data_analysis_using_spss.pdf

DATA ANALYSIS
USING SPSS
Dr. Mark Williamson, PhD
(based on PDF of Andrew Garth, Sheffield Hallam
University)

Purpose
■ The intent of this presentation is to teach you to explore, analyze, and
understand data
■ The software used is SPSS (Statistical Package for the Social Sciences)
– commonly used in social sciences and health fields
– as opposed to other statistical software such as SAS or R, it requires
little to no coding background
■ This presentation is heavily indebted to the work of Andrew Garth (Sheffield
Hallam University) and his full document can be found at the link below:
https://students.shu.ac.uk/lits/it/documents/pdf/analysing_data_using_sp
ss.pdf
■ All the data files used in this presentation can be found at the link below
(download the SPSSDATA.zip):
http://teaching.shu.ac.uk/hwb/ag/resources/resourceindex.html

Outline
■ First, we will look at the Big Picture
■ Next, we’ll define our terms
■ Then, we’ll get set up for working in SPSS
■ Only then will we get into the meat of things, which will
focus on aspects of data analysis
– Descriptive Statistics and Graphs (Exploring our Data)
– Inferential Statistics (Analyzing our Data, and
Interpreting our Results)

The Big Question.
It depends on the nature of the data and
what questions you want to answer
■ How should I analyze my data?
To answer those questions, you need to explore your data.
and select the proper analysis

1. Explore your data
1. Look at data
2. Identify data
3. Graph/Describe data
4. Formulate Question (Hypothesis)
2. Analyze your data
1. Set up hypothesis
2. Check normality
3. Select and run appropriate test
3. Interpret your results
1. Find the Test Statistic, DF, and P-value
2. Determine if significant
3. State if null hypothesis rejected or not
4. Write result
5. Present appropriate plot
Big Picture Steps in Statistical Analysis

Before we can start analysis, we need to get
set up on the basics
■ Defining Terms
■ Working in SPSS

Defining Terms
■ There are two basic data types, each with two sub-types
– Numerical: expressed by numbers
■ Discrete: numbers take on integer values only (number of children,
number of siblings)
■ Continuous: numbers can take on decimal values (height, weight)
– Categorical: expressed by categories (also known as
factors/groups)
■ Nominal: no meaningful order between categories (gender, occupation)
■ Ordinal: categories can be put in meaningful order (agreement, level of
pain, etc.)
■ If data is not used for analysis, it can be labeled as a nuisance or
bookkeeping variable

Defining Terms 2
■ Data can also be paired or unpaired
– Paired: categories are related to one another
■ Often result of before and after situations
(treatments/events)
■ Since each part of the pair is related to each other, this
needs to be considered
■ If there are pairs of higher than 2, this is called repeated
measures
– Unpaired: categories are not related to one another
■ Numerical data can be parametric or non-parametric
– Simply put, parametric data approximately fits a normal
distribution
■ Data are symmetric around a central point
■ “Bell curve”
■ Also known as normally distributed
– Data must be parametric (normally distributed) for many statistical
tests
■ If the data are not parametric, you cannot use the test results
■ If the data are non-parametric (does not fit a normal
distribution), there are non-parametric tests for use, but they
are weaker

Defining Terms 3
Recap
■ Data can be:
– Numerical, categorical, or
nuisance
– Paired or unpaired
– Parametric or non-parametric
(usually must run a test to tell)
Examples
■ Numerical continuous: height, weight, drug
concentration
■ Numerical discrete: number of siblings, number
of drinks in a day, flower petal number
■ Categorical ordinal: time of day (morning, noon,
night), position (assistant professor, associate
professor, department chair, dean)
■ Categorical nominal: flower color, college major,
drug treatment (A, B, C)
■ Nuisance: sample number, subject name, date,
id number
■ Paired: Before, during, and after treatment; pre-
and post-disaster

Defining Terms 4
■ For statistical tests, we use two types of variables:
– Independent Variable- variation does not depend on another variable
■ Usually denoted as X
■ Typically represents what the researcher set up (treatment, group, etc.)
– Dependent Variable – value depends on another variable (the independent one)
■ Usually denoted as Y
■ Represents the variable that the researcher is interested in
■ Output or outcome
■ Almost all statistical tests give three important pieces of information
– Test statistic
■ Variable calculated from sample data and used in hypothesis test
■ Used to determine whether a test was significant or not
– Degrees of Freedom
■ Number of values of quantities that can be assigned to a statistical distribution
■ Should be reported with test results
– P-value
■ Measure of significance for the test statistic
■ Typically 0.05 is the cutoff value

Assessment 1
1. What types of data (categorical [nominal,
ordinal], numerical [discrete, continuous] are
each of the following examples
a) Number of vaccine shots administered
b) Highest level of education attained (high
school, bachelors, masters, PhD)
c) Country of origin
d) Tumor size
2. In the boxplot graph to the right, which axis is
the independent variable plotted on? Which
axis is the dependent variable plotted on?
3. In the table to the right, label each of the
columns as numerical, categorical, or nuisance
Sample # User ID Height Treatment Group
1 34AF001 162.3 1 A
2 67AF001 159.1 1 B
3 78AF001 160.2 1 C
4 22AF001 165.0 2 A
5 13AF001 157.5 2 B
6 49AF001 155.0 2 C

Assessment 1 Answers
1. What types of data (categorical [nominal, ordinal],
numerical [discrete, continuous] are each of the following
examples
a) Number of vaccine shots administered (numerical
discrete)
b) Highest level of education attained (high school,
bachelors, masters, PhD) (categorical ordinal)
c) Country of origin (categorical nominal)
d) Tumor diameter (numerical continuous)
2. In the graph to the right which axis is the independent
variable plotted on? Which axis is the dependent variable
plotted on? Independent on X-axis (Treatment),
Dependent on Y-axis (Inflammation)
3. In the table to the right, label each of the columns as
numerical, categorical, or nuisance
(nuisance, nuisance, numerical, categorical, categorical)
Sample
#
User ID Height Treatment Group
1 34AF001 162.3 1 A
2 67AF001 159.1 1 B
3 78AF001 160.2 1 C
4 22AF001 165.0 2 A
5 13AF001 157.5 2 B
6 49AF001 155.0 2 C

Starting in SPSS : Access
■ You can get access to SPSS using the
CitrixWorkspaceApp for UND
■ Some UND computers also have it
downloaded
■ If all else fails, you can try a free trial
(https://www.ibm.com/account/reg/
us-en/signup?formid=urx-19774)
■ From here on out, I will be using the
following formats
1. White boxes with green border are instructions
in SPSS.
2. These will guide you through how to do the
exploration/analysis I show yourself.
White boxes with purple borders are summaries
1. Orange boxes with red border are
general step outlines
Blue boxes are reminders

Starting in SPSS: Data Format
■ Specifics of format depends on the kind of data
■ Principles that apply in most situations
1. Each case goes in its own row
2. Categorical variables are best represented by
numbers (even though they are not): can be
labeled with Variable Labels option
3. Variable names for the columns are limited in
length, so again can be labeled with Variable
Labels option
4. Multiple groups of subjects should still be set
up with each case having its own row: create a
new variable column and give it the group
label

Starting in SPSS: Entering data
■ There are two ways to enter data into
SPSS
– Manually (entering the data by
hand)
– Loading in a file (data is saved in
some form and can be opened in
SPSS)
■ Let’s try manual first
■ You can look at the data in two ways
– Variable View
– Data View
■ SPSS gives a lot of information, most
which you don’t need
– Ignore what you don’t need
1. Start SPSS from wherever you have it
2. Double click New Dataset at the top left
3. In the box on the right there are 10 people’s names,
type them into the first column
4. You may notice a problem when you get to Peter.
1. Peter has 5 letters in his name, unfortunately
SPSS has assumed all the cases are similar to
the first one and Peter has become Pete.
2. We can alter this by switching to the Variable
View (click the tab at the bottom of the SPSS
window). You should see a row of information
about variable one (var0001), which is where
we are storing these names.
3. Change the Width from 4 to 12.
4. Go back to the Data View and type in Peter
again.
5. Finish typing the names.
5. Go back to the Variable View and change the column
name (variable) to person rather than var00001.
6. Do the same for var00002, replacing it with the name
‘age’.

