Catching the big waves: SURFing at Statistics New Zealand

1
Catching the big waves SURFing at Statistics New
Zealand
NZAMT Conference 25 - 28 September

• Nathaniel Pihama and Deborah Brunning
• Statistics New Zealand

2
What you will see today

• The Statistics New Zealand website
• Case data
• Time series data
• SURF for Schools
• ..and some ideas on how to use them!

3
The Statistics New Zealand Website

• What does it offer?

4
(No Transcript)
5
(No Transcript)
6
Stats NZ Products and Services

• Schools Corner
• Full of resources based on the curriculum.
• Information releases Hot Off the Press
• Full of highlights, commentary, technical notes
  and tables!
• New Zealand in Profile
• Quick stats of New Zealand for 2007
• Analytical reports
• Contain in depth analysis, background and
  technical information
• Table Builder
• Customisable tables of released survey data

7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
Stats NZ Products and Services

• Schools Corner
• Full of resources based on the curriculum.
• Information releases Hot Off the Press
• Full of highlights, commentary, technical notes
  and tables!
• New Zealand in Profile
• Quick stats of New Zealand for 2007
• Analytical reports
• Contain in depth analysis, background and
  technical information
• Table Builder
• Customisable tables of released survey data

12
(No Transcript)
13
Stats NZ Products and Services

• Schools Corner
• Full of resources based on the curriculum.
• Information releases Hot Off the Press
• Full of highlights, commentary, technical notes
  and tables!
• New Zealand in Profile
• Quick stats of New Zealand for 2007
• Analytical reports
• Contain in depth analysis, background and
  technical information
• Table Builder
• Customisable tables of released survey data

14
(No Transcript)
15
Stats NZ Products and Services

• Schools Corner
• Full of resources based on the curriculum.
• Information releases Hot Off the Press
• Full of highlights, commentary, technical notes
  and tables!
• New Zealand in Profile
• Quick stats of New Zealand for 2007
• Analytical reports
• Contain in depth analysis, background and
  technical information
• Table Builder
• Customisable tables of released survey data

16
Battle for the greener city

• an example of using case data from Table Builder

17
The statistical investigation cycle(Wild and
Pfannkuch, 1999)

• Problem the statement of the research questions
  
• Plan planning the procedures used to carry out
  the study
• Data the data collection process
• Analysis the summaries and analyses of the data
  to answer the questions posed
• Conclusion the conclusions about what has been
  learned.

18
Battle for the greener cityan example of
using case data

• Comparing the traveling to work habits of two
  cities Hamilton and Wellington.
• Which city has the greener workers?
• Walking / Running / Cycling
• Public transport
• Carpooling?
• Working at home?

19
Battle for the greener citywhere to find the
data
 
 

• We want a data source that contains information
  about modes of travel to work for our two cities.
• Luckily, we have the 2006 Census of Population
  and Dwellings on Table Builder!

20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
Battle for the greener cityafter a little
clean up on excel..
36
Battle for the greener citywhat we have been
able to do

• Using table builder and excel, we have looked at
• Counts of peoples mode of traveling to work,
  broken down by city.
• Calculated proportions of green travelers for
  each city.
• Whether there appears to be a difference between
  Hamilton and Wellington.
• Where these differences may lie.

37
Unemployment in New Zealand

• an example of using time series data

38
Unemployment in New Zealandan example of time
series data

• How has unemployment changed over time?
• Do qualifications affect unemployment rate?

39
Unemployment in New Zealandwhere to find the
data

• We want a data source that contains information
  about unemployment rate over a long time period
  and has a breakdown by type of qualification.
• Table Builder has that information!

40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48
(No Transcript)
49
(No Transcript)
50
(No Transcript)
51
Unemployment in New Zealandafter a little clean
up on excel..
52
Unemployment in New Zealandwhat we have been
able to do

• Using table builder and excel, we have looked at
• Changes over time in the unemployment rate.
• Differences between qualification groups.

