2. Dr.ChristinePereira
» Purpose and use of SPSS
» Open SPSS
» What is a Variable?
» Defining variables in SPSS
» Entering Data
» Open and save data files
» Import data from Excel
» Handling missing data
ASK at Brunel (2014) 2
3. Dr.ChristinePereira
» SPSS is a Statistical Software Package
» SPSS is a tool
It only does what it’s ‘told’ to do.
It does not think for you
It is not a black box
» You need to know the correct statistics for your
research BEFORE using SPSS.
» If you understand the statistics, then you are ready to
do analysis in SPSS.
ASK at Brunel (2014) 3
4. Click on the Windows
Start Icon
Type SPSS 20 into the
search box
Select SPSS 20
from the list
1
2
3
Dr.ChristinePereira
ASK at Brunel (2014) 4
13. Dr.ChristinePereira
ASK at Brunel (2014) 15
Indicate your level of agreement with this statement:
“Most days, my commute causes me to feel stressed when I
arrive at university”.
14. Dr.ChristinePereira
» Random sample of 200 residents of Uxbridge.
» Asked respondents’ their view on stem cell research
using a 3 pt. Likert Scale (Disagree, No Opinion, Agree).
» Asked respondents’ if they believed global warming
was an important issue using a 5 pt. Likert Scale
(Strongly Disagree to Strongly Agree).
ASK at Brunel (2014) 16
Questionnaire Responses
15. Dr.ChristinePereira
» Random sample of 200 residents of Uxbridge.
» Asked respondents’ their view on stem cell research
using a 3 pt. Likert Scale (Disagree, No Opinion, Agree).
» Asked respondents’ if they believed global warming
was an important issue using a 5 pt. Likert Scale
(Strongly Disagree to Strongly Agree).
ASK at Brunel (2014) 17
Questionnaire Responses
19. Dr.ChristinePereira
» 30 participants were used to investigate the effect of
caffeine on their ability to sleep.
» The 30 participants were randomly assigned to one of 2
conditions: No caffeine (control) or one dose of caffeine
every 3 hours from 9am-6pm.
» The study measured participants ability to sleep by
taking the average number of hours slept per night
over a 2 week period.
ASK at Brunel (2014) 21
Caffeine and Sleep
20. Dr.ChristinePereira
» 30 participants were used to investigate the effect of
caffeine on their ability to sleep.
» The 30 participants were randomly assigned to one of 2
conditions: No caffeine (control) or one dose of caffeine
every 3 hours from 9am-6pm.
» The study measured participants ability to sleep by
taking the average number of hours slept per night
over a 2 week period.
ASK at Brunel (2014) 22
Caffeine and Sleep
21. Dr.ChristinePereira
ASK at Brunel (2014) 23
Participant Condition AvgHoursSlept
1 Control 7.2
2 Caffeine 6.7
3 Caffeine 6.3
4 Control 6.9
… … …
Caffeine and Sleep
25. Dr.ChristinePereira
» 50 participants of varying fitness levels were used to
investigate whether personal trainers make a significant
difference in ones fitness.
» Participants were randomly assigned to one of two
training groups: self training or professional trainer.
» Each participants 1 mile time (in mins) was recorded 2
days prior to the start of the study.
» Each group then followed a specific training regime for
30 days and their 1 mile time (in mins) recorded again.
ASK at Brunel (2014) 27
Fitness Regime
26. » 50 participants of varying fitness levels were used to
investigate whether personal trainers make a significant
difference in ones fitness.
» Participants were randomly assigned to one of two
training groups: self training or professional trainer.
» Each participants 1 mile time (in mins) was recorded 2
days prior to the start of the study.
» Each group then followed a specific training regime for
30 days and their 1 mile time (in mins) recorded again.
Dr.ChristinePereira
ASK at Brunel (2014) 28
Fitness Regime
Pre-test and Post-test
28. Dr.ChristinePereira
ASK at Brunel (2014) 31
Variables
Categorical
Qualitative
Scale
Quantitative
Nominal
(Unranked categories)
Marital Status
Political Party
Eye Color
Ordinal
(Ranked categories)
Satisfaction level
Level of agreement
Not grouped
Height
Weight
Age
No. of cars
No. of students
• In SPSS, data is either Nominal, Ordinal or Scale.
• It is essential to classify data correctly.
- Incorrect classification…
may result in incorrect analyses.
33. Dr.ChristinePereira
ASK at Brunel (2014) 36
1 = tfl (i.e. public transport)
2 = Car
3 = Cycle
4 = Walk
1
2
3
4
2
1
1
1
1
1
Level of
Measurement?
34. Dr.ChristinePereira
ASK at Brunel (2014) 37
Indicate your level of agreement with
the following statement:
“Most days, my commute causes me
to feel stressed when I arrive at
university”.
35. Dr.ChristinePereira
ASK at Brunel (2014) 38
1 = Strongly Disagree
2 = Disagree
3 = I don’t know
4 = Agree
5 = Strongly Agree
Level of
Measurement?
37. Dr.ChristinePereira
ASK at Brunel (2014) 40
1. Enter & define variables from Variable View.
• Variable names
• Must begin with a letter
• No spaces and no special characters (except _ )
38. Dr.ChristinePereira
ASK at Brunel (2014) 41
1. Enter & define variables from Variable View.
• Type
• Should (almost) always be numeric
• SPSS cannot analyse non-numeric data
39. Dr.ChristinePereira
ASK at Brunel (2014) 42
1. Enter & define variables from Variable View.
• Decimals
• How many decimal places do you want to see in
the Data View?
