Transcript Mousing with SPSS - ASSESS SPSS User Group

MOUSING
WITH
SPSS
Useful point and click
Frances Provan, Information Services, Edinburgh University
All at sea
ASSESS YORK 2007
Mousing with SPSS
2
There's a lot in SPSS
ASSESS YORK 2007
Mousing with SPSS
3
Today’s blubber covers

Things I like…
On the fly utilities
Wizards
Open secrets
Old SPSS favourites

All by mouse




ASSESS YORK 2007
Mousing with SPSS
4
SPSS Versions
Version comparison list from SPSS
http://www.spss.com/software_version/


Lists changes between versions
 and

new features.
Goes back to version 6
What’s new for latest version
http://www.spss.com/spss/whats_new.htm

ASSESS YORK 2007
Mousing with SPSS
5
Some slide shorthand

{version no. first appeared}


{14} introduced in version 14
Menu paths
Menu > Sub Menu > Sub menu
 e.g. File > Open > Data…
i.e. choose the File menu,
select Open from the File menu,
then select Data... from the Open submenu

ASSESS YORK 2007
Plus point something I like…
Mousing with SPSS
6
Did you know you could…





ASSESS YORK 2007
Define your own styles?
Create charts from tables?
Web pages from results?
Document your data files?
Get previews when building
chart and tables?
Mousing with SPSS
7
Nice newish graphs…

Dot plots (about time too!) {13}
 Graphs
 Then

ASSESS YORK 2007
choose Simple Dot
Population pyramids {13}
 Graphs

> Scatter/dot…
> Population Pyramid
Panels for ordinary graphs {13}
Mousing with SPSS
8
Dot plot
ASSESS YORK 2007
Mousing with SPSS
9
Population Pyramids
ASSESS YORK 2007
Mousing with SPSS
10
Population Pyramids: categories


ASSESS YORK 2007
Each category -a separate pane
3 level variable
Mousing with SPSS
11
Panels




Lovely idea…
1st in interactive plots
Now Rows and columns {14}
Or Multiple factors


ASSESS YORK 2007
Separate or nested
Great for comparisons
Mousing with SPSS
12
Dot plots: row panels
ASSESS YORK 2007
Mousing with SPSS
13
Dot plots: Column Panels

ASSESS YORK 2007
Panelled in columns
Mousing with SPSS
14
Dot plots: Row and Column Panels
ASSESS YORK 2007
Mousing with SPSS
15
Population Pyramids: Row Panels

ASSESS YORK 2007
Single Row factor
Mousing with SPSS
16
Your own style
Pivot tables & Table looks
 Interactive Charts & Chart looks
 Charts & chart templates
 Edit graph/chart/table to access

 i.e.

Double click on it
Change your style with
 Edit
> Options…
 Charts, Interactive or Pivot Tables tab
ASSESS YORK 2007
Mousing with SPSS
17
Pivot tables


Double click to edit
Change the look
 Format

Change rows, columns & layers
 Pivot

> Pivoting Trays
Keep dimension changes
 Pivot
ASSESS YORK 2007
> Tablelooks…
> Bookmarks
Mousing with SPSS
18
Tablelooks

Default
General
Happiness
Very Happy
Pretty Happy
Not Too Happy

Academic
Is Life Exciting or Dull
Exciting
Routine
Dull
195
98
2
45%
20%
5%
218
338
12
50%
68%
30%
21
61
26
5%
12%
65%
Is Life Exciting or Dull
Exciting
General
Happiness
Very Happy
Pretty Happy
Not Too Happy
ASSESS YORK 2007
Mousing with SPSS
Routine
Dull
195
98
2
45%
20%
5%
218
338
12
50%
68%
30%
21
61
26
5%
12%
65%
19
Chart from a table



Double click on the table
Right click, Create Graph > Bar
Be careful what is selected
 Total
lines look daft
 Use layers to be selective
ASSESS YORK 2007
Mousing with SPSS
20
Bar chart from table
Statistics : Count
Row
General Happiness Very Happy
General Happiness Pretty Happy
300
Values
General Happiness Not Too Happy
200
100
0
Is Lif e Exciting or Dull Dull
Is Lif e Exciting or Dull Exciting
Is Lif e Exciting or Dull Routine
Column
ASSESS YORK 2007
Mousing with SPSS
21
Interactive graphs {long time}






Graphs > Interactive > Bar
A lot of the ‘new’ graph features
already there
Panel variables
Chart looks
Exploratory data analysis
Mmm.. Leopard skin barcharts..

