Essentials of Statistics:
A Tool For Social Research
Joseph F. Healey, Fourth Edition
Chapter 1: Introduction
©2016. Cengage Learning. All rights reserved.
Learning Objectives
Describe the limited but crucial role of statistics in social research.
Distinguish between three applications of statistics and identify situations
in which each is appropriate.
Identify and describe three levels of measurement and cite examples of
variables from each.
©2016. Cengage Learning. All rights reserved.
Chapter Outline
Using Statistics
Why Study Statistics?
The Role of Statistics in Scientific Inquiry
The Goals of This Text
Descriptive and Inferential Statistics
©2016. Cengage Learning. All rights reserved.
Using Statistics
Statistics can be used to:
Demonstrate the connection between smoking cigarettes and experiencing a
stroke.
Track the public approval rating of President Barack Obama over time.
Measure differences in attitudes towards tattoos and body piercings between
people in different age groups.
Statistics are increasingly used in market research, social policy studies, and
the social sciences.
©2016. Cengage Learning. All rights reserved.
Why Study Statistics?
Statistics are essential in the social sciences because they enable ideas to be
evaluated and theories to be tested.
Quantitative Research projects use statistics to analyze numerical data.
Statistics are mathematical techniques used by social scientists to analyze data
in order to answer questions and test theories.
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
A theory is an explanation of the relationships between phenomena.
Theories are general. For example, consider Karl Marx’s theory of
Economic Determinism. Marx believed that economic relations
influenced social and political spheres of life.
Can you test that with statistics? Not easily!
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
A hypothesis is derived from theory. It is a specific and exact statement about
the relationship between variables.
Hypotheses are specific. Psychologist Thomas Pettigrew hypothesized that
group prejudice could be reduced when people were exposed to diverse
others for a period of time under certain specified conditions, such as
those where individuals are treated equally regardless of their race or
ethnicity.
Develop a hypothesis of your own.
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
Hypotheses are stated in terms of variables.
A variable is any trait that can change values from case to case.
Examples: age, income, religious affiliation, attitudes toward some issue
In Pettigrew’s hypothesis (previous slide), the variables included level of
prejudice, amount of time spent with diverse others, and the presence of a
situation where people are treated equally.
What are the variables in the hypothesis that you created?
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
Science distinguishes between two types of variables: dependent variables and
independent variables.
Independent variables are causes.
Dependent variables are effects or outcomes.
In Pettigrew’s hypothesis, an independent variable was the amount of
time spent with diverse others.
The dependent variable was the level of prejudice.
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
To test a hypothesis, we must gather our data or observations.
Things to consider:
How many cases will we select?
How will we measure our variables?
How will we analyze the data?
©2016. Cengage Learning. All rights reserved.
The Role of Statistics in Scientific
Inquiry
The Research Process
After collecting data, we form empirical generalizations through statistical
analysis.
At this point, we would revisit our theory.
Do our results support our theory?
Do our results suggest changes to our theory?
©2016. Cengage Learning. All rights reserved.
The Goals of This Text
Statistics are useful tools, but not ends in themselves. Thus, we will not
take a “mathematical” approach to the subject.
This text will help you to read and critique research literature in the social
sciences.
Learning how to use these tools will help you to carry out your own
research.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Descriptive Statistics
Univariate descriptive statistics summarize the distribution of a single variable.
Percentages, rates, etc.
Data reduction is a process of using a few numbers to summarize many.
An example of a univariate statistic might be the percentage of
students who score an 80 or higher on an exam.
Develop a univariate statistic of your own.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Descriptive Statistics
Bivariate descriptive statistics describe the relationship between two variables.
Multivariate descriptive statistics describe the relationships among more than
two variables.
Measures of association quantify the strength and pattern of a relationship
between two or more variables.
For example, gender, race, and health behaviors are all associated
with life expectancy.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Inferential Statistics
Inferential statistics are used to generalize sample information to its
population.
A population is a collection of all possible cases of interest.
A sample is a carefully chosen subset of a population. It is the set of cases
used for study and analysis.
For example, a hospital has a population of patients. A researcher
can use a probability technique to select a random sample of 20% of
the patients.
Create your own example of a population and sample.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Level of measurement refers to the mathematical characteristics of a variable.
There are three levels of measurement:
Nominal
Ordinal
Interval-ratio
The level of measurement of a variable guides our choice of statistic and
statistical technique.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Regardless of the level of measurement, variables should have categories that
are:
Mutually exclusive (No two cases should fit in one category)
Exhaustive (All cases should fit in a category)
Relatively homogenous (The categories must be similar)
For the variable “political party affiliation,” think of a set of
categories that would be mutually exclusive, and homogenous.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Nominal Variables:
Have categories that are non-numerical and cannot be ranked.
Are the lowest level of measurement available.
Consist of categories that can only be compared by considering how many
cases fall under each one.
Examples of nominal variables include gender and religious affiliation.
The Latin root word “nom” means “name”.
Name five other nominal level variables.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Ordinal Variables:
Have ‘scores’ or categories that can be ranked from high to low.
Can be described as being “more or less” with respect to each other.
Are limited because we do not know the exact distance from one score or
category to the next.
Examples of ordinal variables include social class (lower, working,
middle, upper) and “Likert” items (strongly agree, agree, disagree,
strongly disagree).
Name five other ordinal level variables.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Interval-ratio Variables:
Are numerical and they can be ranked.
Can be ranked from high to low and we can quantify the distance between
different scores.
Sometimes have “true” zero points.
Can be used for all mathematical operations.
Examples of interval-ratio variables include income (measured in
dollars) and number of siblings.
Name five other interval-ratio level variables.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
How to Quickly Determine Level of Measurement
Ask yourself two simple questions:
1. Can the categories of the variable be ranked based on value?
If the answer is no, as in the case of gender, it is nominal.
If the answer is yes, as in the case of age, go to question #2.
2. Can the distance between the categories be quantified?
If the answer is yes, as in age, it is interval-ratio.
If the answer is no, as in level of agreement (Strongly Agree, Agree,
Disagree, Strongly Disagree), it is ordinal.
©2016. Cengage Learning. All rights reserved.
Descriptive & Inferential Statistics
Level of Measurement
Some variables can be measured at multiple levels.
For example, educational attainment can be measured in the number of
years completed (0, 1, 2, 3, ..., 12, 13, etc. – an interval-ratio variable) or it
can be measured in the highest degree received (less than high school, high
school diploma, associates degree, bachelors degree, etc. – an ordinal
variable).
Think of another variable that can be measured at different levels?
©2016. Cengage Learning. All rights reserved.
Summary
The purpose of statistics is to organize, manipulate, and analyze data so that
researchers can test theories and answer questions. Along with theory and
methodology, statistics are a basic tool used by social scientists to enhance
their understanding of the social world.
There are two general classes of statistics. Descriptive statistics are used to
summarize the distribution of a single variable and the relationships between
two or more variables. We use inferential statistics to generalize to
populations from random samples.
Variables may be measured at any of the three different levels. At the
nominal level, we can compare category sizes. At the ordinal level, scores
can be ranked from high to low. At the interval-ratio level, all mathematical
operations are permitted.
©2016. Cengage Learning. All rights reserved.
Basic Terms
Data Measures of association
Data reduction Population
Dependent variable Quantitative research
Descriptive statistics Research
Hypothesis Sample
Independent variable Statistics
Inferential statistics Theory
Level of measurement Variable
©2016. Cengage Learning. All rights reserved.