Lecture 2: Basics of methodology (how one does research)

A research methodology is a strategy for systematically conducting research.

We use this method every day.
- Example: If your laptop broke and you want to buy a new one. You will either do
research online to check what type of laptop you want or if another device like an Ipad
would be a better idea. You can ask your friends that have a MacBook for example if they
like it and see what they have to say about it. In this example, you just used two different
research methods, documented analysis and interviews.
The types of questions that we ask affects the method of data collection that we use.
If you want to figure out what is the best laptop for university students, you will talk to
university students. You do an interview. The goal here is to use the best method suited to
answer the question that you are investigating.
Social science knowledge is based on science, a local organized method of obtaining
information through direct, systematic observation.
Scientific knowledge is based on empirical evidence, information that is directly verifiable.
Journalists report the news, elections, changes in society
We do that too, but we do so through regress research methodology and evaluation of our
research methodology, which is also known as peer review.
Peer review: when other academics review your research and say that it should be published and
made available to the public.
Concepts: Mental constructs that represent some part of the world around us, inevitability in a
very simplified form.
When we start thinking about different ideas we start to refine them, and we figure out what we
want to examine, we start to operationalize them. Therefore, we operationalize a variable. Which
is a process to develop a measure to gauge that variable
Variables: Concepts whose values changes from case to case.
- Example: Gender, Age, Income.
Operationalization: The process of developing the measure to be used in gauging (Make
judgement) a variable.
- Example: Operationalizing violence
There are multiple forms of violence. Individuals tend to oversee other types of violence than
physical violence.
- Have you been hit?
- Are there any bruises?
- Have you had to go to a hospital?
- Have you had to see a doctor?
- Have you had to see a doctor and lie about where the bruises came from?
- Have you been told that you need to lose weight?
- Have you been told that you need to augment your body?
- Have you been called names?
- Have you been yelled at?
If we had only looked at the physical forms of violence, we would’ve missed out on the other
events that have been taking place. When you are looking at a theory or concept, you have to
think how we can refine that to actually be representative of the world around us.
Causation and correlation
Causation: Cause and effect is a relationship in which change in one variable causes change in
another.
You need 3 criteria for that:
1. A must precede B
2. A must cause a change in B
3. There must be no extraneous or spurious variables.
It is difficult to prove causation, so academics use correlation instead.
Academics say “correlation exists when (two or more) variables are related in some way.
Correlation doesn’t mean causation.
Example: In the summer, crime rates go up and so as ice cream consumption. These two
variables are correlated but one does not cause the other.
The weather being nice leads to people going out more, therefore the crime rates increase. That is
the real reason for the crime rates not eating ice cream.

Independent and Dependent variables

Independent (causal variables) cause to have an effect on the other variable (effected) dependent
variables.
Dependent variables (DV) are variables caused by another variable.
Independent variables (IV) are the cause.
Extraneous variables: outcomes that can affect your research that are not part of your study.
(Often times you would want to eliminate these before you say there is a causal relationship.
Spurious variables: A statistical relationship between two variables that would, at first glance,
appear to be causally related, but when closely examined, only appear so by coincidence or due
to the role of another variable.
Theory: states that a certain set of social conditions will cause the other set of social conditions.
Hypothesis: something we want to test when we are doing research.
Example: the increase in the level of inequality (IV) will cause an increase in the crime rate (DV).
Sampling:
1. Stratified sample: if we look at Canada as a whole and start breaking it down by gender,
then by age, then ethnicity, then first language learned, then religion, then sexuality, then
we break it down by province. Population is grouped by key characteristics. A random
selection of groups is made, and a random selection of people is chosen from these
groups. When someone asks you questions down the street they want to know the
characteristics of the person they are asking.
2. Convenience sample: a sample of people who are readily available. Non-representative
and usually lacks external validity. If your teacher asks you to do a survey, you will go to
all your friends and ask them for help in the study. You got people to answer this but they
are not representative of the entire Canadian population.
3. Snowball sample: We use this method when we are investigating a very particular type of
population. Example: furriers, serial killers, vampires, hackers. These are not people that
you can regularly find on the street. You ask someone that is a hacker if they know other
hackers a hacker and you ask if hacker A can give your name to the other hackers and
 they contact you. You ask one person if they know people that belong to let’s say the
bronie subculture then they give you more names and so on.
4. Simple Random Sample: Units have an equal probability of being included in the
sample. Systematic requires a complete list of members of the population and high
response rate to be representative. Example: let’s say we want to figure out what is the
opinion of students on campus regarding crime, you would go to a building and stand at
the door handing people surveys let’s say you get 100 surveys. It won’t be representative,
but it will probably be very close.
Limitations of Social Scientific Research
1. Human behavior is too complex to allow social scientists to predict any individual’s
actions precisely.
Example: why would a person not want to take a vaccine.
The aggregate allows us to predict the general outcome. However, on the individual level,
prediction is more difficult.
2. Because humans respond to their surroundings, the mere presence of a researchers may
affect the behavior being studied
Example: Hawthorne effect: when you watch someone and you are trying to study them, your
presence will affect the way they act.
Social patterns change constantly. What is true in one time or place may not hold in another.
Example: texting while driving was okay before but now its illegal.
Because social scientists are part of the social world they study, objectivity in social research is
especially difficult. When we do research, sometimes we are too connected to what we are
researching that are judgement is clouded. (Author’s Bias)
Subjective interpretation is always an important element in analysis.
Sometimes when you think you saw something it is not actually what you saw. That is why we
need to be critical when examining the social world.

