Week 2 - Assignment 1: Organize Data

Fila64

building-spss-graphs-to-understand-data.pdf

Home >Business & Finance homework help >Management homework help >Week 2 - Assignment 1: Organize Data

Introduction: Graphical Presentation of

Data

In: Building SPSS Graphs to Understand Data

By: James O. Aldrich & Hilda M. Rodríguez

Pub. Date: 2012

Access Date: June 10, 2022

Publishing Company: SAGE Publications, Inc.

City: Thousand Oaks

Print ISBN: 9781452216843

Online ISBN: 9781452270036

DOI: https://dx.doi.org/10.4135/9781452270036

Print pages: 1-18

This PDF has been generated from SAGE Research Methods. Please note that the pagination of the

online version will vary from the pagination of the print book.

https://dx.<wbr>doi.<wbr>org/10.4135/9781452270036

Introduction: Graphical Presentation of Data

1.1 Purpose and Intent of Graphs

The principal purpose of a graph is to answer questions about data. Answers to the questions may then

provide direction for additional analysis or convey information about the data to colleagues or students. The

old adage “a picture is worth a thousand words” is especially applicable when working with large amounts of

data. A graph summarizes a mass of data and is an excellent way to “make visible” answers to questions that

may be hidden in the data. We may even say that graphs give “life” to the data.

The purpose of all statistical procedures is to make data understandable. Properly constructed graphs can

do this in a very direct and straightforward way. In this book, each chapter presents questions concerning

variables from many different databases. The proper SPSS graph is selected, built, and interpreted in a

manner that answers those questions. Following this process you have made the data more understandable.

Whether you are reading a data report at work or conducting a cancer research project, it is a good idea

to be able to interpret large amounts of numerical data. The interpretation of data can mean different things

to different people. For this book, “interpretation of data” means that anyone reading the data is able to

recognize and understand the information contained in the data. This information is then used to answer

questions. Answering such questions will then assist one in making a decision. One way to extract this

information and make the data usable is to build a graph. A well-built graph will provide a clear visual

representation of the data.

Numbers by themselves have meaning. But when they are used to measure and/or count something, they

have more meaning. Throughout this book, the “meaning” may be thought of as information. The ultimate

purpose of a graph is to make a “bunch of numbers” (the data) understandable. One way the graph

accomplishes this is to extract information that is often hidden in the data. This information is then used to

assist one in making a decision. Making an informed (educated) decision involves answering questions.

For example, a new professor designs her first statistics course so that her 90 students can each earn

between 0 and 500 total points. At the end of her first semester of teaching, she questions the amount of

material she attempted to cover in the class lectures. She wonders if she attempted to cover too much or

perhaps too little material. One way to answer her question would be to analyze the data and determine the

proportion of students scoring in various intervals that make up 0 to 500 points.

Let's say that she produces a graph known as a cumulative histogram. The graph reveals that 70 (78%) of the

students earned between 200 and 240 total points. In other words, the graph revealed that more than three

fourths of her students earned less than half of the possible points. Such a finding provides evidence that she

might consider reducing the amount of material for the second semester of her teaching career. Perhaps, she

SAGE

SAGE Research Methods

Page 2 of 17 Introduction: Graphical Presentation of Data

might make a decision to present less material and to slow down, just a bit, with the intention that the students

will increase their learning and therefore earn more total points.

The professor could have relied on basic descriptive statistical analysis to “interpret” or to extract this

information from the data. As there are many statistical approaches designed to make data more

understandable, for instance, she may have elected to organize the data into a table, determine percentile

ranks, and calculate the mean and standard deviation.

This book is about graphs and how to use the graphing features found in the Statistical Package for the Social

Sciences (SPSS) to make sense of data. This book is not about building tables or the calculation of numerous

descriptive statistics. It is about the how and when to produce the appropriate graphical representation as a

way to better understand data and answer pertinent questions. The actual “production” of the graph will be

done by utilizing the power of the SPSS software. Much of this book is devoted to step-by-step procedures

needed to build SPSS graphs. You are introduced to this aspect of our book in Section 1.5 of this chapter.

1.2 Proportional Thinking and Graph Making

It should be obvious that data are usually collected with some purpose in mind. The purpose is that the data

will possess the information needed to answer specific questions that will assist us in making a decision.

Thus, one should always have a clear idea about the questions that we seek to answer by our graph making.

