1 Measurement Scales - University of Colorado （1测量尺度-科罗拉多大学）.pdf

QMIN (2006-02-07) Measurement - 1.1 1 Measurement Scales Statistics operate on a data set. The data set may be viewed as a two dimensional matrix, very similar to a blank spreadsheet found in many contemporary software packages such as Excel. The rows of the data matrix are observations. In neuroscience, observations are usually organisms (humans, rats, mice) but occasionally they may be other phenomena such as cell cultures. The columns of the data matrix consist of attributes—sex, parietal lobe activity in a PET (positron emission tomography) scan, number of bar presses—measured on the observations. This chapter explains measurement scales, the different mathematical classes for the attributes. 1.1 Measurement Scales: Traditional Classification Statisticians call an attribute on which observations differ a variable. The type of unit on which a variable is measured is called a scale. Traditionally, statisticians talk of four types of measurement scales: (1) nominal, (2) ordinal, (3) interval, and (4) ratio. 1.1.1 Nominal Scales The word nominal is derived from nomen, the Latin word for name. Nominal scales merely name differences and are used most often for qualitative variables in which observations are classified into discrete groups. The key attribute for a nominal scale is that there is no inherent quantitative difference among the categories. Sex, religion, and race are three classic nominal scales used in the behavioral sciences. Taxonomic categories (rodent, primate, canine) are nominal scales in biology. Variables on a nominal scale are often called categorical variables. 1.1.2 Ordinal Scales Ordinal scales rank-order observations. Class rank and horse race results are examples. There are two salient attributes of an ordinal scale. First, there is an underlying quantitati