What have you learned about Statistics

Once data has been collected from the research field either as secondary or primary data it has to undergo some statistical analysis to help in deducing an outcome based on the research question. Statistics is an essential discipline that allows the gathering, analysis, presentation, interpretation, as well as the organization of numerical data obtained in both small and large quantities. Through integrations of various analytical components, statistics can be applied to a wide range of research fields. Therefore, the study of statistics entails various elements that are significant in detailing the entire research concept including; inferential statistics, descriptive statistics, hypothesis development and testing, selection of variables or bivariate, and multivariate measures of association. The paper has enumerated the insights of each one of the elements depicting their relations to Topitzes, Pate, Berman, & Medina-Kirchner’s (2016) article on Adverse Children Experiences. In summary, statistical elements are essential in aiding researchers to develop a definite conclusion that is paramount in the decision-making process.
What have you learned about Statistics?
Data is part and parcel of nearly everything in the human life. In most cases, individuals have the desire of knowing about things such as how much, how many, how often; and others. Statistics is an essential discipline that allows the gathering, analysis, presentation, interpretation, as well as the organization of numerical data obtained in both small and large quantities.

The need for performing these processes are to establish an understanding of proportions in fullness from those in the representative data. I have come to learn that statistics entirely deals with numbers and tries to approximate points of accuracy, although in other circumstances it might depend on precision compared to accuracy. Based on that, there are various techniques utilized in the process of visualizing relationships in regards to the data as well as the systematic methods that provide a complete understanding of relationships that are in association with mathematics. Therefore, the study of statistics entails various elements that are significant in detailing the entire research concept including; descriptive statistics, inferential statistics, hypothesis development and testing, selection of variables or bivariate measures of association, and multivariate techniques.
Descriptive statistics are applied in the description of the fundamental aspects of the data under study. They offer clear summaries detailing about the samples and the measures. When used together with the graphical analysis they establish the foundation of virtually all quantitative analysis of data. Descriptive analysis is essential to assist us in identifying if any patterns are resulting from the gathered data at the same time offers grounds for further study. Consequently, descriptive data limits people from making any suppositions about the data collected, but instead it simplifies data for easier interpretation. Moreover, descriptive statistics allows data to be presented in a more simplified and understandable to the user. Stand for the user. For instance, in a study analyzing the Adverse Children Experiences (ACE) in America which has to reflect the whole country means that there is a broad cross-section that has been put under study. Therefore, the outcomes, in this case, must be simplified further with the use of descriptive statistics to make it simpler for other users. An example is the determination of the rate unemployment among the black men which is acquired through a ratio between the numbers of workers to the working-age citizens. The United States’ jobless rate before the Great Recession presented 36.7% that increased to 41.8% in 2011 and lowered to 41% in 2014, while in the Black Americans was 41.5%, rose to 48.9% and later reduced to 45.1% respectively (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). The descriptive data presented makes it easy for any user apart from the primary researcher to interpret and use the information.
Differing with descriptive statistics that defines the features of a single variable, inferential measurements are employed in cases where expectations, approximations, or conclusions have to be established concerning the bigger population related in the sample. However, the inferences face some drawbacks due to errors that arise because of the smallness of the sample used to reflect on the bigger population which we cannot deduce with confidence that the outcomes are accurate and a true reflection of the most significant population. Naturally, inferential statistics is concerned with the analysis of two or additional variables. For instance, in the analysis of the ACEs, a predictor that was used entailed a questionnaire with variances of 10 adversities for children from birth to the age of 18 years. The advertise divided into three categories three types of abuse (physical, verbal and sexual); two neglect types (emotional and physical); and five household dysfunction manifestation (substance abuse, mental illness, divorce, battered mother, and imprisoned family member) (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). Based on the adversities a predictive score measure was developed to included ACE = 0, ACEs = 1–2, ACEs = 3–4, and ACEs ≥ five that was evaluated basing on the face validity, distribution and the prior research (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). Even though the outcomes will present some sense of reliability, the fact that the inferences are made based on some notions there are higher chances of errors about the conclusions made by the entire population.
