Sociology 2206A 001: Research Methods in Sociology PDF

SOCIOLOGY 2206A 001: RESEARCH METHODS IN SOCIOLOGY→ Prof. Jasmine Ha ([email protected]) Subject heading→ course code&sec number ALESSIA SARAH CARINCI +251290...

SOCIOLOGY 2206A 001: RESEARCH METHODS IN SOCIOLOGY→ Prof. Jasmine Ha ([email protected]) Subject heading→ course code&sec number ALESSIA SARAH CARINCI +251290689 WEEK 1 → INTRODUCTION TO COURSE→ SEPT 10 - Textbook→ the art and science of social research - Weekly class exercises→ Short7 answer, 2-3 questions, weekly (20 mins) - STATE CLAIM, THEN SAY WHY - INCLUDE ONLY EXPLANATIONS THAT SUPPORT YOUR CLAIM - BEWARE OF CONTRADICTORY EXPLANATIONS - Extra credit→ WEEK 7 AND WEEK 13 (EASY 4%) WHAT IS RESEARCH - High-quality data - Systematic approach - Generalizable - Replicable - Unbiased/value-free (Value-free → always open to other possible answers) - REFLEXIVITY→ The job of researchers is to be value-free - To be reflexive is to do everything in their power to limit personal perspective/bias theory&expected finding→ “In Canada, Black individuals tend to have lower income than White individuals.” [EXAMPLE OF CASUAL OBSERVATION OR HEARSAY] - THIS STATEMENT IS DESCRIPTIVE - Is this trustworthy? - Systematic observation guided by - Cross-reference impacts of racism and data Because of racism [ABSTRACT CONCEPT], Black individuals tend to have lower income than White individuals in Canada.” - THIS STATEMENT IS EXPLANATORY - CAUSATION=/= CORRELATION DIFF TYPES OF SOCIAL SCIENCE RESEARCH 1.1. QUANT VS QUAL (MEDIUM→ MIXED) 1.2. DESCRIPTIVE VS EXPLANATORY (MEDIUM→ EXPLORATORY) 1.2.1. Descriptive research aims to describe or define the topic at hand 1.2.2. Explanatory research aims to explain why particular phenomena work in the way that they do. 1.3. BASIC VS APPLIED RESEARCH (MEDIUM → KNOWLEDGE MOBILIZATION) 1.3.1. BASIC research gathers information and data on a subject (research for the sake of science) 1.3.2. APPLIED research uses that data to look for answers to questions. Applied research takes the data obtained in basic research and applies it to answer a question and provide a possible solution. (to solve real-world problems) 1.3.2.1.1. 2 opposite directions, goals, purposes  1.3.3. They can overlap→ because you can have many goals in any given project - SELECTIVE OBSERVATION→ Data set does not reflect the research question effectively - ILLOGICAL REASONING→ : STATISTICAL ASSOCIATION =/= CAUSATION THE SOCIOLOGICAL IMAGINATION - “PERSONAL TROUBLES” as “PUBLIC ISSUES” - Agency→ the ability for one to make their own decisions based on their own freewill - Structure→ The broader factors of influence (ex, class, religion, etc) - EX) when to get married (sounds like agency) - Social structures impacted- - College degree=/= high-paying job - Rising housing costs - Therefore premarital sex/mo==nonmzritl cohabitation is more acceptable today WEEK 2 → [SEPT 17] → THEORY AS FOUNDATION FOR RESEARCH 1. RESEARCH FOUNDATIONS a. RESEARCH QUESTION i. It must consist of a research question, a theory, an expected finding, and research design Or subject to verifiable observations ii. If you cannot imagine answers, the question is too broad iii. Must be answerable, b. THEORY Steps of the scientific method ○ A statement (or a claim) which explains. Theory must be a. testifiable→ They can be quant or qual examined b. falsifiable → They can be proved wrong c. Generalizeable → Can explain a broad class of events d. probabilistic → Refer to what is likely, not what is definite i. SHOULD BE ABSTRACT NOT SPECIFIC Theory is a minimum one claim/statement, dont have a name More complex theory→ many related statements ○ ex) 1. Childhood socialization→ level of self control ○ Pt. 2. Level of self control → crime potentials Grand Theory→ becomes a basis for new theories ○ Becomes A PARADIGM c. PARADIGM i. Broad set of taken for granted and often acknowledged assumptions about how social reality is to be defined ii. Positivism 1. The paradigm holding that all knowledge can be confirmed or refuted through empirical observation iii. Postmodernism 1. A paradigm characterized by significant skepticism of claims about the general truths or facts 2. Therefore try to explain how humans arrive at subjective truths SOCIOLOGICAL PARADIGMS 1. Structural Functionalism: ○ Primary Level: Macro (examines society as a whole). ○ Main Question: Does this social phenomenon help maintain stability in society? ○ Sample Theory: Functional Role Theory (focuses on how different parts of society work together to maintain order). 