EC 203 Answers Tutorial 3 Graphical Descriptive Techniques

Ref.: Ch 3&4 (6th ed./7th ed.) and Ch.2&3 (5th ed.)

Questions:
1. According to a Survey by Housing Authority of Fiji in year 2012 around 50000
people aged 18 years or over moved to urban areas from rural location. Of these
people, 10 000 moved for housing reasons, 25 000 moved for employment reasons,
10 000 moved for family reasons and reminder moved for other reasons.

a) Construct a bar chart.

30000
25000 Reasons for Urbanisation in Fiji
25000
# of individuals

20000

15000
10000 10000
10000
5000
5000

0
Employment Housing Family Other
Reasons for Urbanisation
Source: Housing Authority of Fiji (2012)
b) Interpret your bar chart.
Majority of people moved for job reasons etc.
c) If the purpose is to emphasis number of people in each category which graph
would be best? Bar Chart
d) If the purpose is to compare proportion of people in each category which graph
would be best? Pie chart

2. The following bar chart is used to explain the capital investment in Australian
Private and public schools.

I. Interpret Graph.
• The graph presents the capital investment
expenditure per student in Australia by public
and private schools from 2002-2005. The line
graph on the same plot shows the gap between
investment in private schools and public schools.
• Important to note that there is huge
difference between capital investment
expenditure per student in Australia by public
and private schools. Private schools capital
investment expenditure are more than twice of
public schools capital expenditure per student.
• It is worth noting that gap between private
and public schools has been increasing from
2002-2005.
• Capital investment expenditure per
student in private schools are increasing while
for public schools there is no major change.

1
 II. Comment on appropriateness of this graph.
Graphical Excellence:
 labels i.e. exact figures,
 Well planned intervals as values up to 1560 and it stops at 1600 and equal
intervals, keys and clear presentations.
 Displaying both private and public schools on a same graph allows us to
compare between two types of schools.
 Use of expenditure per student is also appropriate rather than writing total;
capital investment expenditure as per student enable us to make
comparison.
 The graph has appropriate title which is quite persuasive to attract readers
“Class Divide”.
 Gridlines
 Keys provided
 Contrast colours for bar so suitable for black and white printing as well.
 Width of the bar is consistent.
 The selected graph is suitable for this type of data as data is time-series
data we can use bar chart or line graph.
Possible improvements:
- Include source of data, horizontal axis title, can use separate graph for gap.

III. Present same information using alternative graphical and tabular descriptive
method.

• Time series chart as date is time series (graph shown below).

• Use of table ok but not best.as we cannot figure out easily the trend in
table.
• Multiple bar chart for expenditure per student in private schools and
expenditure per student in public schools. And on separate chart a line graph form
capital investment gap, as separately we can more clearly observe changes in gap.
Or a separate bar can be created for capital investment gap.
• Use of component bar chart.
Bar total would be equal to expenditure per student in private schools, which
would be sum of expenditure per student in public schools and capital investment
gap. The total of bar should also make sense. We cannot have a component bar of
expenditure per student in private schools plus expenditure per student in public
schools.

2
 2000

1500

1000 917 956 1018

841

500
537 522 540 542
0
2002 2003 2004 2005
Public Gap

3. Major Types of Youth Activity in Samoa (Percentage of youths in 2012)

Male Female
Schooling 60% 60%
Working 18% 15%
Social work 12% 5%
Unemployed 8% 5%
Housework 2% 15%
i) List the variables in above Data.

- Gender and
- Type of Youth Activity.

ii) State measurement scale for the variables defined in part (i).
- Gender: Nominal Data
- Type of Youth Activity: Nominal Data
iii) Using data state whether following statements are True or False.
1) Majority of youth in Samoa are schooling. True
2) Number of male and female youth schooling are equal. False.
We have information about proportion of youth in each activity, but we
do not have information on numbers of youth each activity so we cannot
comment on number of youth in each activity.
3) Percentage of youth unemployed in Samoa are 13%. False.
We cannot sum proportion in two categories and state it as total. For
example, let’s say that there is 100 male and 60 female, So total youth is
160. Now 8% male unemployed that would be 8 males, while 5% female
unemployed this would be 3. Hence total number of youth unemployed is
11. And percentage of youth unemployed is (11/160)*100 = 6.88% which
is not equal to 13%.

