Chapter 1 Statistics and how they are used

1.2 The uses of statistics

shows how descriptive statistics are used to describe data and how inferential statistics are used to reach conclusions from the analysis of the data

1. "To guide the design of an experiment or survey." A statistician should be consulted in the early planning stages so that an investigation can be carried out efficiently, with a minimum of bias. Once data are collected, it is too late to plan ahead. By then, it is impossible to impose an appropriate statistical design or compensate for the last of a randomly selected sample.

2. "To analyze data." Data analysis may take many forms, such as examining the relationships among several variables, describing and analyzing the variation of certain characteristics (e.g., blood pressure, temperature, height, weight), or determining whether a difference in some response is significant.

Example: SMOKING DURING PREGNANCY

A pioneering study of the effects on the newborn infant of smoking during pregnancy was reported by Simpson (1957). She examined data on 7499 patients in three hospitals in and near Loma Linda (California) University and found from the records that pre-maturity rates increases with the number of cigarettes smoked per day. A more recent review of the various studies on this topic is given by the Surgeon General's Report on Smoking and Health (U.S. Department of Health, Education, and Welfare, 1979). The principal conclusion of that report is. "Maternal smoking during pregnancy has a significant adverse effect upon the well-being of the fetus and the health of the newborn baby."

1.5 Clinical Trials

Describes the use of a clinical trial to determine the value of a new drug or procedure

1.4 Sources of Data

Discusses surveys and experiments, two main sources of data, and further classifies surveys as retrospective or prospective and as descriptive or analytical

1.6 Planning of Surveys

Previews some hints on how to maximize the value of survey data

What does statistic mean?

the word statistics has several meanings. It is frequently used to refer to recorded data such as the number of traffic accidents, the size of enrollment, or the number of patients visiting a clinic. Statistics is also used to denote characteristics calculated for a set of data--for example, mean, standard deviation, or correlation coefficient. In another context, statistics refers to statistical methodology and theory. In short, statistics is a body of techniques and procedures dealing with the collection, organization, analysis, interpretation, and presentation of information that can be stated numerically.

1.3 Why study statistics?

Explains how the study of statistics is important for research, for writing publishable reports, for understanding scientific journals, and for discriminating between appropriate and inappropriate uses of statistics.

1.1 The meaning of statistics

Formally defines the term Statistics and illustrated by describing what a statistician does

1.2 The Uses of statistics

It is helpful to distinguish between the two major categories of statistics. Descriptive statistics deals with the enumeration, organization, and graphical representation of data. Inferential statistics is concerned with reaching conclusions from incomplete information-that is, generalizing from the specific. Inferential statistics uses information obtained from a sample to say something about an entire population. An example of descriptive statistics is the decennial census of the United States, in which all residents are requested to provide information such as age, sex, race, and marital status. The data obtained in such a census can then be compiled and arranged into tables and graphs that describe the characteristics of the population at a given time. An example of inferential statistics is an opinion such as the Gallup Poll, which attempts to draw inferences as to the outcome of an election. In such a poll, a sample of individuals )frequently fewer than 2000) is selected, their preferences are tabulated, and inferences are made as to how more than 80 million persons would vote if an election were held that day: Statistical methods provide a logical basis for making decisions in a variety of areas when incomplete information is available. Here are some examples of scientific questions to which the application of statistical methodology has been useful:

1.7 How to succeed in statistics

Offers some tips on getting the most out of class and other resources.

1.1 The meaning of statistics

One way to understand statistics is to consider two basic questions: (1) what does the term statistic mean? and (2) what do statisticians do? One we have the answers to these questions, we can delve into how statistics are used.

Example: The Multiple Risk Factor Intervention Trial (MRFIT)

Paul (1976) reported on a national study of the primary prevention of coronary heart disease. The study's approach was to determine whether the risk of coronary disease in middle-aged men can be significantly reduced through intervention. This intervention entailed simultaneously reducing their serum cholesterol levels, treating any high blood pressure, and encouraging the men to stop smoking. The 7-year trial involved 20 clinical centers and 12,866 subjects, all initially health but at high risk for coronary disease. At random, halt the men were assigned to be followed through the intervention program, and the other half through their usual medical care, which included annual physicals and lab tests. The report of the results was prepared by the MRFIT research group and appeared in the "Journal of the American Medical Association (1982;248:1465-1477). Investigators observed that the risk factor levels declined in both groups. Furthermore, during the 7-year follow-up period, the mortality rates for coronary heart disease (CHD) were 17.9 deaths per 1000 for the intervention group and 19.3 deaths per 1000 for the untreated group. This was a non-significant difference, and the lack of a positive result has generated considerable discussion. There may be more plausible reasons for this outcome:

What do Statisticians do?

a statistician is usually a member of a group that works on challenging scientific tasks. Frequently engaged in projects that explore the frontiers of human knowledge, the statistician is primarily concerned with developing and applying methods that can be used in collecting and analyzing data. He or she may select a well-established technique or develop a new one that provides a unique approach to a particular study, thus leading to valid conclusions. Specifically, the statistician's tasks are as follows:

Learning Objectives

After studying this chapter, you should be able to 1. Define statistics 2. List several reasons for studying statistics 3. Distinguish clearly between a.) descriptive and inferential statistics b.)Surveys and experiments c.)retrospective and prospective studies d.)Descriptive and analytical surveys 4. Define bias 5. Describe the purpose and components of a clinical trial

3. "To present and interpret results." Results are best evaluated in terms of probability statements that will facilitate the decision-making process. Mainland (1963:3) defines statistics as the "science and art of dealing with variation in such a way to obtain reliable results." The art of statistics is especially pertinent to this task and involves skills usually acquired through experience.

The interpretation of statistics is both an art and a science. When results are said to be Significant, the statistician is making a probability statement. He or she is saying that the differences are most likely real differences rather than chance or random differences. Because probability is often linked with sample size, large samples can yield seemingly impressive results. However, when the statistician scrutinizes the sample more closely, she or he may realize that, even though the results appear impressive, they have little or no practical value. An excellent example of the effects of a large sample is presented in Chapter 13 and involves the Pearson Correlation technique. Statisticians can also help with another key aspect of interpretation: Are the results applicable to other groups? Subject selection, especially the nonrandom selection of patients with specific study eligibility requirements, poses a particular problem. The statistician can help determine if the results from a published study are applicable to a somewhat different group of patients. For example, a clinical study on drug efficacy might find certain effects in the study group, but would those same effects be found in an elderly population? Are studies on cardiovascular disease, which have historically focused on men, applicable to women? Are bone density studies, in which women have been studied most extensively, applicable to men? Although statisticians may not be able to give definitive answers, they should be able to five objective advice or consider prior to making a judgement