2. Conflict Theory: ○ Primary Level: Macro (focuses on large-scale structures and power dynamics). ○ Main Question: Does this phenomenon benefit some groups at the expense of others? ○ Sample Theory: Critical Race Theory (explores how race and power create inequality in society). 3. Rational Choice Theory: ○ Primary Level: Macro, meso, and micro (can be applied to individuals, groups, or societies). ○ Main Question: Does this social phenomenon maximize benefits and minimize costs for individuals or groups? ○ Sample Theory: Time Allocation (looks at how individuals decide to use their time based on rewards and costs). 4. Symbolic Interactionism: ○ Primary Level: Micro (focuses on individual interactions and meanings). ○ Main Question: How do people experience and interpret face-to-face interactions in their everyday lives? ○ Sample Theory: Looking-Glass Self (how we see ourselves based on how we think others see us). These paradigms provide different lenses through which sociologists can analyze social issues, helping to reveal varying dimensions of human behavior, power structures, and social stability. d. UNIT VS LEVEL OF ANALYSIS - “On average, students in Ontario have better mental health than students in other provinces.” - Unit of analysis→ province - Level of analysis→ Meso → Unit and level of analysis should match. - ECOLOGICAL FALLACY→ Insights from macro-level to predict micro-level outcomes -A nation does not account for each individual under their location - misleading and incorrect - REDUCTIONIST FALLACY→ Study individuals on the micro-level & draw a conclusion on the macro-level - Outcome did not account for the individual perspectives of the final decision or why it changed 2. THEORY AND EMPIRICISM a. CONCEPTS & VARIABLES i. Research= Theory + Empiricism (Data) ii. Research Methods- ways to obtain data to answer research question 1. Surveys 2. Experiments 3. Ethnography 4. interview/focus group 5. Material based methods (content analysis) 6. Evaluation research iii. Theory- Data relationship 1. Deductive research 2. Inductive research Deductive research Inductive research - Start with theory - Start with some clues, some data, some - Develop hypothesis/proposition situations - Figure out a way to test it using empirical data - Develop a hypothesis/theory from that b. RELATIONSHIPS c. HYPOTHESIS/PROPOSITION WEEK 3 → [SEPT 24] → PROCESS 1. THEORY AND HYPOTHESIS ○ Elements of theory THEORY→ a set of related statements/claims about something 1. theory= relationships among concepts CONCEPT: name of a social artifact or phenomenon 1. abstract&very short: one or two words a. Example; Crime, Fmily, Mental Health ○ VARIABLES Variables operationalize concepts so they can be concretely measured. A variable must vary, i.e., have a set of attributes (or values).\ 1. Example: a. Concept: Poverty b. Unit of analysis: Household c. Variable: Household income relative to the poverty line d. Two possible values: i. 1.“Above the poverty line” ii. 2. “On or Below the poverty line” VARIABLE X UNIT OF ANALYSIS 1. Operationalize the concept “poverty” on a different unit of analysis? Unit of analysis: Individual Variable? Possible values? Unit of analysis: Country Variable? Possible values? ○ OPERATIONALIZATION CONCEPT → DEFINITION → VARIABLE (S) Concept: social class Unit of analysis : individuals Definition A: The honor or prestige associated with a person in a society B: The level of wealth a person have Etc VARIABLES A) More options  ○ 1. Personal income, education, occupation prestige ○ 2. Education, household occupational prestige, household income ○ 3. Etc. B) Even More Options ○ Income ○ Income and assets There are many ways to operationalize a single concept Diff. units of analysis Diff. definitions Diff. Variables USE THEORY TO HELP Specific definition Specific variables ○ Your decision to apply theory as is or modify it ○ HYPOTHESIS : A testable statement of a relationship between 2 variables INDEPENDANT VARIABLE→ THE CAUSE DEPENDANT VARIABLE (THE EFFECT) Research and Null hypothesis PROPOSITION→ a statement about relationship between 2 variables, but not for empirical testing ○ TYPE OF THEORY AND HYPOTHESIS DESCRIPTIVE THEORY EXPLANATORY THEORY EXPLORATORY THEORY ○ —--------------------------------------------------------------------------- HYPOTHESIS OF DIFFERENCE (general, guiding) HYPOTHESIS OF ASSOCIATION (cause and effect, analyzing relationship) CAUSAL HYPOTHESIS(core independent and dependant variable, prominent cause and effect) PROPOSITION 2. CONCEPTUALIZATION ○ RELATIONS AMONG 2 CONCEPTS EX) “access to counselling services improves students’ mental health” 1. Negative or positive relation? a. POSITIVE! ○ 1. RELATIONSHIP AMONGST 3 CONCEPTS SPURIOUSNESS- real relations?  