ASSESS YORK 2007
Tastefully tacky
Mousing with SPSS
22
Grrraphs - to the leopard skin




ASSESS YORK 2007
Double click to open
Select object
 right clicking on bar
 select all bars
Choose Fill button
 Pattern
 .bmp image
Or choose
 Format >
chart properties
 Filled Objects tab
Mousing with SPSS
23
Chartlooks & Chart templates


Available when editing
Interactive Graphs


Format > Chartlooks…
Chart builder
File > Save Chart Template
 File > Apply Chart Template


Use your own style by default

ASSESS YORK 2007
Edit > Options
Mousing with SPSS
24
Grrraph to Dante
ASSESS YORK 2007
Mousing with SPSS
25
Builders

Custom tables {12?}


Analyze > Tables > Custom Tables..
Chart builder {14}

Graphs > Chart Builder…
 Still
new
 Not sure I like them yet…
ASSESS YORK 2007
Mousing with SPSS
26
A Customized table

With academic table look…
Is Life Exciting or Dull
Exciting
General
Happiness
Very Happy
Pretty Happy
Not Too Happy
ASSESS YORK 2007
Mousing with SPSS
Routine
Dull
195
98
2
45%
20%
5%
218
338
12
50%
68%
30%
21
61
26
5%
12%
65%
27
A built stacked bar chart
ASSESS YORK 2007
Mousing with SPSS
28
Exporting Output


File > Export
Export as:

HTML - web pages {7}




Word {11.5}






ASSESS YORK 2007
Anyone can read it
‘clean’ HTML
separate picture files
All in one file
Big files.
Powerpoint (see later) {13}
Excel (tables only) {11.5}
PDF {15}
No copy and paste
Mousing with SPSS
29
Controlled output export

Can specify:
Amount of output:
 All output (includes hidden stuff)
 All Visible output
 Only selected objects
 Charts, documents or both
 Image types for charts
 Output types, as above


Use in conjunction with OMS {14}


ASSESS YORK 2007
Utilities > OMS Control Panel
Utilities > OMS Identifiers
Mousing with SPSS
30
Export to MS Powerpoint {13}

Not everything translates, but
you get:
 Pivot
tables
 charts {14}
 Maps
 Trees

ASSESS YORK 2007
Used it for some slides…
Mousing with SPSS
31
Table straight to powerpoint
Case Processing Summary
Cases
Valid
N
General Happiness *
Is Life Exciting or Dull
* Region of the
United States
ASSESS YORK 2007
Missing
Percent
971
64.0%
N
546
Mousing with SPSS
Percent
36.0%
Total
N
1517
Percent
100.0%
32
Crosstabs output for:


Analyze > Descriptive Statistics >
Crosstabs
Variables
Row is happy
 Column is life
 Layer variable is region




ASSESS YORK 2007
Count & column statistics from Cells
Chi-squared tests from Statistics
Ticked Clustered bar chart
Mousing with SPSS
33
Case Processing Summary
Cases
Valid
N
General Happiness *
Is Life Exciting or Dull
* Region of the
United States
971
Missing
Percent
64.0%
N
546
Percent
36.0%
Total
N
1517
Percent
100.0%
Case processing
ASSESS YORK 2007
Mousing with SPSS
34
General Happiness * Is Life Exciting or Dull * Region of the United States Crosstabulation
Region of the
United States
North East
Is Life Exciting or Dull
General
Happiness
Very Happy
Count
% within Is Life
Exciting or Dull
Pretty Happy
Routine
40
43.0%
121
17.9%
5.3%
28.2%
99
159
6
264
53.2%
71.0%
31.6%
61.5%
7
25
12
44
3.8%
11.2%
63.2%
10.3%
186
224
19
429
100.0%
100.0%
100.0%
100.0%
52
33
1
86
48.6%
22.8%
8.3%
32.6%
51
89
3
143
47.7%
61.4%
25.0%
54.2%
4
23
8
35
3.7%
15.9%
66.7%
13.3%
107
145
12
264
100.0%
100.0%
100.0%
100.0%
63
25
0
88
44.7%
19.5%
.0%
31.7%
68
90
3
161
48.2%
70.3%
33.3%
57.9%
10
13
6
29
7.1%
10.2%
66.7%
10.4%
141
128
9
278
100.0%
100.0%
100.0%
100.0%
Count
% within Is Life
Exciting or Dull
Total
Count
% within Is Life
Exciting or Dull
South East
General
Happiness
Very Happy
Count
% within Is Life
Exciting or Dull
Pretty Happy
Count
Crosstabs table
% within Is Life
Exciting or Dull
Not Too Happy
Count
% within Is Life
Exciting or Dull
Total
Count
% within Is Life
Exciting or Dull
West
General
Happiness
Very Happy
Count
% within Is Life
Exciting or Dull
Pretty Happy
Count
% within Is Life
Exciting or Dull
Not Too Happy
Count
% within Is Life
Exciting or Dull
ASSESS YORK 2007
Total
Count
Mousing
withIs Life
SPSS
% within
Exciting or Dull
Total
Dull
1
Count
% within Is Life
Exciting or Dull
Not Too Happy
Exciting
80
35
Chi-Square Tests
Region of the
United States
North East
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear
Association
N of Valid Cases
South East
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear
Association
N of Valid Cases
West
Pearson Chi-Square
Value
94.279(
a)
70.060
59.027
Asymp. Sig.
(2-sided)
df
4
.000
4
.000
1
.000
4
.000
4
.000
1
.000
4
.000
429
52.875(
b)
43.738
37.470
264
51.772(
c)
Likelihood Ratio
39.841
4
.000
Linear-by-Linear
29.150
1
.000
Association
N of Valid Cases
278
a 1 cells (11.1%) have expected count less than 5. The minimum expected count
is 1.95.
b 2 cells (22.2%) have expected count less than 5. The minimum expected count
is 1.59.
c 2 cells (22.2%) have expected count less than 5. The minimum expected count
is .94.
Chi-squared tests
ASSESS YORK 2007
Mousing with SPSS
36
Barchart 1
ASSESS YORK 2007
Mousing with SPSS
37
Barchart 2
ASSESS YORK 2007
Mousing with SPSS
38
Barchart 3
ASSESS YORK 2007
Mousing with SPSS
39
On the fly

