Statistics

• Author: Alison Weir
• Pages: 814-836

814

A. INTRODUCTION

Statistics is the analysis of data. Statisticians assist other professionals in

the production of trustworthy data by analyzing and modelling data to

make the meaning clear so that practical conclusions can be drawn from

the information. The role of statistics is to provide mechanisms for the

extrapolation of information in sample data to the real world, as well as to

develop measures of reliability for all such extrapolations. These measures

of reliability are particularly important in law when the trier of fact, wheth-

er judge or jury, is tasked with weighing the evidence presented at trial.

All disciplines use statistics:

 Medical researchers use statistically designed experiments.

 Global warming researchers use statistical predictive models with

missing data.

 Many industrial decisions are based on statistical systems that have

replaced hunches and guesswork.

 Government statistical agencies investigate issues to guide public policy.

A lack of understanding of the basic principles of statistics among the

general public has led to signif‌icant statistical misinformation being pre-

sented in the media and on the Internet. Questionable statistical applica-

tions are easy to f‌ind even in the research literature of many disciplines.

Care should always be taken to use statistical information from proper

sources. When in doubt, ask a statistician.

CHAPTER 23

Statistics

Alison Weir

Statistics 6 815

B. WHAT IS DATA?

Data in its most basic form (known as raw data) is a collection of individ-

ual pieces of information. Datum can be a number, word, measurement,

categorization, even description. Data, correctly analyzed, increases know-

ledge and informs decisions.

1) Data Can Be Categorical or Interval

Categorical data places subjects into categories. For example: gender

(M/F). Interval data is measured or counted. For example: number of sib-

lings, or height.

a) Categorical Data Can Be Nominal or Ordinal

Nominal data has no proper order. For example, gender could be presented

(M/F) or (F/M). Ordinal data has an order, but the order is not truly num-

eric. For example, Likert scales, where respondents rate statements on a

scale of 1-to-7, with 1 being “strongly disagree” and 7 “strongly agree,” are

ordinal data.

b) Interval Data Can Be Continuous or Discrete

Continuous data are real numbers. Practically speaking, any number from

any type of measurement is a real number. For example, length, area, and

volume are all continuous. Discrete data are numbers from any f‌inite set

of numbers. Countable data (for instance, the sequence 2, 4, 6, . . .) and

categorical data are also discrete.

Continuous Discrete

Interval Interval Categorical

Nominal Ordinal

Data

 23.1. Types of data. (All f‌igures and tables have been created by the author,

unless otherwise stated.)