4. Type of Data
Qualitative
(Categorical)
Type of
Categorization
One
categorical
variable
Goodness of fit
X2
Two
categorical
variables
Contingency
Table X2
Quantitative
(Continous)
Type of
Question
Relationships
Number of
Predictors
One
Measurement
Continuous
Primary
Interest
Degree of
Relationship
Pearson
Correlation
Form of
Relationship
Regression
Ranks
Spearman's’ r
Multiple
Multiple
Regression
Differences
Number of
Groups
Two
Relation
between
Samples
Independent
Two Sample t Mann-Whitney
Dependent
Related Sample t
(paired t tests)
Wilcoxon
Multiple
Relation
between
Samples
Independent
Number of
Independent
Variables
One
One Way
ANOVAs
Kruskal-Wallis
Multiple
Factorial
ANOVAs
Multivariate
Analysis
Dependent
Repeated
Measures
Friedman
6. Types of Data
• Key Points
– Categorical data - nominal and can be counted.
– Numerical data may be ordinal, discrete, or
continuous, and are usually measured.
– VAS measurements are ordinal data.
6
7. Types of Data
• Qualitative
– Data which is descriptive
and characterizes an
event and may include
an intangible measure of
worth or quality.
7
8. Types of Data
• Quantitative
– Data which is measured
via a numerical scale.
8
10. Categorical Data
• Observations are grouped in categories,
counted, and sorted accordingly.
• When there are only two categories or choices
the data is referred to as binary or
dichotomous.
10
11. Categorical Data
• Examples
– Gender
• Male
• Female
– Type of operation
• CABG
• Hysterectomy
• Cholecystectomy
• Appendectomy
11
12. Categorical Data
• Examples
– Type of ICU Admission
• Medical
• Surgical
• Injury
• Illness
– Adverse/Untoward Event
• NPPE
• Positioning nerve injury
• Transfusion reaction
• PONV
12
14. Ordinal Data
• Data in which a relative value or ranking can
be applied.
– Can be viewed as a hybrid between categorical
and numerical data.
– The true measure of the data is not tangible but it
does have an essence that is more than just
descriptive.
14
15. Ordinal Data
• Recording observations
– Typical some type of numerical
system is applied.
• Numbers
• Roman numerals
– Scoring can also be letters or
symbols
• A, B, C, D
• +, ++, +++, ++++
• The advantage of a numerical
system
– Data can undergo
nonparametric statistical
analysis.
• In a nutshell, using a parametric
statistical analysis on ordinal
data.
15
16. Numerical Data
• Quantitative Data
– Discrete measurments
– Continuous measurements
• Discrete data
– Can only be a whole
integer
• You cannot have half a
person
• Continuous data
– Can take any value
• CBC values
• Cardiac Index
16
17. Numerical Data
• There can be further division of Continuous
Data.
– Interval data
– Ratio data
17
18. Numerical Data
• Interval Data
– Location of the zero value is arbitrary and not a
true zero point.
• Celsius temperature, Dates
• Ratio Data
– Simply stated this data has a true zero reference
point.
• Kg, m, in., lb, Kelvin temperature
18
20. A frequent tool used in Anesthesia
• VAS
– Can measure, pain,
PONV, anxiety, patient
satisfaction.
– When using the 100 mm
scale some researchers
use erroneously this
data as continuous data.
• Is everyone’s pain the
same?
20
23. Changing Data Scales
• Smoking status can be recorded as
smoker/non-smoker (categorical data), heavy
smoker/light smoker/ex-smoker/non-smoker
(ordinal data), or by the number of cigarettes
smoked per day (discrete data).
• MI – ischemia or no ischemia, or the extent of
ST segment depression in mm.
23