Lecture 1 - Introduction to Quantitative Methods 09-04-2024 PDF

Quantitative Research Methods in Political Science Lecture 1: Introduction to Quantitative Research Methods Course Instructor: Michael E. Campbell Course Number: PSCI 2702(A) Date: 09/05/2024 ...

Quantitative Research Methods in Political Science Lecture 1: Introduction to Quantitative Research Methods Course Instructor: Michael E. Campbell Course Number: PSCI 2702(A) Date: 09/05/2024  Course Description Purpose of course: to arm you with knowledge required to conduct research using quantitative research methods Overtime, you will be introduced to more complex techniques/formulas to analyze data A mixture of theory and practical knowledge (divided between lecture and tutorial) You will learn the systematic processes underpinning empirical research methods in political science Learning Objectives Learning-Outcomes 1. Understand the purpose and advantages of social scientific research 2. Comprehend foundational concepts and operations associated with empirical data analysis 3. Effectively interpret and evaluate data 4. Utilize various statistical techniques used for data analysis and hypothesis testing  Format Lectures versus Tutorials Weekly Lectures: Thursdays from 9:35 to 11:25 Lectures: focused on foundational concepts, theories, and formulas Tutorials: Before or after Tutorials: focused on use of lecture, depending analytical software (SPSS) on group  Office Hours and Communication E-mail TAs with Instructor: Michael Campbell Instructor Office Hours: TBA questions or Instructor E-mail Address: [email protected] concerns T.A.: Rohit Samaroo T.A. E-mail Address: [email protected] T.A.: Kaia Goodhope Cc Instructor on e- T.A. E-mail Address: [email protected] mail  Course Materials Required Textbook Downloads Textbook: Healey, Joseph F. Christpher Donoghue, and Steven 1. SPSS Analytical Software Prus. 2023. Statistics: A Tool for Social Research, 5th ed. Toronto, ON: Cengage. 2. Varieties of Democracy Data Textbook is on reserve at library Links for each on Additional Readings: found on Brightspace ARES reserves via-Brightspace or Carleton Library Website  Grading Breakdown Grading Breakdown Late Penalties 1. Tutorial Attendance (10%) No due date 5% late penalty / day 2. Assignment #1 (10%) 10 October (11:59PM) without valid reason for extension 3. Midterm Exam (25%) 17 October (in-class) 4. Assignment #2 (20%) Any assignment 5 December (11:59PM) submitted 7-days late will 5. Final Exam (35%) not be graded (without TBA (during exam period) valid reason for extension)  Introduction to Quantitative Research Methods  Research: “any process by which information is systematically and carefully gathered for the purpose of answering questions, examining ideas, or testing theories” (Healey, Donoghue, Prus 2023, 10). What are Quantitative research make use of statistical Quantitative analysis Research Statistics: “a set of mathematical techniques Methods? used by social scientists to organize and manipulate data for the purposes of answering questions and testing theories” (Healey, Donoghue, Prus 2023, 10). Think of statistics as tools that will tell you specific information about your data Systematically means that quantitative research follows a series of predetermined steps  The Scientific Method The Scientific Method 1. Identify the problem (Dependent Variable) 2. Hypothesize a cause (Independent Variable) 3. Define the concepts (what are we talking about?) 4. Gather the empirical data (measurement) 5. Test the hypotheses 6. Reflect on theory 7. Publicize the results Components of The Scientific Method 8. Replicate the analysis were developed by several thinkers, but empirical science can be traced to  Natural Vs. Social Sciences The Natural Sciences: The Social World: Studies natural Is unpredictable phenomena (physics, biology, chemistry) Subject to more ‘Noise’ Is predictable less ‘Noise’ Deals in Facts Less control than natural sciences The Social Sciences: Studies society and behavior Data are less Deals in Facts & Values valid/reliable  Facts vs. Values  Q1: At what temperature does sand turn Facts vs. to glass? Values  Answer: Between 1700 °C and 2000 °C OR… Facts = What Is (Empirical/Objective)  Q2: What was voter turnout in among Canadians who were eligible to vote in the 2021 Federal election? Values = What Ought to Be  Answer: 17.2 million people (or roughly (Normative/Subjective) 62.6% of the eligible electorate) These are Questions of Fact  What happens if are questions normative?  Q3: Should/Ought sand turn into glass at 1700 °C to 2000 °C? Facts vs. Values  Answer: Why even ask? (The results will never change because it Cont’d is based in the natural sciences…)  Q4: Should/Ought we increase voter turnout?  Answer: Requires us to make a value judgement – i.e., “what is an appropriate level of voter turnout?” (Based in the social sciences…)  Value Judgement: a choice between things we believe are right or wrong Quantitative Research Methods = minimize personal opinions/biases as much as possible Opinions are empirically testable (do data support your beliefs?) Results of research should not reflect opinions, opinions should reflect results of research Goal of quantitative methods is to make the research process as objective as possible Will never be completely objective...because we must make Value Judgements e.g., conceptualization and operationalization, selecting variables to represent concepts, etc. Value Judgments  The Role of Statistics in Social Science Theory: a statement about Research the relationship between phenomena The Wheel of Science Hypothesis: a statement about the relationship between variables Observations: what we see when we study our data Generalizations: summary patterns observed in our Source: W. Wallace. 1971. The Logic of Science in data concerning Sociology. Sourced from Healey, Donoghue, and Prus. 2023. expectations about Statistics a Tool for Social Research and Data Analysis. 