53
The First SURF a Synthetic Unit Record File for
Schools

54
Overview

• SURF???
• What is it?
• How and why did we make a SURF?
• What teachers and students can do with the SURF

55
What is a Unit Record File?

• Other names
• Data set
• Unit Record Data set
• Microdata

56
Confidentialised Unit Record Files

• Confidentiality methods include
• Categorical Data
• Global recoding
• Local recoding
• Numerical Data
• top/bottom coding,
• capping,
• rounding,

57
From CURF to SURF
58
What is the SURF?

• Data from 200 synthetic respondents.
• Target population is those aged 15-45 in paid
  employment.
• 7 variables

59
What does the SURF look like?
The first 7 of 200 complete unit records
60
SURF- the variables
61
How to start SURFing?

• The gender gap (Level 3 and 4)
• Do more females have higher qualifications than
  males?
• Is this different from how it was in the past?
• Am I average? (Level 4)
• What defines the average person?
• Under pressure? (Level 5)
• Are people who have never been married different
  from married people?
• Equal Pay! (Level 6)
• Are males and females paid equally?
• Money for nothing (Level 7)
• Investigating hours worked by employees in a
  company
• Should I do a degree? (Level 8)
• Investigation into whether getting a degree helps
  improve earning power
• http//www.stats.govt.nz/schools-corner

62
PPDAC cycle
Problem
Conclusion
Plan
Analysis
Data
63
Task -Money for Nothing

• A large company is concerned that it has too many
  employees who do not work a 40-hour week.
• You have been hired to investigate the working
  patterns of the employees.

64
Problem

• The company needs summary statistics for the
  hours worked by employees each week.
• The companys database has data on the hours
  worked by its employees

65
Plan

• Take a representative sample of 35 employees.
• Calculate appropriate sample statistics and draw
  appropriate statistical graphs.
• Write a conclusion. Is the companys concern
  valid?

66
Data

• SRS, systematic, or stratified on age group
  (10-year groups?)

67
SURF and CURF Analysis
68
Further Analysis- Hours by Gender
69
Conclusion for Money for Nothing

• According to the SURF analysis
• majority (gt75)of males work 40 or more hours
  per week
• At least 25 of females work 40 or more hours per
  week
• The lower quartile for males is higher than the
  median for females
• A high proportion of men do work a 40 hour week,
  but the company should consider the reasons why
  such a small proportion of women work a 40 hour
  week.

70
SURF vs CURF

• SURF
• majority (gt75) of males work 40 or more hours
  per week
• At least 25 of females work 40 or more hours per
  week
• CURF
• Majority (gt50) of males work 40 or more hours
  per week
• At least 25 of females work 42 hours or more per
  week

71
A look at some other pairs of variables - Income
by Hours
72
Related variables Hours by Marital Status
73
Related variables Hours by Age Group
74
How school friendly is SURF????

• SURF ? Excel spreadsheet
• Records are in random order
• First 30 records could be used for manual data
  analysis
• Use Excel How???????

75
Filter function
76
Pivot Tables
77
Pivot Tables
78
Box plots !!!!

• QUARTILE function
• SORT by Marital

79
Regression and Residuals

• Trend line in a scatter plot
• Good for quick visual check
• Provides equation R-sq
• But no residuals

• Plot the data (XY scatter)
• (tidy plot up)
• Add Trendline
• Chart menu gt Add trendline
• Options tab

80
Regression and Residuals

• Using Excel functions
• SLOPE(), INTERCEPT(), RSQ()
• Copy cell ref into formula bar

81
Regression and Residuals

• - Easy to create predicted values and residuals
• (can copy formula and use )

82
Regression and Residuals

• - Plot the residuals

83
What can the SURF do for us?
Privacy Security Confidentiality
Availability Utility Relevance
84
Resources

• More resources can be obtained from
• http//www.stats.govt.nz/schools-corner
• Or to register for your free copy of the SURF for
  Schools CD, please email us at
  education.services_at_stats.govt.nz