Visual bander {12}


Define Variable properties


Data > Copy Data Properties
Automatic recode {forever}

ASSESS YORK 2007
Data > Define Variable Properties
Copy data properties


Transform > Visual Bander
Transform > Automatic Recode
Mousing with SPSS
40
Visual Bander {12}


Interactive tool to categorise data
Cut points
Manually defined
 Equal ranges
 Equal counts
 Using mean and standard deviations



ASSESS YORK 2007
Labelling, either automatic or manual
Love combined recode & labelling
Mousing with SPSS
41
Define Variable Properties {?}


Easier than Variable View
Use it to:
 Type
labels
 View whole definition
 Copy definitions to and from
other variables
ASSESS YORK 2007
Mousing with SPSS
42
Automatic Recode {forever}





ASSESS YORK 2007
String to numeric
Large numeric codes
Doesn’t miss out values
Sorts out messy codes
Keeps coding to use again
{recent}
Mousing with SPSS
43
Data file stuff

Showing data file information
Display data file information >
external file
 Display data file information >
Working file


Data file comments


ASSESS YORK 2007
Utilities > Data File Comments
Multiple open datasets {14}
Mousing with SPSS
44
File information
File Information
Source
E:\Program Files\SPSS14\1991
U.S. General Social Survey.sav
Type
SPSS Data File
Creation Date
16-SEP-2002 11:12:10
Label
None
File Contents
Data Type
Case
N of Lines of Documents
Data
Information
ASSESS YORK 2007
391
Variable Sets
Yes
Trends Date Information
None
Multiple Response
Definitions
Yes
Data Entry for Windows
Information
None
TextSmart Information
None
Clementine Information
None
N of Cases
1517
N of Defined Variable
Elements
43
N of Named Variables
43
Weight Variable
None
Compressed
Yes
Mousing with SPSS
45
Variable Information
Variable Information
Name
sex
race
Position
Label
Format
Column Width
Alignment
Missing Values
1
Respondent's Sex
Nominal
F1
8
Right
2
Race of Respondent
Nominal
F1
8
Right
3
Region of the United
States
Nominal
F8.2
8
Right
4
General Happiness
Ordinal
F1
8
Right
0, 8, 9
5
Is Life Exciting or Dull
Ordinal
F1
8
Right
0, 8, 9
6
Number of Brothers and
Sisters
Scale
F2
8
Right
98, 99
7
Number of Children
Ordinal
F1
8
Right
9
8
Age of Respondent
Scale
F2
8
Right
0, 98, 99
9
Highest Year of School
Completed
Scale
F2
8
Right
97, 98, 99
10
Highest Year School
Completed, Father
Scale
F2
8
Right
97, 98, 99
11
Highest Year School
Completed, Mother
Scale
F2
8
Right
97, 98, 99
12
Highest Year School
Completed, Spouse
Scale
F2
8
Right
97, 98, 99
13
R's Occupational Prestige
Score (1980)
Scale
F2
8
Right
0
14
Occupational Category
Ordinal
F8.2
8
Right
15
R's Federal Income Tax
Ordinal
F1
8
Right
0, 8, 9
F1
8
Right
0, 8, 9
region
happy
Measurem
ent Level
life
sibs
childs
age
educ
paeduc
maeduc
speduc
prestg80
occcat80
tax
ASSESS YORK 2007
usintl
Mousing with SPSS
16
Take Active Part in World
Affairs
Ordinal
46
Value Labels
Value Labels
Value
sex
race
region
happy
life
ASSESS YORK 2007
Label
1
Male
2
Female
1
White
2
Black
3
Other
1.00
North East
2.00
South East
3.00
West
0(a)
NAP
1
Very Happy
2
Pretty Happy
3
Not Too Happy
8(a)
DK
9(a)
NA
0(a)
NAP
1
Exciting
2
Routine
3
DullSPSS
Mousing
with
8(a)
DK
47
Data File Comments



