Chapter 2 Research Methods PDF
Document Details
Uploaded by HealthyNeon
University of British Columbia
Tags
Summary
This is a textbook chapter on research methods, covering topics such as the scientific method, variables, and different types of research. It describes how researchers conduct studies, focusing on research processes and variables.
Full Transcript
Chapter 2: Research Methods Unit textbook 2.1 Limits of Intuition and Experience problem w basing conclusiosn on intuition → our ideas “feel right” bc we tend to be overconfident in what we know and how well we think we understa...
Chapter 2: Research Methods Unit textbook 2.1 Limits of Intuition and Experience problem w basing conclusiosn on intuition → our ideas “feel right” bc we tend to be overconfident in what we know and how well we think we understand events our biased intutitons tend to discount cases that don’t match what we think 2.2 The Scientific Method: Testing Theories w Data theory-data cycle → developing a theory abt what ppl do and collecting data compared w the theory data either confirm or disproves theory hypotehsis → prediction about what will happen based on theory data → observations form study, usually in numerical form, collected from ppl at certin times or in certain situations replication → study has been conducted more than once on a new sample of participants and found the same results journals → scientists share their scienfitic research in specialzied scientific publications part of publication process involves peer review → other scientists who are experts in that area of research identify the manuscript’s virtues and flaws 2.3 Variables: The Building Blocks of Research Chapter 2: Research Methods 1 variable → something of interest that can vary from person to person or situation to situation meaured variables → height or braking time, observed and recorded in some numeric form manipulated variable- levels are controlled by the researcher b assigning diff partiicpants to diff levels of that variable ex. researcher could manipulate a variable such as degree of intoxication by assigning some people to drink alcohol and others to drink a nonalcoholic beverage only used in experiments 2.4 Operationalizing Measured Variables operationalizing a measured variables means turning a variable into a number self-report → ppl asked to describe themselves on scale from 0 to 1 direct observation → count how many times ppl do a behaviour or record time researchers can rely on technology to operationalize variables not easily observale ex. measuring levels of intoxiation (blood sample); brain activation choice of how to operationalize a variable made for practical reasons 2.5 Surveys: Descriptive Research Based on Self-Report descriptive research → focus on one measured variable at a time with the goal of describing what is typical what do ppl do on average? descriptive research based on self-report, takes form of survey survey research provides concise summaries on a lot of ppl sample → group of ppl participating in research Chapter 2: Research Methods 2 random sampling → ensures we can generalize the sample to full population of interest sample are part of larger group, which is population of interest tell scientists what ppl are doing not why they are doing it 2.6 Descriptive Research Based on Naturalistic Observations and Case Studies naturalistic observation → psychologists observe behvaviour of animals or ppl in their normal, everyday worlds and environments goal → observe wo interferring w thier ususab behaviour can be good measurement of behaviour bc ppl may not always accurately self report what they do case study → naturalistic observatinon but only studied on one individual or very small group at a time bc it’s so rare 2.7 Correlational Studies: Measuring Variables to See How They Are Related correlational research → measure two or more variables to understand research between them data from correlational study presented on scatterplot scatterplot → each dot represents a study participant x axis → one variable (explanatory) y axis → another variable (response) positive correlation (x goes up, y goes up) negative correlation (x goes up, y goes down) zero corelation (no systematic relationship btwn two measured variables) 2.8 Correlations Reveal Relationships but Are Not Enough to Support Causal Claims Chapter 2: Research Methods 3 stronger the correlation → better prediction but it does not allow us to say one variable causes another ex. we see that men who are dependant on alcohol exhibit more agression to wives, we cannot say alcohol caused men to be agressive bc to be convinced one variable causes another, 3 criteria: 1. the two variables must be correlated 2. we must know for certain which variable came first in time 3. must be no reasonable alternative explanations for pattern a. third variable problem → a correlation btwn two variables is explained by influence of third variable 2.9 Experimental Research: Manipulating a Causal Variable to Observe Its Effect experimental variable → conducted in a way that can support causal statments such as: alcohol leads to aggression feeling wealthy reduces generosity researcher manipulates a variable that is hypothesized to be causal → researcher assess effect of that manipulation on one or more measured Chapter 2: Research Methods 4 variables (assign participants tohv one level of variable or other; take drug or dont take drug) independant variable → hypothesized cause dependant