Often such questions revolve around the idea of percentages, proportions, and probability.1 Let's use an

example that places an emphasis on this aspect of a database. A young woman is trying to decide if she

wants to attend a campus party. She hopes to meet a nice young man as a potential boyfriend. A basic

question for this young lady is as follows: Given the gender makeup of the party, what is the probability of

meeting a potential boyfriend?

She is able to look at the list of 100 names of those attending. She wants to use these data (the 100 names)

to help decide whether to attend the party. Just looking at the list of names gives her some information about

the proportion of males that will be at the party. She may have an intuitive feeling about the party's gender

makeup but little understandable information.

Since the data are discrete (having the two categories of male and female), we could produce a bar graph to

give us a clear picture of the proportion of males and females. Most would agree that the proportion of males

attending the party would be something for this young woman to consider when deciding whether to go or not.

For instance, the bar graph might show that there are 80 females and 20 males on the party list; therefore,

there is a .80 probability that any encountered person would be female. We could also say that 80% of those

people encountered at this party would be female. We could all agree that such information would be useful

when one considers her purpose in attending the party. If her purpose is to meet a young man as a potential

boyfriend, then perhaps, she should forget the party, stay at home, and study statistics.

SAGE

SAGE Research Methods

Page 3 of 17 Introduction: Graphical Presentation of Data

Being able to visualize proportions (hence probabilities) as demonstrated in the party example represents

only one of many graphing purposes. Another purpose for a graph might be that we wish to determine if a

distribution of continuous values are approximately normally distributed.2 Or perhaps, we wish to see how

two variables change together. We could also determine if the values of a variable change in some systematic

manner over a period of time. These are only a few of the many uses of graphs. Later chapters will inform the

reader as to the purpose of the graph and the type of graph required to answer the question at hand.

Proportional-type thinking is also related to inferential statistical concepts. When proportions are calculated

on samples that were obtained through random techniques, another dimension is opened. In looking at the

above party example, that new dimension is whether similar parties would be expected to have approximately

the same proportion of females and males. If proper research practices were followed, then we can use the

sample proportions to estimate the likelihood that those same proportions would exist at a similar party.

There are several considerations for the graph maker when deciding the best way to use the power of SPSS

to answer questions about a database. This is the topic of the following section.

1.3 Decisions for the Graph Maker

What is the best approach when seeking to extract useful information from a large mass of data? What type

of graph should be built to answer specific questions about the data? How can the extracted information be

used to assist one in making decisions? What is the best way to organize data so as to answer specific

questions? These are the types of questions about data answered in this section. Since this book uses SPSS,

the first step in understanding data is to make sure that its format is compatible with SPSS.3 Once this is

accomplished, we can let SPSS do the difficult and time-consuming work involved in producing a graph. Once

SPSS has been selected as the software, there are other considerations before one attempts to produce a

specific graph.

1.3.1 First Consideration: How Were the Variables Measured?

The first consideration when attempting to graph data is whether the measurement of the variable was

discrete or continuous. When using SPSS, you will not see data classified as discrete or continuous, but

you will see a column heading called “Measure” in the “Variable View” Screen. Measure is short for “level

of measurement.” It is here that the person entering the data indicates how the data were measured. If

you are entering the data into the SPSS software yourself, you must understand the meaning of “level of

measurement” as it relates to the numerical values associated with your observations. The data used in this

textbook already have these levels of measurement specified, so this point becomes academic.4 When using

this text, you will only have to decide if the data are discrete or continuous. The SPSS databases used in this

text will have already specified the data as being measured at one of three levels of measurement (scale,

SAGE

SAGE Research Methods

Page 4 of 17 Introduction: Graphical Presentation of Data

ordinal, and nominal).5 Those variables measured at the scale level are considered as continuous data.

Variables specified as being measured at the nominal or ordinal level should be considered as discrete.

Variables measured at the continuous level are not based on distinct categories. Such variables can be

recognized when an infinite number of points between each interval are possible. To understand the concept

of infinite points, think of the variable of weight. If it was observed that an individual weighs 165 pounds,

we can see that there are an infinite number of points between 165 and say 166. One could weigh 165.25

pounds or a more sensitive scale could record 165.2567 pounds and so on. One of the common methods

used to understand a large database of continuous variables is to construct a histogram. SPSS constructs

intervals for the continuous data and records these on the x-axis. Next it counts the number of observations

within these intervals, which are read from the y-axis. Such graphs are distinct from the bar graph in that the

intervals are connected on the x-axis and are not separate categories as with the bar graph.