Starting in SPSS: Saving data
■ Graphs and analyses will not be
saved unless you save them
specially
■ Save often
■ It is good practice to have multiple
copies of data (especially when
working on original data)
1. To save the names and ages from the previous slide, choose Save from
the File menu. Call it people and put your name at the end of the word
(ex. peopleAnderson).
2. You can save anywhere you want by using the Look in: and selecting the
appropriate location
3. To save graphs or analyses, we need to do an analysis first
1. Click on the Analyze menu and choose Descriptive Statistics, then
Descriptives.
2. The button between the two windows let you choose the variables
to be analyzed, in our case the choice is simple, just click the
center button to move the age variable over to the right then click
OK.
3. SPSS should display the results in a separate window, you will see
this appear in front of the Data Editor and a new button will
appear on the Windows task bar at the bottom of your screen. The
new window has a title, have a look in its title bar at the top of its
window.
4. Look at the output. If you want to save results like this, you have
to save it separately.
Reminder: the data needed for the
tasks to follow are at:
https://teaching.shu.ac.uk/hwb/a
g/resources/resourceindex.html

Starting in SPSS: Looking at data
■ Seeing what data looks like is the first step to data
analysis
■ It gives a broad-overview in what is going on
■ Again, each row is a different sample, while the columns
show the value of different variables for that sample
■ Looking at the data tells you a lot of big-picture things
– How many samples there are
– How many variables there are
– The types of variables and their values
– If there is any missing data
■ We will examine some data collected by an Occupational
Therapy student, looking at how age affected OT
students’ participation in discussion in class.
■ She counted how many times each student contributed
orally in a period totaling 12 hours of classes. The
students were from the 1st and 2nd years of the course
and were classed as young if under 21 and mature if 21
or over, making 4 groups altogether.
1. Open up Studentss in SPSS
1. choose the File menu and select Open->
Data (will need to search for wherever
you downloaded the sample files)
2. Take a look at the data and answer the
following questions.
1. What is each column telling you?
2. Which group is which?
3. How many students were in each group?
4. Do older students contribute more
frequently in class discussion?

Starting in SPSS: Exploring the Data
■ When analyzing data, it is necessary
to know what variable is what
■ Dependent variable:
– depends on the factor
– Is usually numerical
– In our case, it is ‘speaks’
■ Independent variable (Factor):
– Is the groups that the different
samples are grouped into
– Is usually categorical
– In our case, it is ‘group’
1. Click on the Analyze menu->Descriptive
Statistics->Explore.
2. Transfer the speaks variable to the
Dependent list and the group variable to
the Factor list and then click OK.
3. Take a look at the results.

Descriptives
group Statistic Std. Error
speaks M1 Mean 33.09 7.303
95% Confidence Interval for Mean
Lower Bound16.82
Upper Bound49.36
5% Trimmed Mean 32.16
Median 31.00
Variance 586.691
Std. Deviation 24.222
Minimum 2
Maximum 81
Range79
Interquartile Range 34
Skewness .677 .661
Kurtosis .185 1.279
M2 Mean 46.91 10.964
Lower Bound22.48
Upper Bound71.34
Median 34.00
Variance 1322.291
Minimum 19
Maximum 148
Range129
Skewness 2.475 .661
Kurtosis 6.939 1.279
Y1 Mean 9.67 2.101
Lower Bound5.04
Upper Bound14.29
Median 8.00
Variance 52.970
Minimum 0
Maximum 21
Range21
Skewness .245 .637
Kurtosis -1.248 1.232
Y2 Mean 16.50 3.845
Lower Bound7.80
Upper Bound25.20
Median 12.00
Variance 147.833
Minimum 4
Maximum 40
Range36
Skewness 1.292 .687
Kurtosis .542 1.334

Using descriptive statistics
■ It is hard to read out the various
descriptive statistics from graphs
■ Instead, we can calculate them and
spit out numbers in tables: such as
medium, mean, interquartile range,
and, standard deviation
■ Measures of central tendency, or
‘average:
– Mean: all values are summed
and divided by the number of
values
– Median: middle value
– Mode: the most common value
■ Measures of spread:
– Interquartile range
– Standard Deviation
1. Go back to the studdentsss file
2. Got to Analyze menu, select Descriptive
Statistics, then Explore. The dependent list
refers to the quantity we are measuring, in this
case, the number of times people speak. In
the factor list we put the factor that we are
investigating, in this case "agegroup".
3. From the output find the Mean and Median of
each group. The mean and median are both
forms of average, do they seem to agree?
Part A-3d

Assessment 2
1. When formatting data in SPSS,
should each sample be put in its
own row?
2. Will SPSS automatically save
results and graphs?
3. What is the mean, median, and
mode of the dataset to the right?
Number of
Siblings
2
1
1
2
3
5
10
2
4
1

1. When formatting data in SPSS,
should each sample be put in its
own row? YES
2. Will SPSS automatically save
results and graphs? NO
3. What is the mean, median, and
mode of the dataset to the right?
3.33, 2, 2
Number of
Siblings
2
1
1
2
3
5
10
2
4

Descriptive Statistics and Graphs
(Exploring our Data)
■ A large part of data analysis is exploring your data and
understanding more about it, both by visually graphing it
and generating statistics such as means
■ This section will go over a variety of the basic approaches

Rules for Exploring Data
■ Discipline
– If you discipline yourself by doing each of these things each time you look at your data, you will
develop the skill to intelligibility see the data
– This will give you the freedom to analyze the data without struggling to comprehend even the most
basic understanding of the data
– Computers are fast but dumb, so they rely on you to supply the intelligence to make sure the results
are useful
■ Rules
1. Look at data: open up the file and look at the raw data (or, if the data is too large, a subset)
2. Identify data: for each column determine what type of data it is
a) If it is numerical, is it continuous or discrete?
b) If it is categorical, how many categories and is it nominal or ordinal?
c) Or if it is not useful, call it a nuisance variable?
d) Are their any variables that may be paired?
3. Graph/Describe data: for each variable or set of variables (comparison), graph and run descriptive
statistics
4. Write Research Question: Write out in a clear sentence what each comparison is trying to test

Rules: Example with Plant data
Sample Treatment Growth Rate Leaf Number
1 Control 20.1 5
2 Control 27.5 6
3 Control 23.2 5
4 Control 19.8 4
5 Phosphorus 45.6 5
6 Phosphorus 33.4 4
7 Phosphorus 42.2 6
8 Phosphorus 47.7 7
9 Nitrogen 32.5 4
10 Nitrogen 27.3 5
11 Nitrogen 24.6 5
12 Nitrogen 30.0 5
1. Look at the
Data
2. Describe Each
Variable
3. Graph/Stats
each Comparison
4. Write Research
Question
Nuisance Categorical
Nominal
Numerical
Continuous
Numerical
Discrete
• Is there a difference in [plant]
growth rate across nutrient
treatments?
• Is there a difference in [plant] leaf
number across nutrient
treatments?
• Is Leaf number related to growth
rate?

Mean vs. median
1. Open a new file. (File->New->Data) We are going
to type in a few figures.
2. Put the following numbers in the first column
(7000, 7000, 7000, 7000, 7000, 7000, 7000,
7000, 7000, 100000).
3. Give the column the title ‘Salaries’ (you need to
click onto the Variable View for this
4. Back in Data View you may want to alter the
column width by dragging the vertical bar next to
the variable name.
5. The numbers represent the annual salaries of the
10 permanent employees of a small (mythical)
private clinic. Which is the director’s?
6. Run Descriptive Statistics->Explore to find the
mean and the median. If you were the union
negotiator for the employees of the clinic which of
the two average salaries would you quote to the
press? If you were the owner of the clinic which
might you quote?
7. Find the inter-quartile range and the standard
deviation. Can you sketch what the Boxplot would
look like? Create the Boxplot on SPSS if you like.
Part A-3
Summary: Mean vs. Median - both are types of
average. The mean is based on all the data
values, however because of this it is prone to
being unduly affected by outliers in the data,
most noticeably when the sample is small. The
median however is largely unaffected by one or
two extreme outliers, even in small samples, it is
simply the middle value.

Standard deviation
■ What is the Standard Deviation (S.D.)
really measuring?
■ What can it tell us about our data?
■ Let’s take a look at some data
– The table below shows the
German, Geography and IT
results of a group of ten
students.
1. Open the file std dev example in SPSS
2. Use the Descriptive Statistics->Descriptives
to fill out the table below
3. Which set(s) of figures has the largest
range?
4. Which set(s) of figures has the largest
number in it?
5. Which set(s) of figures contains the smallest
number?
6. Which set of figures has the largest
minimum?
Part A-4
German Geography IT
MEAN
MAX
MIN

Standard deviation 2
■ Given the figures for mean, maximum and minimum it is hard to differentiate between the German and
IT figures, the mean, (arithmetic mean) of the figures is the numbers all added together then divided by
the number of numbers.
■ However it gives no indication of the distribution of the marks within the sets of figures. To do this we
could graph the three sets of figures and see if that helps us (later we will create bar charts, for now just
look at these).
■ Look at the three graphs above. Which two do you think are most similar?
■ Possibly Geography and IT but it is rather subjective. They do seem to have less variation in the values
than the German results.
Part A-4

■ Question: How can we asses in a
fair, unambiguous way, which of
three has the least widely deviating
set of numbers?
■ Answer: Use the Standard
Deviation.
■ The standard deviation of a set of numbers is a measure of how widely values are dispersed
from the mean value. It can be calculated manually, or SPSS can calculate it for you.

