Title: GRAPHICAL DESCRIPTIVE STATISTICS FOR QUALITATIVE, TIME SERIES AND RELATIONAL DATA
1- GRAPHICAL DESCRIPTIVE STATISTICS FOR QUALITATIVE,
TIME SERIES AND RELATIONAL DATA
2Reasons To Collect Data
- Obtain Input to a Research Study
- Measure Performance
- Assist in Formulating Decision Alternatives
- Satisfy Curiosity
- Knowledge for the Sake of Knowledge
3Random Variables
- Random variables are phenomena or characteristics
that are not known in advance - Data are observations of random variables
4TYPES OF DATA
- Qualitative
- Result to a survey question is non-numeric
- Categorical Data
- Ordinal (Ranked) Data
- Quantitative
- Result to a survey question is a number
- Ratio Data Has an absolute 0 point, e.g. age
- Interval 0 is simply another number e.g.
degrees Fahrenheit
5Graphical TechniquesQualitative Data
6Plot of Frequency Distributions
- Objective is to develop a frequency distribution
table - Hand Count
- Excel Approach
- COUNTIF command
7Pie Charts
- Determine the relative frequency for each
category - Apportion sectors (wedges) of a 360 degree circle
proportionately
8Example
- Frequency Distribution of Origin of Car
Manufacturer -
Frequency American 16 Asian 20 European
4
Relative Frequency .40 .50 .10
9Pie Chart
10Bar Charts
- Bars show the frequency or relative frequency of
the observations - Consider the same example
11Frequency Bar Chart
12Relative Frequency Bar Chart
13EXCELPie Charts
- Example -- 66 people surveyed and asked the color
of their car -- Their choices - Red
- Blue
- Black
- White
- Other
- These are recorded in cells A2 to A67 of a
spreadsheet
14Determining Frequency DistributionsStep 1
Record Responses in a Column
15Step 2 Type Categories in another Column
16Step 3 Use COUNTIF to Determine Frequencies
17Creating a Pie Chart
18Creating a Pie Chart - 2
19Creating a Pie Chart - 3
20Creating a Pie Chart - 4
21Creating a Pie Chart - 5
22Completed Pie Chart
23Editing Options
- Enlarge
- Put Labels Inside Wedges
- Change Colors
- Add Text
- Etc.
24EXCELBar Charts
- What we call Bar Charts, Excel calls Column
Charts - The steps are similar
- If you already have a pie chart, you can convert
it to a bar chart as shown on the next slide
25Creating a Bar Chart From a Pie Chart
26Resulting Bar Chart
27Line Charts for Time Series Data
- Time series -- Values vs. time
- Dow Jones vs. Day
- Sales vs. Quarter
- Population vs. Year
- Typically depicted as line charts
- In Excel if you already have a bar chart you can
convert it to a line chart using the method on
the last slide - Otherwise we do the following
28Example -- Sales at Epencil.com
29(No Transcript)
30(No Transcript)
31Note In this dialogue box, there are many
editing features
32(No Transcript)
33Can now edit figure Resize Delete Gray
Background Delete Lines Etc.
34Scatter Diagrams -- Showing Relationships
Between Variables
- A scatter diagram shows the relationship between
two quantitative variables as a plot of a series
of points (observations) - (Grade vs. Study time)
- (Sales vs. Advertising )
- (Production vs. Resources)
- Types of relations that can be detected
- Linear (Positive or Negative)
- Least Squares Line -- Best line through points
- Nonlinear
- No Relation
35Example
- Sales of Pencils at Epencil vs. Advertising
spent during the week
36(No Transcript)
37(No Transcript)
38(No Transcript)
39In this dialogue box there are many editing
features
40(No Transcript)
41Can now edit figure Resize Delete Gray
Background Delete Lines Etc.
42Determining Trend
- We can have Excel put the best straight line or
other curve (parabola, etc.) through these points
so that we can easily observe trend -
432. Select Add Trendline from the Popup Menu
44(No Transcript)
45How to Lie With Statistics
- Graphs can be used to accurately portray data
- However, sometimes a graph can be distorted in
such a way as to skew the information it conveys - Examples
- Fatter or different shaped bars
- Stretched or condensed axes
- No indication of a break on the axis
- Unlabeled axes
46Review
- How to construct by hand and by Excel
- Pie Charts for Qualitative Data
- Bar Charts for Qualitative Data
- Line Charts for Time Series Data
- Scatter Diagrams to Show Relationships Between
Variables - How Graphs Can be Deceptive