Discrete variables are made up of separate and distinct categories. An example of a discrete variable is

gender (nominal), where all observations must fall into one of two unique categories, male or female. Another

example could be responses on a survey where one has several choices such as military ranks of general,

colonel, captain, sergeant, and private. In this example, the respondent must fall in one of the five categories,

which measure the level (amount) of military rank (ordinal). At this point, it is instructive to think in terms of

a bar or pie graph when attempting to understand the distributions of discrete variables. With discrete data,

each distinct bar or pie slice represents a category. Each category is represented either by a rectangular box

or a slice of the pie. It must be mentioned that it is sometimes desirable to use a bar graph to represent

continuous data. SPSS accomplishes this by organizing the continuous data into separate categories by

specifying ranges. A bar graph can also be used to compare the means (continuous data) of a number of

groups. An example would be to compare two university classes on the mean number of points earned on an

exam. In spite of these exceptions, the primary use of the bar graph is to represent discrete data.

In many of the later chapters, you will learn to build graphs that use a combination of discrete and continuous

data. At this time, we do not attempt to describe all the various combinations of discrete and continuous data

and the appropriate graphs for these situations. You will learn these procedures when we describe the types

of graphs and how they are used. Additional guidance in choosing the correct graph for various levels of

measurement and to provide answers for specific questions is provided in Appendix D (Graph Selection: Type

and Purpose).

To summarize this section: When one wishes to make a graph to discover information contained in a mass of

data, it is necessary to first establish whether the data are discrete or continuous.

1.3.2 Second Consideration: What Are We Trying to Understand About

the Data? What Is the Question? What Information Must the Graph

Communicate?

SAGE

SAGE Research Methods

Page 5 of 17 Introduction: Graphical Presentation of Data

When confronted with a large amount of data, we might be seeking information about a single variable. In this

case, we are not concerned with how the values of one variable may relate to the values of other variables. In

other words, we are concerned with what the distribution of values “looks like.” If we have a discrete variable,

such as the previously discussed gender variable, then “looks like” can refer to the number (or percentages)

of males and females in the database. A bar graph would encourage thinking in a proportional way by drawing

a graph, making it easy to visualize the relationship between the two categories for the single variable of

gender. Another method might be to construct a pie graph. The pie graph is especially useful with more than

two categories in that it makes it easier to see the proportions between the distinct categories and the whole.

We will address such issues in greater detail when they are discussed in later chapters that describe how to

produce such graphs.

If we are investigating a single continuous variable, we might choose to build a graph known as a histogram

to see what the data “looks like.” The histogram could provide evidence as to whether the data approximates

the normal distribution.6 Such graphing of continuous data could also discover extreme values (outliers) that

should be dealt with before additional statistical analysis is undertaken.

If the question we are seeking to answer involves two or more variables, we need a graph that depicts their

relationship. With two variables, we need to display a bivariate relationship.7 A scatter plot–type graph will

enable the graph maker to display the relationship between two variables. In this application, the x-axis and

y-axis both represent a value for each variable. Another example of graphing two variables would be putting

the results of a 2 × 2 crosstab analysis into a stacked bar chart, making it easier, for most people, to visualize

the proportions.

Another question might be the interpretation of the results of a multivariate analysis, where it is necessary

to graph three or more variables. You will soon learn how to build a graph representing, say, two continuous

and single discrete variables. Another use would be to use a line graph to display an interaction or lack of

interaction between two independent variables.

Later chapters discuss all the above graphing possibilities (and more) with detailed guidelines as to which

graphs answer specific questions about the data.

1.3.3 Third Consideration: Who Will Use the Information and What Is

Their Level of Statistical Knowledge?

Another consideration would be the type of audience. By this we mean that graphs prepared for a meeting of

members of the American Statistical Association might take on a different appearance than graphs prepared

for an introductory course in statistics. Whoever the audience, graphs should always be clear and concise

representations of the data and should be readily understandable.

SAGE

SAGE Research Methods

Page 6 of 17 Introduction: Graphical Presentation of Data

1.3.4 Fourth Consideration: Is the Graph Intended to Guide Further

Research or Will It Be Used for Presentation to Others?