■ Let’s work out the standard deviation of the
numbers in each column from the std dev
example
– Higher Standard Deviation values indicate a
greater spread of values
– Lower Standard Deviation values indicate a
tighter spread of values
1. Use Descriptive Statistics then Frequencies from
the Analyze menu.
2. Select the three variables (get German,
Geography and Information Technology (IT) from
the left into the right pane).
3. Click the “Statistics” button and select the
Standard deviation as well as mean, maximum
and minimum, then click “Continue”.
4. Before pressing OK on the Frequencies dialog
box, uncheck the option to display frequency
tables then click OK.
5. Compared the standard deviations.
1. Which set of figures, German, Geography
or IT, is the least spread out?
2. Of the two subjects with the same mean,
and the same range, which varies least?
3. Which of the three sets of figures, German,
Geography or IT varies most?
Part A-4
Summary: Range, IQR & SD are all measures of
spread. Only the SD takes all the data values into
account, however this leaves it open to problems
similar to the mean, i.e. a tendency to be swayed
inordinately by extreme values. The range is
extremely sensitive to outliers, since it is based
only on the smallest and largest values. The Inter
Quartile Range is again based on only two values,
the upper and lower quartiles, these are on each
end of the middle half of the data, therefore less
effected by extremes.

Assessment 3
1. In the data to the right, which subject
had the highest average score?
had the most variation in score?
Which had the least?
3. What are the 4 rules for exploring
data?
Exam Scores
Subject N Mean Standard
Deviation
Art 10 95 3.3
Spelling 10 70 5.8
Math 10 67 3.5
Science 10 84 12.3
Social Studies 10 89 2.1
Physical Education 10 98 1.2

had the highest average score?
Physical Education
had the most variation in score?
Which had the least?
Science, Physical Education
3. What are the 4 rules for exploring
data?
1. Look at the Data
2. Describe Each Variable
3. Graph/Stats each Comparison
4. Write Research Question
Exam Scores
Subject N Mean Standard
Deviation
Art 10 95 3.3
Spelling 10 70 5.8
Math 10 67 3.5
Science 10 84 12.3
Social Studies 10 89 2.1
Physical Education 10 98 1.2

Graphs
■ Graphs serve two purposes
– Quickly visualize data during data exploration
– Present results of significant statistical analyses

Types of Graphs to be covered
Type of Graph Data Type Usage Basic Example Another Example
Histogram Single numerical variable Data exploration
(determining normality)
Heights of freshman
students
Tooth number of apex-
predator dinosaurs
Boxplot Single numerical variable;
single numerical variable +
categorical variable
Data exploration,
presenting non-parametric
t-tests/ANOVA
Heights of freshman
students; Heights of
students by grade
Weights of apex-predator
dinosaurs; Weight of apex-
predator dinosaurs by
geological period
Bar Chart Single numerical variable +
Presenting Parametric T-
test/ANOVA results
Heights of students by
grade
Tooth number of sharks by
species
Scatterplot Two numerical variables Data exploration,
presenting correlation
results
Heights and weights of
students
Weights and top swimming
speed of sharks
Line Charts Two numerical variables
(one usually time)
Data exploration Heart rate over time Ounces of coffee drank by
students over time
Multiple Line Charts Three or more numerical
variable (one usually time,
rest on same scale)
Data exploration Various concentrations of
nutrients in bloodstream
over time
Ounces of various
caffeinated beverage
drank by students over
time
Pie graph Single numerical variable
(proportions) + categorical
variable
Data exploration Percentage of students
across grades
Percentage of different
caffeinated beverages
drank in a month

Histogram and Normal Distribution
■ Histograms can be used to look at
the distribution of data
■ This is important for determining if
the data is parametric or not
1. Open the file Reconstructed male heights 1883 in SPSS.
2. This file contains data that is similar to that from which the
table you have seen was derived. The file contains 8585
heights, measured in inches.
3. We are going to create a histogram from the values in the
variable called hgtrein
4. From the menus choose Graph->Chartbuilder.
5. A dialog box will come up, choose OK.
6. In the bottom section Choose Histogram and double click the
first image
7. Drag the hgtrein (Heights in inches - reconstructed) variable
over to the box representing the horizontal (X) axis of the graph.
8. Click OK and wait to see the graph in the output viewer. You
should see a normal (bell shaped) pattern to the distribution of
the data.
9. To see a normal curve superimposed on the graph go back to
the Create Histogram dialog box (from the menus Graph,
(Legacy,) Interactive, Histogram) then click on the Histogram
tab and tick the "Normal curve" check box, then Click OK.
10. Are these data Discrete or Continuous?
Reminder: if data is parametric, it will
approximate a normal distribution (bell
curve) when viewed as a histogram. Many
statistical tests can only be used if the
data is parametric

Histogram 2
Radiologist example:
■ The file Radiologist dose with and without lead
combined.sav contains data gathered to assess the
effect of a lead screen to reduce the radiation dose to
Radiologists hands while carrying out procedures on
patients being irradiated.
■ In the trials the lead screen was placed between the
patient and the radiologist, the intended effect was to
reduce the radiation dose to the radiologist, however
there were fears that working through the screen would
lengthen the procedure. We want to answer two
questions with this data, one about the hand dose and
the other about the length of time the examination
took.
1. Open Radiologist dose with and without lead combined file in
SPSS
2. Look at the data, the variable called "screen" is the variable
that lets you discriminate between procedures carried out with
or without the lead screen. If there is a 1 in the screen variable
column it means the procedure was carried out with the screen
in place, if not the value is 0.
3. We can use this discriminatory variable to create two
histograms at once, by using it as a panel variable.
4. The variable we are interested in is the dose to the radiologists'
left hand, the left-hand would be nearest the patient so we will
concentrate on the left-hand dose variable.
5. Draw histogram using the left-hand dose variable (lhdose)
6. Go to the Groups/Point ID tab and click the Rows panel
variable
7. Drag the discriminatory variable (Lead or No Lead) as the panel
variable.
8. What do the histograms show us about the data?
9. If you have time draw a similar histogram using the extimmin
variable. Does this back up the fears about the increase in
examination time?
Summary: Histograms are for displaying continuous data,
e.g. height, age etc, the bars touch, signifying the
continuous nature of the data. The area of the bars
represent the number in each range, the bars are usually
of equal widths but this need not always be the case.
Histograms should be clearly labelled and the units of
measure displayed. The use of Histograms compared to
Bar Charts is summarized after the section on Bar Charts.

Drawing boxplots
■ Boxplots are a great way to visualize
data between varies groups
■ Requires: numerical dependent variable
and a factor with 2 or more groups
■ For paired data, you can draw boxplots
straight from the graph menu
1. Go back to in studentsss
2. Choose Chartbuilder under Graphs
3. In the bottom section Choose Boxplot and double
click the first image
4. Drag the speaks variable to the y-axis and the year
variable to the x-axis.
5. Look at your boxplots. Can you see an asterisk or
circle beyond the whiskers? In SPSS an asterisk
represents an extreme outlier (a value more than 3
times the interquartile range from a quartile). A circle
is used to mark other outliers with values between
1.5 and 3 box lengths from the upper or lower edge of
the box. The box length is the interquartile range.
6. Which number on your data screen does the most
extreme outlier correspond to? (SPSS gives a bit of a
hint here!) Why is it an extreme outlier?
7. Look at the boxplots, which group has the highest
median? What does this tell you about the groups?
8. Look at the boxplots, which group has the highest
interquartile range (IQR)? What does this tell you
about the groups? Refer to a glossary to review IQR
Summary: Boxplots are good for seeing the range
and level of data and highlighting outliers. The
box shows the IQR (Inter Quartile Range) and the
bar in the box shows the median. Boxplots should
be clearly labelled with the units of measure
displayed.

Bar Charts
■ Bar charts and histograms look similar at first
■ there is however a definite difference in the type of
data each is designed to show and this subtle
difference is an important one if you are using
them in your research.
■ Bar charts are for non-continuous data, i.e. data in
categories that are not related in any order.
■ Histograms are for displaying continuous data
■ the graph can be edited after it is drawn, just
double click on the graph and then click into the
labels you wish to alter
1. open the file shoetypes in SPSS
2. this file contains data about the type of shoes
worn at the time the data were gathered and
number of pairs owned by a sample of 100
people. We can use SPSS to analyze the data by
using bar charts among other methods.
3. Graphs->Chartbuilder->Select Bar
4. Drag the footwaretxt variable to the x-axis then
click OK. The graph above should appear.
5. Try again but this time, include a Rows Panel
variable, then drag the gendertxt variable over
Panel box and see what happens.
Summary: Bar charts are for non-continuous data e.g. the number of
people from each of five towns, the bars do not touch. Bar charts should
be clearly labelled and the units of measure displayed. Bar charts and
Histograms look similar, however the type of data they should be used on
is different. In a Histogram the bars touch each other, this denotes the
continuous nature of the data being displayed. Bar charts should be used
for discrete data. If you aren't sure about the difference between
continuous and discrete data look it

Percentages
■ Percentages are often used in bar charts
■ General formula for calculation
percentages
100 × the individual value ÷ the total of
the values
■ If percentages span across all values, the
total needs to sum to 100% across all
groups
Summary: Percentages show
proportions, it should be clear
what they are percentages of.
1. You can very quickly create summary percentages using the
"frequencies" command, for example in the shoes file,
2. What percentage of subjects were wearing each type of shoe?
3. Analyze->Descriptive Statistics->Frequencies
4. Add footware to the variables list
5. Does the percentage of footwear types differ in the different
gender grouped?
6. Lets get SPSS to do everything twice, once for males and once
for females, we can do this using the split file command.
Choose Data->Split file. Now calculate the percentages again
as you did before.
7. The output should now be split into two groups, one for Male
and one for Female. Tables like this are rarely in the ideal
format for inclusion in a dissertation or paper but can be
copied and pasted into a word processor and manipulated
there.
8. Remove the split once you have done with it. If you leave it on
you may get some strange results. Choose Data, Split file. Then
select the "Analyse all cases" option, then click OK.
9. Don't forget to switch this feature off when you don't need it!

