GGR277 – EXAM REVIEW
(Multiple choice, two long answer case studies)
Introduction to Research
This is an ontological question. What is the social world?
Is it a ‘thing’ that people experience?
Is it created through people’s actions?
Definition: The branch of philosophy dealing with the nature of “being”.
In other words: the question of ‘WHAT IS?’
What is the social world?
OBJECTIVISM: Social phenomenon have an existence independent of social actors or their
perceptions Is it a ‘thing’ that people experience? Is it fixed and out there?
CONSTRUCTIVISM: Social phenomenon and their meanings are continually created by social
Is it created through people’s actions? Not pre-existing and fixed but created through “ppls”
This is an epistemological question. How can the social world be studied?
Can the methods of the natural sciences be applied to the social world?
Do social sciences need unique methods?
Definition: The branch of philosophy concerned with what constitutes knowledge and how
knowledge is acquired. (Likes to reflect on past facts and test facts, usually natural science)
In other words: how do we know about that which exists in the world?
How can the social world be studied?
POSITIVISM: Can the methods of the natural sciences be applied to the social world?
Following the model of the natural sciences:
Meant to provide an explanation of human behaviour
Anything that can be confirmed by the senses: sight, hearing etc can be accepted as
It is a process of fact gathering to test and produce theories
Science must be ‘value free’ (objective knowledge that different people can validate at
different time and place) Clear distinction between scientific statements (what is known/knowledge) and
normative statements (assessments of moral acceptability/feelings – which can’t be
confirmed by the senses)
INTERPRETIVISM: Do social sciences need unique methods?
Critique of positivism –
Role of social scientists is to grasp the subjective meanings of people’s actions
Therefore, subject matter of social sciences is fundamentally different from natural
sciences – and needs different method of study
Social science meant to provide interpretation of human behavior
How do people in a social group/setting understand their actions and the actions
- Individuals are continuously interpreting the symbolic meaning of their environments including
the actions of others.
- Symbolic interactions requires the student to catch the “process of interpretation” through
which actors construct their actions
What is a classroom? (Objectivist/Positivist style)
What are its rules and regulations?
What are its procedures?
Is there a division of labor?
How are these social practices enforced?
What happens if you don’t follow them?
Will the classroom go on without you?
What is a classroom? (Constructivist/Interpretivist style)
What are the social interactions that produce the norms of a classroom?
Who is involved in creating these social norms?
How do these norms change with time? Vary with place?
Soft Constructionism There may be an objective social reality, but our understandings of it are constructed to justify
forms of domination.
How does this space make this room a classroom?
Objectivist/Positivist: facts and fixed knowledge (can be tested within different time and place
and still yield the same results)
Constructivist/Interpretivist: experience and interpretations, sometimes based on feelings
Why does theory matter to research?
First: what is a theory?
“an explanation of observed regularities or patterns”
Middle Range Theories
How are theory and methods related?
The question of ontology and epistemology
The chicken vs. the egg debate
Should research test established theories?
Should research provide the basis for constructing theories?
Should research just be fact - finding projects?
Deductive Method Inductive Method
Research tests established theories Research provides info to construct theory
Theory Identifying a phenomena to be studied
Hypotheses Gather or examine data
Data Collection Construct concepts or theories
Hypothesis confirmed or rejected Gather more data
Revision of theory.. Clarify concepts or theories......
Iterative approach: Moving back and forth
between data and theory
Observation/Findings Theory Research Questions
Start from general area of interest
Narrow down what you’re interested in
Natural Science Uniquely Social Science
Ontological approach Objectivist Constructivist
Epistemological approach Positivist Interpretivist
Every research project is based on some theoretical foundation.
The more aware you are of yours, the better your proposals and research will be.
Respect for Persons: humans should not be treated as mere objects or means to an end. Must
have consent, be informed, confidential and allowed to withdraw. Deceptions: must include
Concern for Welfare: avoids physical, emotional, mental, spiritual, economic and social harm
of individuals, communities and organizations.
Justice: burdens and benefits of research should be spread evenly across society. No person
or group should be exploited (for another group) or excluded from its benefit.