variable → measured variable; hypothesized effect random assignment → a procedure used in experimental research in which a random method is used to decide which participatns will be in which group allows researchers to assume that people in each gorup are similar on average at beginning of study expermental group → group where active ingredient or treatment present control group → group in which active ingredient or treatment is absent ideally, control and experimental group hv same experience during day w exception of one variable being manipulated Chapter 2: Research Methods 5 why researchers want participatns in both conditions to think they are being manipulated placebo condition → help researchers separate physoological effects from ppl’s expectations 2.11 Random Assignment vs Random Sampling random → smthg selected or assigned wo any bias random sampling → method for selecting participants who will be in study every person in population of interest hasequal chance of being selected can be used in survey, observaional study, correlational study, or experiment random assignment → method to assign participants to diff levels of the indpendant variable only used in experiments each person in study has equal chance of being in one experimental condition or another 2.12 Comparing and Contrasting Different Methods Chapter 2: Research Methods 6 researchers may start w descriptive study to find out what is typical → move to correlational method to establish relationships btwn variables → use controlled experiments to establish causation 2.13 Asessing Construct Validity: How Well Were the Variables Operationalized? to assess construct validity → how well a variable has been manipulated Chapter 2: Research Methods 7 ask how accurately the operationalziations used in a study capture the variables of interest what questionaire did they use? which behaviours did they observe? reliability → to which degree a measure yields constant results each time administered how well independant variable was manipualted 2.14 Asessing External Validity: Are the People Studied Representative of Broader Population? external validity → the study can generalize to the population of interest random sampling whethehr ersults from one popoulatin of interest can generalize to another 2.15 Asessing Internal Validity: Can We Rule Out Most Plausible Alternative Explanations? internal validity → ability to rule out most plausible alternative explanations study is experiment rather than correlational study sometimes experiemnts are confound → experiemtnal groups accidentlaly differ on more than just the independant variable poor internal validit bc confounds become alternative explanations 2.17 Using Data to Describe People: Central Tendency and Standard Deviation descriptive statistics → summarize participants’ differing responses in terms of what was most typical and how much ppl’s responses varied from the average frequency distribution → bar graph in which possible scores on a variable are listed on the x-axis from lowest to highest, and the total number of people who Chapter 2: Research Methods 8 got each score is plotted on the y-axis three ways to describe central tendancy: mean → average median → middle value mode → msot common value (most number of repeats) measures of variability: standard deviation → how much a batch of scores varies around its mean finding the distance between each individual score and the mean and then computing the average of these distances to measure size of an effect: r (correlation coefficient) and d (standard deviation) 2.19 Statistical Significance descriotive statistics merely summarize bath of scores from study effect sizes → descrie strength of a correlation or degree of difference between groups in experiment inferential statistics → use sample to infer what is true about broader population statistical significance testing → apply rules of logic and probablity to estimate whether results obtained in sample came from particular population ex. if sample randomly selected, assume sample is good estimate of the rest of the population of interest or maybe it was fluke, esp if sample size was smalle ven if randomly assigned null hypothesis → assumption there is no relationship between variables null hypothesis significance testing → assuming nothing is going on (no significant effect) Chapter 2: Research Methods 9 researchers reject null hypothesis if sample’s result only rarely come from null hypothesis population (reject that nothing is going on aka smthg is going on) sample’s reuslt would happen less than 5 percent of the time if the null hypothesis is true p < 0.05 when researchers reject null hypothesis → result is statistically sginficant when effect size is large (strong correlation or verly large group difference) → more likely to be statistically significant meta analysis → researchers find all studies that tested the same variables and average them to estimate the effect size of entire body of studies 2.20 Replication and High Quality Science Chapter 2: Research Methods 10 2.21 Ethical Principles for Research w Humans Chapter 2: Research Methods 11 2.22 Ethical Princniples for Research w Animals replacement: alternatives to using animals in research when possible ex. use computer simulations refinement: researchers should modify experiemntal procedures to minimize animal distress reduction: adopt experimental designs that require fewest animal subjects possible Chapter 2: Research Methods 12