SOC202H1 Study Guide - Fall 2018, Comprehensive Midterm Notes - Sampling Distribution, Null Hypothesis, Test Statistic
SOC202H1
MIDTERM EXAM
STUDY GUIDE
Fall 2018
1
SOC202H1F
Notes for Session 1
15 May 2017
THE BIG PICTURE
Descriptive statistics
Describe some aspect of the world (univariate descriptive statistics) – data reduction
Describe how things are related (bivariate/multivariate descriptive statistics) – measures of association
(cannot prove by themselves how variables are causally related, however, give clues)
Inferential statistics
Draw a conclusion about the wider world from a smaller part of it—a sample drawn from a larger
population
Is a difference we observe with data likel a real differee, or could it probably have just arisen by
chance?
- This is an era of unprecedented quantitative data
- Statistical savvy is increasingly in demand in both the private and public sectors
- Math is used in this course b/c mathematical expressions offer a precision that is hard to
achieve in everyday English
o However, a very few of us communicate in mathematical form
BASIC CONCEPTS IN STATISTICS
• Cases – entities from which quantitative data are gathered (people, groups, provinces,
countries)
• Population – every case representing a given group or
• Sample – selection of cases drawn from the population
• Variables – traits that can change values from case to case (e.g. age, gender, social class, form of
government)
o Things that vary from one thing to another
Important distinctions for variables
• Independent vs. dependent variables – their role in causal theories
o Value of independent variable varies
o The value of dependent variable is dependent on the independent variable
find more resources at oneclass.com
find more resources at oneclass.com
2
o A hypothesis – formal statement that an X variable has a particular type of relationship
(positive or negative) with a Y variable. Hypotheses are specific expectations derived
from general theories
▪ For example, we might propose that being older (age = X) is associated with
getting higher marks in a U of T class (final mark = Y)
• As people get older, they tend to get higher marks (hypothesis)
• Students = unit of analysis
• Age and final mark are things that vary from student to student
• Level of measurement
o Nominal – scores are labels only, they are not true numbers
• Scores are different from each other but cannot be treated as inherently meaningful numbers.
for e…
o Gender – 1 = female, 2 = male
o Immigrant status – 1 = Canadian born people, 2 = Foreign-born people
o These numbers do not represent anything meaningful as they are just numbers that
separate one from the other
▪ We can easily reverse the scoring of the variables and this would not make any
meaningful difference, thus they are not considered meaningful numbers
o Ordinal – scores have some numerical quality and can be ranked
▪ First two are categorical variables
• Scores can be ranked from high to low or from more to less
• If you can distinguish between the scores of a ategorial ariale usig ters suh as ore,
less, higher or loer the ariale is ordial
• Survey items that measure opinions and attitudes are typically ordinal
o Interval-ratio – scores are true numbers
• Scores that are actual numbers and have equal intervals between them
o Examples:
▪ Age (in yrs)
▪ Income (in $)
▪ Number of children – the difference between 4 and 5 kids is one unit, the
difference between 4 and 12 kids is one unit
• Equal intervals
Ordinal vs. Interval-ratio variable – though both variables can be ranked, unit increases or decreases do
not have clear mathematical meaning for ordinal variables. The numbers associated with an ordinal
variable are merely codes signifying ranks
Do you agree or disagree that
smoking should be banned on
campus?
Do you agree or disagree that
smoking should be banned on
campus?
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Describe some aspect of the world (univariate descriptive statistics) data reduction. Describe how things are related (bivariate/multivariate descriptive statistics) measures of association (cannot prove by themselves how variables are causally related, however, give clues) Draw a conclusion about the wider world from a smaller part of it a sample drawn from a larger population. This is an era of unprecedented quantitative data. Statistical savvy is increasingly in demand in both the private and public sectors. Math is used in this course b/c mathematical expressions offer a precision that is hard to achieve in everyday english: however, a very few of us communicate in mathematical form. Independent vs. dependent variables their role in causal theories: value of independent variable varies, the value of dependent variable is dependent on the independent variable. 1: a hypothesis formal statement that an x variable has a particular type of relationship (positive or negative) with a y variable.