Qualitative and Quantitative Research

Qualitative: the nonnumerical examination and interpretation of observations for the purpose of
discovering underlying meanings and patterns of relationships.
Types of research we would employ here are observational and interviews.
Quantitative: the numerical representation and manipulation of observations for the purpose fo
describing and explaining the phenomena that those observations represent.
Surveys and experiments and it is more statistically orientated

Mixed Methods: when you use triangulation which is the three types of methods below it
enables you to look at the phenomena you are looking at.
- Surveys
- Interviews
- Document analysis
When you apply three types of methodologies we call that triangulation.
Ethics
When we conduct research, it is important to conduct ethics.
When we conduct research, we must ask how may the research affect other people's lives?
1. Do you have informed consent?
2. Are they free to participate?
3. Do they know what they are getting into?
4. Explanation of the purpose of the research, and a description of any foreseeable risks of
participation.
5. Description of any benefits to the participants. (sometimes they receive money or gift
cards)
6. List of resources available in the case of adverse effects. Emotional, psychological, or
medical resource.
7. Researchers should offer to answer questions about the study.
8. Right to terminate their participation at any time. You have to inform them as a research
that they can do so.
Studies that affected people’s lives and had an impact on how researchers conduct their
ethical reviews.
- Milgram Experiment
 - Zimbardo’s Standford Prison Experiment
- Tearoom Trade
Ethics is the most important issue in scientific research!

Qualitative Research
Interviews: 3 main methods used in the interview process
1. Structured
- Around 5-15 mins each.
- Each respondent is asked the same question
- Quantitative in nature.
- Example: you go to a mall someone asks you general questions, like how old are you?
What is your level of education? How much household money comes in? questions about
what they are looking at? Maybe the mall wants to know your opinion on the stores or
thoughts about elections.
2. Semi-structured
- Most used structure of interviews
- Around 30-60 mins sometimes longer depending on the participant not longer than 3
hours.
- Use a set of questions but allows respondents to guide the interview in areas they think are
important.
3. Unstructured
- No predetermined questions and the interview proceeds conversationally
- Last for days or weeks. You might meet with them on a regular basis, lets say 2 hours
every Friday. Let’s say you want to study a specific culture.
The thing the three types have in common is the studying and analyzing of texts that gets really
time consuming.
Interviews and relations of power
There is an unequal power balance between the interviewer and the interviewee.
1. Who is conducts and analyzes the interview?
 2. Whose voices are included in the interviews?
3. How do the experiences of the interviewees get reported?
We put this under reflexivity, giving voice to the person being interviewed.
Strengths and challenges of interview methods
Time consuming
Can you get the trust of the person you are trying to interview?

Strengths(优势):

- Deriving how people make sense of their worlds.

- Giving voice to groups often silenced by others.
- Allowing methods to understand the nuances and follow up on unexpected findings.
- In qualitative research we are trying to find the deeper meaning of what is going on in the
deeper social phenomena. Unlike experiments which try to prove hypothesis.
Challenges:
- Drawing the research sample. Research sample is small, people are worried that the
information may be used against them, they drop out.
- Analyzing copious amount of data (very time consuming) for every hour of interview you
are doing probably 5-6 hours of data analysis.
Another major method is Participant Observation
Involves active participation in daily life activities of those he or she is observing, using
processes of induction as opposed to deduction.
Covert: those in the field are not informed of the researcher’s status (completely undercover).
Semi-covert: only some people involved are aware.
Open: everyone is aware of the researcher’s status
Difficulties with observational research
- It may be dangerous (if you want to study sex workers or biker gangs and they find out
you are lying to them)
- Is it ethical? To tell people or not that you are observing them and justify that you not
telling them benefits that research other than harm it.
- How will you affect the behavior of the group (like the Hawthorne effect)
 - It is very time-consuming (if you want to study toddlers during soccer matches)

Unobtrusive(不显眼的) methods

Content analysis:
- The analysis of texts, movies, tv shows, magazines, blogs, etc.
- Can be qualitative and quantitative.
Secondary analysis
- Use existing data, libraries, government document, churches, information records,
newspapers, magazines (university students do) use other people’s research to inform your
own. You look at the same data but reanalyze it.
Case study
- Indept investigation of one or few phenomena.
- You want to study a police agency during a natural disaster in a developed world. The best
example would be hurricane Katrina, you do not have many events like this so you will
conduct an in-depth study of one phenomenon then you would apply it to others.
- A kitchen sink study because you throw everything at this research. Whatever you find
you will use to help you out.

