A quantitative analysis method, meta-analysis is largely utilized to categorically evaluate existing research work and arrive at feasible conclusion with regards to the research conducted. Meta analysis is undertaken in the medical, social science, business, education, ecology and other arena. In the medical arena, results of a meta-analysis can be effectively utilized to anticipate the outcome of a specific treatment or to understand the underlying implications of a disease. The advantage of meta-analysis is relatively larger and involves a combination of quantitative assessment pertaining to largely intricate and challenging research activities. The characteristics of the outcome of a specific research based on hypothesis that are extensively assessed are vital to understand the implication of meta-analysis. Also, lack of clarity in recognizing the impact of existing research could generate flawed results. Meta-analysis that has been arduously executed emerges as a valuable tool with regards to research based medicine. What renders meta-analysis as a most sought after method of quantitative analysis is its ability to be integrated with various other analytical tools.
Meta-Analysis of Observational Studies
Based on the aims and the research design of this study, meta-analysis was chosen as the method of choice in answering the research question. Meta-analysis is defined as a non-experimental quantitative research method used to pool together data obtained from two or more experimental or observational studies which have similar or related hypothesis (Bruce, Pope and Stanistreet, 2008; Anderson, 2010). As suggested by Polit-O'Hara and Beck (2010), systematic reviews were previously basically narrative until recently when statistical methods are now used to synthesise findings from multiple studies. Meta-analysis is useful in enabling objective conclusions to be made from a variety of findings which are sometimes conflicting (Villari, La Torre and Leyland, 2005) and it also increases the probability that an observed association is true, as sample size is increased and confidence interval narrowed (Polit-O’Hara and Beck, 2010). However as with any research method, meta-analysis that is not properly conducted either due avoidable bias or inappropriate combination of studies could be misleading (Sutton, 2000).
The choice of a meta-analysis as the quantitative research method was informed by availability of observational studies which when pooled together make a larger sample size thus increasing power and enabling generalisations regarding the relationship between optimal and non - optimal birth weights and the risk of type 2 diabetes. Other factors which influenced the choice of a meta-analysis is that cohort studies which make up most of research on birth weight and the risk of disease in adulthood would normally take a long time to conduct if it were to be a primary research and this would not be possible within the time frame available for this study. Furthermore, the rising prevalence of type 2 diabetes (Chen, Magliano and Zimmet, 2012; Herman and Zimmet, 2012; Campbell, 2011) suggests the urgent need for interventions based on synthesis of already available evidence in this regard.