Introduction to Research Methodology
• causal relationships: a change in A causes, produces or forces a change in B
◦ 3 necessary conditions for causality:
‣ association: there is statistical evidence of a relationship
‣ direction: there exists a cause effect (X --> Y) relationship between the 2 variables
‣ absence of spuriousness
Approaches to social research - will be asked on the exam
1. experiments
A. formally testing the effect of A on B (investigating causes for a phenomenon)
B. a controlled environment + control group > groups should be homogenous
C. experiments are best at/for proving causality
D. example: testing if chewing gum during a test improves scores
2. surveys (oral or self-administered questionnaires to a large group of people)
A. increased ability for generalization --> large sample needed
B. large sample is easier to get through a survey due to greater reach
C. aimed to ask about a certain topic/variable --> opinions --> con: self-reported behavior
D. example: how many people have voluntarily donated blood?
3. field research
A. observing individuals > avoids reactivity
B. can observe unknown things --> viewing actual behavior/reactions in real life situations
C. more flexible since not as much planning needed (just need to go out and watch)
D. example: an anthropologist who wants to understand and learn about a culture
4. available data
A. data that is pre-existing but not collected
B. making inferences about developments over time or in the past
C. written records, letters, diaries, newspapers, government documents
D. example: using a website for volunteers' data to find information on volunteers
Scientific research is guided by:
• empiricism: way of thinking or understanding the world that (in)directly uses our 5 senses
◦ such as observing how (what they say, body language) someone orders an ice cream
• objectivity: being free from emotion, conjecture (opinions based on incomplete info.) or personal
bias
◦ independent observers should agree on same conditions/conclusions = if achieved leads to
intersubjective testability
• control: the techniques and procedures used (in research) to minimize biases and maximize
trustworthiness
,Cumulative process of science
theory (general) --> hypothesis (specific) --> observations --> empirical generalizations/propositions
• theory: logically interconnected set of proportions that show how or why a relationship occurs
◦ theorizing vs. speculating
‣ theorizing: answering a why question
‣ speculating: forming a theory without evidence
• empiricism: understanding the world through our 5 human senses
• objectivity: viewing the world as it is, without opinions or background information
◦ basically impossible because we're human, we don't know all the biases we have
Hypothesis and Proposition - be able to distinguish between the two
• proposition: a general statement
• hypothesis: a specific statement that can be tested
◦ an unconfirmed, expected relationship between 2 or more variables
◦ look for objects that can be tested
◦ what are they testing for? what does the question want to know?
• example:
◦ proposition: people perform better when others are there watching them
‣ very broad > which people? perform better at what? who is watching them?
◦ hypothesis: soccer players perform better when journalists are watching them
‣ specific > you know who's performance we're observing. you know what activity we are
interested in. and we know who is watching.
Variable vs. concept
• variable: an empirical manifestation of the social phenomenon --> the measure itself
◦ variables can be categorical (qualitative) or metric (quantitative)
• concept: a theoretical description of the social phenomenon --> what is measured
Types of hypothesis:
• bivariate hypothesis: an expected relationship that has 2 variables
• multivariate hypothesis: an expected relationship with 3 or more variables
◦ mediator: a third variable intercepts the relationship between 2 variables
• age (X) has an affect on having the opportunity to re-
enter the labor market (Z) which in turn affects actually
entering the labor market (Y) = full mediation
• the orange line makes this model a partial mediation: a
relationship between 2 variables is simultaneously direct
and indirect
, ◦ moderator: the effect of X on Y differs based on a third variable
the willingness of someone to buy ice cream (X) and actually
buying ice cream (Y) is dependent on the weather/temperature of
that day (Z)
• conditional effects:
◦ buffering effect: the moderating variable has a negative
effect on X and Y --> colder temperatures (Z) mean less
people are willing to (X) and actually do buy ice cream (Y)
◦ intensifier effect: when the moderating variables has a
positive effect on X and Y --> warmer temperatures (Z)
mean more people are willing to (X) and actually do buy ice
cream (Y)
◦ spuriousness: there appears to be a relationship between 2 variables, but it is explained by
the 3rd variable as this is the common cause
due to higher temperatures (Z) there are more ice creams
purchased (X) and more swimsuits bought (Y) - you may think
that more ice creams purchased means more swimsuits
purchased, but it's due to the common cause Z
◦ relative importance or (total effect) of the independent variables (multiple causality) on the
dependent variable
the effect of the weather/temperature (X1) on buying ice
cream (Y) is stronger than flavor variety (X2) - people buy
more ice cream based on the weather than based on flavor
variety
Key elements of research design:
• unit of analysis: the entities (objects or events) being studied - what/who you want to make
statements about in your research
◦ (1) objects: individuals, groups, organizations, communities, nations etc.
◦ (2) events: social roles, positions, relationships etc.