Guest Lecturer: Dr. Dean Sharpe (Research Ethics Board Manager)
Nuremberg Code (1947)
· WWII crimes against humanity
Declaration of Helsinki (1964)
· World Medical Association, drug trials Belmont Report/Common Rule (1979)
· Research scandals (e.g., Tuskegee syphilis study)
Tri-Council Policy Statement (1998, 2010) & MOU
· Canadian research council guidelines
Tri-Council Policy Statement, 2 Ed. (TCPS-2, 2010)
Research ethics: key principles and issues
• Respect for human dignity
– Autonomy . . . e.g., consent
– Welfare . . . e.g., privacy, confidentiality
– Justice, fairness, equity . . . e.g., vulnerability
• Risks versus benefits
System of research participant protection
• Prior review of “protocols”: Office of Research Ethics (ORE) and Research Ethics
• 5 members, women & men
• 2 expertise in relevant disciplines, fields, methods
• 1 knowledgeable in ethics
• 1 no affiliation with the institution
• 1 knowledgeable in relevant law (biomed research)
University of Toronto: 3 boards
• “Social Sciences, Humanities & Education” (& management, law, engineering, . . .)
• Health Sciences
• HIV (for HIV-related protocols)
Research Ethics Culture: Integral Part of Scholarly Process
Excellence in research & excellence in research ethics go hand in hand; not about
• Mandated by research funding bodies • Researchers: Take possession, conception to completion: expert on
groups/topics/methods -> expert on consent/confidentiality; budget for it, have models
on hand, supervise/educate…push back if ill informed
• Reviewers: informed, principles based, tightly reasoned, collegial tone…open to
• Myth that ethics/scholarship totally separate: compelled to comment if
groups/topics/methods unclear, contradictory; expertise/experience/supervision
• Community-based research…conception to completion: consultative, iterative…explicit
agreements on principles
Research Ethics Culture: Proportionate Approach
Delegated: minimal risk, on par with daily life (but see risk matrix) ~90% of protocols in SSH
• Undergrad: Delegated Ethics Review Committees
• Grad & faculty: review by 1 REB member
Full REB: Greater than minimal risk (but see risk matrix)
Continuing: annual renewal, amendment, completion
Proportionate Review & “Risk”
Group vulnerability: diminished autonomy . . . Informed? Free?
• Physiological (e.g., health crisis, service dependence)
• Cognitive/emotional (e.g., age, capacity, recent trauma)
• Social (e.g., stigma, under the table, undocumented)
Research risk: probability & magnitude of reasonably foreseeable, identifiable harm
• Methods invasiveness & data sensitivity
• Physiological (e.g., new diagnoses, side effects)
• Cognitive/emotional (e.g., stress, anxiety)
• Social (e.g., dismissal, deportation, reporting, subpoena)
Review Type by Group Vulnerability & Research Risk
Research Risk Group vulnerability Low Med
Low Del. Del.
Med Del. Full
High Full Full
Research Ethics Issues: Free & Informed Consent
Quality of relationship from first contact to end
• Emphasis on process: not signature on paper; not jargony; not contractual/legalistic (I
the undersigned… I understand that..I understand that..I understand that..)
• Group-appropriate, plain language: who researcher is, affiliation, what they’re studying,
what participation would involve, voluntariness, confidentiality…(check readability)
• Variations, as appropriate, with clear rationale:
– Verbal (literacy, criminality, cultural appropriateness), phone, web
– Age-appropriate assent, alternate (e.g., parental) permission
– Deception & debriefing
– Admin consent, community consultation, ethics approval
Research Ethics Issues: Privacy & Confidentiality
Some projects: name participants, attribute quotes; most projects: protect personal info
• Consider collection, use, disclosure—life of project
• Recruitment: e.g., snowball, distribution/disclosure?
• Data collection: e.g., notes/recording; 1-on-1/groups
• Data management plan:
– identifiers (collected/separated/de-linked?)
– safeguards (double locking/passwords/encryption?)
– retention/destruction (sensitivity, richness, standards of discipline? Not simply:
When will you destroy…)
• Publication: pseudonyms, generics, aggregates • Limits: duty to report (abuse, suicidality, homicidality), subpoena (criminality)
Research Ethics Issues: Conflict of Interest
Commercialization, investment… but typically
Role-based: concurrent dual roles with power over
• e.g., researcher + instructor/minister/manager
• real or perceived, should inform REB and participants of non-research aspects
• may have to manage—e.g., not recruit directly, stay blind to participation until after
• May have to abandon one interest
Research Ethics Issues: Inclusion/Exclusion Criteria
Equity, justice—fair distribution of benefits/burdens
• justify basis for including/excluding
• students sometimes have trouble with complex constructs (e.g., sex/gender/sexual
State consistently throughout protocol sections & appendices (e.g., recruitment, consent)
Research Design: Criteria for Evaluating Social Research
Would the same results be achieved if the same measurement technique were administered
several times to the same research subject?
Are other researchers able to repeat the study and get the same results?
Do the findings of a research study have integrity?