Students height→ maths knowledge 1. Cofounder for both concepts= SCHOOL GRADE LEVEL ○ 2. RELATIONSHIP AMONGST 3 CONCEPTS -UNCOVERING THE MECHANISM MEDIATION: concept 3 is essential for all MODERATION: Relationships between concepts 1 & 2 vary by the levels of concept 3 EXAMPLE: POVERTY - Define “poverty” - Unit of analysis: person, household, country, etc - Write sentence describing the nature or meaning of concept - “Have less”, “have-not”, “minimum”, “basic needs” - Operationalize “poverty” - how would you measure “poverty” - `to measure (w/ data): how can you tell if a person is poor using your definition - Indicator: a household will be classified as “poor” if they have the value of X in the variable Y. - Possible values of the variable must allow you to distinguish between poor and not poor households  ○ VARIABLE TYPES ○ -THIS INCLUDES SCALE -this may include all values ex) days in week -properly ordinal variables, but may be treated as interval/ratio - VARIABLE TYPE IN SURVEYS Using scale (smallest/largest) Ask people to write number - VARIABLE TYPES - Researchers may choose to modify the possible values in a variable - More detailed→ Less detailed - Survey question: What is your total household income after tax? (Please enter the amount in CAD) - Income as an interval/ratio variable - Income as an ordinal variable - Less than $35k ~ “Poor” - Between $35k and $249,999 ~ “Not poor, not rich” - $250k and above ~ “Rich” - MORE COMPLICATED VARIABLE COMPOSITIONS - What we’ve done so far: - One concept (poverty) → one variable (income) - In reality: - One concept→ Multiple dimensions→ Multiple variables [COMPOSITE VARIABLE ] - COMPOSITE VARIABLE→ one that averages a set of variables to measure 1 concept - ABSTRACT- 1-2 WORDS SEVRAL DIMENSIONS ABSTRACT [1-4 WORDS] SEVERAL VARIABLES 3. VALIDITY & RELIABILITY [QUANTITATIVE] ○ RELIABILITY→ how dependable? Will the measure yield the same results when used over and over again? MEASURES OF RELIABILITY 1. Internal reliability (FOR COMPOSITE VARIABLES) 2. Intercoder reliability 3. Precision 4. Robustness tests a. Split-half method  b. Test-retest method c. Pilot testing ○ VALIDITY→ how accurate? Do the measurement capture the concept being studied? MEASURES OF VALIDITY - INTERNAL validity - Face validity - Criterion-related validity - Content validity - Construct validity - EXTERNAL VALIDITY - Representativeness - How “real” is the study Construct→ based on theory and logical reasoning, creating expectation→ “if i am measuring this, it should constantly correlate with other things” Criterion→ gold standard of measure → nurse records height, criterion involves validity and authority of measure Construct→ do many items correlate (is it a construct) → UNDERLYING CONCEPT→ WEALTH Face→ is it feasible and practical  ○ [QUALITATIVE] CREDIBILITY: Do the respondents argue about how the researcher is presenting and interpreting their words? DEPENDABILITY: are the findings consistent? Would the findings be the same in different circumstances or contexts? WEEK 4 → [OCT 1] → 1. RESEARCH ETHICS 2. RESEARCH ETHICS REVIEW BOARD -researchers considered “agents of knowledge” - general public AKA- audience WHAT IS RESEARCH ETHICS? - ETHICS = moral system - Deliberate harm & unintended consequence - Unethica;l research does not have to come from a “bad person” - Research ethics have everything to do with participation - NO NEED FOR ETHIC APPROVAL→ when researchers observe social activities in a public space - This- requires NO PARTICIPATION - Research methods that do not require human participation? → medical, public material, historical public archive - 3 ETHICAL PRINCIPLES/considerations [respect, beneficience, justice] 1. RESPECT a. Noone can be forced to participate in a project b. VULNERABLE POPULATIONS- unable to give INFORMED CONSENT i. Freedom to say yes or no to participating in a research study ii. All possible risks and benefits must have been properly explained iii. Informed consent- written, verbal, implied, Continuous(participants can leave anytime) 1. Underage 2. Language barrier? 