California. The dependent variable is asthma status and the independent variables are the risk factors (smoking status, alcohol use, education level, income level, years since immigration, occupation, and healthcare access).
The covariates include age, gender, marital status, housing type, and caring neighbors. The socioecological theory will be the main conceptual framework for this study. The research design is the quantitative correlation approach. I will analyze secondary data collected by the California Health Interview Survey. Using the SPSS software, I will conduct both descriptive and inferential statistical analyses. The inferential statistical analysis will include both binary and multiple logistic regressions. This study is important to state-based health agencies and health professionals involved in asthma management because its findings could be used to develop health programs that target minority populations in California. It also has the potential of promoting positive social change by increasing asthma awareness and improving the health status of immigrant populations in California.
Instrumentation and Operationalization of Constructs
Instrumentation of Constructs
As mentioned in this study, the data will be obtained from the California Health Information Survey (CHIS). It is among the largest health interview survey in America and is conducted on an annual basis to provide information about different health topics (UCLA, 2017). Although the CHIS publish health data relating to different years, the data used in this study is CHIS 2009-2016. Looking at the appropriateness of the CHIS data to the current study, I find that it is relevant and specific to the topic under investigation because the dataset provides health data relating to different ethnic and racial groups in California. A focus on African immigrants as one cohort in the study is ideal for my analysis because the current study focuses on this ethnic group as the target population. The dataset is also appropriate to the current study because asthma is a health topic investigated in the research data. Other health issues surveyed in the dataset include diabetes and obesity (UCLA, 2017). The inclusion of asthma as a relevant health issue in the dataset and the provision of health data relating to immigrants bring my attention to the appropriateness of the data to the current study.
The findings contained in the CHIS document are freely available to the public. Therefore, there is no special permission from the researcher needed to use the instruments. Based on the availability of the health data outlined by CHIS, different people, including policymakers, state agencies, and community organizations, find the resource useful in improving the health outcomes of their subjects.
The reliability of the CHIS dataset is enshrined in the fact that the formulation of data was guided by the data quality objectives of the total survey error perspectives (UCLA Center for Health Policy Research, 2017). The data were also developed after undergoing several rigorous methodological processes. For example, to address non-coverage bias, the CHIS data were compared with other surveys. For example, comparisons were made with data findings from Medicaid, and no significant differences were reported, in terms of validity and reliability. Benchmarking is also another strategy that was used in the development of the CHIS database to maintain a high level of credibility for the data. Here, the CHIS data was benchmarked against three other surveys: Agency for Healthcare research and quality, National Center for Health Statistics, and the National Health Interview Survey. No significant methodological weaknesses were reported (UCLA Center for Health Policy Research, 2017). Additionally, a nonresponse analysis was undertaken to make sure the information presented in the CHIS database was credible and reliable. In the process, the researchers did a follow-up interview with people who did not respond to the study through telephone interviews and evaluated their findings by comparing their responses with at least 40 health indicators used to assess the health status of the responsive group. The studies showed insignificant differences between the responsive and unresponsive groups (UCLA Center for Health Policy Research, 2017).
The validity of the research data obtained in this study is also addressed by specific data quality assurance measures. For example, to make sure the target population fits the criteria of African immigrants, the study involved multiple language inclusivity as a communication method for maximizing the inclusivity of linguistically diverse populations (UCLA Center for Health Policy Research, 2017). Some of the languages used in the study included Spanish, English, Chinese, and Vietnamese (UCLA Center for Health Policy Research, 2017). The validity of the study was also supported by the use of maximum call attempts where each of the households chosen received up to 14 phone calls (UCLA Center for Health Policy Research, 2017). The calls were placed at different times of the day and different days of the week. When the respondents were unresponsive or inconvenienced by the same phone calls, they were rescheduled (UCLA Center for Health Policy Research, 2017).
Well-trained interviewers and high-quality interviews were also used as instruments for upholding the study’s validity. The interviewers were trained for up to 18 hours to make sure they asked the right questions and recorded the responses appropriately (UCLA Center for Health Policy Research, 2017). They were also constantly monitored and supervised for quality control. Westat, Inc., which has a national reputation for quality, oversaw the entire process (UCLA Center for Health Policy Research, 2017). Generally, a review of the reliability and validity of the values outlined in the research study reveals that both metrics are desirable because the CHIS data acts as a model for collecting state and local health data. Its use in advanced sampling and application methodologies also adds to the same view because such evidence has been used to influence key policy changes in different areas of research (Mertz et al., 2014). Its credibility is also supported by the fact that the database contains information that would appeal to key stakeholders in the health sector. These published reliability and validity values show that the sources of data are relevant to the current research.