Scatterplots
■ Used when data are paired: each
point on a diagram represents a pair
of numbers
■ A better description is that you use
scatter plots when comparing two
numerical variables (Unlike a
numerical and categorical like in box
or bar)
■ Scatter plots are used to detect
correlation
– Correlation is not causation
– Strong, weak, or no correlation,
– Positive or negative
1. Open the file Step in SPSS
2. These data come from an experiment to see whether subjects could
perform more step exercises in a fixed time in a group or on their
own. A physiotherapy student collected them as part of a third year
project.
3. Look at the data; you will see that the columns are of equal length,
this is another indication that the data are paired.
4. We are going to draw a scatterplot for these two columns with the
number of steps done individually on the x-axis.
5. Graphs->Chartbuilder->Scatter/Dot.
6. Drag individ to X-axis and group to Y-axis.
7. Do the points appear to form a line?
8. If they do is it a clear, quite thin line or more like a cloud?
9. Does it slope up or down from left to right?
10. Look at your answers and decide if there is a strong, weak or no
correlation. Is it positive or negative?

Scatterplots 2
1. Adding more information to the previous plot
2. Go back to the Chartbuilder.
3. To add a linear fit line, select the Total box
under Linear Fit Lines, or select the second
scatterplot image.
4. Does the line match what you predicted?
Summary: Scatter plots are used to
show paired data, where for example
one person is tested under two
circumstances, each individual will
have a pair of readings. In this
example a scatter plot can be used to
indicate changes between the
performance in different
circumstances. Scatter plots are also
typically used to show correlation.
Scatter plots should be clearly labelled
and the units of measure displayed.

Line graphs
Line graphs are useful in time-based designs
Typically consists of a numerical variable over time
Example: Oxygen used walking description
■ The variables in the file are: vo2 Volume of O2
ml/min vco2 Volume of CO2 ml/min hr Heart
Rate beats per minute seconds time in seconds
from start of procedure
■ The protocol employed to take the
measurements consisted of:
– 5 minutes rest, to achieve baseline values
for heart rate and enable the subject to
get used to the equipment, followed by:
– 10 minutes exercise, (walking at a self-
selected speed) followed by:
– a second 5 minutes rest, to ensure
baseline values return to the norm for the
subject. This is important when
interpreting the graph we are about to
draw.
1. Open the file Oxygen used walking in SPSS
2. The data is just part of a large dataset collected by a student researching
the effect of tibial malunion on oxygen expenditure during exercise.
3. For our purposes the data gives us a good example of a variable
changing over time. The file contains the data from just one subject.
4. From the menus choose Graphs->Chartbuilder->Line
5. Drag the Heart Rate (HR bpm) onto the Y-axis and the Time (time in
seconds) onto the X-axis.
6. Look at the graph. It is easy to see when the subject started and stopped
walking!
7. The increase looks massive, but it is because the graph used a false
origin (not set at zero). We’ll want to redraw and label better.
8. Go back to Chartbuilder
9. Click on the “Titles” tab and switch the text from Automatic to Custom
and add an appropriate name, such as “The effect of exercise on heart
rate.”
10. Click on the Y-axis. Under the Scale Range, Switch off the Automatic
feature. Set the minimum to 0 and the maximum to 100. Press OK.

Multiple Line graphs
Part A-8
■ More than one line can be plotted at once, as
long as the time variable is consistent
■ Example: Children looked after
– The data are from the Department for
Education and Skills
– gives figures for children looked after by
Local Authorities in England.
1. We will use the older graphing system and the data in the file called
Children looked after
2. The variable names may look a bit strange at first, go to the Variable view.
3. Graphs->Legacy Dialogs->Line->Multiple
4. Select the option for Multiple lines and Summaries of separate variables.
Then press "Define".
5. Transfer the variables "Boys 1-4" and "Girls 1-4" to the top box and the
"year" to the Category Axis then click OK.
6. The graph that appears should let you answer the following questions;
7. In the 11 years covered by the data do the numbers of girls and boys
aged 1 to 4 looked after by Local Authorities in England appear to
increase?
8. Are the number of boys and girls in the age group 1 to 4 staying in
roughly the same proportion, i.e. do they seem to increase or decrease
together?
9. Now plot the data for the 16 and over age group, can you see any
difference between the girls and boys?
Summary: Line graphs are ideal for showing the
changes in a variable as another alters, e.g.
changes over time. The independent variable goes
on the x-axis and the dependent variable goes up
the y-axis. More than one line is often shown on
the chart allowing comparisons. Line graphs
should be clearly labelled and the units of
measure displayed.

Pie charts
Part A-9
■ Pie charts are ideal for showing
proportions and summarizing data
■ They can be made using raw data or
pre-aggregated data
1. Open the file shoetypes in SPSS again
2. Graphs->Chartbuilder->Pie.
1. Put footware in the left bar (footwaretext should be in the
Set color bar on bottom)
2. Click the Rows panel variable, then drag gendertxt to the
panel bar. You should get two pie charts, one for each
gender, this might help identify any differences between
the gender groups in their choice of shoes.
3. Open the file hip patient numbers in SPSS.
1. This is a simplified version of the NHS hip fracture
discharge data for 1997 to 1999 for England for patients
aged 65 and over.
4. Drag the “Trust Cluster” variable to the “Slice By” box, this will
tell SPSS to make each slice of the pie represent one type of
trust (Small/medium acute, Large acute, Very large acute, Acute
teaching, Multiservice).
5. Drag the “Patient 97” variable to the “Count” box and press OK
Summary: Pie charts, are used to show proportion,
e.g. the number of votes cast for each party in an
election. The pie should add up to 100% of the
observed data. Pie charts should be clearly
labelled and the units of measure

Assessment 4
1. In the boxplot to the right, label the
letters with the appropriate term
a)
b)
c)
2. For the three histograms to the
right, label them as parametric
(normally distributed) or non-
parametric
3. For the scatterplots to the right,
label the correlation as:
a) Strong, Weak, or None
b) Positive, Negative, or None
a b
c

1. In the boxplot to the right, label the
letters with the appropriate term
a) Interquartile Range
b) Median
c) Extreme Value / Outlier
2. For the three histograms to the right,
label them as parametric (normally
distributed) or non-parametric
parametric, non-parametric, non-parametric
3. For the scatterplots to the right, label
the correlation as:
a) Strong, Weak, or None
b) Positive, Negative, or None
None-None, Strong-Negative
a b
c

Inferential Statistics (Analyzing our Data)
• If we want to draw conclusions about an
entire population from our sample, we enter
the realm of inferential statistics
• This section will go over a variety of the
basic tests

Guidelines of tests
■ You ought to be interested in using statistics to make as accurate mathematical
inferences about the complexities of reality to make the world a better place
■ The statistics only tell you as much as you put into them, and again, they are only
mathematical representations
■ It is up to you to be as disciplined as possible in setting up your data and analyzing it
in such a way as to best get at the truth of the world
■ The following are my strong suggestions of how to go about analyzing data: think of
them like football drills: you need to master the basics to be any good at answering
questions with statistics
Part B-2

Rules of Analysis
Part B-2
A. Explore your data (outlined in first section)
1. Look at data
2. Identify data
3. Graph/Describe Data
4. Formulate Question
B. Analyze your data
1. Set up hypothesis (null and alternative)
2. Check normality
C. Interpret your results
4. Write result

Analyze your data: Set up Hypothesis
■ When running a statistical test, there are two hypotheses being tested
– Null Hypothesis: the default, or ‘boring’ state
■ Typically ‘no change’, ‘no difference’, or ‘no relationship’
– Alternative Hypothesis: something else happening
■ Construct the two hypotheses based on your question from the data exploration step
■ Example 1
– Question: Is there a difference between male and female shark body length?
– Null Hypothesis: There is no difference in shark length by gender.
– Alternative Hypothesis: There is a difference in shark length by gender.
■ Example 2
– Question: Is there relationship between the cups of coffee consumed during studying
and exam grade?
– Null Hypothesis: There is no relationship between cups of coffee and exam grade.
– Alternative Hypothesis: There is a relationship between cups of coffee and exam grade.

