1/30/’25

Chapter 1: Introduction to Statistics

Social Statistics (SOC 241-04) | Spring 2025

Instructor: Jisoo Cho
 Brightspace
- No more Blackboard

- We use Brightspace for:

- Course
announcements
- Material distribution
- Assignment
submission
- Exam taking
- Grade posting

- If unavailable → Email
assignments directly

- Tech Support: Hunter

College IT Helpdesk (
https://hunter.cuny.edu/i
nformation-technology/s
Foundation
■ Research helps us answer questions, test ideas, and understand
the world around us. Social scientists use different methods to
collect and analyze information, and that’s where statistics
comes in.
■ Numbers alone don’t tell the full story. Without statistics, data is
just a collection of numbers. Learning how to analyze data
helps us find patterns, draw conclusions, and make informed
decisions.
■ Research → A way of systematically gathering information to
explore questions and test theories.
■ Quantitative Research → A type of research that collects
information in the form of numbers.
■ Data → Measurable facts or information that researchers analyze.
■ Statistics → A set of tools that help organize, analyze, and
The Scientific Process
 Theory
■ Theories help us explain why things happen. In social
science, a theory connects causes and effects to
understand human behavior.
■ Example: What causes racial prejudice? Over 50 years
ago, social psychologist Gordon Allport proposed that:
– Prejudice decreases when people from different groups
interact as equals and work together toward shared
goals.
• Prejudice increases when groups compete for resources
like jobs or housing.
 Theories help us test ideas by identifying relationships
between variables → gives us a structured way to explore
society.
Variable
.■ Anytrait that can change values from case to case.
– The independent variable (X) is the cause.
– The dependent variable (Y) is the effect.