Just add text to dialog box
Time stamped
Use comments to:
Describe where data has come from
 Keep codebook with the data
 Document changes to data file




ASSESS YORK 2007
Can print to output
Documents command is useful
Remains with your SPSS data file
Mousing with SPSS
48
Comments box
ASSESS YORK 2007
Mousing with SPSS
49
The wizards

Date and time {13}


Restructure data {11.5}


Transform > Date/Time
Data > Restructure…
ODBC & read text
File > Open Database
 File > Read Text Data..


Sample Wizard

ASSESS YORK 2007
Analyze > Complex Samples
Mousing with SPSS
50
Date/Time wizard {13}



ASSESS YORK 2007
Date format very useful
Do you know how difficult it
used to be to calculate age from
date of birth
Loads of things you could only
do with syntax before.
Mousing with SPSS
51
Restructure wizard {11.5}





ASSESS YORK 2007
‘long' data files into 'wide' files
'wide' data files into 'long' files
cases become variables,
variables become cases
Indexing
Great for repeated records
Mousing with SPSS
52
Identify Duplicate Cases {12}



Data > Identify Duplicate Cases…
Filter out duplicates
Create indicator to use elsewhere
 E.g. Data > Select Cases…
to delete duplicates


Creates indexes
One stop shop

ASSESS YORK 2007
easier than sort cases & aggregate
Mousing with SPSS
53
Identify duplicate cases
Indicator of each last matching case as Primary
Valid
ASSESS YORK 2007
Duplicate Cas e
Primary Cas e
Total
Frequency
384
102
486
Percent
79.0
21.0
100.0
Mousing with SPSS
Valid Percent
79.0
21.0
100.0
Cumulative
Percent
79.0
100.0
54
For the enquiring mind

Online help

Help > Topics or Tutorial or Case Studies

Help button on every dialog box
 Help
about the procedure
 Details on choices

Right click (or Mac, Control-click)

To see what options you have
 Context

ASSESS YORK 2007
sensitive menus
To get a bit of background
Mousing with SPSS
55
And so much more…
$casenum





that order
$casenum
$casenum
$casenum
System variable - Current case order, Copy to record current order, Sort by copy to return to that order
Use with Transform > Compute

Merge/Match Files

Count
Merging two data files, IN for case source, BY matches,
Data > Merge Files > Add Cases or Add Variables




Use with Transform > Compute
Counts across range of variables, Good for Multiple Response, Also non response (missing values)
Transform > Compute

System
- Current
casecase
order,order,
Copy Copy
to record
current
Systemvariable
variable
- Current
to record
Frequencies

 Files
 Merge/Match
current
order,
Sort
order,
byIN
Sort
copy
by
tosource,
copy
return
to
return
that
order
to thatcase
order
Merging
two data
files,
forSort
case
BY
matches,

System
variable
-toUser
Current
order,
current
order,
by
copy
to&return
that order

Lists
absolutely
everything,
Systemto
missing
values,

> Merge
Files
> Add Cases >
or Add
Variables
DataUse
with
Transform
Compute
Copy to record current order, Sort by
Merge/Match
Files toforthat
copy
to

Merging
two return
data files, IN
case order
source, BY matches,
> Compute
Transform
Merging
two
data
files,
IN
for
case
source, BY matches,
 Frequencies
Aggregate

Data
> Merge two
Files >data
Add Cases
or Add

Merging
files,
INVariables
for case

Use
with
Transform
>
Compute

Data
>
Merge
Files
>
Add
Cases
or
Add
Variables
Lists
absolutely
everything,
User
&
System
missing
values,
Each
separate value
Count
source,
BY
matches,