Analyze your data: Check Normality
■ Many tests can only be run with data that is parametric (normally
distributed)
■ Check normality by histogram, QQ-plot, and test for normality
■ Usually try two or three, as each gives some different information
– I prefer using histograms and QQ-plots, as the test for normality is
strict and most data isn’t neat enough to pass
■ Determining if something is normally distributed via graph inspection is
partly an art (you have to get good at looking at the graphs)

Analyze your data: Tests for Normality
■ Histogram
– Bars should approximate the bell curve if
it is normally distributed
– Doesn’t have to be perfect
■ QQ plot
– In this plot, the normal distribution is a
straight line
– If normally distributed, the points should
cluster around the straight line
– Should not have ‘tails’
■ Test of normality
– Statistical test
– Kolmogorov-Smirnov standard
– Shapiro-Wilk for small sample size
– Sig. column (p-value) interpreted as if
more than 0.05, from normal distribution
– If less, then not from normal dist.
1. Open tests for normality file in SPSS
2. For Histogram
1. Graphs->Chartbuilder->Histogram
2. Variable under investigation to the horizontal
(try both Random number and Normally
distributed)
3. Select normal curve
3. Q-Q plot
1. Analyze->Descriptive Statistics->Q-Q Plots
2. Can run both variables at the same time
4. Test for normality
1. Analyze->Descriptive Statistics->Explore
2. Put variables to check under the Dependent
list box
3. Click to select Normality plots with tests
Part B-9
Reminder: every test has a test statistic
and a p-value. The p-value tells you if the
test statistic is significant

Degrees of Freedom
Test Statistic P-value

Assessing normality
■ If the data is not normally distributed, you can try transformations
– These only work for numerical data, and some only work for certain kinds of
numerical
– Most common:
■ Log transformation
– Can’t use on variables that include zero or negative numbers
■ Square root transformation
– Can’t use on variables that include negative numbers
■ Doing this in SPSS
– Can play around with Transform->Compute Variable
– This can create new columns of variables based on transformations
– (We won’t be transforming data in this presentation, but it is useful to know in the
future)
■ Then, check for normality (histogram, Q-Q plot, etc.) on the transformed data,
– If it looks normally distributed, use the transformed data in the analysis
– If not, try another transformation
– If still not, will have to use non-parametric Reminder: Non-parametric tests are
weaker, so we only use those tests if we
cannot use parametric ones

Does the data fit a normal
distribution (or close
enough)?
Great! Use a
PARAMETRIC TEST.
Can the data be
transformed to fit a
normal distribution?
Great! Use a
PARAMETRIC TEST.
Use a NON-
PARAMETRIC TEST.
Y
Y
N
N

Analyze Your Data: Select Appropriate
Test
■ Depends on the variables and the types of questions you
want to answer
■ Whether the data is numerical, categorical
■ How many categories there are in the categorical variable
■ Whether the data is paired or not
■ Whether the data is parametric or not

Interpret Your Results
– Generally picking them out of a results table
– If the test does not give degrees of freedom (DF), use number of samples instead (N)
– If the p-value is below a certain threshold (usually 0.05), it is significant
– If the p-value is above, it is not significant
– Null hypothesis rejected if significant p-value for the test statistic
4. Write result
– Answer the question stated from the data exploration
– Include the test statistic, p-value, and degrees of freedom
– Also include type of test
– Example 1: Female sharks were significantly larger than males sharks (two-tailed T-test, F=5.67, p-value=0.0024, DF=19)
– Example 2: There was no relationship between the number of cups of coffee drunk and exam score (Pearson Correlation,
F=1.23, p-value=0.5863, DF=24)
– Don’t typically plot anything for a non-significant test
– Simple tests like t-tests can just have the written results; more complex analyses like correlation or ANOVA should get a plot

Assessment 5
1. Based on the Q-Q plot to the right, would
you consider the data normally
distributed or not?
2. Based on the normality test to the right,
would you consider the data normally
distributed or not?
3. A variable in a dataset is assessed for
normality and found to not be normally
distributed. However, a logarithmic
transformation of the data is normally
distributed. Can you use a parametric
test?

1. Based on the Q-Q plot to the right, would you
consider the data normally distributed or
not?
No, the points don’t follow the straight line
very well at all
2. Based on the normality test to the right,
would you consider the data normally
distributed or not?
No, significant p-value means it likely does
not follow a normal distribution
3. A variable in a dataset is assessed for
normality and found to not be normally
distributed. However, a logarithmic
transformation of the data is normally
distributed. Can you use a parametric test?
Yes, on the transformed data

Types of tests
■ 1 categorical variable + 1 numerical variable
– Categorical variable is non-paired and group number is one:
■ One Sample T-test
– Categorical variable is non-paired and group number is two:
■ Parametric: T-test
■ Non-parametric: Mann-Whitney Test
– Categorical variable is paired and group number is two:
■ Parametric: Paired T-test
■ Non-parametric: Wilcoxon Signed Ranks Test
– Categorical variable is non-paired and group number is greater than two:
■ Parametric: ANOVA
■ Non-parametric: Kruskal Wallace Test
– Categorical variable is paired and group number is greater than two:
■ Parametric: Repeated Measures ANOVA
■ Non-parametric: Friedman test
■ 2 numerical variables
– Correlation
■ Parametric: Pearson correlation (usually)
■ Non-parametric: Spearman rank-order correlation
■ 2 categorical variables
– Chi Square test

One Sample T-test
– Categorical variable is non-paired and
group number is one
■ This is when you have a single numerical variable
you are interested in and want to know if it is
different from some value
– Is the average height of basketball players
greater than 6.2 feet?
– Is the infant mortality rate in a certain
county less than 2 death in 1000?
– Is the effectiveness of treatment of a new
drug any different from zero?
1. Easy, since all you have is one variable
2. Histogram and maybe boxplot
2. Check normality
1. Histogram, QQ-plot
1. Null: the variable is no different from a certain
value
2. Alternative: it is different
1. Student’s T-test
2. If non-parametric, mumble, mumble Mann-
Whitney
5. Interpret results
1. Null rejected or failed to reject?
2. What does it mean for your question
3. Write it out

One Sample T-test
■ Example: Women’s Height
– data on the heights of women of
different ages (women, age, height).
– Focus on just the first column of
young women (women from ages
20-24)
– Question: Is the average height of
women ages 20-24 different from
155cm?
1. Open waheig2 file in SPSS
2. Explore data and check normality
1. We know that it is for this
3. Define null and alternative hypothesis to question.
1. Null= there is no difference
2. Alt= there is a difference (younger women different than
155cm)
4. Run t-test (2-sample T-test)
1. Analyze -> Compare Means -> One-Sample T-test
2. Sam20-24 goes into the Test variable box
3. Set Test value to 155
5. Interpret Results
1. See next page
2. What is the test statistic, degrees of freedom, and p-value?
3. Did you reject or fail to reject the null?
4. What does it mean for the question?
5. Are young women on average different from a height of
155cm? How so?
Part B-2

1. Find the Test Statistic, DF, and P-
value
• t=7.533
• DF=29
• P-value<0.0001
• P-value < 0.05
• Significant
3. State if null rejected or not
• Reject Null
4. Write result
• Young women were
significantly taller
(mean=162.5) than the value
of 155 cm (1-sample t-test,
t=7.533, DF=29, p-
value<0.0001).
• N/A
Degrees of
Freedom
Test Statistic
P-value

T-test
■ 1 categorical variable + 1 numerical
variable
– Categorical variable is non-paired and
group number is two
■ This is when you have a single numerical
variable you are interested in and want to
know if it is different between two groups
– Are men taller than women?
– Does treatment A reduce mortality
more than treatment B?
– Are there more sharks attacks on the
East Coast or the West Coast?
1. Histogram of numerical variable
2. Boxplot of numerical variable grouped by
the categorical variable
2. Check normality
1. Null: there is no difference between groups
2. Alternative: there is a difference
1. Parametric: T-test
2. Non-parametric: Mann-Whitney
3. Write it out
Part B-2

T-test: Examples
■ Parametric Example: Women Height
– data on the heights of women of
different ages (women, age, height).
– This is not paired data, these are 60
different women not the same 30
measured twice with 30 years between!
– Question: Is there a difference in height
between the two age groups of women?
1. Open waheig2S file in SPSS
1. We know that it is for this
1. Null= there is no difference
2. Alt= there is a difference (younger women taller)
4. Run t-test (2-sample T-test)
1. Analyze -> Compare Means -> Independent-Samples T-test
2. All heights goes in to Test variables
3. Age range goes in to Grouping variable
1. See next page
5. Are the two groups of women different?
6. If so, how are they different (which group is taller?)
Part B-2

value
• F=0.094
• DF=58
• P-value=0.016
• P-value < 0.05
• Significant
• Reject the null
4. Write result
• Younger women, age range of
20-24, are significantly taller
than older women of an age
range of 50-54 (2-tailed T-test,
F=0.094, DF=58, P-
value=0.016).
• N/A
Degrees of
Freedom
Test Statistic
P-value

