Displaying Two Categorical Variables: Clustered Bar Charts and Stacked Bars, Slides of Economic statistics

Download Displaying Two Categorical Variables: Clustered Bar Charts and Stacked Bars and more Slides Economic statistics in PDF only on Docsity!

Graphical presentation: displaying

two categorical variables

Displaying two categorical variables

At the end of this session you will

• Learn how to produce clustered bar charts
• Discuss how to choose an appropriate

variable as a clustering variable.

• Learn about the use of stacked bars and

when they should be used.

Clustered bar charts

Household actively operates land for crop production

Percent Yes No N/A

80

60

40

20

0

Region

Missing Urban Rural

Is there anything odd with this graph?

Can you explain how the % are calculated?

Missing values

value labels croppro

1 "Yes"

2 "No"

3 "N/A".

recode region

(sysmis = 3)

(else = copy).

In order to find out exactly what is happening we can give a

‘value’ to the system missing in both categories.

variable labels region

'Region'.

value labels region

1 "Urban"

2 "Rural"

3 "Missing".

Household actively operates land for crop production

Percent Yes No N/A

80

60

40

20

0

Region

Missing Urban Rural

32

36 32 25

71

19

59

22

Household actively operates land for crop production

Percent Yes No N/A

80

60

40

20

0

Region

Urban Rural Missing

19

59

22

32 36 32 25

71

Before dealing with missing After dealing with missing

Would it be better to change the cluster variable? Household actively operates land

for crop production

Region

Urban Rural Missing

Percent

Yes

No

N/A

GRAPH

/BAR(GROUPED)=PCT BY region BY croppro /MISSING=REPORT.