4) Greater proportion of male are involved with social work compared to

proportion of female in social work. True

4. The numbers of defective items produced by a machine over the last 25 days are
as follows:

19 6 15 20 17 16 17 12 15
29 23 17 7 10 14 14 27 22
8 5 23 19 9 28 5

3
 Class Limits Frequency relative Cumulative Cumulative
0 up to 5 2 frequency
0.08 frequency
2 relative
0.08
5 up to 10* 5 0.2 7 frequency
0.28
10 up to 15 5 0.2 12 0.48
15 up to 20 7 0.28 19 0.76
20 up to 25 3 0.12 22 0.88
25 up to 30 3 0.12 25 1
Total 25 1
*Class”5 up to 10” contains observations up to including 10 but 5 is not
included as it is included in first class. The other classes are defined
similarly. This notion is used throughout the chapter. Note it can also be
written as 5 -10 etc.
a) Fill above class frequency distribution. Filled above.
b) Calculate relative frequency, Cumulative frequency, Cumulative relative
frequency for this data. Filled above.
c) Construct a frequency histogram for these data.
Histogram of # of defective itmes produced by machine
8 7

6 5 5
# of days

4 3 3
2
2

0
5 10 15 20 25 30 # of defective items

d) Comment on the histogram, its skewness and modality.

- Its unimodal most of the days between 15-20 items were defective.
- Not symmetric as f left and right side are not identical.
- Per day at least 5 items were defective. Etc.

e) Calculate the proportion of days when 15 or less defective items were

produced.
0.48
f) Calculate the proportion of days with more than 20 defective items
produced.
0.24
g) Calculate the proportion of days with between 10 and up to 25 defective
items produced.
0.60

4
 h) Construct an ogive. Discuss usage of this graph.

1
O G I VE 1.00

0.8 0.88

0.76
0.6

0.48
0.4

0.28
0.2

0.08
0 0
0 5 10 15 20 25 30
# OF DEFFECTIVE ITEMS PRODUCED

Interpretation: 48% of days less than 15 defective items were produced,

while 52% of days more than 15 defective items were produced. Etc. this
graph is useful to illustrate running total. Etc .
Note: below ogive graph illustrate cumulative relative frequency, we can
also use ogive to illustrate cumulative frequency, shape will be same
however horizontal axis would be cumulative frequency.

5.
100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%
140 145 150 155 160 165 170 175

Weight (kgs)

i) Name the graph shown above. Ogive

ii) State label for vertical axis. Cumulative relative frequency
iii) If weight above 145kg is regarded is overweight, state what percentages
are overweight. 90%
iv) State at least two required improvements in above graph to achieve
graphical excellence.
- Appropriate title
- correct axis labels.
- Vertical grids, labels
- starting horizontal axis from 140 is not a problem, as it is important to
have consistent intervals as above.

5
6. Application of cross-classification Table: A random sample of 5030 individuals
was taken by department of statistics in Fiji. Out of these 5030 individuals, 1624
were unemployed and rest employed. Out of these 5030 individuals, 1130 were
residing in rural areas and rest in urban areas. 465 individuals out of all individuals
were both unemployed and residing in rural areas.
a) State the variables in above Data.
1) Geographic location: (Rural/urban)
2) Employment status: (employed/unemployed)
b) Create a contingency table for information above let Geographical location
(Rural/Urban) in rows of table.
Employed Unemployed Total
Rural 665 465 1130
Urban 2741 1159 3900
Total 3406 1624 5030
c) What percentages of individuals are unemployed in rural areas? In which
location unemployment rate is high.
Rural Unemployment Rate = 465/5030*100 = 9.24%
Urban Unemployment Rate = 1159/5030*100 = 23.04%
In urban the unemployment rate is high.

d) What percentage of rural population is unemployed?

Rural Unemployment Rate = 465/1130*100 = 41.15%
e) What percentage of urban population is unemployed?
Urban Unemployment Rate = 1159/3900*100 = 29.72%
f) Where would you recommend government to allocate resources to reduce
unemployment?
The government shall prioritize rural area in reducing unemployment. due
to the fact that relatively a greater percentage of the rural population is
unemployed. more individuals are unemployed in urban, because urban
population are high, since both region are of unequal size decision based on
relative aspects need to be considered rather than based on absolute
numbers. In case, If government solve problem in urban more individuals
will migrated to urban contributing to problem of urban drift.

7. A professor of economics wants to study the relationship between annual income

(in $000s) and education (in years). A random sample of eight individuals is
taken and the results are shown below.

Education (x) 16 11 15 8 12 10 13 14
Income (y) 58 40 55 35 43 41 52 49

6
 Draw scatter plot and comment on relationship.

Scatter Diagram for Education and Income

70
income in 000's
60
50
40
30
20
10
annual inc
0
0 5 10 15 20
Years of Education

The scatter plot illustrate that there is positive linear relationship between years
of education (x) and annual income (y). Positive relationship, since both
variables are moving in same direction. Also it appears line pattern hence
relationship is linear. M Bhatt