3. Diminished mental capacity a. refugees 4. Power dynamics a. ex) if prof asks b. If freshman, is underage i. Highschool students - majority are underage-parentds and legal guardians signing informed consent c. Refugees- researchers are apart of the organzation that is helping them VOLUNTARY PARTICIPATION? - Your professor asks for volunteers to participate in her research project during class. Ethical or not? UNETHICAL (PUBLICLY ASKING FOR VOLUNTEERS) - Offer CAD $500 cash for each participant in a 30-minute experiment. Ethical or not? UNETHICAL (too much money is considered power imbalance) - Must rationalize the compensation in direct correlation with the expenses required to participate  CONTINUOUS CONSENT EXAMPLE -there is no explanation as to what is asked from participants - duration of time - “random selection of prisoners or guards” DECEPTION - Researchers conceal the real research questions from consent form - ex) “study on racism vs study on social groups from different backgrounds” - ETHICAL- people will act accordingly and become overly conscious if they know the study is about racism 2. BENEFICIENCE a. Do no harm b. Minimize possible harms i. Anonymity 1. Ideal, but often not possible in social research 2. Are online surveys always anonymous a. Solution → De-identify data ii. Confidentiality 1. But can these promises always be kept? a. Researchers do not have the legal right to refuse to cooperate with law enforcement in order to protect research subjects i. Court & legal enforcement can seize information about illegal activities b. Social scientists can be called to testify in crim/civil cases 2. Sometimes ethical/legal requirement is to not maintain confidentiality a. US states- child abuse, self harm 3. Digital data lost, stolen, or hacked  4. Deductive disclosure risk a. Create bigger categories to falsely display c. Maximize possible benefits i. Articulating social problems in an attempt to create new findings 3. JUSTICE a. Research participant group is different than those who will benefit from results of their participation i. EX) 1. Studying people who cannot afford to say no, and give benefits to another group 2. Study people SIMPLY because they are easily available, in a compromised position, or easily manipulated. a. CONSEQUENCE? i. Law suits, if no law suit, researchers can get away with it TEAROOMS-HUMPHREYS STUDY - Where ethics draw the line in study→ found license plates to find home addresses - Deception 2 - told these people that they were ”randomly selected” - Caused harm with no clear gain as the research were outing people (VIOLATING CONFIDENTIALITY) OTHER STUDIES TO KNOW: - The Nazi experiments (Nuremberg Code) - Tuskegee experiment - Wichita Jury study - Milgram experiment - Zimbardo experiment - Tearoom Trade study - Be familiar with the study design, what was wrong, implications, and a general historical timeline RESEARCH ETHICS REVIEW BOARD (REB) - A researcher submits a research protocol for ethic approval before starting data collection - Research question - Intended methods and procedures - Target population - Recruitment methods - Possible risks & Benefits - Possible injustice - In the US (textbook), the equivalent is the Institutional Review Board (IRB) REB - Human subjects research - What constitutes research w/ human subjects - Question remains under debate, research changes overtime WEEK 5 → [OCT 8] → SAMPLING STRATEGIES 1. NON-PROBABILITY SAMPLING MIDTERM REVIEW 2. POPULATION & SAMPLE - Target population - Population parameters - Census - Sampling frame - Sample: a subset of population selected for a study - Sampling: process of deciding who gets included in sample - Target pop. : group about which social scientists attempt to make generalizations about - Can be very specific- western graduate students - Can be abstract- youth, people in general - Non human populations- tiktok, corporations, legal documents - Unit of analysis→ I will need to collect data from ______ to answer my research question. - Target pop. of —---------- Is- POPULATION - : census is a study that includes data on every member of a population - Population parameter: Represents “TRUE