40. Dr.ChristinePereira
ASK at Brunel (2014) 43
1. Enter & define variables from Variable View.
• Label
• IMPORTANT!!! Make good labels!
• Short description of the variable
• This is what will be written on all graphs & tables
41. Dr.ChristinePereira
ASK at Brunel (2014) 44
1. Enter & define variables from Variable View.
• Values (i.e., Value labels)
• IMPORTANT!!! - Tell SPSS what all the codes represent!
• This is what is written on all graphs & tables
43. Dr.ChristinePereira
ASK at Brunel (2014) 46
1. Enter & define variables from Variable View.
• Missing
• Missing data can be coded too
• Chose a number not in the data, like 99 or 999
• Replace all missing values in your data with this code first
44. Dr.ChristinePereira
ASK at Brunel (2014) 47
1. Enter & define variables from Variable View.
• Measure
• Level of Measurement (Nominal, Ordinal, Scale)
• By default ALL new variables are unknown
• YOU must choose the appropriate measure
45. Dr.ChristinePereira
ASK at Brunel (2014) 48
2. Enter data from Data View.
• Variable names
• Will become column headings in Data View.
49. Dr.ChristinePereira
ASK at Brunel (2014) 52
Locate a folder
to save it in:
On your H drive OR
On your pen drive
Name it appropriately
Extension .sav
50. Dr.ChristinePereira
ASK at Brunel (2014) 53
Everything done in
SPSS is shown here,
in outline form.
Shows output for ALL analysis run in SPSS
Keeps a log of all activity of open data files
Saved with the extension .spv
A data file was saved as
SPSS Workshop Example.sav
and logged in the output file.
56. Dr.ChristinePereira
» Log onto Blackboard
» Go to the Organisation Academic Skills
» In the left column, under Workshop Presentations
Click on Statistics and SPSS
» Find the SPSS Workshop
Download the Excel File: “CommutingSurvey.xlsx”
» Save this file somewhere you can find it later.
» Do not open it. You cannot import the file if it’s open.
ASK at Brunel (2014) 59
57. Dr.ChristinePereira
» Can be .xls OR .xlsx
» Variable names CAN be imported too
Must be in row 1 of the worksheet
They will be imported to the Variable View
» Data will appear in SPSS Data View
ASK at Brunel (2014) 60
59. ASK at Brunel (2014) 62
Existing
data file
Dr.ChristinePereira
60. Dr.ChristinePereira
ASK at Brunel (2014) 63
Choose Excel file type to
see your .xlsx
Find the folder
where you
saved the fileSelect the
Excel file
3
1
2
65. Dr.ChristinePereira
» Any blank (i.e., missing) data is automatically
considered missing by SPSS.
It will not be included in the analysis.
This means you DO NOT have to code missing data
» Why code missing values if you don’t have to?
Sometimes data accidently gets deleted. If missing values are
coded then you know a blank space shouldn’t be there.
There is more than one reason data is missing and you want
to distinguish between them.
+ E.g., A participant chose not to answer a survey question or question
was not applicable to them – you may want to record these as
different kinds of missing data.
ASK at Brunel (2014) 68
66. Dr.ChristinePereira
ASK at Brunel (2014) 69
There are 3 missing
values for
TravelTimemin
• Replace missing values with a code
• Use a code that will not occur for the variable
• Let’s use -1, as negative time is not possible
69. Dr.ChristinePereira
ASK at Brunel (2014) 72
4
Enter the
code as the
new value
Click Add.
MISSING -> -1
will appear here
5
6
Click Continue
then Click OK
7
71. Dr.ChristinePereira
ASK at Brunel (2014) 74
Remember…
From Variable View
Define the missing values
8
• Discrete (whole numbers)
• Enter missing value codes used for the variable (e.g., -1)
• Click OK
72. Dr.ChristinePereira
ASK at Brunel (2014) 75
This is only a code – it tells SPSS what value(s) represent
missing values.
It does not replace missing values with the code for you!
73. Dr.ChristinePereira
» Variables go in columns
» Categorical data should be coded first
» If you import data from Excel, make sure to:
Put variable names in the first row
Code your categorical data first, then import codes
» If you choose to code missing data:
First, replace all missing values with the code
Second, define the missing value code in the Variable View
ASK at Brunel (2014) 76
74. Dr.ChristinePereira
» SPSS Survival Manual, 4th Edition (2010) by Julie Pallant.
» SPSS Survival Manual, 5th Edition (2013) by Julie Pallant.
» Discovering Statistics Using SPSS, 3rd Edition (2009) by Andy Field.
» Discovering Statistics Using SPSS, 4th Edition (2013) by Andy Field.
ASK at Brunel (2014) 77
Indicate your level of agreement with the following question:“Most days,my commute causes me to feel stressed when I arrive at university”.(1 = Strongly disagree, 2 = Disagree, 3 = I don’t know, 4 = Agree, 5 = Strongly Agree)
Indicate your level of agreement with the following question:“Most days,my commute causes me to feel stressed when I arrive at university”.(1 = Strongly disagree, 2 = Disagree, 3 = I don’t know, 4 = Agree, 5 = Strongly Agree)
Indicate your level of agreement with the following question:“Most days,my commute causes me to feel stressed when I arrive at university”.(1 = Strongly disagree, 2 = Disagree, 3 = I don’t know, 4 = Agree, 5 = Strongly Agree)
Indicate your level of agreement with the following question:“Most days,my commute causes me to feel stressed when I arrive at university”.(1 = Strongly disagree, 2 = Disagree, 3 = I don’t know, 4 = Agree, 5 = Strongly Agree)