Chapter 2: Graphical Descriptive Techniques I Flashcards
Methods of descriptive statistics
Graphical
Tabular
Numerical
Descriptive statistics
Arranging, summarizing and presenting a set of data in such a way that useful information is produced.
Factors that determine appropriate statistical method
Most importantly:
Type of data
Information that is needed
Objective of statistics
To extract information from data
Variable
Some characteristic of a population or sample (that can change within the population)
Values of a variable
Possible observations of the variable
Data
The observed values of a variable (plural)
Singular = datum
Types of data
Interval
Nominal
Ordinal
(Also ratio but for statistics ratio = interval)
Interval data
Real numbers
Quantitative or numerical
Intervals between values are meaningful
All calculations are valid
Data may be treated as ordinal or nominal
Nominal data
Categories
(Generally assigned numbers for analysis purposes, but numbers could be randomly assigned, order does not matter)
Qualitative or categorical
Only calculations based on frequency or percentage of occurances allowed
Cannot be treated as other kind of data
Ordinal data
Appears nominal except the order or the values has meaning (scales). Order of values matter
Any code that preserves the order will produce the same result. Intervals are not meaningful!
Calculations based on ordering process are valid
Data may be treated as nominal (not interval)
Calculations for interval data
All calculations permitted
Nominal data
No calculations permitted - averages are meaningless. Percentages of occurances (frequency) is meaningful
Ordinal data
Only permissible calculations involve ranking in order (median)
Hierarchy of data
In order of permissible calculations From interval (all computations allowed) to nominal (no calculations allowed but frequencies)
High level data may be converted to low level data (but only if necessary) but not vise versa
Frequency distribution
Table presenting categories of nominal data and their frequencies
Relative frequency distribution
Lists categories and the proportion with which each occurs
Graphical presentation of nominal data
Bar chart and pie chart
Both good for simple presentation of numbers associated with categories IF it enhances understanding
Bar chart
Generally used to display frequencies
Pie chart
Generally used to show relative frequencies (as always showing parts of a whole)
Graphical presentation of ordinal data
Same techniques as nominal (bar charts and pie charts)
Bar charts- bars in ascending or descending ordinal values
Pie charts- wedges typically arranged clockwise (ascending or descending)
When to use frequency and relative frequency tables and bar and pie charts
1) objective is to describe a single set of data
2) data type is nominal or ordinal