Interpreting Results
■ Look at table
■ Find the various values:
– Test statistic:
– Degree of freedom or number of samples
– P-value: tells you whether to reject or fail to reject the null hypothesis
■ Typically, the cutoff value is 0.05
■ So, if value is <0.05, Reject the Null (and Retain the Alternative)
■ If value >0.05, Fail to Reject the Null, so retain the Null over the Alternative
– (another way of saying, “based on our statistical test, there is no evidence that reality is
anything other than the null hypothesis”)

T-test: Examples 2
■ Non-Parametric Example: Student
Contribution
– The file has all the numbers
representing the number of times each
student contributed in the variable called
“speakn” and the age group in the
variable called “grp”
– Each row of this data represents a
student, the number in the “speakn”
column is the amount they contributed
and the number in the “grp” column tells
us their age and year grouping.
– The middle column is just some text to
help you see which group is which, if you
go to variable view you will see the “grp”
variable labels similar to the ones
explained in the previous task
– Question: Do mature first year students
contribute more than young first year
students?
1. Open studentsss file in SPSS
1. Should not be normally distributed
1. Fill out yourself
4. Run t-test (Mann-Whitney)
1. Analyze -> Nonparametric Tests-> Legacy Dialoges->2
independent samples
2. Speakn goes in the Test Variable
3. Age goes in the Grouping Variable
1. Need to define groups (1=Year1 young, 2=Year1 mature)
4. Make sure the Mann-Whitney test is ticked (under Test Type)
1. See next page
5. Are the two groups of students different? If so, how?
Part B-2

value
• U=23.500
• DF=n/a
• P-value=0.007 (Exact)
• P-value < 0.05
• Significant
• Reject the null
4. Write result
• Mature students (group 2)
spoke significantly more than
young students (Mann-Whitney
Test, U=23.500, N=23, P-
value=0.007 with exact
significance).

Paired T-test
■ 1 categorical variable + 1 numerical
variable
– Categorical variable is paired and
group number is two
■ This is when you have a single numerical
variable you are interested in and want to
know if it is different between two groups,
but the groups have a integral relationship
(they are paired)
■ Often ‘before’ and ‘after’ type data
– Is heartrate different before and after
exercise
– Is the number of bear attacks in parks
lower after preventative measures
have been implemented?
– Does treatment reduce symptoms?
2. Boxplot of numerical variable grouped by
the categorical variable
2. Check normality
1. Parametric: Paired T-test
2. Non-parametric: Wilcoxon test
3. Write it out
Part B-2

Paired T-test: Examples
■ Parametric/Non-par Example: Student Steps
– data in this file come from an
experiment to see whether subjects
could perform more step exercises in a
fixed time in a group or on their own
– Paired data often occur in ‘before and
after’ situations. They are also known as
‘related samples’. These data are paired,
it’s the same person doing step
exercises under two different conditions.
– Question: Is there a difference in the
number of exercises completed in a fixed
time for students alone versus in a
group?
1. Open Step file in SPSS
1. Whether it is normally distributed or not, try both ways.
4. Run Both Parametric and Non-Parametric
1. Parametric: Analyze->Compare Means->Paired Samples T-test
1. Both group and individ goes in Test Pair list
2. Non-Parametric: Analyze-> Nonparametric Tests -> 2 related
samples
1. Both group and individ goes in Test Pair list
2. Make sure Wilcoxon is selected in Test Types
3. Also include Descriptive Statistics
1. See next page
5. Are the two groups different?
Part B-2

value
• t=3.503
• DF=11
• P-value=0.005
• P-value < 0.05
• Significant
• Reject the null
4. Write result
• Subjects had a tendency to
complete more steps under
group conditions than under
• individual conditions. (Paired
Samples T-test, t=3.503,
DF=11, p-value=0.005).
• N/A
Degrees of
Freedom
Test Statistic P-value

value
• Z=-2.631
• DF=n/a
• P-value=0.002
• P-value < 0.05
• Significant
• Reject the null
4. Write result
• Subjects had a tendency to
complete more steps under
group conditions than under
• individual conditions. (2-tailed
Wilcoxon signed ranks test,
Z=-2.631, n=24, p = 0.009).
• N/A

Correlation
■ 2 numerical variables
■ This is when you have two numerical variables and
you want to see if there is a relationship between
the two (positive, negative)
– Is there a relationship between drug
concentration and inflammation level?
– Is there a relationship between length and
weight in trout?
– Is there a relationship between the number of
guns and violent crime rates in a city?
■ Remember, correlation is not causation
1. Determine which variable is dependent and which is
independent
2. Histogram of dependent variable
3. Scatterplot (independent on x-axis, dependent on y-
axis)
2. Check normality
1. Histogram, QQ-plot, of dependent variable
1. Null: there is no relationship between the variables
2. Alternative: there is a relationship
1. Parametric: Pearson Correlation
2. Non-parametric: Spearman Rank-correlation
3. Write it out
Part B-2

Correlation: Examples
■ Parametric Example: Women Height
– file contains data from a student project
on the effect of heat on hip stretches.
– The first column gives the subject’s
height, and the second column gives 54
the increase in hip extension after
stretching exercises.
– (Other columns relate to the discomfort
experienced, and the stretch and
discomfort when heat is used; for our
purposes those are nuisance variables)
– This is paired data (measurements
taken under two conditions)
– Question: Is there a relationship
between the subject’s stretch increase
and height
1. Open Heathip file in SPSS
1. Determine which is dependent and which is independent
2. Whether it is normally distributed or not
3. Plot scatterplot (height on x-axis and stretch (without heat) on y-axis):
Graphs->Interactive Scatterplot
4. Run Appropriate test (Try both)
1. Parametric
1. Analyze->Correlate->Bivariate
2. Height and stretch go in Variable
3. Make sure Pearson is checked under Correlation Coefficients
4. Also check that Two-Tailed is set up and Flag significant correlations
2. Non-Parametric: same thing but check “Spearman” instead of Pearson
1. See next page
5. Is there a relationship? If so, what type (positive/negative) and how strong?
Part B-2

■ If a result is not significant, it is no
necessary to include the test
statistic and p-value
■ Should not graph results
value
• Pearson Correlation=-0.548
• N=10
• P-value=0.101
• P-value > 0.05
• Not Significant
• Failed to reject Null
4. Write result
• There was no correlation
between Height and Stretch
Increase in subjects.
• N/A

Correlation Notes
■ Looking for correlation is different from looking for increases or decreases
■ Correlation does not necessarily mean a causal relationship. Just because two
values appear to go up and down together does not mean one is causing the other.
■ The Pearson’s coefficient is designed primarily for looking at linear relationships.
Two variables can be related, but if the relationship is not linear, Pearson’s
correlation coefficient is not an appropriate statistic for measuring their association.
■ The number of observations as with other statistics effects the significance.
Part B-5

P-values a summary
■ "P-values do not simply provide you with a Yes or No
answer, they provide a sense of the strength of the
evidence against the null hypothesis.
■ The lower the p-value, the stronger the evidence.
■ Once you know how to read p-values, you can more
critically interpret journal articles, and decide for
yourself if you agree with the conclusions of the
author. " - TexaSoft, (1996-2001)

Chi-Square
■ 2 categorical variables
■ This is when you have two categories (simple case is that
the categories have two groups in each, but doesn’t need to
be the case)
■ End up getting frequencies of each category class and then
generation overall ratios
■ Tests for whether the categories are independent or not;
can be set up against many null frequencies
■ Non-parametric, so there isn’t a parametric/non-parametric
dichotomy
– The data are assumed to be a random sample.
– The expected frequencies for each category should
be at least 1.
– No more than 20% of the categories should have
expected frequencies of less than 5."
■ Examples:
– Is hair color independent of gender?
– Are the ratios of expected genetic crosses of pea
plants independent of the observed genetic
crosses?
– Is coffee type (caffeinated, decaf) independent of
mood (happy, sad)?
2. Check normality
1. Not applicable
1. Null: categories are independent
2. Alternative: categories are not independent
1. Chi-Square
3. Write it out
Part B-2

Chi Square: Examples
■ Parametric Example: Male/Female Ratio
– Does the ratio of males to females in
each school in SHU reflect the overall
ratio in the university? (or put another
way is there a larger than expected
number of one gender in some schools?)
– The data we have available are from a
survey of students done in 2001.
– You will see that the data is all numeric.
If you want to know what the numbers
represent you can look under the
Variable View to find out, but this isn't
necessary for our purpose. The crosstab
system automatically labels the output!
– Question: Does each school in the SHU
have male/female ratio that reflect the
overall ratio?
1. Open Students data 2001 file in SPSS
2. Explore data
1. Analyze->Descriptive Statistics->Crosstabs
2. Put Gender under Row(s)
3. Put School under Column(s)
4. Examine Crosstabulation table
1. Get the ‘expected’ values
2. Go to Crosstabs dialog box
3. Click Cells button then select ‘Expected’ under the counts section
4. Run Appropriate test
1. Analyze->Descriptive Statistics->Crosstabs
2. Click Statistics button, then select “Chi-Square”
1. See next page
Part B-2

value
• Chi-Square=635.561
• DF=8
• P-value<0.0001
• P-value < 0.05
• Significant
• Reject the null
4. Write result
• There is a significant
difference in the
representation of the sexes
across the schools (2-tailed
chi square test, chi-
sq=635.561, df=8, p-
value<0.0001).
• N/A

