Class Notes
(783,221)

Canada
(480,539)

University of Guelph
(25,380)

SOAN 2120
(387)

Scott Schau
(113)

Lecture

# Feb.25 2013 Notes.docx

Unlock Document

University of Guelph

Sociology and Anthropology

SOAN 2120

Scott Schau

Winter

Description

SOAN*2120 – Week 7 LEC 1
Asia Barclay
Monday, February 25, 2013
MIDTERM:
- Wednesday Feb 27; 50 mins (may be 45 mins)
- 1.5 mins/question, multiple choice
- 29 questions, 5 pages
- Answer on the paper as well as on the scantron (hand in both, OTHERWISE YOU
GET ZERO!)
- NO calculators (but for the final exam we will)
- Chapters 1-7, everything up to the lecture that occurred just before the break
- Equal focus between lecture and textbook
- Big concepts that you would have gotten from reading the chapter overall (no little
details)
- We will begin Chapter 3 on Friday (therefore this is obviously not on the midterm)
- Usually gets a class average of 71% for the midterm
A. SIMPLE RANDOM SAMPLE (SRS)
- Every member of a population has an equal chance of being included in the sample
- Only possible to give every member of a population an equal chance of being selected if
we can identify every member of the population
Failure to do so is a major source of bias
- Two populations: theoretical (defined for selection) and sample (actual population that
it represents)
- This is because our sampling frame is not always the same size as our population
- Our sample population then becomes different from our theoretical population, which can
be a problem TABLE OF RANDOM NUMBERS
- I.e. If you wanted to select 30 individuals out of a class of 690 students for a survey, you
take the class list and select 3 numbers because the number “690” has “3” digits in it
- Table must be used consistently
- Discard numbers greater than 690, and discard repeated numbers
- Never re-use numbers for surveys, because otherwise it’s not random anymore
- See appendix in textbook for table of random numbers
B. SYSTEMATIC SAMPLING
- Every nth person is selected
n = population/sample size
- Potential catch (source of error):
Population may be listed in such a way that might affect the sample
C. STRATIFIED RANDOM SAMPLING
- One way to reduce sampling errors
- Purpose of stratification: ensure that different subgroups are represented in the correct
proportions
Goal in stratification is not to make comparisons across subgroups
- Divide population into strata
E.g. Divide population into men and women
Still doing random sample, but within each of the strata
D. SAMPLES OF CONVENIENCE (accidental samples)
- Biased
- E.g. If a professor is interested in a principle of learning theory as it applies to all college
graduate students, but only

More
Less
Related notes for SOAN 2120