The instrumentation of the constructs was done by carrying out a telephone survey. The constructs were measured by assigning unique codes and values to them. These indices helped the researcher to identify differences in the participant’s responses. For example, when analyzing the respondents’ housing type, homeowners were assigned a value of “1,” while renters were represented with a value of “2.” At the same time, the study’s constructs were assessed using frequencies and expressed as a percentage of the total population (UCLA Center for Health Policy Research, 2017). Again, if I use the housing example, the number of respondents who owned or rented home was expressed as a percentage of the total respondents because the database revealed that 58.4% of them owned a home, while 41.6% of them rented one (UCLA Center for Health Policy Research, 2017). Therefore, the constructs were measures using frequencies, percentages, and codes.
The CHIS database was used to collect health data across different ethnic and racial groups in California. In the past, it has been used to collect health data from all 58 counties in the state of California (UCLA, 2017). However, there are cases where the researchers have oversampled specific areas within the state that are heavily populated (such as Los Angeles and San Diego). The reliability and validity of the findings developed from the past use of CHIS have been confirmed through the involvement of large and diverse samples (UCLA, 2017). In other words, past users of the data established that the samples used in the dataset were representative of the ethnic and racial diversity of the state, particularly because the findings could be used to answer specific and important health questions about different ethnic and racial groups in the state.
The CHIS data has sufficient instrumentation to answer the research question. For example, I have established that the dataset contains health data about different ethnic and racial groups in California. This one instrumentation is useful in answering my research question, which focuses on one ethnic group – African immigrants. Other aspects of the instrumentation used to develop the CHIS are its consistency, flexibility, and adaptability. These attributes mean that the findings included in the report can be used to investigate new health issues and emerging trends among specific racial or ethnic groups. Lastly, the inclusion of spatial and geographic data in the CHIS dataset is another aspect of its instrumentation that would help to answer the research question because I can narrow down to a specific locality (California), which is at the core of my analysis. This type of instrumentation is often unavailable in national health data (Liamputtong, 2013).
Operationalization of Constructs
The research variables identified in the CHIS database were numerous though I utilized only a few variables that are related to my research questions. In this study, the dependent variable will be asthma and independent variables will be the associated risk factors (smoking status, alcohol use, education level, income level, occupation, and healthcare access). The covariates will be age, gender, marital status, housing type, and caring neighbors. In the CHIS database, race emerged as one variable that helped to define the target population – African immigrants. It was denoted by the term Self-reported African American (SRAA). Asthma was denoted by code ASTCUR. The process of measuring the variables will be done empirically and quantitatively.
Dependent variable
Asthma was assessed by asking participants questions about whether they had ever been diagnosed with the condition, what their current health status was (relative to this condition), and whether they had been taking medications to treat it (UCLA Center for Health Policy Research, 2017). The respondents were also asked to state whether they had an attack within the past 12 months and whether they had received a health management plan from a doctor (UCLA Center for Health Policy Research, 2017). The exact questions asked were, “Has a doctor ever told you that you have asthma,” “do you still have asthma,” and “during the past 12 months, have you had an episode of asthma or an asthma attack” (UCLA Center for Health Policy Research, 2017, p. 3).
Questions relating to whether a doctor had diagnosed them with the condition were represented with the code AB17. Information relating to daily medication to manage asthma was denoted by the code AB18, while the code AB40 denoted data relating to the status of asthma. Asthma will be operationalized for this study by simply evaluating whether patients had been diagnosed with this disease in the last 12 months, or not. Therefore, data relating to medication and treatment would not be included in the analysis because the study is (largely) descriptive in nature and is meant to understand whether there is a relationship between the condition and any of the variables identified in the study. In other words, by understanding the number of diagnoses in the last 12 months, it would be possible to deduce a relationship between the occurrence of the condition and the independent variables identified in the sections below (Bakk, Tekle, & Vermunt, 2013).
Independent variables
The independent variables include smoking, alcohol use, education level, income level, occupation, and health care access.
