Estimations, Predictions and Strengths

Strengths and Limitations of Statistical Analysis

After reading this section, summarize the key strengths and limitations of statistical analysis in an organized list and post to your Learning Journal.

Statistical analysis uses both existing values as well as the statistics calculated from raw data to draw conclusions about the data set as a whole. Statistical analysis considers all of the data points from the data set, not just a cross-section or representative sample of the data. Through this science of collecting and analyzing, trends and patterns can be spotted and large amounts of data can be processed.

What are some of the strengths of statistical analysis? A partial list follows:

· Research findings that have been replicated on different populations and subpopulations can be used to make general conclusions about the population as a whole. For example, a research study of unemployment conducted on countries in the European Union can be combined with comparable studies conducted in North America, Asia, South America, and other parts of the world to make a general statement about unemployment around the world.

· Research findings on random samples can also be synthesized to create general conclusions about that data as a whole. For example, a study on the consumption of candy bars can be combined with a study on the consumption of potato chips to draw a conclusion about the consumption of “junk” foods.

· Because of the availability of statistical software, data analysis is relatively less time consuming than conducting new experiments to obtain new data. Being able to use statistical analysis to cull new information from existing data is a great time-saver.

· Some data collection methods, such as telephone interviews, allow data to be obtained relatively quickly. This information can then be retained and analyzed in multiple ways as new ideas for using that information occur to researchers.

· When studying large numbers of people or large sets of data, statistical analysis is extremely useful as it allows the data to be sorted and categorized in flexible ways. Measures of central tendency, dispersion, and position can help summarize the overall data; graphical displays will reveal outliers and the significant patterns can be more easily observable.

· The nature of the process for statistical analysis leads to unbiased results which can then become credible sources of information. While people (such as those in power or who are funding a project) may have different agendas for wanting data to appear a certain way or for certain conclusions to be drawn, statistical analysis does not bring these biases to the data. Additionally, the results of research are independent of the researcher.

· The numerical data from statistical analysis are precise and quantitative. As the saying goes, “The numbers don’t lie.”

· Statistical analysis is a way to test and validate existing theories about how and why certain phenomena occur.

· With statistical analysis, you can test hypotheses on existing data before running a new experiment. This enables the researcher to expend a minimum amount of time and money to at least see if the hypothesis is on the right track.

· A researcher can develop an analysis of the data that eliminates the extraneous influences of unrelated variables and, thus, allows one or more cause-and-effect relationships to be credibly established.

· Statistical analysis is an extremely useful tool for obtaining data from which quantitative predictions can be made. For example, if a trend is identified through statistical analysis, the data can be extrapolated to predict where the next data points in the data set are likely to fall.

While these strengths for statistical analysis make it an attractive option for working with data, it is important to be aware of the limitations of this type of analysis as well. Some limitations follow.

· The information produced or conclusions drawn from statistical analysis might be too abstract or general. If a researcher is looking for information or data on a specific, local situation, the general information drawn from a large national or international study may not be applicable. For example, knowing the national average for calories consumed by an individual will not enable you to say with any reliability what the average calorie consumption for a 15-year-old boy in your local school district will be.

· Confirmation bias is the tendency of people to gather and remember only the information that supports their current beliefs. It is sometimes said that people hear what they want to hear and see what they want to see. In the world of statistical analysis, a researcher might miss out on a trend or pattern because he or she is focused on testing a particular theory or hypothesis rather than thinking about what other, new questions should be asked.

· If, through the analysis, the researcher chooses categories that do not reflect the categories of the target group, the data could be skewed. Theories could be proposed and conclusions could be drawn about the target group, based on the data, which are not actually true for that group.

Navigating Statistical Analysis

Explore the following articles and documents that discuss the strengths and weaknesses of statistical analysis. What advice would you give to others on interpreting and evaluating statistical claims? Share your thoughts on the Discussion Board; then read the advice posted by your peers and respond to one that you find particularly insightful.

Resource	Description
Quantitative and Qualitative Research	Consider and compare the strengths and weaknesses between quantitative and qualitative research presented in this article by UXMatters.
Disadvantages of Statistical Analysis	Read this article on the four essential limitations of statistics. How might you argue the limitations described in this article?
Small Studies	Consider the strengths and limitations of small studies, as described in the ER Journal. How do they compare to the strengths and limitations of large studies?
Navigating Statistics	This webpage on misleading statistics describes how to navigate and make sense of statistical claims, describing the simultaneous value in statistical analysis and the pitfalls.