Talk:Statistics/Introduction/What is Statistics

General observations
I hope you will consider the following criticism as constructive.

As the opening paragraphs of a book, neither the flow nor the content are sufficiently "racy" to capture the attention of a putative student. The traditional opening remarks found in many if not all statistics textbooks refer to the etymology of the word "statistics". Such discussion is easily understood and gently eases the reader into the subject matter.

The reasons for doing statistical studies are much more varied than garnering biometrics from large animals, and at the very least a selection of contrasting applications should be included in the book's opening preamble.

The English is not sufficiently straightforward in places to allow smooth reading of the text. By this I mean that one has to pause periodically and work out what the author intended to say. J.Morgan. 6 Feb 2006

=An Outline=

I think it would be best if we came up with a good general outline for a "What is Statistics?" section. It would give us a clear train of thought from beginning to end, improve the flow of the section, and address any overall issues we may have with the section. I am just throwing out a first draft, please edit as necessary as we transition the actual content to match the outline.

Additionally, it would be a good idea to think about the questions which might be asked in a quiz on this section. Here are some possible questions:


 * "From where do we get the word statistics?"
 * "Who invented the term "standard deviation"?"
 * "I give you a list of numbers and ask for their sum. This is an example of ___________ statistics."
 * "Name three of the five(/six/seven) central concepts of statistics."
 * "Give a real-world example where it would be appropriate to use statistics."

Outline

 * Some Loose Examples
 * I always found those first pages in maths textbooks which laid out one or two real-world examples were very helpful in alleviating the more numbers-heavy aspects of the rest of the book. It would also be good if these examples covered the gamut of statistical topics (data description, probability, distribution, surveys/samples, correlation/causation, etc.) without too much overlap. They also can provide nice tie-ins to later chapters ("In the opening chapter, we asked how you would conduct a test to determine a drug's effectiveness ...")
 * Medicine: Drug testing = samples, null hypothesis, placebo effect
 * Business: Sales data report = mean/median/mode, standard deviation, bar graphs
 * Sports: Proposition bet = probability
 * Politics: Conducting a poll = sample size, bias, margin of error
 * Technology: Widget maker error = tolerance, yield, min-max analysis
 * A General Definition
 * Again, as this is the general definition contained in the introduction, we should be leery of any attempt to compartmentalize it ("Statistics includes Bayesian statistics, which does such and such, and also Gaussian statistics, and also, and also" ad infinitum), to overgeneralize it ("Statistics is the study of why nothing in life is certain"), and especially to underrepresent it ("Statistics are used a lot in the worlds of business, politics, and technology," with the implication that no other fields use statistics "a lot.")
 * Etymology
 * Does anyone have any interesting stories about the etymology of statistics? Or the history of statistics as it relates to its etymology? When did the word itself come into popular use? I find the fact that its modern derivation is German in origin to be particularly amusing.
 * History
 * The history should be 4 paragraphs at most, in my opinion, for two reasons. First is that this book, while attempting to encompass many facets of statistics, is more primarily concerned with the mathematical treatment of it - in short, this is not the History of Statistics book. (And there are many good references for its history.) Secondly, most of the concepts to be covered in this book have been discovered and enumerated in the past 200 years, suggesting that the topic's history is not so illustrious.
 * Also, these paragraphs should focus less on the details of the concepts, and more on what particular aspects of statistics were being advanced at this time.
 * Good Example - "In the 1860s, British scientist Francis Galton began using statistics to study hereditary traits, and helped form the concept of regression to more accurately identify and predict the relationships between things."
 * Paragraph 1 - Pre-1600 - Summarize all statistical efforts prior to the modern adoption of statistics in the 17th century. My own personal suggestion would be to focus on the taxation and collection efforts of the ancient world, with a culmination in the Domesday Book.
 * Paragraph 2 - 1600-1800, Early efforts to modernize statistics: Graunt, Pascal, Bernoulli, Bayes.
 * Paragraph 3 - 1800 - 1900, Gauss, LaPlace, Galton, the bell curve, Social Statistics
 * Paragraph 4 - modern day advances, Pearson, Spearman, Fisher, Neyman
 * "For additional reading, consider Stephen Stigler's Statistics on the Table: The History of Statistical Concepts and Methods."


 * Statistics as a subset of mathematics
 * By this I mean we should also explain that statistics is a lead in to many modern mathematical concepts: probability, distribution, samples, surveys, the bell curve, estimation, and data analysis. We shouldn't try to get into too many practical applications (that's covered in a later part of the introduction), nor should we get bogged down in definitions or examples. It should be sufficient to say that in answering the question "What is statistics?", part of the answer is "Statistics is an important part of mathematics."
 * Up Ahead
 * Just like it says, telling them what lays ahead within the Introduction chapter. Also, what's the deal with navigation within the Wikibooks? Is there some sort of navigation code available? Because right now all of these sections are just dead ends from the book perspective.Kthejoker 17:21, 7 March 2006 (UTC)