5th Empirical Research Example Theory: the more informed an Hypotheses: the electorate, the higher the level of more likely they household internet are to participate access, the higher politically the level of voter turnout Empirical Generalizations: in countries where Observations: there are more study data and households with make observations access to the (identify internet, there relationships, tends to be higher patterns, etc.) levels of voter  The Value of Statistics in Political Science Good social science is both empirical and normative – because it deals in both facts and values Common argument (but wrong): ‘social science is not a science because science must be value free’ The better we answer questions, the better equipped we are to understand the world Quantitative methods are not better than qualitative methods, but provide reliable picture of reality A note on Methods vs. Methodology: Methods: tools we use to conduct research (think about different tools in a toolbox) Methodology a concern with the logical structure and procedures of scientific inquiry Therefore, methods are tools that researchers use to collect and analyze data, and methodologies are justifications made by the researcher for the use of these tools Variables and Levels of Measurement Introductio n to Variables Characteris Variables have three primary characteristics: tics of 1. Response categories must be mutually Variables exclusive 2. Response categories must be exhaustive 3. Response categories should be homogenous  Mutual Exclusivity Response categories must Not Mutually Mutually Exclusive not overlap Exclusive 18-24 18-24 Each observation should 23-34 25-34 belong to only one 34-44 35-44 category 45-54 45-54 Above 54 Above 54 Example: a survey question asks you to Categories on the left overlap. This is corrected identify which age group by ensuring the categories do not overlap. you fall into  Exhaustiveness A variable must Not Exhaustive Exhaustive encompass all possible Red Red categories or value Blue Blue If observations are left Green Green unclassified, variable is Yellow Yellow not exhaustive Other Example: a survey asks On the left, not all colors are accounted for. This respondents for their is corrected by adding “other” category. favorite color  Homogeneity Categories should be Not Homogenous Homogenous consistent, measuring the Ford Ford same characteristics or attributes Suzuki Suzuki Honda Honda Categories represent the Tomato Other same underlying concept – to ensure consistency Other and comparability On the left, a tomato is unrelated and does not measure the same concept as other categories. Example: a survey asks This is corrected by eliminating that category.  Discrete and Continuous Variables Discrete Variables Continuous Variables Variables whose basic Variables whose subunits can subunits cannot be divided be subdivided infinitely Will always be a whole Can have decimal points (but number may not) Example: the number of Example: time can be people living in a household measured in hours, minutes, (there won’t be 1.7 people seconds, etc. (10 minutes = living in a house) 600 seconds = 0.17 hours)  Levels of Measurement  Levels of Measurement Level of measurement of a variable: “The mathematical nature of variables under consideration” There are three levels of (Healey, Donoghue, and Prus 2023, 22). measurement: 1. Nominal (least precise) In other words, the level of 2. Ordinal measurement refers to the degree 3. Interval-Ratio (most of precision with which a variable precise) measures the empirical characteristic it is supposed to. We can determine the level of measurement by looking at a variable’s response categories. Nominal Level of Measurement Nominal variables classify observations into categories The categories are different, but cannot be more-or-less / higher-or-lower than another Therefore, the categories and cases cannot be ranked Categories can only be counted and compared  Nominal Variable Example Religious Affiliation Ontario Area Codes Religion Type Frequncy Area Code Frequency Christianity 10 613 10 Judaism 22 753 22 Islam 15 683 15 Hinduism 19 437 19 Buddhism 7 365 7 Other 12 Other 12  Ordinal Level of Measurement A higher level of measurement than Question: How would you rate your shopping nominal level variables experience at the Carleton Bookstore? Response Frequency Contain scores or categories that Category can be ranked from high to low Very Good 10 Somewhat Good 25 Categories/cases can only be described as “more or less” or “higher or lower” Neither Good nor 7 Bad Often found in public opinion Somewhat Bad 27 surveys Very Bad 36 Cannot distinguish distance between  Interval-Ratio Level of Measurement The highest and most precise level of measurement Variables measured at this level have equal intervals – i.e., scores are equidistant Example: income – difference between $1.00 and $2.00 is $1.00 Ratio level variables have a naturally occurring zero – interval variables do not All mathematical operations are possible with interval-ratio variables (addition, subtraction, division, etc.)  Interval-Ratio Variable Examples Ratio Interval Respondent Hours Spent Day Temperature in Studying Per Celcius Week Monday 0 (temp. still exists Michael 0 (real absence of and can go lower) time) Tuesday 3 Brittany 1 Cenk 2 Wednesday 15 Tim 3 Thursday 7 Muhammad 4 Friday 2 Amanda 5 Hank 6 Saturday 3 Aisha 7 Sunday 1 Aki 8 Characteristics of Levels of Measurement  Levels of Measurement Summary Level of Examples Measurement Mathematical Measurement Procedures Operations Permitted Nominal Gender, race, Classification into a) counting number religion, martial categories of cases in each status category of the variable b) Comparing sizes of categories Ordinal Socioeconomic Classification into All operations above, status, attitude and categories plus as well as opinion scales ranking of categories judgements of with respect to one “greater than” and another “less than” Interval Age, number of All of the above, plus All of the above, plus children, income description of all other distances between mathematical scores in terms of operations (addition,