STATISTICS: A set of methods and rules for organizing, summarizing, and interpreting information.

 

 

POPULATION: A set of all the individuals of interest in a particular study.

 

 

SAMPLE:  A set of individuals selected from the population, as representative of the population in a particular study.

 

 

PARAMETER: Usually a numeric value that describes a population.

 

 

STATISTIC: Usually a numeric value that describes a sample.

 

 

DATA: Measurements of observation (singular: datum).

 

 

A data set is a collection of measurements or observation.  A datum is a single measurement (also called score, or raw score)

 

 


DESCRIPTIVE STATISTICS

Statistical procedures used to summarize, organize, and simplify data.

 

 

INFERENTIAL STATISTICS

Techniques that allow us to use sample data to make statements about the population.

 

 

Sampling error

The discrepancy (error) that exists between sample data (a sample statistic) and the properties of the population (parameters)

 


SCALES OF MEASUREMENT

 

NOMINAL SCALE

A scale that has no numeric properties. The categories or values on the scale differ by name only (e.g. male-female, red, black, blonde hair color)

 

 

ORDINAL SCALE

Values on the scale are meaningful in terms of a rank order of some quantity (e.g. high, medium, low, in popularity)

 

 

INTERVAL SCALE

The values on the scale reflect intervals that are equal in terms of size (e.g. temperature).  There is no absolute zero point on the scale (where the zero indicates absence of the attribute measured)

 

 

RATIO SCALE

Has equal intervals and a true zero point (e.g. the number of errors on a proofreading task during a 10-minute period).

 

 

The type of measurement scale determines the appropriate statistic used when analyzing data.

 

FACTORS AFFECTING YOUR CHOICE OF A SCALE OF MEASUREMENT

 

A nominal scale yields the least information. An ordinal scale adds some crude information. Interval and ratio scales yield the most information.

 

The statistical tests available for nominal and ordinal data (nonparametric) are less powerful (sensitive) than those available for interval and ratio data (parametric).