1-way ANOVA
– Categorical variable is un-paired and group
number is greater than two
different between multiple groups
– Is there a difference in grade point average
between Freshmen, Sophomores, Juniors,
and Seniors?
– Is there a difference in soil moisture
retention between 5 types of soil plot
designs?
– Is there a size difference between 4
different species of owls?
■ Important thing, the ANOVA test doesn’t tell you
which groups are different, only that there is a
statistical difference
■ Need to do a Post Hoc test to determine what the
difference is
2. Boxplot of numerical variable grouped by the
2. Check normality
1. Parametric: One-Way ANOVA
2. Non-parametric: Krustal Wallace Test
2. Post Hoc test
4. Write it out
Part B-2

One Way ANOVA: Examples
■ Parametric Example: Teaching Methods
– An experimenter is interested in evaluating
the effectiveness of three methods of
teaching a given course.
– A group of 24 subjects is available to the
experimenter
– This group is considered by the
experimenter to be the equivalent of a
random sample from the population of
interest.
– Three subgroups of eight subjects each are
formed at random; the subgroups are then
taught by one of the three methods. Upon
completion of the course, each of the
subgroups is given a common test (exam)
covering the material in the course
– Note: the Method is set up with numbers
(1,2,3) but is actually categorical
– Question: Is there a difference in scores
between the three methods?
1. Open anova one way example file in SPSS
2. Explore data and test for normality
1. Histogram of scores
2. Boxplots of scores by method
1. Analyze->Compare Means->One-Way ANOVA
2. Score in Dependent List
3. Method in Factor
1. See next page
4. If reject, null, run post hoc and determine the difference.
1. Go back to One Way ANOVA dialog box
2. Choose Post Hoc -> Tukey
Part B-2

• F=6.053
• DF=23
• P-value=0.008
• P-value < 0.05
• Significant
• Reject the Null
4. Write result
• There was a significant difference in
teaching methods (1-way ANOVA,
F=6.052, DF=23, p-value=0.008). Method
3 had the highest exam scores
• Barplot

Notes on ANOVA
■ Basically, ANOVA answers the question “Is there a significant difference between the samples
(is any one different from the others)?”
■ If there is not (Sig. >0.05) then there is no need to go any further
■ If there is then you might want to know which sample(s) is different from each other.
■ A supplementary (Post-hoc) test is carried out to investigate differences between the samples.
■ Selecting a post test is not simple; generally, to compare groups with each other choose the
Tukey test.)

One Way ANOVA: Examples
■ Non-Parametric Example: Teaching Methods
– The data are really three different sets of
scores, one set for each group, so when
we test them for normality, we need to
remember this, if we treat them as one
group then any differences between the
groups might lead us to thing that the
data aren’t normally distributed when
the data from each group is
– It is the normality of each group that
matters
– Question: Is there a difference in scores
between the three methods?
1. Open anova one way example file in SPSS (again)
2. Explore data and test for normality (the CORRECT way)
1. Normality: Analyze->Descriptive Statistics->Explorer
2. Put Score in Dependent list box, then click on the Plots button
3. Click to select Normality plots with tests (if p-value below 0.05 in any of the groups,
then go non-parametric)
4. Pretend that it was the case and try non-parametric (just less power)
3. Define null and alternative hypothesis to question (SAME AS BEFORE)
1. Analyze->Non-Parametric Tests->Legacy Dialogs -> K Independent Samples
2. Score in Test Variable List
3. Method in Grouping Variable (define groups to 3 using the Define Range button)
1. See next page
1. Go back to One Way ANOVA dialog box
2. Choose Post Hoc -> Tukey
Part B-2

■ Notice that the nonparametric test still says that there is a significant
difference between the groups (p=0.018) however it isn't quite as
well convinced as the more sensitive ANOVA. This is a good
illustration of the minor penalty that you pay for the more rugged 75
nonparametric tests, they are less likely to catch a small effect that
does exist, i.e. they are less powerful.
■ Run Post-Hoc test like before (Tukey)
■ So to recap; generally scores would be better treated by
nonparametric methods. In this example we did find them to be
normally distributed and used them as an example in applying a one
way ANOVA and its nonparametric equivalent, the Kruskal-Wallis test.
Finally, the two tests agreed but we noticed a slight difference in how
certain they were.
• DF=2
• P-value=0.018
• P-value < 0.05
• Significant
• Reject the Null
4. Write result
• There was a significant difference in
teaching methods (Kruskal Wallis
Test, Chi-Square=8.077, DF=2, p-
value=0.018). Method 3 had the
highest exam scores.
• Boxplot

Repeated Measures ANOVA
– Categorical variable is paired and group
number is greater than two
different between multiple groups that are paired,
usually the same subject
– Is there a difference in grade point average
between students their Freshman,
Sophomore, Junior, or Senior year?
– Is there a difference in soil moisture retention
1, 2, 3, 4, 5, or 6 years post treatment?
– Is there a difference in heartrate after 1, 2,
and 3 cups of coffee?
■ Repeated Measures ANOVA like extension of the
paired t-test, like 1-way ANOVA like extension of the
independent samples t-test
2. Boxplot of numerical variable grouped by the
2. Check normality
1. Histogram, QQ-plot, Test for Normality
2. Do these by group
1. Parametric: Repeated Measures ANOVA
2. Non-parametric: Friedman Test
2. Post Hoc test
4. Write it out
Part B-2

Repeated Measures ANOVA: Examples
■ Parametric Example: Jumping
– The following data shows the results of
an experiment where subjects jumped
three times.
– Each subject jumped three times, the
height was recorded, the column
labelled Jump1 has each subjects first
jump in it, the column labelled Jump2
has each subjects second jump in it and
so on.
– Question: Is there a difference in energy
between the three jumps?
1. Open Three Jumps file in SPSS
2. Put the three variables containing the energies (jump 1-3) in Dependent Box
3. Define null and alternative hypothesis to question
1. Analyze->General Linear Model->Repeated Measures
2. Repeated Measure; define factors dialog should appear -> put 3 in number of
levels, as there were three jumps, then click Add button
3. Click define, highlight the three jump variables and send them into the box with the
question marks in and click OK
1. See next page
1. Analyze->Compare Means->Paired-Samples T Test (no more than three levels)
Part B-2

■ Lots going on
■ Two are of importance; Mauchly's Test of Sphericity and
Tests of Within-Subjects Effects.
■ The first one, Mauchly's Test of Sphericity, tells us which
line of the second one to read (don't worry to much, just
whatever) if sig value greater than 0.05, use first line in
the Tests of Within-Subjects Effects,
■ If not, use Greenhouse-Geisser
value
• F=7.233
• DF=2
• P-value=0.002
• P-value < 0.05
• Significant
• Reject the Null
4. Write result
• There was a significant in
energy used between the three
jumps (Repeated Measures
ANOVA, Spericity Assumed,
F=7.233, DF=2, p-value=0.002).
• Barplot

Repeated Measures ANOVA: Examples
■ Non-Parametric Example: Walking
– One group of subjects walked in three
conditions; no crutches, elbow crutches
and axillary crutches
– the energy used was measured
(indirectly by looking at the oxygen used)
– Notice that there is no discriminatory
variable in this data set, this is because
there is only one group, each person
was measured in all three conditions.
– Question: Is there a difference in energy
used between the three methods?
1. Open enerjy and cruches file in SPSS
2. Put the three variables containing the energies in Dependent Box
3. Define null and alternative hypothesis to question
1. Analyze->Non-Parametric Tests->Legacy Dialogs->K-related samples
2. Put all the three variables in the Test Variable Box
1. See next page
1. NONE FOR NOW
Part B-2

value
• DF=2
• P-value=0.670
• P-value > 0.05
• Not Significant
• Fail to Reject the Null
4. Write result
• There was no significant
difference in energy used
between the three jumps
walking measures.
• N/A

Mixed Designs
■ Can combine things for more complicated analysis
■ We can think of the repeated measures ANOVA as an extension of the paired t-test,
and the One-way ANOVA as an extended version of the independent samples t-test
■ A mixed design is when we have for example two groups of subjects measured
repeatedly, e.g. two treatments, each treatment group being measured before and a
couple of times after treatment
■ Beyond the scope of this Presentation
■ Professional approach -> just use generalized linear mixed model with the
appropriate distribution
Part C-7

Assessment 6
1. A Non-parametric paired test was
performed one two different heart
rates of the same individual. Based
on the results to the right, was there a
significant difference between the two
heart measurements? If so, how?
2. The number of bears and the number
of marshmallows was counted at
various campsites. Is there a
correlation between bears and
marshmallows? If so, what is the
strength and direction?
3. Individuals from three different animal
groups (Sharks=1, Dinosaurs=2,
Bears=3) were rated for coolness. A
one-way ANOVA was performed, along
with a Post Hoc Test. Was there a
significant difference in coolness
across the three groups?