◦ to identify the UoA, ask what or who is being described, compared to or analyzed
‣ always choose the smallest level being studied
◦ UoA in multilevel data: the dependent variable often defines the UoA
‣ example: the gross domestic product of countries has a positive effect on the level of
optimism of the citizen
• unit of observation: how you gain information about what you want to analyze
◦ doesn't have to be the same as the unit of analysis
• causal relationships: a change in A causes, produces or forces a change in B
◦ 3 necessary conditions for causality:
‣ association: there is statistical evidence of a relationship
‣ direction: there exists a cause effect (X --> Y) relationship between the 2 variables
‣ absence of spuriousness
Approaches to social research - will be asked on the exam
1. experiments
A. formally testing the effect of A on B (investigating causes for a phenomenon)
B. a controlled environment + control group > groups should be homogenous
C. experiments are best at/for proving causality
D. example: testing if chewing gum during a test improves scores
2. surveys (oral or self-administered questionnaires to a large group of people)
A. increased ability for generalization --> large sample needed
B. large sample is easier to get through a survey due to greater reach
C. aimed to ask about a certain topic/variable --> opinions --> con: self-reported behavior
D. example: how many people have voluntarily donated blood?
3. field research
A. observing individuals > avoids reactivity
B. can observe unknown things --> viewing actual behavior/reactions in real life situations
C. more flexible since not as much planning needed (just need to go out and watch)
D. example: an anthropologist who wants to understand and learn about a culture
4. available data
A. data that is pre-existing but not collected
B. making inferences about developments over time or in the past
C. written records, letters, diaries, newspapers, government documents
D. example: using a website for volunteers' data to find information on volunteers
Scientific research is guided by:
• empiricism: way of thinking or understanding the world that (in)directly uses our 5 senses
◦ such as observing how (what they say, body language) someone orders an ice cream
• objectivity: being free from emotion, conjecture (opinions based on incomplete info.) or personal
bias
◦ independent observers should agree on same conditions/conclusions = if achieved leads to
intersubjective testability
• control: the techniques and procedures used (in research) to minimize biases and maximize
trustworthiness
,Cumulative process of science
theory (general) --> hypothesis (specific) --> observations --> empirical generalizations/propositions
• theory: logically interconnected set of proportions that show how or why a relationship occurs
◦ theorizing vs. speculating
‣ theorizing: answering a why question
‣ speculating: forming a theory without evidence
• empiricism: understanding the world through our 5 human senses
• objectivity: viewing the world as it is, without opinions or background information
◦ basically impossible because we're human, we don't know all the biases we have
Hypothesis and Proposition - be able to distinguish between the two
• proposition: a general statement
• hypothesis: a specific statement that can be tested
◦ an unconfirmed, expected relationship between 2 or more variables
◦ look for objects that can be tested
◦ what are they testing for? what does the question want to know?
• example:
◦ proposition: people perform better when others are there watching them
‣ very broad > which people? perform better at what? who is watching them?
◦ hypothesis: soccer players perform better when journalists are watching them
‣ specific > you know who's performance we're observing. you know what activity we are
interested in. and we know who is watching.
Variable vs. concept
• variable: an empirical manifestation of the social phenomenon --> the measure itself
◦ variables can be categorical (qualitative) or metric (quantitative)
• concept: a theoretical description of the social phenomenon --> what is measured
Types of hypothesis:
• bivariate hypothesis: an expected relationship that has 2 variables
• multivariate hypothesis: an expected relationship with 3 or more variables
◦ mediator: a third variable intercepts the relationship between 2 variables
• age (X) has an affect on having the opportunity to re-
enter the labor market (Z) which in turn affects actually
entering the labor market (Y) = full mediation
• the orange line makes this model a partial mediation: a
relationship between 2 variables is simultaneously direct
and indirect
, ◦ moderator: the effect of X on Y differs based on a third variable
the willingness of someone to buy ice cream (X) and actually
buying ice cream (Y) is dependent on the weather/temperature of
that day (Z)
• conditional effects:
◦ buffering effect: the moderating variable has a negative
effect on X and Y --> colder temperatures (Z) mean less
people are willing to (X) and actually do buy ice cream (Y)
◦ intensifier effect: when the moderating variables has a
positive effect on X and Y --> warmer temperatures (Z)
mean more people are willing to (X) and actually do buy ice
cream (Y)
◦ spuriousness: there appears to be a relationship between 2 variables, but it is explained by
the 3rd variable as this is the common cause
due to higher temperatures (Z) there are more ice creams
purchased (X) and more swimsuits bought (Y) - you may think
that more ice creams purchased means more swimsuits
purchased, but it's due to the common cause Z
◦ relative importance or (total effect) of the independent variables (multiple causality) on the
dependent variable
the effect of the weather/temperature (X1) on buying ice
cream (Y) is stronger than flavor variety (X2) - people buy
more ice cream based on the weather than based on flavor
variety
Key elements of research design:
• unit of analysis: the entities (objects or events) being studied - what/who you want to make
statements about in your research
◦ (1) objects: individuals, groups, organizations, communities, nations etc.
◦ (2) events: social roles, positions, relationships etc.
◦ to identify the UoA, ask what or who is being described, compared to or analyzed
‣ always choose the smallest level being studied
◦ UoA in multilevel data: the dependent variable often defines the UoA
‣ example: the gross domestic product of countries has a positive effect on the level of
optimism of the citizen
• unit of observation: how you gain information about what you want to analyze
◦ doesn't have to be the same as the unit of analysis