Cyber research
- New information technology and the internet provide researchers up-to-date resources.
(Instagram, Twitter, Tik Tok, Reddit, etc.) it is similar to content analysis but a brand new
way of conducting research.
Quantitative Research
Two major methods are surveys and experiments
One of the primary methods is the survey
Survey: a method for collecting information by asking members a population a set of questions
and recording their responses. Collect huge amounts of data quickly.
Use interview sand questionaries. Measures attitudes, opinions, beliefs, values, and behaviors to
determine if there is a relationship between two or more attributes of interest.
Example: we want to trach student opinions of crime on campus we do that in September, this
only tells us about their opinion on crime in September.

This is Cross Sectional(横截面的): it compares different groups in one point in time.

- People’s opinions and thoughts can change over periods of time.

- It is important to identify the respondents and the way they differ from one another.
Example: could engineering students see things differently than social science students (history
students)
Questionnaire surveys present the questions on paper on the computer screen. Respondents read
and answer at their own pace.
Advantage:
- Very low in cost
Disadvantage:
- Lower response rate. How do you find a representative sample?
- How can we know if we have a representative sample if we are only asking engineering
and social science students?
- Will people respond to your survey?
- Will people tell the truth about their behaviour and opinions?

Longitudinal research(纵向研究): sometimes we do this during surveys. Maybe we want to

ask students about crime in September and April over a 10-year stretch. Studying prisoners on
parole over a long period of time. 5-10-20 years.
- Diaries
- Interviews
- Health and employment records
Disadvantage:
- Time consuming
- Sample attrition (people drop out of research) people move
Experiments (most controlled out of all research)
The primary purpose is to establish casual relationships between variables under highly
controlled conditions.
3 characteristics
1. Requires consenting participants
2. 2. Manipulation of one or more independent variable (needs an experimental and control
group for comparison. to compare if there is an effect between the IV and DV)
3. Random assignment (mitigates effects of extraneous variables)
Laboratory Experiments
- Allow for more control of stimuli, tasks, information, or situations faced by participant
- Laboratory settings also for random assignment of participants, manipulation of the IV,
measurement of the DV, and control of extraneous variables
Study: we ask university men, put them in a room and show them violent pornography.
Hypothesis: violent porn causes men who watch it to be more aggressive towards women.
What we may do is put up post or put a poster out (recruitment)
Once we have our participants, we split them up to the controlled group and experimental
group (random assignment)
Operationalizing violence and aggression
The experimental group was shown violent porn, and the controlled group was shown and
innocuous movie (Disney).
We introduced the variables; how do we measure now the effects of that variable.
He made the subject write a story after watching the movie. The individuals that watched the
violent pornography were more likely to have violent imagery in their stories.
Another study was conducted the same way but instead of writing a story, the subjects were put
in a room with a female (often referred to as confederates because they are aware of what is
going on). The subject had to teach the female graduate student something (ex:cards) they can
actually punish the female. The ones that watched the violent pornography are more likely to
punish the females. Therefore, proving that men that watch violent pornography are more likely
to be aggressive towards females.
Experiments: Strengths & Weaknesses
Strengths:
- High levels of internal validity due to the ability to control or offset all factors other than
the independent variables that might affect the dependent variable.
Weaknesses:
 - Is it ethical to show university student males pornography?
- Is it realistic to do so? (who watches pornography and then writes a story about it?)
- Can we generalize from such a small sample?
- Can you really control all the relevant factors? (how do we know if what the experimenter
think is violent is what other find violent?) (What we may think is violent may not be
violent for others.)
Quasi-experimental or Field Research
- Field experiments occur when investigators manipulate variables in a natural, non
laboratory setting. They use real situations for experimental purposes.
- Hard to control assignment to groups.
- High external validity compared to lab settings.
- Participants less conscious of their status as experimental participants.
- Little control over extraneous variables.
- There is no control when it comes to the real world.
- The strength is that it is a real-life experiment.
Meta Analysis and Aggregate Data Research
- Meta-analysis is a statistical technique that allows the researcher to combine the
quantitative results from all previous studies on a question.
- Meta-analysis is useful for making sense of large numbers of studies on the same
phenomena that may produce divergent findings. (This is useful for us when there is a
large number of studies on the same phenomenon that has divergent findings.)
- We want to test the relationship between unemployment rates and property crimes.
- We are saying that the higher the unemployment the more property crimes.
- We will answer that question by looking at governmental statistics, social indicators,
uniform crime reports.
- Meta-Analysis can be used to detect relationships and trends.
- The problem can be “can we trust and depend on the agencies taking the count or are they
biased?”
- Police officers get more funding when the crime rate goes up if it goes down it means less
funding. We have to be careful on where this information come from.