After being introduced to a variety of topics in any subject, it is worthwhile to look back and analyze the value of the introductory content. Although every learning experience has some value, there is no doubt that some experiences are more valuable than others. A survey of statistical concepts is no different than any other subject. There are some concepts that will prove immediately fruitful in home or work life. Other concepts either need more attention or are just generally less useful in a wider range of non-statistics-based settings.
One of the most useful concepts in statistics is hypothesis testing. According to Mario Triola (2008), hypothesis testing is the “standard procedure for testing a claim about a property of a population” (p. 386). It provides the same basis for statistical analysis as the scientific method does for experimental science. In other words, hypothesis testing is valuable not because of its ability to answer difficult statistical questions, but because it stands as a foundation for other forms of statistical analysis to build upon. Additionally, using proper hypothesis testing early in statistical analysis ensures that further work is not analyzing the wrong data or trying to prove a claim that cannot be proven.
The second most valuable concept was the process involved in determining the necessary sample size. An appropriately sized sample is necessary to detect the statistical significance of an outcome. Conversely, too large a sample results in wasted resources, whether human or financial (Sedlak, Zeller, & Doheny, 2002). Having the skill to determine the necessary sample size beforehand is invaluable when the time comes for data collection. In my own work experience, some methods of data collection are resource intensive, and do scale well when targeted at the entire customer population. The ability to choose an appropriate sample size before testing begins will allow me to better utilize resources and use newly available resources to run other kinds of analyses.
One of the least valuable learning experiences in my introduction to statistics was the quick overview of data collection methods. This overview included quick explanations of various methods such as surveys and focus groups. This concept is important, but the sparseness of the information made this information slightly more useful than a dictionary look up. More than likely, this concept is better covered in more detailed in another subject. Fortunately, that’s really the only concept I found lacking. I originally thought I would not have much use for contingency tables. These tables compare two variables and determine if they are independent (Triola, 2008). This seemed very limited until I realized that correlation and regression require independent variables. It then occurred to me that since correlation and regression are an important part of my own work, that contingency tables are a necessary component. Therefore, something that I originally considered to by of low value, turned out to be the foundation of an extremely valuable concept.
Sedlak, C., Zeller, R., & Doheny, M. (2002). Determining Sample Size. Orthopaedic Nursing, 21(2), 63. Retrieved May 11, 2009, from Health Source: Nursing/Academic Edition database.
Triola, M. F. (2008). Elementary statistics (10th ed.). Boston: Pearson.