Book Report # 59: Meaning from Data: Statistics Made Clear Part 2.
Begin: 9/2/2024
Finish: 1/5/2025
Title: Meaning From Data: Statistics Made Clear Part 2.
Author: Professor Michael Starbird
University of Texas at Austin.
Why I choose to read this book:
Data are value-free and useless or actually misleading until we learn to interpret their meaning appropriately. Statistics provides the conceptual and procedural tools for drawing meaning from data. This book helps me understand statistics so I could teach and tutor students attending the Merced College Statistics class.
What I learned from This Book:
Law:
In the law, there are many places where statistical data and inferences are used to help in making legal decisions. The law provides many opportunities in which statistics play a significant role in the dispensation of justice. A statistical anomaly know Simpson’s paradox , that is a situation where both subprograms indicate that women are being treated better. The Simpson’ paradox is an example of possible effect of a lurking variable. In this case, the lurking was the existence of the subprogram.
Democracy:
Simply count the number of first place votes – Plurality voting. It works great if there are two candidates. Other voting methods: second-place preferences, Broda Count and pair-wise sequential voting. Arrow’s impossibility theorem makes us realize that the concept of democracy choice is an intrinsically problematic issue.
Sports:
Statistics about sports are fun. They also help us understand sports and appreciate and evaluate the success of individuals and teams. Comparing standard deviations away from the mean is a method of normalizing the comparisons over different eras. In a sense, it measures how well a person performed relative to his contemporaries.
Risk – War and Insurance:
Insurance is an industry based on probability. the whole concept of insurance is base on statistics. We can view insurance is a game of chances.
Real Estate – Accounting for Value:
Multiple linear regression is a technique by which we can approximate or summarize a situation where there are several explanatory variables influencing the response variable.
Misleading, Distorting and Lying:
Outliers can have so large an effect on the mean that merely stating the means gives a distorted view of the actual distribution. Sometimes a much better measure of the center would be the median which is not affected by outliers. Biased samples, some intentional, some not, are common. Using sample that is not representative of the whole population gives a distorted views. The wording of questions in a survey can influence results. Virtually any news source is biased, in the sense that its contents are chosen for interest. Frequently, the interest in a story comes from being rare and being bad. Any new source is biased but we must realize that an unbiased news reporting system would be dull. Graphs and phrasing can be distorted. Another way to have accurate graphs that are misleading is to draw the height of the graphical symbol but is drawn to look like a three dimensional thing. Extrapolating trends mindlessly can give ridiculous conclusions. World-record running times cannot continue a linear trend because there are physical limits to how fast people can possibly run. People often start with a true correlation but then derive a false causal relationship from it. Lurking variables often underlie such misconception.
Social Science:
Factor analysis is a statistical technique that tries to find whether data comprising a number of variables can be summarized or explained by a small number of factors.
Health Care Systems:
Everyday we make decisions about our personal health; it is one of the most important things that we think about. Our state of health is often measured by data. These data present to us a picture of our well-being. Part of our decision making about such questions as whether to take cholesterol-lowering medication is done by comparing our numbers to those reported in studies that have been conducted with a large number of people. Quack medicines appear to work because of the phenomenon called regression to the mean: All ill person usually expect to return to his or her mean health situation.
Economics:
Economics is certainly one of the most common arenas in which we think of statistical data as being centrally important. Data mining refers to the process of looking at an existing collection of data to find patterns or trends. Forensic accountants have used the Benford’s law to detect fraud. We must be careful to avoid the pitfall of drawing inappropriate conclusion from data mining pitfalls.
Sciences:
Statistics and statistical analysis of data are obviously central in all aspect of science. Advances in empirical science depend on drawing deductions from data. A scientific theory is tested by comparing experimental results to predictions of the theory. Using carefully executed statistical capture-recapture methods, scientists can estimate quantity as diverse as the population of tigers in a jungle, the volume of water in a lake and the size of natural gas deposits in the ground. Statistical analysis of experimental data is key to validating or invalidating a scientific theory.
Statistic Everywhere:
Statistical reasoning can contribute to decisive arguments in matters that seem very difficult to resolve, and even issues that don’t appear to have any statistical components to them at all. Monte Carlo method, which involves using random process by a computer to generate thousands of scenarios, enabling statistical techniques to derive the distribution of the behavior of a system. Any attempt to reduce data to a formulaic adherence to following test is likely to be misleading and can often produce nonsensical arguments. Statistics usually comprise of two parts: 1) Organizing, describing, and summarizing a collection of data when we know all the data. 2) Inferring information about the whole population when we have data about on a sample of population. When we make an estimate of the value of a feature of the whole population we have to describe our accuracies: 1) How close is our estimate to the correct value? 2) How confident are we that our estimate is in fact that close?
How will this book contribute to my success upon release:
This statistic book allows me to understand statistics so I could tutor students attending the Merced College statistics class. This book has improved my analytical and critical thinking skills. This information can be shared with communities which I hope to volunteer my teaching, tutoring and mentoring services.