Measurement or construct validity:
· Does a particular indicator actually measure what it is supposed to measure?
Internal validity: · How confident can you be that the independent variable really does have an impact on the
· Are the study’s findings applicable outside of the research setting? How far can the results be
Four Types of Research Design:
Experimental - is an orderly procedure carried out with the goal of verifying, refuting, or
establishing the validity of a hypothesis.
- experimental design manipulates and independent variable to determine its influence
on a dependent variable
- experimental group/ treatment group: independent variable is manipulated on
- control group: those not exposed to the treatment
- Control over research environment
- Quasi-Experiments: Researcher waits for phenomena or changes to be brought by
people not doing the research (natural); this allows gathering of data like they do in experiments
Ex. earthquakes in a particular city
Cross-Sectional - type of observational study that collects data from a population, or a
representative subset, at one specific point in time (SNAPSHOT). (no manipulation of
- can use: structured interviewing, structure observations
Longitudinal - is a correlational research study that involves repeated observations of the same
variables over long periods of time — often many decades.
- Panel Study: People studied at two different occasions
ex. children at 11 doing a follow up interview every 2 years
- Cohort Study: People sharing the same experience
ex. born the same year, graduating from a particular school at the same time (same ppl
may not be selected for the sample each year)
Problems: some may die, move or withdraw from study
Case Study - is a descriptive, exploratory or explanatory analysis of a person, group or event.
ex. a state or province, like a study of rise in cohabitation in Quebec - Critical Case: researcher clearly specifies hypothesis, a case is then chosen to allow
further understanding of the circumstances regardless of whether the hypothesis stands or not.
- Extreme/Unique Case: common focus in clinical studies
ex. study of growing up in Samoa
- Revelatory Case: investigator gets to analyse a previously inaccessible phenomenon,
like previously evidence becomes accessible
Transferability: which parallels external validity
· Do the findings apply to other people and other contexts?
Credibility: which parallels measurement and internal validity
· How believable are the findings?
Dependability:which parallels reliability
· Are the findings likely to be consistent over time?
Confirmability: which parallels replicability
- Would another investigator reach the same conclusions?
Types of questions:
Personal Age, income, status, occupation etc. Rely on memory to answer
About others Bad idea. Best to interview person directly instead of asking
Entity or Event Act as an informant: may be biased in their observations
Attitudes Used in structured interviews and questionnaire research
Beliefs Asked about moral, political, social or religious beliefs
Knowledge Testing respondents on their knowledge in a certain area
· Definition: respondents can reply however they wish.
People can answer in their own terms Time consuming to record variety of
Allow unanticipated answers answers
Taps into a person’s knowledge base More work for respondents; might
Useful in exploring new/changing ideas decrease response rate They can help generate fixed-choice Must code answers, time consuming
answers Answer may be lost during
Transcription may increase in data
· Definition: respondents have a fixed set of answers to choose from.
Answers are easy to process Answers may lack authenticity, or not
Enhance the comparability of answers cover all the options
Help clarify question for respondents May irritate respondents: if they can’t find
Usually quick and efficient a suitable answer + largely impersonal
Circumvent post-coding Provide all possible answers: common
Reduce bias, no need of interpretation compromise to list the most likely to be
from interviewer chosen and leave the rest in the “other
category” + avoid overlap ex: 30-40, 40-50,
50-60 (confuses participants)
A form of closed question, presents respondents with one or more scenarios and asks for their
response if confronted by the matter.
Ex. Choice between paid and unpaid work, gender questions: should men or women be
- Key stage for quantitative research
- To make sense of unstructured and unorganized data from interviews / open ended questions
o Themes and categories of behaviour must be derived from the given data
Job that people like
11 codes: 1 pay, 2 feeling of accomplishment, 3 control of work, 4 pleasant work, 5 security etc. In quantitative research:
· Codes are the tags used to assign the data on each variable to a category of the variable in
question. Numbers are usually assigned to each category to allow easier computer processing.
In qualitative research:
· Coding is the process in which data are broken down into component parts, which are then
(Pre - Coding vs. Post – Coding)
Tips on creating questions
- Avoid long questions
- Avoid ambiguous terms (ex. Very often, quite often, not often etc) instead ask about frequency
- Avoid double-barrelled questions (ex. How satisfied are you with your pay and working
- Avoid generalization, too broad of questions, questions with negatives, minimize technical
Types of Surveys & Interviews
· Surveys / Questionnaires
o Paper & pencil or Virtual / Online or Telephone or Face by face
- Absence of interviewer effects - Greater risk of losing data
(doesn’t influence the - Cannot: probe, ask a lot of