If the intention of the graph builder is to use the graph to guide students or other users in the selection

of further statistical approaches, then the use of the SPSS Chart Editor is not required. The basic graph

produced by SPSS will usually provide the information required to answer their questions. On the other hand,

a graph intended as part of a larger presentation to students and/or colleagues would require one to “edit” the

graph using the SPSS Chart Editor. The editing process, which uses the Chart Editor, is covered in detail in

the later chapters.

1.4 Databases Used in This Book

In all cases, the databases you will use to build the graphs are found in the SPSS Sample file that was

downloaded with the software. In general, each database selected for use is briefly described when it is

introduced. However additional information on specific databases is available by using the SPSS Help menu

as follows:

• Open SPSS.

• Click Help, then click Topics, and type sample files.

• Click list topics, then click sample files.

• Click Display.

The data used to illustrate various graphs at the beginning of each chapter were obtained from the authors’

years of teaching intermediate-level statistics classes. You are not required to have access to these data.

1.4.1 Access to Databases Used to Build Graphs

Before we can examine the major features of the Chart Builder in the next section, we must open a database.

The procedure shown below will open one of the databases (accidents.sav) that was included with the SPSS

software. You will use many of the databases that are found in the SPSS Sample files location. You should

make a note to mark this page since you will use these procedures in almost every chapter.8

• Start SPSS if it is not already running.

• On the SPSS Menu, click File, select Open, and then click Data.

• In the Open data window, click the down arrow in the Look in box and scroll to and click the C drive.

• Double click Program Files.

• Scroll to SPSSInc, then double click SPSSInc.

• Double click PASW Statistics 18 (currently Version 20 shows 18, and you may see other version

numbers as well).

SAGE

SAGE Research Methods

Page 7 of 17 Introduction: Graphical Presentation of Data

• Double click Samples.

• Double click English.

• A window opens in which you will see a long list of available files.

• Click accidents.sav.

• Click Open (Data Editor opens). We now have the required database and may open the Chart

Builder.

• Click Graphs, and then click Chart Builder.

• Chart Builder window opens as shown in Figure 1.2. Make sure that Gallery is selected—if not, click

Gallery.

Following the above procedures can sometimes be tricky. It may become necessary to exercise some

patience, but we assure you that the procedure is the correct way to access SPSS Sample files. Be sure to

double click on the files as specified.

1.5 SPSS Graph Production: Using the Chart Builder

The Chart Builder feature available in SPSS plays a major role in this book. Detailed procedures on building

simple to complex graphs while using the Chart Builder are provided throughout this textbook. The purpose

of this section is to provide you with a basic understanding about how to use the SPSS Chart Builder. Don't

worry, you won't have to memorize everything in this section, as you are given step-by-step procedures when

building graphs in the later chapters. We do, however, highly recommend that you take this time to become

familiar with the Chart Builder. It won't be long before you will easily produce graphs to better understand

large amounts of data. Let's proceed with the assumption that you have never used the SPSS Chart Builder

feature.

When you click Graphs on the SPSS Main Menu, you are given three options for building a graph: (1)

Legacy Dialogs, (2) Graphboard Template Chooser, and (3) Chart Builder. The Legacy Dialogs option is one

of the first methods developed by SPSS, some people feel comfortable in using this earlier method. There

is nothing wrong with the Legacy approach; however, it does take more statistical knowledge (and time) to

use it correctly. The next option Graphboard Template Chooser came along a few years later and offered

more automatic steps in building a graph. The final option Chart Builder is the most recent version, and in

the opinion of the authors, it is the most efficient. Therefore, we have chosen the Chart Builder as the option

described in this book. Therefore, all step-by-step instructions in this book begin and end with the Chart

Builder.

Chart Builder is the first available option when clicking Graphs on the SPSS Main Menu (Figure 1.1). This is

the option used in all chapters of this textbook.

SAGE

SAGE Research Methods

Page 8 of 17 Introduction: Graphical Presentation of Data

Figure 1.1 SPSS Main Menu Showing How to Open the Chart Builder

The following section provides details on the Chart Builder window that opens once you click on Chart Builder

in the drop-down menu as shown in Figure 1.1.

1.5.1 A Description of the Chart Builder Window

Figure 1.2 shows the basic Chart Builder window that opens once Chart Builder is clicked. If the window you

see is slightly different from the one shown in Figure 1.2, then look at the horizontal row of buttons in the

middle of the window to make sure that Gallery is highlighted, which means it is active and does not need to

be clicked. If one of the other buttons (Basic Elements, Groups/Point ID, and Titles/Footnotes) is highlighted,

the window's appearance will not match the one shown in Figure 1.2, and you must then click the Gallery

button.