Smoking status
Smoking status was assessed by asking participants questions about whether they had smoked 100 or more cigarettes in their lifetime, or smoked cigarettes every day, some days, or not at all. The multiple sets of data were denoted by the codes AE15 and AE15A respectively. Collectively, the CHIS database categorized this data into three types: number of cigarettes per day, current smoking status, and current smoking habits. The codes NUMCIG, SMKCUR, and SMOKING represented the three sets of data, respectively (UCLA Center for Health Policy Research, 2017). Smoking status will be operationalized for this study by evaluating the current smoking status of the respondents. In other words, the variable will be assessed by evaluating whether the respondents are currently smoking, or not. Therefore, the relationship between asthma and smoking will only be established by evaluating two states: whether a person smokes, or not.
Alcohol use
Alcohol use was assessed by asking participants to respond to one question – “In the past 12 months, about how many times did you have five or more alcoholic drinks in a single day” (UCLA Center for Health Policy Research, 2017, p. 10). No other questions were asked by the interviewers to explore this variable (UCLA Center for Health Policy Research, 2017). Alcohol use will be operationalized for this study in the same way. Evidence will be examined to establish how many respondents had alcohol in the last 12 months and the data will be used to evaluate its relationship with asthma status. The operationalization of this variable is the only way to explore its relationship with the dependent variable because the CHIS database used one way to obtain data relating to alcohol consumption. Therefore, the operationalization of this variable is limited to this method.
Education level
Education level was assessed by asking participants questions about their highest level of educational attainment. This variable was denoted by the code AHEDC_P1 in the CHIS dataset and the exact question asked was, “What is the highest grade of education you have completed and received credit for?” Additionally, each respondent was asked to state whether they had a formal education, reached grade 9-11, reached grade 12 H.S Diploma, and whether they had attended a college or vocational school. As part of efforts to ascertain the respondents’ educational levels, they were also asked to state whether they had associated, bachelor’s, master’s, or PhD degrees (UCLA Center for Health Policy Research, 2017). The education variable will be operationalized for this study by examining how asthma incidences vary across four tiers of education: PhD (or its equivalent), master’s degree, bachelor’s degree, high school diploma, and “below high school.”
Income level
Income level was assessed by asking participants to state whether their household income supports someone living in the US (but not in the household), or not. The respondents were also asked to state whether their household income is equal to (or more than) 200% Federal Poverty Level (FPL). Additionally, they were asked to inform the interviewers of the total number of adults whose household income was less or equal to 300% FPL. Lastly, the respondents were asked to state their spouse’s earnings in the last month and their earnings in the same month. In compiling the data obtained, the CHIS data provided information relating to the total household income. The income level will be operationalized for this study by categorizing it into two main groups: respondents who earn equal or less than 200% FPL and respondents whose household incomes are less or equal to 300% FPL.
Occupation
Occupation was assessed by asking participants questions about whether they usually work, their employment status, and their main occupation. The exact questions asked included, “what is the main kind of work you do” and “on your main job, are you employed by a private company, the government, or are you self-employed, or are you working without pay in a family business or farm” (UCLA Center for Health Policy Research, 2017, p. 8). The occupation will be operationalized for this study by evaluating whether asthma is more predominant in the management, business, computer, engineering, law, construction, military, and health care fields, which were highlighted in the study and cited by the respondents. Evidence will be gathered to assess whether there is a pattern in the occurrence of asthma across these respective fields, as done in the work of Morgan et al. (2014).
Health care access
Health care access was assessed by asking participants to explain their usual sources of health care services, whether they had trouble in finding an affordable health care plan, and whether anyone had helped them to find a suitable healthcare plan. The interview process also sought their views regarding the relationship with the person who helped them find an affordable health care plan, and whether they had visited an emergency room in the last 12 months. Health care access will be operationalized for this study by evaluating whether the respondents experienced a delay in getting health care services in the past 12 months, or whether they were unable to get the services in the first place. Therefore, the study will focus on understanding whether there is a relationship between the number of people who are unable to get health care access and the incidence of asthma.
Covariates
The covariates include age, gender, marital status, housing type, and caring neighbors.
