Instructor Resource
Treadwell and Davis, Introducing Communication Research: Paths of Inquiry, 4e
SAGE Publishing, 2020
Lecture Notes
Chapter 8: Generalizing from Research Results: Inferential Statistics
Learning Objectives
8-1 Explain the concept of inferential statistics and how they differ from descriptive
statistics.
8-2 Describe the normal curve and its significance to inferential statistics.
Chapter Summary
This chapter begins by explaining the concept of inferential statistics and how they differ from
descriptive statistics. Next, the chapter describes the normal curve and its significance to
Annotated Chapter Outline
I. Introduction
A. The language of curves
1. Inferential statistics assume a normal distribution of values in a population.
B. Skew
1. Skew is where the “tail” of distribution is.
Instructor Resource
Treadwell and Davis, Introducing Communication Research: Paths of Inquiry, 4e
SAGE Publishing, 2020
II. Generalizing from Data: Inferential Statistics
A. Inferential statistics
1. Inferential statistics help us generalize (make inferences) about a wider
population from a smaller sample.
B. The normal curve and the central limit theorem
1. In the normal curve or bell curve, data are symmetrical.
2. The central limit theorem states distribution of averages will be approximately
C. The normal curve, z scores, and return of standard deviation
1. The area under the standardized normal curve allows us to calculate the
probability of obtaining any given result.
2. Under a normal curve.
D. Calculating probabilities based on the normal distribution
1. To get the exact probability, we can go to tables of z scores.
E. z scores, hypotheses, and decision making
1. Conventionally, we use a .05 probability cutoff to decide whether to accept or
Instructor Resource
Treadwell and Davis, Introducing Communication Research: Paths of Inquiry, 4e
SAGE Publishing, 2020
3. The calculated range of possibilities is called the confidence interval.
4. For a sampling distribution (the distribution of the sample results), the standard
F. Confidence level and sample size
1. There is a trade-off among confidence level, standard deviation, and sample size.
III. Testing for Differences Between and Among Groups
A. The t-test
1. The t-test compares the mean scores of two groups on a variable to determine the
B. t-test: formula
1. t-test is based on the differences in means for a variable common to two groups
but also considers the range of scores for each group and each group’s size.
C. t test: example
1. Degrees of freedom (df) is an estimate of the number of independent pieces of
Instructor Resource
Treadwell and Davis, Introducing Communication Research: Paths of Inquiry, 4e
SAGE Publishing, 2020
D. t-test for dependent samples
1. Where test groups consist of the same individuals, we have dependent samples
E. Analysis of variance
1. ANOVA
2. Compares values for three or more groups to determine if they differ.
F. ANOVA: formula
F = variance between groups / variance within groups
G. ANOVA: example
IV. Testing for Relationships Between and Among Variables
A. Correlation
1. Correlation is used to assess relationships between variables.
B. Regression
1. Linear regression predicts a specific value for one variable (outcome or
criterion variable).
Instructor Resource
Treadwell and Davis, Introducing Communication Research: Paths of Inquiry, 4e
SAGE Publishing, 2020
V. Two Final Decisions
A. Accept or reject my findings?
1. Statistical testing offers you the probability, not certainty.
B. If it’s significant, is it significant?
1. Significance in a general sense is a matter of disciplinary judgment or in the case
of academic publications—editors’ judgments.