Let's go through some of the features and options available when you open the Chart Builder window. At this

time, our intention is to only familiarize you with the major features available in the Chart Builder window. You

will learn to build graphs, using the Chart Builder window, starting with the next chapter.

SAGE

SAGE Research Methods

Page 9 of 17 Introduction: Graphical Presentation of Data

Figure 1.2 Chart Builder Window Showing Key Features and Control Points

1.5.2 Panels in the Chart Builder Window

There are three major panels in the Chart Builder window titled as follows: (1) Variables, (2) Chart preview

uses example data, and (3) Choose from. Let's examine each of these panels.

The “Variables:” Panel: The Variables panel provides a list of all the variables in the opened

database. The level of measurement is also specified by the icon that precedes the variable's

name. The first icon shown in the Variables panel is for Age category, and it represents data

measured at the ordinal level. Therefore, if you select this variable for graphing you will only

use graphing procedures appropriate for discrete data. The next variable, Gender, is represented

by an icon resembling three balloons that denotes nominal data. Once again, for graphing

procedures, one must consider this as discrete data. The third variable in this panel, Accidents,

is preceded by an icon looking like a small ruler. A variable shown with this SPSS icon indicates

that it was measured at the scale level (interval and ratio). Data measured in this manner and

accompanied by this icon are treated as continuous data for graphing purposes. We summarize

our icon descriptions as follows:

• Icons for discrete variables are and

SAGE

SAGE Research Methods

Page 10 of 17 Introduction: Graphical Presentation of Data

• Icon for a continuous variable is

The lower portion of the Variables: panel also provides a brief description of the categories for

discrete (nominal and ordinal) data. You obtain these descriptions by clicking on each variable.

This feature can be very useful when deciding on variables to include in your graph-building

projects.

Figure 1.2 shows that the accidents.sav database has four variables available for graphing, two

discrete (Age category and Gender) and two continuous (Accidents and Population at risk).

The “Chart preview uses example data” Panel: SPSS uses the words “uses example data”

to emphasize that the graph shown in the panel, during the graph-building process, does not

represent your data. It is only a very rough approximation of what your finished graph will look

like. In later chapters, we shorten this panel's name to “Chart preview.”

It is in this panel that much of your basic chart building takes place. At this time, we do not attempt

to show the many different appearances of this panel. Basically, this panel provides a template for

various types and styles of graphs that are selected from the third panel (“Choose from:” panel).

Briefly, the procedure is to select a graph and style from the Choose from panel, drag it to the

Chart preview panel, then drag appropriate variables onto the shown graph. The procedures for

using this panel, as they relate to specific graph-building projects, are detailed in all the later

chapters.

The “Choose from:” Panel: Direct your attention to the Choose from: panel in Figure 1.2. Notice

that “Bar” has been highlighted; this was accomplished by clicking on “Bar.” When the word “Bar”

is clicked, a specific set of bar styles appears to the right. If a different graph type had been clicked

(highlighted), then you would see different set of graph styles to choose from. Clicking the Bar

graph gives you eight style choices. Notice that some are hidden by our informational balloon. It

is from these eight different styles that you choose one and then click and drag it to the Chart

preview panel directly above. Later chapters will show styles appropriate for answering specific

questions about a particular database. Once the style has been moved to the Chart preview

panel, you will then drag variables onto various places on the graph to continue the graph-building

process.

1.5.3 Element Properties Window

The Element Properties button is located in the center-right portion of the window shown in Figure 1.2.

Clicking this button opens the Elements Properties window (Figure 1.3), which provides many ways to make

changes during the graph-building process. The use of the features accessed through the Element Properties

window will be illustrated as the need arises. In most cases, this window opens automatically when performing

operations in the Chart Builder window. If it does not open automatically, then clicking the Element Properties

button in the Chart Builder window offers an alternative way to gain access to its features.

SAGE

SAGE Research Methods

Page 11 of 17 Introduction: Graphical Presentation of Data

Once you have made changes to your graph from the Element Properties window, remember to click the

Apply button as shown in Figure 1.3. Clicking the Apply button takes you back to the Chart preview panel

that incorporates your changes.

Figure 1.3 Element Properties Window

1.5.4 Final Stages When Using the Chart Builder

In the row of buttons at the bottom of the Chart Builder window, you will most frequently use OK and Reset.

