Theories of the Relationship Between the Homelessness and Mental Illness

On determining the correlation between homelessness and mental illness, we utilized the quantitative methodological approach. This paper will focus on the quantitative research approach to look at the relationship between variables and also test the theories of the relationship between the homelessness and mental illness.

Methods Used

The Cross-sectional approach was the first research design used in analyzing the relationship between homelessness and mental illnesses. Cross-sectional approach attempted to determine the number of people who were homeless based on their characteristics. Additionally, surveys and interview-based cross-sectional approach were used in obtaining data from the people who were homeless directly and indirectly by getting data provided by the volunteers and the general public (Morse, 2016).

Additionally, this research design considered diverse variables such sociodemographic variables, for example, mental health of the homeless people like mental disorders and consumption of drugs and alcohol. Also, the homeless history was also an important socio-demographic variable used, and to evaluate these variables, instruments such as the surveys and structured, open and semi-structured interviews were used (Morse, 2016).

Data Collection

We searched for surveys that estimated the prevalence of personality disorder, major depression, psychotic illness and alcohol dependence among the homeless people between the year 1966 and 2007 ("Prevalence of Mental Disorders among the Homeless in Western Countries: Systematic Review and Meta-Regression Analysis," n.d.). Moreover, the computer-based literature indexes and relevant journals by hand was searched. For the computer-based, keywords on being homeless such as the roofless, homeless and shelter. And the keywords relating to psychiatric illnesses, we used mental, depress, alcohol and drug. Guidelines were followed in getting the data as the articles which were non-English were then translated.

The studies done were to meet the following criteria in order to be considered accurate for data collection: Classification of the survey which was less than 50% of response rate was not included in the study of the population selected, for example, a sample of the homeless people in which direct interviews were not conducted on the selected population who were diagnosed with psychiatric disorders. And rather case note review and self-report was used in obtaining the data. Moreover, the data in which the study selected only juvenile or elderly people to ensure the data is accurate and uniform. The individual interviews were the basis of the reports that were included and one study that was identified which interviewed families was also excluded.

Data Analysis

Variance arcsine transformation was used in calculating the prevalence estimates because we used inverse variance weight when we dealt with binary data that had the low prevalence. Cochrans Q (x2-value and p-value) which describes the percentage of variations was used in estimating the heterogeneity among the studies. Where heterogeneity was high, (I2>75%), random effects models were used to produce study weights which reflected study variation and hence providing close to equal variation.

Sampling Strategies

Various sampling strategies were employed in these surveys. For example the consecutive sampling and the random methods. 950 individuals who are from shelters homeless were selected, and the others were selected from settings such as the soup kitchens, day centers, streets and other shelters. One study had a response rate of 85% and above, and the other was above 50%. Clinical examination was used without the use of the diagnostic instruments while diagnosing the psychotic illness, major depression, alcohol dependencies and personality disorder.

Results

The final sample of the research which had 25 studies was published between 1978 and 2006. This study included a total of 6,773 individuals who were homeless. In the survey of those who were of mixed gender samples, 80% of them were men and the rest women. Mental illness was reported in 20 surveys that were done, with a random effects prevalence being 10.3%. Additionally, all the estimates found ranged from 2.0% to 42.0% but with substantial heterogeneity between the estimates ("Prevalence of Mental Disorders among the Homeless in Western Countries: Systematic Review and Meta-Regression Analysis," n.d.

Moreover, higher prevalences were observed in the studies which were small, and the results obtained in different regions were also analyzed and categorized accordingly. In the individual variable meta-regression analyses was done and the surveys in which the interviewer was mentally ill it had a higher rate of prevalence of psychosis, that is, p=0.042, b=0.08 and the standard error was se[b] =0.04. But where the participation rate was lower <84%, it had lower prevalence, that is, b=-0.07, se[b] =0.013, p=0.091, and where the study had 100 or more participants it also had lower prevalence. In the meta-regression model which it included those characteristics, the rate of the participant only remained significant. Lower prevalence rates were associated with lower participation rates.

References

The Prevalence of Mental Disorders among the Homeless in Western Countries: Systematic Review and Meta-Regression Analysis. (n.d.). Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2592351/#pmed-0050225-b037

Morse, J. M. (2016). Mixed method design: Principles and procedures (Vol. 4). Routledge.