Study Guides (380,000)
CA (150,000)
UTSG (10,000)
SOC (700)
SOC101Y1 (200)

SOC101Y1 Study Guide - 20 Guide: Sampling Frame, Convenience Sampling, Response Bias

Course Code
Robert Brym
Study Guide

of 2
voluntary response sample of a population (people who choose themselves for a
survey) are not representative as they might be highly motivated about the specific
subject of the survey.
representative sample is a group chosen so their characteristics closely match those
of the population.
󲍻 in a survey, we get different responses from voluntary response sample and
representative sample as the former isn’t accurate
a convenience sample is easiest to reach.
so for an accurate survey, neither the population nor the one conducting the survey has
to choose volunteers for the survey, they have to be chose at random.
󲍻 they constitute the probability sample.
a sampling frame is a list of all the people in a population.
a randomizing method minimizes inequality in the selection of people. (it’s a sampling
there is a 2.5% margin of error for the results gotten.
if the margins of error overlap in the responses of a survey we can’t conclude.
a properly conducted survey wouldn’t have overlapping margins of error.
Types of surveys:
a mail questionnaire is a form containing questions mailed to the respondents.
a face-to-face interview survey, questions presented to the respondent during a
a telephone survey, questions a presented to the respondent over the phone.
Threats to validity:
Undercounting due to an imperfect sampling
Nonresponse due to respondents not answering all the questions
Response bias when respondents do not answer the questions accurately
Try to avoid wording effects, questions have to be simply phrased, neutral, specific
survey data is useful to discover relationships between variables.
a contingency table is a cross classification of cases by at least 2 variables.
(e.g. cross classifying people by gender and height 󲍻 2 variables)
soc101 lecture 20 research methods II : survey methods!
2 types of variables: independent variable and dependent variable.
to read the contingency table we have to find the 100% which will give us the label of
the percentage of the independent variable. (the 100% is always aligned with the label if
the percentage of the independent variable)
the higher the percentage difference the stronger the correlation between the 2
variables. ( if the percentage difference is > 0 then there is an association)
the total frequency in a contingency table give us the total number of people for the
independent variable.
if there is an association we want to find out if there is causality.
in order for that to happen we need to remove the irrelevant variables.
3 conditions to causality:
there has to be an association between the variables (if one one weavers so does the
the cause has to happen first
when we introduce a controlled variable a possible explanation, the original
association does not disappear. (if it does disappear or it gets lower, then it’s spurious 󲍻
the original association is not causal)
regression analysis is a graph that provides similar information as a contingency
table. we need to draw the line (y = ax + b)
soc101 lecture 20 research methods II : survey methods!