# Variable (Possible) Values/ Categories

1 Sex Male, Female, Non-Binary
Any positive numbers and zero (0, 5,
2 Age (in number)
30…)
Temperature
3 Any numbers (-5, 0, 30, 80)
(˚F)
4 Education level High School or less, College or more…
Crime rate (per Any positive numbers and zero (0,
5
100,000) 3.5, 124.2…)
 Hypotheses
■ Hypothesis: a clear, testable statement that predicts how two
variables are related. It helps turn broad theories into something we
can measure and test.
■ Theories are broad; hypotheses make them specific. They tell us
what to look for in our research.
■ Example (Testing Allport’s Theory):
– Theory: Equal-status contact reduces prejudice.
– Hypothesis: "Students in diverse, cooperative study
groups will show less prejudice than those in competitive
groups."
■ Before Testing, We Must:
– Clearly define key terms (What is prejudice? What is equal-
status contact?)
– Review past research to refine ideas
 Observations
■ Once we have a hypothesis, the next step is to test it with real data.
This means designing a study, collecting information, and analyzing the
results.
■ Testing Allport’s Theory
• We place participants from different racial groups into cooperative or
competitive situations.
• We measure prejudice levels before and after using a survey.
• Our goal: See if cooperative contact reduces prejudice.
■ The Role of Statistics
• We collect a lot of data—potentially hundreds or thousands of
responses.
• Statistics help us organize, analyze, and interpret this data.
• Without statistics, quantitative research is impossible.
■ But there are limits as well!
• Good data → meaningful results.
 Empirical Generalizations
■ Finding Patterns in Data
• Patterns in the data that apply beyond a single study.
• Do results apply equally to different groups (e.g., age,
gender, education)?
■ Example: Testing Allport’s Theory
• Does cooperative contact reduce prejudice for all groups?
• If younger people show more change, theory may need
refining.
■ Why It Matters
• Helps improve and expand theories.
• Makes research more useful and applicable.
• Generalizations turn single studies into broader insights
about society.
Either direction is good!
■ Research can begin with a theory or an observation.
■ Starting with THEORY → OBSERVATION (Deductive
reasoning)
– Example: Allport’s theory says cooperative contact
reduces prejudice.
– Test: Researchers place diverse groups in cooperative
tasks and measure prejudice levels before and after.
■ Starting with OBSERVATION → THEORY (Inductive
reasoning)
– Example: A researcher notices younger people seem
more open to diversity.
– Question: Why? Further study leads to a theory about
age and openness to attitude change.
 Unit of Analysis
• The main entity being analyzed in a study.
• Determined by the research question or hypothesis.
• In a dataset, each row represents a unit of analysis.
• The unit of analysis determines what is being studied
—getting it wrong can lead to misinterpretation of data.
• Examples:
• Does poverty level affect educational attainment level?
→ Neighborhoods; individuals
• Does crime victimization vary by gender? → Individuals
• Does GDP influence civic engagement? → Countries
Discrete Variables Continuous Variables
■ Can only take whole numbers ■ Can be infinitely subdivided into
(no decimals or fractions). smaller units.
■ Examples: ■ Only interval-ratio level variables
– Number of siblings (1, 3, but can be continuous
not 1.5) ■ Examples:
– Number of students in a – Time in a race (10.7 sec,
classroom (25, 30, but not 10.732 sec, etc.).
25.5)
– Height of a person (5.8 feet,
– Number of pets owned (1, 2, 5.8123 feet, etc.)
3… but never 1.3)
– Time spent studying (2.5
– Number of text messages hours, 2.75 hours)
sent in a day
– Temperature (98.6°F, 72.4°F)
– Number of cars in a parking
lot – Distance traveled (10.3 miles,
10.345 miles)
– Number of goals scored in a
soccer match – Weight of a package (2.45 kg,
 Descriptive Statistics
■ What do they do? Descriptive statistics organize,
summarize, and reveal patterns in data.
■ Univariate (One Variable)
• Describes a single variable.
• Examples: Averages, percentages, graphs; Instead of
listing 10,000 incomes, use an average or graph to
summarize.
■ Bivariate/Multivariate (Two or More Variables)
• Shows relationships between variables.
• Measures of association reveal strength and
direction of relationships.
• Example: More study time → higher grades (positive
 Inferential Statistics
■ Help us make generalizations about a population (The total
collection of all cases in which the
■ researcher is interested) based on a sample (A carefully
chosen subset of a population).
■ Why Are They Needed? Populations are too large to study
entirely (e.g., all U.S. voters). Instead, we analyze a sample
and use inferential statistics to draw conclusions.
■ Example:
• A poll surveys 1,500 people to predict how 130 million
voters will vote.
• If done correctly, the sample accurately represents the
larger population.
■ Inferential statistics allow researchers to study a small group
and make reliable claims about a much larger group.
The level at which a variable’s value is measured or
 Nominal Level
■ Categories with no numerical value or ranking.
■ Nominal variables label and classify but do not indicate order or
quantity.
■ Cannot be added, divided, or ranked. Only counted and compared.
■ Mutually Exclusive → Each case fits into only one category.
■ Exhaustive → Covers all possible responses.
■ Homogeneous → Categories must be meaningful and relevant.
Variable (Possible) Values
Sex Male, Female, Non-Binary
Race/Ethnicity White, African American, Hispanic…
Yes/No (Dichotomous Variable: A variable that has only
Are you a student?
two categories.)
Religious Affiliation Christian, Muslim, Jewish, etc.
 Ordinal Level
■ Categories that can be ranked (high to low, more to less).
■ Ranks matter, but the exact difference between ranks is
unknown.
■ Cannot use addition or precise calculations.
■ For example, if a job satisfaction scale ranks responses as: (1)
Very Dissatisfied to (5) Very Satisfied → 5 (Very Satisfied) is higher
than a 2 (Dissatisfied), but we cannot say it's 2.5 times more
satisfied because the gaps between ranks are not necessarily
equal.Variable (Possible) Values
Socioeconomic Status High, Medium, Low
5-point Likert Scale (Very Low, Low, Neither Low nor
Feeling of Safety
High, High, Very High)
Less than high school, high school, college,
Education Level
graduate…
 Interval-Ratio Level
■ Equal intervals → Differences are consistent (e.g., 10 to
20 is the same as 30 to 40).
■ True zero → A value of zero means nothing exists (e.g.,
zero income = no income).
■ All math operations apply → Can add, subtract,
multiply, divide.
■ Examples:
• Age (20 years is exactly 2 years more than 18).
• Number of siblings (0, 1, 2, 3… with equal steps).
• Income ($0 means no income; $50,000 is twice $25,000).
• Test scores (0 means no correct answers, and scores
increase equally).
For Next Class:
■ Read chapter 2
■ Bring your notebook and pen/pencil
■ We need calculator by 2/4
■ Quiz format (not a test)
with no time limit. You
cannot save and return
later. Complete in one
sitting.
■ Two attempts are given
→ Highest grade counts.
– If you don’t do well the
first time, you can try
again before the
deadline.
– After your first attempt,
you’ll see which
questions were
incorrect, but not the
correct answers →
gives you a chance to
think critically before
retrying.
■ When entering “interval-
ratio,” type it exactly as