VALUE” of the population - Not feasible in social research: - Time, resource, frequency (every 5,6,10, 15) - Limited number of questions SAMPLE - : Subset of population - Save time and resource - Ask questions to relevant to research interest - Sampling: action to draw sample CASE STUDY Sampling strategy OBSERVED VALUE = TRUE VALUE + SYSTEMATIC ERROR + RANDOM ERROR - Types of errors - 1. Systematic errors: cannot be estimated, only discuss direction of bias - CANNOT BE ESTIMATED WITH STATISTIC - 2. Random Errors: unbiased, estimated using statistics - Representativeness: sample MIRRORS popultion -Random sample→ handpicking 8 people , bias due to being drawn to specific demographics - Random number generator→ unbiased, would be a random sample PROBABILITY SAMPLING - IDENTIFY: - Target pop. - Desired sample size - The sampling frame - Select sampling process - Simple random/systematic - Cluster - stratified RANDOM ERRORS SYSTEMATIC ERRORS - ex) 1 - SIMPLE RANDOM SAMPLING - Each individual has the same probability of being selected & each pair of individuals has the same probability of being selected. - Sampling process: - 1. Obtain the sampling frame - 2. Generate a set of random numbers and select individual corresponding to the selected numbers - Why do this?  - Easy, but the sample may be nonrepresentative due to pure chance - ex) 1B→ SYSTEMATIC SAMPLING - Each individual has the same probability of being selected. - Sampling process: - 1. Obtain the sampling frame - - 2. Decide on the sample size = population size - 3. Random select the first case, then select every nTh case in the list. - Why do this? - Easy, but consecutive cases will not be selected. - Order in the sampling frame may create bias - ex) 2. CLUSTER SAMPLING - SAMPLING PROCESS - 1. Divide the target population into clusters (e.g., cities within Canada, classrooms in a high school) - 2.Select clusters randomly - 3. Get sampling frame for all selected clusters - 4. Select individuals randomly from the selected clusters - Why do this? - Improved feasibility, lower cost - ex) 3. STRATIFIED SAMPLING - SAMPLING PROCESS - 1.Obtain the sampling frame - 2. Divide the target population into population is divided into strata (e.g., gender, social class) - 3. Select individuals randomly from all strata - 4.Number of selected individuals reflects the proportions from each stratum - Why do this? - Prevent samples from becoming non-representative due to pure chance - Oversample for small groups RECAP - WEIGHTING - How much sample members “count” when producing estimates - The oversampled group should receive less weight than other members of sample - ex) SIMPLE RANDOM SAMPLING was used to create a sample of 20 students from a total of 200 students in our class - How much weight should we assign to each student in the sample? - Sampling method: simple random , no oversampling - weight= - Each respondent “speaks for ” 10 PPL IN POPULATION - OVERSAMPLE - In a class of 100 students- we want a stratified sample of 10. - Stratified by international students status (90% domestic, 10% international) & by gender (male 45%, female 55%) - PROBLEM: - Stratification by international status: - How many domestic students? - How many international students? - Stratification by gender: How many respondents in each? - Male domestic - Female domestic - Male international - Female international - EX) - In a class of 100 students, we want a stratified sample of 10. - Stratified by international student status (90% domestic, 10% international) & by gender (male 45%, female 55%) - The solution: Sample 2 international students instead of 1 - International students are oversampled by a factor of 2 - Sample size = 11 (9 domestic + 2 international) Each international student should have smaller weight, specifically 1/2 = 0.5 times the weight of - domestic students in the sample In a class of 100 students, we want a stratified sample of 10. ○ Sample size calculation: Oversampling: ○ International students are oversampled by a factor of 2. This means that international students will have  a higher representation in the sample compared to their actual proportion in the population. Weighting: ○ ○ This weight reflects the fact that international students are oversampled, so their contribution is reduced by multiplying their weight by 0.5 to account for the oversampling. 