February 25th, 2021
SOCI2151H Week 6: Quantitative Data Analysis
Reading: Chapter 14 & 15 (optional)
Part One: Coding Data
Coding: convert data into numerical format (a numerical representation assigned to a
variable
● Eg. “gender” 1 = Male 2 = Female
We must reduce the variety of idiosyncratic items of information into a more limited set
of attributes composing a variable
● Eg. “occupation”
○ 1 = managerial
○ 2 = clerical
○ 3 = semi-skilled
The more complex the grouping of the variables being coded, then such complexity of
grouping could become unworkable when trying to analyze them
● The more coded variables not grouped together under large headings can be
combined individually and get a truer analysis of findings in research
Aim of coding: be detailed and complete
Coded categories of variables must be...
1. Mutually exhaustive: one definition for each
2. Mutually exclusive: only one category can be checked off (no overlap with other
answers)
Coding Scheme: Keeping a Code Book
● Your primary guide throughout the coding process (otherwise, it can/will become
confusing)
● Your guide during analysis: to locate variable and interpret them
○ Tells you where to find variables (potentially hundreds!)
○ Tells you exactly what the code represents
● Use exact wording from your questions in your codes
Part Two: Types of Quantitative Analysis
1. Univariate Analysis
● Examining the distribution of ONLY one variable
, 2
● The use of frequency distributions analysis. Create a table that shows the
number of cases observed for each of the attributes in one variable
● Purpose: frequency distributions summarize the distribution of the variable. We
count the number of observations that fall into each category of the variable. It
shows the distribution of frequencies of each category (attribute) of the variable
○ Eg. 40% were under the age of 30 60% were 30 or older
Valid percent: excludes missing data (those who didn't do the survey)
● Only valid cases count for analysis
We can then compare each category and combinations of categories:
● Gives us a much more complete picture of what it is we’re measuring
If variables are “discrete” (nominal or ordinal levels)
● 45 males and 55 females participated in the study
● 45% of people vote for the conservative, 42% liberal, 12% NDP and 1% other
If the variable is “continuous” (interval or ratio levels)
● The participants’ age range from 18-25 (average age of 22)
● The median income of the participants is 25 000
We can then use this analysis to present into averages, called the central tendency of a
variable
● How dispersed or how spread out the frequency distribution of all responses
● Typical value of a variable we have discovered
○ Pie charts help describe the sample
Measures of Central Tendency
● Mean: the sum divided by the number of elements in the sample (average)
● Median: the midpoint of the distribution. The value of the observation that lies
halfway through the distribution
● Mode: the category that has the most people (or observation)
● Sum: a single number (an attribute) represents all the detailed data collected in
regard to that specific variable
Standard Deviation of a Variable
● Dispersion: refers to the way values are distributed around some central value
(the range, which is the standard deviation)
● Standard Deviation: the index of the amount of variability in a set of data
○ Represents the average squared difference of all the observations from
the mean
SOCI2151H Week 6: Quantitative Data Analysis
Reading: Chapter 14 & 15 (optional)
Part One: Coding Data
Coding: convert data into numerical format (a numerical representation assigned to a
variable
● Eg. “gender” 1 = Male 2 = Female
We must reduce the variety of idiosyncratic items of information into a more limited set
of attributes composing a variable
● Eg. “occupation”
○ 1 = managerial
○ 2 = clerical
○ 3 = semi-skilled
The more complex the grouping of the variables being coded, then such complexity of
grouping could become unworkable when trying to analyze them
● The more coded variables not grouped together under large headings can be
combined individually and get a truer analysis of findings in research
Aim of coding: be detailed and complete
Coded categories of variables must be...
1. Mutually exhaustive: one definition for each
2. Mutually exclusive: only one category can be checked off (no overlap with other
answers)
Coding Scheme: Keeping a Code Book
● Your primary guide throughout the coding process (otherwise, it can/will become
confusing)
● Your guide during analysis: to locate variable and interpret them
○ Tells you where to find variables (potentially hundreds!)
○ Tells you exactly what the code represents
● Use exact wording from your questions in your codes
Part Two: Types of Quantitative Analysis
1. Univariate Analysis
● Examining the distribution of ONLY one variable
, 2
● The use of frequency distributions analysis. Create a table that shows the
number of cases observed for each of the attributes in one variable
● Purpose: frequency distributions summarize the distribution of the variable. We
count the number of observations that fall into each category of the variable. It
shows the distribution of frequencies of each category (attribute) of the variable
○ Eg. 40% were under the age of 30 60% were 30 or older
Valid percent: excludes missing data (those who didn't do the survey)
● Only valid cases count for analysis
We can then compare each category and combinations of categories:
● Gives us a much more complete picture of what it is we’re measuring
If variables are “discrete” (nominal or ordinal levels)
● 45 males and 55 females participated in the study
● 45% of people vote for the conservative, 42% liberal, 12% NDP and 1% other
If the variable is “continuous” (interval or ratio levels)
● The participants’ age range from 18-25 (average age of 22)
● The median income of the participants is 25 000
We can then use this analysis to present into averages, called the central tendency of a
variable
● How dispersed or how spread out the frequency distribution of all responses
● Typical value of a variable we have discovered
○ Pie charts help describe the sample
Measures of Central Tendency
● Mean: the sum divided by the number of elements in the sample (average)
● Median: the midpoint of the distribution. The value of the observation that lies
halfway through the distribution
● Mode: the category that has the most people (or observation)
● Sum: a single number (an attribute) represents all the detailed data collected in
regard to that specific variable
Standard Deviation of a Variable
● Dispersion: refers to the way values are distributed around some central value
(the range, which is the standard deviation)
● Standard Deviation: the index of the amount of variability in a set of data
○ Represents the average squared difference of all the observations from
the mean