1. A Non-parametric paired test was
performed one two different heart rates
of the same individual. Based on the
results to the right, was there a
significant difference between the two
heart measurements? If so, how?
No significance.
2. The number of bears and the number of
marshmallows was counted at various
campsites. Is there a correlation
between bears and marshmallows? If
so, what is the strength and direction?
Significant correlation. Strong-Positive
3. Individuals from three different animal
groups (Sharks=1, Dinosaurs=2,
Bears=3) were rated for coolness. A
one-way ANOVA was performed, along
with a Post Hoc Test. Was there a
significant difference in coolness across
the three groups?
Significant difference.
Assessment 6
Answers

Part D: Other Analyses
■ Reliability and Sensitivity
– Reliability: how consistent measurements are
■ Inter-rater reliability deals with the issue of reliability between different people
(raters).
■ Intra-rater reliability deals with whether one rater is consistent, i.e. when they re-
look at the same subjects do they rate them in a similar way again.
– Sensitivity: how much measurements can detect true effects
Part D-0

Reliability: Examples
■ Imagine that a student wants to find out if a
certain exercise can improve performance.
■ To measure performance they decide to use a
simple measured jump. However to be sure
that he can sensibly repeat the measures
after the exercise regime has been completed
he wants to estimate the reliability of his
measurement method.
■ To get round the problems of (XYZ), use
Intraclass Correlation Coefficient
■ The coefficient will tell us how much
agreement the two measurements have
■ Can also use Cronbach’s Alpha, another
measure of reliability (Note that a reliability
coefficient of .70 or higher is considered
"acceptable" in most Social Science research
situations using Cronbach's Alpha)
■ Alpha also works for more than 2 measures
1. Open ICC and Cronbachs alphs file in SPSS
2. Calculate the ICC
1. Analyze->Scale->Reliability Analysis
2. Put Jump 1 and 2 into the Items box and click
statistics
3. Tick the Intraclass Correlation Coefficient
3. Calculate the Alpha
1. Done at the same time by the analysis
Part B-2

■ The Intraclass Correlation Coefficient (ICC) in this case is
0.962 we use the single measures because the figures we
fed SPSS were raw measurements not an average of
several attempts. This value, 0.962 shows a considerable
amount of agreement!
value
• Coefficient=55.606
• DF1=5
• DF2=5
• P-value<0.0001
• P-value < 0.05
• Significant
• Reject the Null
4. Write result
• There was a significant
correlation between jump
measurements (ICC=55.606,
DF=5,5, p-value<0.0001
• N/A

Inter-rater agreement using Kappa
■ Do two measurement agree, or two
measurers (raters)?
■ Between any two raters, even if they
just guessed, there would have still
been a degree of agreement just due
to chance
■ The Kappa statistic takes this into
account
Part D-3
1. Open radiologist eg from p403 file in SPSS
2. Analyze->Descriptives->Crosstabs
3. Put Radiologist 1 in Rows
4. Put Radiologist 2 in Columns
5. Click statistics and choose Kappa

value
• Kappa=0.473
• N=85
• P-value<0.0001
• P-value < 0.05
• Significant
• Reject the Null
4. Write result
• There was moderate agreement
between raters (Kappa=0.473,
N=85, p-value<0.0001).
• N/A

Calculating sensitivity and specificity of
a diagnostic test
■ The table below is a 2x2 cross tabulation (contingency table) representing the
findings of a diagnostic test when compared to the actual disease state. I.e. a
comparison of what the test indicated and the real facts.
■ The four cells TP, FP, FN & TN would have in them the number in each category, they
will total the number of cases investigated.
■ The crosstabulate command in SPSS or the pivot table feature in MS Excel can
calculate the matrix values.
Part D-4

Calculating sensitivity and specificity of
a diagnostic test 2
■ Sensitivity=TP/(TP+FN)
■ Specificity=TN/(FP+TN)
■ Prevalence=(TP+FN)/(TP+FN+FP+TN)
■ Positive Predictive Value =TP/(TP+FP)
■ Negative Predictive Value =TN/(FN+TN)
■ Positive Likelihood=SENS/(1-SPEC)
■ Negative Likelihood=(1-SENS)/SPEC
■ Overall Accuracy = (TP + TN)/(TP + FP +
FN + TN)
Part D-4

Example
■ Sensitivity = TP/(TP+FN) = 45/(45+5) = 45/50 = 0.9
■ Specificity = TN/(FP+TN) = 35/(15+35) = 35/50 = 0.7

Odds Ratio
■ Information and examples based on Explaining Odds Ratio (by Magdalena
Szumilas)
■ Odds Ratio (OR): Measure of association between exposure and outcome
■ Odds that an outcome will occur given a particular exposure, compared to absence
of that exposure
■ Uses: odds of disease/disorder given certain exposure (health characteristics,
medical history, treatment, environmental factor, etc.)
– OR=1 (exposure does no affect odds of outcome)
– OR>1 (exposure associated with higher odds of outcome)
– OR<1 (exposure associated with higher odds of outcome)

Odds Ratio Calculation
■ Odds Ratio (OR) =
■ OR = a/c / b/d = ad /bc
a=Number of exposed cases
b=number of exposed non-cases
c=number of unexposed cases
d=number of unexposed non-cases

Odds Ratio Example
■ Can do by hand
■ In the study, 186 of the 263 adolescents
previously judged as having experienced
a suicidal behavior requiring immediate
psychiatric consultation did not exhibit
suicidal behavior (non-suicidal, NS) at six
months follow-up. Of this group, 86 young
people had been assessed as having
depression at baseline. Of the 77 young
people with persistent suicidal behavior
at follow-up (suicidal behavior, SB), 45
had been assessed as having depression
at baseline.
■ What is the OR of suicidal behavior at six
months follow-up given presence of
depression at baseline?
■ Determine the following numbers
a: Number of exposed cases (+ +) = ?
b: Number of exposed non-cases (+ –) = ?
c: Number of unexposed cases (– +) = ?
d: Number of unexposed non-cases (– –) = ?

Odds Ratio Example Cont.
Q1: Who are the exposed cases (++ = a)?
A1: Youth with persistent SB assessed as having
depression at baseline
a=45
Q2: Who are the exposed non-cases (+ – = b)?
A2: Youth with no SB at follow-up assessed as having
depression at baseline b=86
Q3: Who are the unexposed cases (– + = c)?
A3: Youth with persistent SB not assessed as having
c: 77(SB) –45(depression) = 32
Q4: Who are the unexposed non-cases (– – = d)?
A4: Youth with no SB at follow-up not assessed as having
d: 186(NS) –86(depression) = 100
Odds Ratio (OR) = a/c / b/d = ad/bc
= 45*100 / 86*32 = 1.63
Odds of persistent suicidal behavior is
1.63, higher given baseline depression
diagnosis compared to no baseline
depression

Assessment 7
1. Two scientists observed the number of
cups coffee I drank each day over a two-
week period. A Kappa statistic (table to
the right) was calculated to see how well
their observations agreed with each other.
Was the agreement significant? If so,
what was the strength of the agreement.
2. On the right is a contingency table for the
test of a certain disease (fear of empty
coffee cups). What is the Sensitivity and
Specificity of the test?
3. What is the odds ratio for the following
information:
Number of exposed cases= 45
Number of exposed non-cases= 66
Number of unexposed cases= 87
Number of unexposed non-cases= 92
Actual Disease State
Test for Disease Disease No Disease
Positive Test TP: 37 FP: 3
Negative Test FN: 13 TN: 18

1. Two scientists observed the number of cups coffee I
drank each day over a two-week period. A Kappa
statistic (table to the right) was calculated to see how
well their observations agreed with each other. Was
the agreement significant? If so, what was the
strength of the agreement.
Yes, moderate strength
2. On the right is a contingency table for the test of a
certain disease (fear of empty coffee cups). What is
the Sensitivity and Specificity of the test?
Sensitivity = TP/(TP+FN) = 37/(37+13) = 0.74
Specificity= TN/(FP+TN) = 18/(3+18) = 0.857
1. What is the odds ratio for the following information:
Number of exposed cases= 45
Number of exposed non-cases= 66
Number of unexposed cases= 87
Number of unexposed non-cases= 92
OR= a/b / c/d = ad /bc = 45*87 / 66*92 = 0.645
Actual Disease State
Test for Disease Disease No Disease
Positive Test TP: 37 FP: 3
Negative Test FN: 13 TN: 18

Summary
■ SPSS is a very useful tool in analyzing data that does not depend on lots of computer
coding background
■ Analyzing data helps answer important questions we have about science, medicine,
business, society, etc.
■ The three major steps are:
– Explore Your Data
– Analyze Your Data
– Interpret your results
■ For more information on SPSS, start by checking out the resources below
– https://www.ibm.com/products/spss-statistics/resources
– https://www.spss-tutorials.com/

data_analysis_using_spss.pdf

More Related Content

What's hot (20)

Similar to data_analysis_using_spss.pdf (20)

Recently uploaded (20)

data_analysis_using_spss.pdf