2. Postsurvey Weighting Response Rate Formula: Nonresponse Rate: Systematic Errors: Nonresponse can create systematic errors, such as when a specific group (e.g., people without access to personal computers) systematically fails to respond to a survey. ○ Example: Access to personal computers → nonresponse to an online survey. Adjusting for Nonresponse: Postsurvey weighting can be used to increase or reduce the weight of specific respondent groups to account for nonresponse. This ensures that the sample more accurately reflects the population. 3. Postsurvey Weighting Example Scenario: Based on Census data, we know that 20% of the population consists of older adults. However, due to nonresponse bias, only 10% of the respondents in the sample are older adults. Adjustment: Since older adults are underrepresented, their weight in the sample should be increased. The weight adjustment factor can be calculated as:  ○ Interpretation: The older adults in the sample should have twice the weight of younger adults in order to correct for their underrepresentation. 4. Why Probability Sampling? Unbiased Sampling: In probability sampling, every member of the population has an equal chance of being selected. This means that the sample is equally likely to underestimate or overestimate the population parameter. Sampling Error: The difference between the sample estimates and the population parameters is due to random chance. ○ Systematic errors are avoided through proper sampling methods. ○ Random sampling errors still exist but can be estimated statistically. 5. Random Sampling Error Randomness in Samples: If we draw 1,000 random samples, each sample will likely yield 1,000 different results. Reality: In practice, researchers can only draw one sample for their research project, which may or may not reflect the true population. Error Estimation: Statistical distributions (such as a normal distribution) are used to estimate errors in the sample, helping to quantify the uncertainty associated with the sampling process. 6. Margin of Error The margin of error represents the amount of uncertainty in an estimate. It equals the distance between the sample estimate and the boundary of the confidence interval. ○ Confidence Intervals: Common levels of confidence are 95% and 99%. Gallup Poll Example: ○ According to a Gallup poll, 43% of Americans approve of the president's performance. ○ The margin of error is 3 percentage points, with a 95% confidence interval. ○ Translation: We can be 95% confident that the true approval rating of the president is between 40% and 46%. ○ Confidence Interval Calculation: - - MARGIN OF ERROR & SAMPLE SIZE - Larger sample=smaller margin of error - ex) - - - - NONPROBABILITY SAMPLES - Individuals in the sample are not randomly selected - Systematic errors: impossible to quantify, only possible to predict the direction of bias - Not representative of POP. - Low generalizability - Can a probability sample be non-representative? - Yes!!, by chance [aka stratified sampling] - Might draw a sample completely diff. Than population - WHY NONPROBABILITY SAMPLES - Diversity of representative samples make detecting cause and effect relationships more difficult - ex) how many people would response to cash incentives? Many experiments use samples of college students - WHY NONPROBABILITY SAMPLES? - Gather better info on these samples - Qualitative research and in-depth examinations of subgroups - Non-probability sampling - Convenience - Select anyone willing to participate - cheapest/easiest method - Systematic errors → selection bias (most of time its loved ones) - Purposive - Researcher articulates reason (cases based on key features) - access/quality - Typicality - Extremity - Importance - Deviant cases - Contrasting outcome - Key differences  - Past experience/intuition - Sequential - Collecting additional data based on findings from previous data collected - Key informants - Sampling for range - saturation - Snowball - Starts w 1 respondent who meets the requirements for inclusion - 1 PERSON GIVING CONTACT TO MORE CONTACTS OF DESIRED GROUPS - Ask to recommend other people to contact - Useful for studying “hidden populations” - Hidden populations examples? - Gender identity (20-30 yrs ago) - Occupations (sex work) - MINIMIZE SYSTEMATIC ERROR→ 1. RANDOM SELECTION 2. ADVANCES STRATIFIED SAMPLE 3. OVERSAMPLING - MINIMIZE RANDOM ERROR→ 1. INCREASING SAMPLE SIZE