Clicking OK causes all previous instructions to be incorporated into the graph and immediately produces the

finished basic graph that appears in the Output Viewer of SPSS. If the Output Viewer does not appear on the

screen after clicking OK, then click Output found on the Task bar at the bottom of your computer screen.

Once your graph appears in the Output Viewer, there is an opportunity to further modify this basic graph. We

refer to this graph as “basic” since there are many ways it can be embellished and improved by using the

Chart Editor as described in the following section.

Before leaving the Chart Builder, we describe the use of the Reset button found at the bottom of the Chart

Builder window. This handy feature clears all prior work from the Chart preview panel (Figure 1.2). We

SAGE

SAGE Research Methods

Page 12 of 17 Introduction: Graphical Presentation of Data

recommend this, unless you are building another graph of the same type and style as the one you may have

just built.

1.5.5 Chart Editor and Properties Window

The Chart Editor window is key when building clear and concise graphs that effectively communicate the

information needed to understand the data. The Chart Editor window (see Figure 1.4) opens when you double

click the basic graph in the Output Viewer.

In this section, our intention is to briefly introduce you to some of the editing features made possible by using

the Chart Editor. Specific instructions on how to use the Chart Editor are provided in later graph-building

exercises. You will utilize features of the Chart Editor and its associated Properties window (see Figure 1.5)

to make changes that you see immediately as you continue with your graph-building project.

Figure 1.4 Chart Editor Window With Some Useful Points Tagged

Figure 1.4 shows the Chart Editor window with a few of the key features and control points emphasized with

information bubbles. As before, our intention here is to only familiarize you with the Chart Editor window. You

will learn much more about its use as you progress through various graph-building procedures.

SAGE

SAGE Research Methods

Page 13 of 17 Introduction: Graphical Presentation of Data

Figure 1.5 Properties Window Showing the Categories Tab Clicked

Figure 1.5 shows the Properties window (not to be confused with the Element Properties window shown

in Figure 1.3). Normally the Properties window opens with the Chart Editor. If it does not, then click the

Properties icon as shown in Figure 1.4. Depending on the particular graph-building project and the request

you make by clicking items in the Chart Editor, the Properties window takes on many different appearances.

You will be directed on a case-by-case basis to click various tabs, which then provide access to the features

needed to make the desired changes.

At this point, you may feel a little overwhelmed by the many details associated with graph building. However,

we assure you that after making a few graphs as described in the following chapters, the process becomes

very natural and intuitive—SPSS has done an excellent job in this regard.

1.6 Definition of Key Terms and Concepts

Variable—A characteristic that describes some physical or mental aspect of the individual, group, or

SAGE

SAGE Research Methods

Page 14 of 17 Introduction: Graphical Presentation of Data

inanimate object. The key point is that variables can vary and can be expressed as a particular numerical

value or as falling in a unique category. The following are some examples:

• Gender as the observed variable could vary in two ways: male or female (discrete).

• Mental state could be a variable and vary in five ways: very anxious, anxious, neutral, relaxed, and

very relaxed (discrete).

• Number of errors on a written test can be considered as a variable. If there were 10 questions, then

such a variable could vary from 0 to 10 errors (discrete).

• Make of automobile owned or leased can be considered as a variable (discrete).

• Weight of an individual can be considered as a variable (continuous).

• Speed of an airplane can be considered as a variable (continuous).

• How high one can jump can be considered as a variable (continuous).

Discrete Variable—A characteristic that describes some physical or mental aspect of the individual, group,

or inanimate object that has been observed. The term discrete describes how the variable is measured

or counted. Discrete variables vary in a manner so that the characteristics being measured fall in unique

categories. Such categories must be mutually exclusive, which means that any observation must fall in one

and only one category. The categories must also be inclusive, which means that there must be a category

for every possible observation. Examples of discrete variables from the list above are gender, mental state,

number of errors, and make of automobile.

Continuous Variable—A characteristic that describes some physical or mental aspect of the individual,

group, or inanimate object that has been observed. The term continuous describes how the variable is

measured. Continuous variables vary by taking on any one of a large number of measures (often infinite).

Examples of continuous variables from the list above are weight, speed, and how high one can jump.

Category—A natural grouping of the characteristics of a discrete variable. Think of the type of automobile as

a variable having categories such as Porsche, Ferrari, and Maserati. The automobile possessing “Porsche”

characteristics is counted or placed in a category titled Porsche.