Age
Age was assessed by asking participants to respond to one question – “are you between 18 and 29, between 30 and 39, between 40 and 44, between 45 and 49, between 50 and 64, or 65 or older” (California Health Interview Survey, 2016, p. 3). This covariate (age) will be operationalized for this study by evaluating the mean and median ages associated with asthma attacks. Stated differently, evidence will be sought to establish if asthma incidences commonly occur among respondents within a given mean or median age, or whether a group of respondents are less likely to be suffering from the attacks because of their age.
Gender
Gender was operationalized by asking respondents one question – “are you male or female” (California Health Interview Survey, 2016, p. 2). The participants could only answer in three ways: “male,” “female,” or “refused.” This categorization of responses means that there was a tripartite conception of the research question (Clarke, Sastry, Duffy, & Ailshire, 2014). Gender will be operationalized for this study by evaluating how many males and females suffer from asthma. In other words, the analysis will involve investigating whether there are any gender discrepancies in the reporting of asthma incidences. Thus, this variable will be a dichotomous one.
Marital status
Marital status was assessed by asking participants one question – “are you now married, living with a partner in a marriage-like relationship, widowed, divorced, separated, or never married” (California Health Interview Survey, 2016, p. 5). Marital status will be operationalized for this study by finding out whether the respondents were married, or not. Like age, the analysis will be dichotomous. Thus, all other forms of unions (or living arrangements) will be ignored to create a dichotomous understanding of whether marital status affects the relationship between the dependent and independent variables. Therefore, only two outcomes could emerge in the study: marital status has influenced the relationship with the dependent variable, or not.
Housing Type
Housing type was assessed by asking participants the question – “do you live in a house, a duplex, a building with three or more units, or in a mobile home” (California Health Interview Survey, 2016, p. 6). They were also asked to state which type of residency status they held in their homes. Two options were available: renting or owning a home. Housing type will be operationalized for this study by finding out if there is a relationship between the different types of household ownership (renting or owning a home) and the dependent variable, which is asthma. This analysis will be undertaken to understand whether there is a correlation between homeowners and asthma incidences, or whether there is a relationship between the number of people who rent a house and asthma. This analysis would aid in understanding whether housing types affect asthma incidences (Harring & Blozis, 2014).
Caring neighbors
Caring neighbors, as a covariate, was assessed by asking participants to respond to this question – “tell me if you strongly agree, agree, disagree, or strongly disagree with the following statement: people in my neighborhood are willing to help each other” (California Health Interview Survey, 2016, p. 6). Participants were also asked to respond to two statements: “people in this neighborhood do not get along with each other” and “you can count on adults in this neighborhood to watch out that children are safe” (UCLA Center for Health Policy Research, 2017, p. 9). Caring neighbors will be operationalized for this study by examining the number of people who trust their neighbors. This variable will be operationalized this way because trust is a significant indicator of people’s ability to care for their neighbors (Depaoli, 2013; De Mutsert, Sun, Willett, Hu, & Van Dam, 2014).
References
Bakk, Z., Tekle, B., & Vermunt, J. (2013). Estimating the association between latent class membership and external variables using bias-adjusted three-step approaches. Sociological Methodology, 43(1), 272-311.
Clarke, P., Sastry, N., Duffy, D., & Ailshire, J. (2014). Accuracy of self-reported versus measured weight over adolescence and young adulthood: Findings from the national longitudinal study of adolescent health, 1996–2008. American Journal of Epidemiology, 180(2), 153-159.
De Mutsert, R., Sun, Q., Willett, W., Hu, B., & Van Dam, R. (2014). Overweight in early adulthood, adult weight change, and risk of type 2 diabetes, cardiovascular diseases, and certain cancers in men: A cohort study. American Journal of Epidemiology, 17(9), 1353-1365.
Depaoli, S. (2013). Mixture class recovery in GMM under varying degrees of class separation: Frequentist versus Bayesian estimation. Psychological Methods, 18(1), 186-219.
Harring, R., & Blozis, S. (2014). Fitting correlated residual error structures in nonlinear mixed-effects models using Sas Proc Nlmixed. Behavior Research Methods, 46(1), 372-384.
Morgan, J., Collins, C., Plotnikoff, C., Callister, R., Burrows, T., Fletcher, R.,… Lubans, D. (2014). The “Healthy Dads, Healthy Kids” community randomized controlled trial: A community-based healthy lifestyle program for fathers and their children. Preventive Medicine, 61(1), 90-99. Web.