Count

Checking
duplicates
(Superceded
Count
Analyze
>Descriptive
> Frequencies
Merge/Match
Files Statistics

Counts across range of variables, Good for Multiple
Aggregate

Counts
across
range
ofIN
variables,
Good
forBY
Multiple
Response,
Merging
two
data
files,
forLAST
case
source,
matches,
Response,
Also
non
response
(missing
values)
now),
FIRST
and
within
preChecking
duplicates
(Superceded
now),
FIRST
and
LAST
within
pre-sorted
groups,
Data
>
Merge
Files
>
Add
Cases
or

Counts
across
range
of
variables,
Also
non
response
(missing
values)

Data
> Merge
Files
> Add Cases or Add Variables
variables
(command
only)
string
Transform
>
Compute
sorted
string
Condenses
Transform
>groups,
Compute
Add
Variables
data
by
any
variable, Recently
addedvariables
{14}, Automatic Also
matching in,
New
Good
for
Multiple
Response,
non
Count
Frequencies
datasets in SPSS session
(command
only)
Count

across range
of variables,
for missing
Multiple values,
Response,
response
(missing
values)
DataCounts
Lists
absolutely
everything,
User
&Good
System
> Aggregate..

Frequencies
Also non
response
(missing
values)

Counts
across
range
of variables,
Good for Multiple Response,
Each
separate
value
Condenses
data
by
any
variable,
Lag
Also
non
response
(missing
values)

Transform
>
Compute
Transform
> Compute

Lists
absolutely
everything,

Analyze
>Descriptive
Statistics
> FrequenciesUser &

Transform
> Compute
Frequencies
Recently
added
{14},
Automatic
System
missing
values,
Each
separate
Aggregate
gets
next
in
sequence,
Great
forand
Frequencies


Lists
absolutely
everything,
User
&now),
System
missing
values,
Checking
duplicates
(Superceded
FIRST
LAST
matching
in,
New
datasets
in
SPSS
value
Each
separate
value
selecting
duplicates,

Lists
Uservariables
& System
missing values,
withinabsolutely
pre-sortedeverything,
groups, string
(command
only)
session
Each
separate
value


Analyze
>Descriptive
Statistics
>
Frequencies
Condenses
data
by any variable, Recently
added {14},

Analyze
>Descriptive
Statistics
>
Aggregate
with
Data
>Statistics
Sort
Cases
or
Use
Analyze
>Descriptive
>
Frequencies
Automatic
matching
in,
New
datasets
in SPSS
session
 Data
>duplicates
Aggregate..
Frequencies
Aggregate
Transform
> Compute
Data > Aggregate..

Checking
(Superceded now), FIRST and LAST
Lag


System variable - Current case order, Copy to record current order, Sort by copy to return to
$casenum







Count


Each
separate
value > Compute
Use with
Transform
Merge/Match
across Files
range
of variables, Good
for Multiple
Also non response (missing
Counts
Analyze
>Descriptive
Statistics
> Response,
Frequencies
Merge/Match
Files
values)






















ASSESS YORK 2007

Checking
within
pre-sorted
duplicates
groups,
(Superceded
string variables
now), FIRST
(command
and LAST
only)
Aggregate
Lag








gets
next
in
sequence,
Great
forvariables
selecting
duplicates,
within
pre-sorted
groups,
string
(command
only)
Condenses
data
by
any
variable,
Recently
added
{14},
Checking
duplicates
(Superceded
now), FIRST
and
LAST
gets next
in
sequence,
Great
for selecting
duplicates,
Mousing
with
SPSS
Use
with
Data
> Sort
Cases
or
Transform
>added
Compute
Condenses
Automatic
matching
data
by
any
in,
New
variable,
datasets
Recently
in SPSS
session
{14},
within
pre-sorted
groups,
string
variables
(command
only)
Use with
Data
> Sort
Cases
or
Transform
> Compute
Automatic
matching
in,
New
datasets
in
SPSS
session
Data > Aggregate..
Condenses
data by any variable, Recently added {14},
56
Those were my
Mousing
Tips
Fallen asleep?
ASSESS YORK 2007
Mousing with SPSS
58
Useful urls

ASSESS web site
http://www.spssusers.co.uk/

SPSS web site:
http://www.spss.com/


help system, on-line manuals
SPSS mailing list,
http://www.listserv.uga.edu/archives/
spssx-l.html
ASSESS YORK 2007
Mousing with SPSS
59
ASSESS YORK 2007
Mousing with SPSS
60
ASSESS YORK 2007
Mousing with SPSS
61