Continuous Distribution—Plainly speaking, a continuous distribution is just a “bunch of numbers” that

resulted when something was measured at the continuous level. We use various types of statistical analysis,

including graphs, to make sense of such distributions. Histogram and line graphs are the most commonly

used graphing techniques to describe continuous distributions.

Discrete Distribution—Could also consist of a bunch of numbers, but the numbers take on a different

meaning. Using the gender variable as an example, we could assign the value label of 1 to the male category

and 2 to the female category. This being the case, you now have a column of numbers consisting of 1s and

2s that you are trying to understand. You could use SPSS to produce a bar graph.

Levels of Measurement—A classification method used to define how variables are measured. For most

statistical work, there are four possible ways to measure variables: (1) nominal, (2) ordinal, (3) interval, and

SAGE

SAGE Research Methods

Page 15 of 17 Introduction: Graphical Presentation of Data

(4) ratio. When using SPSS, you will find that interval and ratio have been combined and are referred to

as scale. For purposes of graph making, only two methods of measurement need to be distinguished: (1)

continuous, which includes interval and ratio (SPSS's scale) and (2) discrete, which includes SPSS's nominal

and ordinal levels of measurement.

Independent Variable—The independent variable is manipulated and has the freedom to take on different

values. It is the presumed cause of change in the dependent variable in experimental work. In observational-

type studies, it is often referred to as the predictor variable. This definition hinges on the idea that knowledge

of the predictor variable will facilitate the successful estimation of the value for the dependent variable.

Dependent Variable—This variable can take on different values; however, these values are said to “depend”

on the value of the independent variable. In experimental work, we test whether the manipulation of the

independent variable results in a significant change in the dependent variable. In observational studies,

the value of the dependent variable can be better predicted by knowledge of the value of the independent

variable.

Horizontal Axis—In this book, the term horizontal axis is used infrequently, as the preferred terminology is

the x-axis. Both these terms will always refer to the horizontal axis of the chart you are building. During certain

SPSS operations, you will find that the vertical axis is referred to as the x-axis. When this happens, we revert

to the term horizontal axis so as to avoid confusion.

Vertical Axis—The vertical axis rises perpendicular to the x-axis (horizontal) and is most frequently referred

to as the y-axis.

Percentages, proportions, and probabilities may, in a general sense, be thought of as possessing the same

type of information. Percentages use a base of 100, and therefore, any observed value represents a portion

of 100 and can be expressed as a percentage of the total (100%). Probability uses a base of 1, and therefore,

any observed value represents a portion of 1 and can be expressed as a decimal or fraction. The words

proportion and probability actually convey the same information. Therefore, we say the proportion of females

may be .80 or that the probability of a female is .80. Expressed as a percentage, we would say that 80% of

the group is female.

More details regarding the normal distribution are provided in Chapter 8 on histograms.

For information on how data are inputted into the SPSS format (not required in this textbook) consult Using

SPSS: An Interactive Hands-On Approach by Cunningham and Aldrich, 2011, Sage.

If you wish to refresh yourself on the meaning of “levels of measurement,” we refer you to Using SPSS: An

Interactive Hands-On Approach by Cunningham and Aldrich.

One note of caution is that the levels of measurement specified in the databases found in SPSS Sample files

are not always correct. We will only use such databases when absolutely necessary and when it does not

affect the graph building procedure. You should be alert to incorrect levels of measurement as you continue

to develop your graph making skills and you explore other databases in the SPSS Sample files.

The normality aspect of continuous data is discussed in more detail in Chapter 8 on histograms.

When graphing a bivariate relationship, we have a single point on the graph that represents two values, one

SAGE

SAGE Research Methods

Page 16 of 17 Introduction: Graphical Presentation of Data

for x and one for y.

As you proceed through the chapters, you will be required to repeatedly use the same databases from the

SPSS Sample files. You may wish to save the most used databases to your computer, making access much

quicker. A chapter-by-chapter listing of all SPSS databases used in this book is given in Appendix C.

http://dx.doi.org/10.4135/9781452270036.n1

SAGE

SAGE Research Methods

Page 17 of 17 Introduction: Graphical Presentation of Data

http://dx.doi.org/10.4135/9781452270036.n1

Introduction: Graphical Presentation of Data

In: Building SPSS Graphs to Understand Data