A critical section of statistics is the hypothesis development and theory testing. Hypothesis testing entails a form of statistical guessing that involves establishing a question, collection of data followed by a reflection of what the information reveals to us on how to proceed. In a typical assumption test, theories are reliably proclamations regarding the population. There are two constant principles when it comes to measurable theory testing. The premise of the trial is referred to as the invalid theory presented by H0. The null hypothesis depicts that there are no differences between a surmised population mean and a specimen mean. Therefore, the alternative theory could only be sustained by dismissing the null hypothesis. The invalid theory can only be rejected if the differences between the approximated mean and the example mean are significant, but when the difference is less, there is no dismissal. For instance, in the ACE study the hypothesized that “men in the study would report: a) high rates of individual and cumulative ACE exposure relative to the original ACE study; b) poor health status particularly when compared to population or alternative sample norms; and c) linkages between ACE exposure and health and between ACE exposure and barriers to current employment” (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). All the hypothesis tested positive. For instance, the participants in the research were reported to have a higher likelihood of mental or behavioral health challenges about the alternative samples or national norms. In particular, it was realized that the rate current smoking rate for the study sample (55.8%) was way above the national averages for an entire number of adults (19.0%). At the same time above adult men (21.6%), as well as for the Black Americans (19.4%) as presented by the CDC (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). Therefore, the hypothesis tested in this case was proved to be significant.
Apart from that, the measures of association denote a broad range of coefficients that are used to assess the statistical strength of the association on different variables of interest; these measures of association can be detailed in various ways based on the analysis. On this grounds, the statistics that are known to describe or establish an inference concerning a single distribution are known as the univariate statistics. Despite the fact that the univariate statistics makes the grounds for many other forms of statistics, no issue is related to the measures of association within variables that can be solved by analysis of single variables. Since the ACEs study considered a broad range of variables to construct a conclusion based on their relationships, the use of single variable analysis was limited. Therefore, to measure relationships among some variables, it is required to move to another level of statistics examination of bivariate statistics which in general examines two variables.
Cross-classification tables that are used in the determination of dependence and independence for variables and events is an example of the bivariate statistics. Moreover, a test that is meant to establish the differences between two proportions could also be regarded as a form of bivariate statistics. Besides that, other forms of bivariate statistics methods that are essential for the measurement of association are the nominal or ordinal variables where lambda represents the nominal while gamma and Kendal’s Tau-b represents the ordinal variables. For instance, in the ACEs study, the researchers made use of the Kendal’s Tau bivariate correlations that exposed that the ACE index was substantially related, in the predictable direction, with each associated mediators – total health, drug issues, current smoking, anxiety, and depression. The sample analysis illustrated that the ACE index had a significant association with both distal study results that is the history of incarceration and history of employment. As a result “All statistically significant relations reached α< 0.05 level of probability, while the majority reached α< 0.01” (Topitzes, Pate, Berman, & Medina-Kirchner, 2016).
However, when the variables under examination exceed three variables, so as to deduce the measure of association, many researchers will move beyond bivariate statistics to multivariate statistics. In this case, the relationships amongst some variables are simultaneously evaluated. For instance, the multivariate statistics was used to deduce the relationships amongst the multivariate regressions that were adjusted for the research covariates ending up with results as follows. “The ACE index was significantly associated with each of the study’s health indicators and employment-related measures, i.e., general health (p = 0.002), current smoking (p = 0.010), drug problems (p < 0.000), depression (p < 0.000), and anxiety (p < 0.000), history of employment problems (p = 0.001), and history of arrest or incarceration (p = 0.001)” (Topitzes, Pate, Berman, & Medina-Kirchner, 2016). Based on the results, in all incidences the escalation of the ACE scores also meant an increase in employment-related and health-related problems. Convincingly, bivariate and multivariate statistics are not only useful for statistical reasons but stretch to an extended functionality in social science research.
After a period of statistics overload, it has come to my realization that statistics is more than numbers. In regards to the article information acquired, statistics, besides being used in its quantifying manner it is essential in determining and estimating the rate at which some social factors impact on our daily lives. The various elements of statistics that is: descriptive statistics, inferential statistics, hypothesis development and testing, selection of variables or bivariate measures of association, and multivariate techniques are critical in establishing a clear conclusion in any given research study. The descriptive data entails the simplification of crude data through sequential organization making it easier for interpretation. Inferential statistics serves as an extension of the descriptive data where it offers predictions and inferences from the data acquired. The development of a hypothesis and theory testing is important in all research studies so as to determine the relation of the findings to the research question. Lastly, the measures of association are essential for the evaluation of the relationships between the variables under study which end up proving the hypothesis as well as help in the decision-making process.
References
Topitzes, J., Pate, D., Berman, N., & Medina-Kirchner, C. (2016). Adverse childhood experiences, health, and employment: A study of men seeking job services. Child Abuse & Neglect, 61, 23-34. http://dx.doi.org/10.1016/j.chiabu.2016.09.012