Surviving SPSS in COMM - George Mason...
Transcript of Surviving SPSS in COMM - George Mason...
-
Surviving SPSS
http://dataservices.gmu.edu/workshops/spss
http://dataservices.gmu.edu/software/spss
Debby KermerGeorge Mason University LibrariesData Services Research Consultant
-
Mason Data Services
http://dataservices.gmu.edu
Wendy MannJoy Suh
Debby Kermer
Geospatial Analysis and
Mapping
Finding & Archiving Data,Data Management Plans
Qualitative & Quantiative Methods, Analysis, & Software
-
http://dataservices.gmu.edu/software/spss
-
Download the Dataset
Go to: http://dataservices.gmu.edu/workshops/spss/
Find SPSS I then click Datafile
Complete the form and read the user agreement
Click Submit and Download Data
Save the file where you can find it
Open the file (by default “PST 11‐05c.sav”)Double‐Click or File | Open… (if SPSS is open)
-
Download the Dataset
Go to: http://dataservices.gmu.edu/files/pew.sav
The file will download
Double‐Click or File | Open… (if SPSS is open)
-
Open the Dataset
Click the Workshop Files Icon:Or, go to the "W" drive.
Choose SPSS Workshop Files
Open “PST 11‐05c.sav”Double‐Click or File | Open… (if SPSS is open)
-
Data View
-
Data ViewLook at the
Data Window
Confirm you are inData View
Where are Variables?
Where are Cases?
-
Navigating Data View
Variables are in columnsFind variable q2Make q2 wider
Cases are in rowsDetermine the
# of cases(scroll down)
-
Value = What is stored and analyzed; the "actual" value
Values & Labels
Label = What is displayed; the characteristic or group
Press or use View | Value Labels
Value → Label1 → “English”2 → “Spanish”
-
Switching Views
#1: Click on the Tabsor
#2: Double-Click on the Blue Header area• Variable View• Data View
From Data View, double‐click q2 to switch views
-
Variable View
-
Variables & Values
Name – "Nickname"Label – Variable LabelValues – Value Labels
-
Variable TypesMeasure
Level of MeasurementAvoids incorrect analysis
Scale (Interval/Ratio)
Ordinal
Nominal
String / Nominal
TypeStorage & Analysis Type
Numeric = For numeric or categorical (label the values)
String = For essay text; only keeps text up to the Width
-
Missing Values
System Missing. (a period)
User Missing7 = “N/A”8 = “Don’t Know”9 = “Refused”
-
Q.2
Name: q2Type: NumericLabel: “Generally, how would you say things are these days in your life -- would you say that you are very happy, pretty happy, or not too happy?”
Width: 1Measure: OrdinalValues:
1 = “Very happy”2 = “Pretty happy”3 = “Not too happy”9 = “Don’t know/Refused”
-
User Missing Values
Add 9 as a User Missing value for q2
-
Data View Display
Decimals – # of decimals shown (in output, too)
Column – Width of the column
Align – Justification of values in column
only affects the look in Data View
-
Analyze Menu
-
Descriptive Statistics
How happy are people with their life?
Analyze | Descriptive Statistics | Frequencies
-
Finding Variables
Find q2
Type nameVERY FAST
q2
Right Clickto re‐sort
or
-
Choosing Variables
Click thearrow
Move q2 to the box then click OK
or
Double‐Click the name
-
Output Window / Viewer
Outline
-
Output TablesVariableName Variable Label
Value Value Label
-
Edit | Options: General
-
Edit | Options: Output
-
SPSS Syntax
-
Syntax
Syntax
-
Keeping Syntax
A. Copy from Output OR
Double-Click to allow copying. Paste into Syntax Window.
B. Press Paste in Dialog
Pasting does not “Run” the command. So, run it yourself.
-
Choosing Multiple Variables
Analyze | Descriptive Statistics | Frequencies…Select q3a to q3j Then, Click Paste
Hold Ctrl click on each
Press Shiftclick on last
-
"Running" Syntax
1. Highlight what to "Run" 2. Press the "Play" Button
-
Q.3a‐j
-
Reproducible Variable Properties
-
Defining Variable Properties
-
Editing Syntax
Copy
Paste
Then Run!
Original
-
Goal
Is there a relationship between happiness and having pets?
Happiness q2
1 Happy = 1 & 2 0 Not Happy = 3
Have pets q75
1 Has Pets = 10 No Pets = 2+
Use Edit | Go to Variable…
-
Windows
-
WindowsSyntax.sps
Data.sav
Output.spv
Unsaved ActiveInactive
-
Frequencies & More
For the Q3‐Series, get: Mean Median SD Min Max Range
Don't forget to choose Paste, then Run.
Analyze | Descriptive Statistics | Frequencies
-
Tip
Double-click the table to open it in a separate windowClick Pivot | Transpose Rows and Columns
-
Finishing Touches
Make the column wider (same as Excel)
-
SPSS Syntax
Starts with a COMMAND, ends with a period.
-
COMMENTCOMMENT The following is for SPSS with Syntax.
OR
* The following Is for SPSS with Syntax.
**********************.**** Leave a Comment ***.
-
Transform Menu
Working with Variables
-
Compute
Goal: Make an index of topical happiness
Transform | Compute Variable…
New Variable Name:happy_score
Summary Function: MEAN
-
Avoiding Frequency Tables
-
Goal: Recode q76_1 into pets
Step 1: Descriptive Statistics on q76_1
Step 2: Plan new values
Recode
Other Pet
Dog
No Pet
1
2
3
-
Goal: Recode q76_1 into pets
Transform | Recode into Different Variables…
Recode
-
Labeling
-
Labeling with Syntax
Data| Define Variable Properties…
-
Variable Properties
Other Pet
Dog
No Pet
1
2
3
-
Data Menu
Working with Cases
-
Weighting# of people a case represents, accounts for the over‐ or under‐representation of a group
In “Readme PST 11‐05.txt”The WEIGHT variable is the weighting variable
Data | Weight Cases…Look for or Select:
Weight cases by: weight
-
Selecting [a Subset of] Cases
Goal: Do analysis on only those who have a pet.
Data | Select Cases…
ignored cases 0 or missing value
included casesevery other value
Filtering
-
Splitting a File for Comparison
Goal: Compare happiness among those with pets
Data | Split File…
Choose Compare groupsMove gender to box
Turn off Grouping
Add group as a factor for each analysis
Do analysis for one group, then for the other
-
Status Bar
Shows status of:
Filter
Split
andWeighting
-
Return to Normal
Turn off Grouping Turn off Filtering
Data | Select Cases…Data | Split File…
-
Finishing Up
-
DataSets
-
If there are ERRORSTry to Fix the 1st Error only
That is often the cause of the following errors
Actually read the error messageThey can be long, but are often informativeIt may offer suggestions
Use the message to identify the exact spotLook for the variable name or text given
-
More Troubleshootng
Are "Transformations Pending"? Changes to data require an EXECUTE.
Does the command end with a period?
Are there blank lines in the middle of a command?Blank lines act as periods
Are your variable names correct?
Do the variables have the values you expect?
-
Save Files
File | Save As…
Data File & Output File &Syntax File
To Exit, click “Yes”
-
Saving Output
Click and press theDelete Key
-
Questions
http://dataservices.gmu.edu
http://dataservices.gmu.edu/software/spss/
c b n a© 2015 by Debby Kermer,
Mason Library Data ServicesThis work is licensed under the c Attribution-NonCommercial-ShareAlike International License: http://creativecommons.org/licenses/by-nc-sa/4.0/