Professional Documents
Culture Documents
Variables
Types of variables
Level of Measurement
Errors of Measurement
Univariate, bivariate, and multivariate data sets
Variables
The numbers which we observe and record in a data matrix
may have very different meanings and this fundamentally
affects the inferences we can draw from them.
A variable is a characteristics or property of a person, an
object or a situation, comprising a set of different values or
categories.
For example, height, weight, age, blood type, gender,
nationality
Types of Variables
Blood type
Record in terms of kind –
Gender
Qualitative Variables
Nationality
Height,
Weight Measured in units –
Quantitative Variables
Age
A. Qualitative variable – Categorical
Variables
Categorical variables (CVs) arise when individuals are
allocated to categories. Country of birth is a CV as is the
highest level of education.
Numbers can be used to code the categories into which
individual falls but those numbers are nothing more than
codes.
Qualitative variable – Categorical
Variables
However, if for example, respondents are asked to say
whether they strongly agree’’, ‘agree’, or ‘disagree’ or ‘strongly
disagree’ with some proposition, it is not unusual for equally
spaced scores such as 1, 2, 3, 4 to be assigned to these
categories, thus appearing to turn them into metrical variables.
Ordered Categorical Variables e.g. level of education
Unordered Categorical Variables e.g. gender
B. Quantitative Variable – Metrical
Variable
Metrical variables are those which can be recorded on some
kind of scale, like response times, lengths or examination
sores, where the interval b/w two values has a quantitative
interpretation.
Continuous Variable can take any value in an interval e.g. ?
Discrete
Continuous
Discrete Variable
Variable that can take only a discrete set of integers or whole
numbers i.e., the values are taken by jumps or break.
It represents a count data.
Example:
Ordinal
Interval
The units in which they are measured are arbitrary - money may be
measured in dinars or dollars, for example.
They are termed ratio level variables because ratios are independent of
the units of measurement.
For example, a camera which costs 3 times as much as another when
priced in dollars will still be 3 times as costly if the ratio is expressed in
any other currency.
Ratio Scale Data Example
RULER: inches or centimeters
YEARS of work experience
Absolute error
Error has both magnitude and direction and error in statistics does not mean
mistake which is a chance inaccuracy
Errors of Measurement
Biased error
Accurate measurement
Biased Error
When the deviations, i.e. the excess and defects, from the
true value tend to occur equally often – random error or
accidental errors
Accurate Measurement
Benjamin Disraeli,
Former British Prime Minister
A Word of Caution
Lecture 2: Variables and Levels of
Measurement