sadc course in statistics graphical summaries for quantitative data module i3: sessions 2 and 3
TRANSCRIPT
SADC Course in Statistics
Graphical summaries for quantitative data
Module I3: Sessions 2 and 3
2To put your footer here go to View > Header and Footer
Learning ObjectivesStudents should be able to:
• Present data in a histogram – and interpret data when shown a histogram
• Present data in a boxplot – and interpret data in a boxplot
• Recognise advantages and limitations – of each method of presentation
• Explain what is gained and lost from data summary• Interpret graphical summaries to answer questions
– concerning proportions, – extremes, – medians – and quartiles
• Resolve simple problems with graphical displays– when real data (not text-book examples) are used
3To put your footer here go to View > Header and Footer
Session Overview
Activity 1: Presentation to introduce the sessions
Activity 2: Demonstration on histograms in Excel
Activity 3: Practical covering the same ideas– CAST Chapter 2.1– Histograms– Population pyramids
Activity 4: Practical on boxplots and percentage points– CAST Chapter 2.2– Boxplots in Excel
Activity 5: Presentation continued
4To put your footer here go to View > Header and Footer
Activity 1- Case studies• Case studies used in these sessions are:
– Rice Survey– Zambia Rainfall Data – The Swaziland Crop Cutting Survey
• The were introduced in Modules B1 and B2 – so for many this will be a reminder.
• The rice survey and the Swaziland Crop Cutting– were already used in Session 1 of this Module.
5To put your footer here go to View > Header and Footer
Rice surveyUsed repeatedly
in slightly different forms
In CAST as shown here
In demonstration
And in Excel
Qualitative and quantitative variables to analyse
6To put your footer here go to View > Header and Footer
Zambia Rainfall data
• Farmers are migrating from Southern Zambia– citing climate change as the reason– they can no longer grow the crops as before
• A local NGO – acknowledges climate change in general, – but believes improved farming practices is more
important– wherever farmers locate
• They question the evidence– for climate change in the pattern of rainfall– as it affects farming strategy
7To put your footer here go to View > Header and Footer
Data analysed in these sessions
Total rainfall from 1 Jan to 31 March – called SeasonTot
Number of rain days in the same period - Scount
8To put your footer here go to View > Header and Footer
Original daily values are also available
Not needed for analysis here, but used for checking
Also useful if further questions posed
9To put your footer here go to View > Header and Footer
Zambia rainfall continued:
• Here we use the annual data – on the final worksheet– preparing data for analysis is done in Module I2
• We have access to the raw data – for checking purposes – and in case other variables are needed on a
later occasion
10To put your footer here go to View > Header and Footer
Swaziland Crop Cutting Survey
• Annual survey– Of agricultural holdings– And areas planted– Then yields from a crop-cutting exercise
• Data from 2005 made available
11To put your footer here go to View > Header and Footer
Person-level data
Ages will be analysed in these sessions
12To put your footer here go to View > Header and Footer
Yield data
The dry weights will be analysed
Analysis overall and just for maize
The zero yields cause a slight problem
13To put your footer here go to View > Header and Footer
Activity 2 – demonstration of histograms
• Before the practical (which is activity 3)• Follow the demonstration• On histograms and population pyramids in Excel
• It shows• Use of Excel• But keeping control – you must remain in charge• And not be limited by the software
• So it shows how to resolve problems• As well as how to use Excel
14To put your footer here go to View > Header and Footer
Activity 3 – practical with CAST and Excel
• Using CAST to understand histograms
• Then use Excel to construct them• Being observant – as was shown in the demonstration• And keeping control
• Remember the aims• Are more to understand statistics• Rather than to practice with Excel• Excel is just the tool
15To put your footer here go to View > Header and Footer
An example with CAST
Practice with small data sets – as shown here and also with large data sets
16To put your footer here go to View > Header and Footer
Population pyramid in CAST
How close to this display can you get with Excel?
Interpret this display
17To put your footer here go to View > Header and Footer
Activity 4: Boxplots and data summary
• Now do the demonstration
• And these two practical exercises
• Then return to the remaining slides
• For a discussion of the key points
18To put your footer here go to View > Header and Footer
Topics for the class discussion
• What was interesting?
• What did you discover
• What was difficult?
• What needs further discussion?
19To put your footer here go to View > Header and Footer
Boxplots and histograms – CAST page 2.2.3
Are you clear how boxplots and histograms relate?
20To put your footer here go to View > Header and Footer
Boxplots show outliers – CAST page 2.2.4
And this makes them very useful for data exploration as well as summary
21To put your footer here go to View > Header and Footer
You can calculate percentage points
• The formula
r * (n + 1)/100
• for the r’th % point
• Should now hold no fears
• For example– when there are 11 observations– and you want the median– use 50 * (11+1)/100 = 6– the median is the 6th highest in the sorted data
• If necessary look again at practical 2
22To put your footer here go to View > Header and Footer
Practical problems with real data
• This always happens
• You saw a problem – with the rainfall data– and with the crop yields
• The solution always involves being observant
• You can follow guidelines for a good analysis
• But not always simply obey rules
• Become a data detective instead!
23To put your footer here go to View > Header and Footer
Learning ObjectivesAre you now able to:• Present data in a histogram
– and interpret data when shown a histogram
• Present data in a boxplot – and interpret data in a boxplot
• Recognise advantages and limitations – of each method of presentation
• Explain what is gained and lost from data summary• Interpret graphical summaries to answer questions
– concerning proportions, – extremes, – medians – and quartiles
• Resolve simple problems with graphical displays– when real data (not text-book examples) are used
24To put your footer here go to View > Header and Footer
These sessions were largely on graphical summaries
The next sessions consider numerical summaries of the data