Research Methods Week 5 PDF

Research Methods Week 5 Measuring Behavior Scales of measuring behavior - Nominal - Ordinal - Interval - Ratio - Each extends on previous Nominal scales - Distinct category - “This is that” - Means name - Least amount of info on behavior,...

Research Methods Week 5 Measuring Behavior Scales of measuring behavior - Nominal - Ordinal - Interval - Ratio - Each extends on previous Nominal scales - Distinct category - “This is that” - Means name - Least amount of info on behavior, least precise - Ex. study examines if women are more likely to give her number to a man with a dog. Nominal category is “with or without dog” Ordinal scales - Ordered categories - Can't quantify magnitude of difference - Ex. “top 10 scariest movies of all time” Interval scale - Numerical ordered intervals, same distance apart but no clear 0 - Defining feature: no clear 0 (score 0 another point on the scale, doesn’t mean “nothing”) - Can’t do magnitude comparison (ex. twice as many) - Most used in psychology research - Ex. temperature being 0 isn’t no temperature, it means cold Ratio scale - Have a true zero- complete absence of something - Can have ratio comparison (ex. Twice as likely) - Ex. number of words recalled- can have 0 number of words recalled Mean, Median and Mode - Measures of central tendency: number that tells us what’s in the middle - Mean: average, add all numbers and divide by 10 - Median: middle values, order all numbers see what’s in middle  - Mode: number that occurs the most Reliability - Consistency of a measure (repeatability) - Lack can lead to errors in conclusions or claims Validity - Truth and accuracy - Does the test measure what it claims to measure - If the measure is reliable, doesn’t mean valid - If a measure is valid, it’s automatically reliable Sample - take a part of population - Well defined and unbiased - Representativeness can determine or limit inferences - Reasons: economics, time, manageability and control - Population: everyone Types of sampling - Non probability: not random - Purposive: purposely choosing a particular characteristic, problems with researcher bias Snowball: referral and networking (friend of a friend)- problems with sample bias - Probability: random - Stratified: population divided in subgroups - Proportionate stratified: random from each subgroup based on proportion in population - Cluster: naturally occurring groups Validity of research - Internal validity: truth in the study- IV explains the behavior of DV (ex. If you have a dog, you are more likely to get girls) - External validity: truth in real life Research Methods Week 6 Confounds History - Unplanned, uncontrolled event outside that affects DV  - More likely in longitudinal - Ex. COVID Maturation - Natural process in participant’s life that impact dependant variable - Ex. development - More likely in longitudinal studies Selection bias - Bias and unequal ways of grouping subjects - Groups must be equal in all ways except for Independent variable - Previous experience/culture of subjects impacts studies - Executive Monkey test: shocked monkeys, one could pull a lever to stop shock, other couldn’t, found executive monkey developed ulcers. Wanted to study if executives of a company got more ulcers. Found selection bias through monkeys pre-tested to see which pulled leveler faster, faster pulled the lever were executive monkeys. Study replicated and found opposite results Attrition - Participant mortality- when they drop out of study - Problematic if there is a systematic pattern to people dropping out - Sample at start no longer equivalent to sample at end of study External validity - Results of study generalize to context outside of research context - Does study applies to real world - Externally valid studies generalize to other population, environments, points in time - Must generalize to population; socioeconomic status, ages, education level Other population - Can’t generalize to entire population, but can with a varied sample - Study found 68% from US, 96% from western countries - Students were WEIRD: western, educated, industrialized nations, rich, democratic - Can generalize when doing research with basic processes (ex. Vision, hearing) Other environments - Generalize to outside environments outside of lab - Ecological validity: need to apply basic principles to real life situations (ex. Attention and eyewitness testimony) Other times - Generalize to other time period - Conformity study- Asch, perceptual judgment study on lines to test if confederate says wrong answer participant will follow - Social factors that influence conformity: one person didn’t conform, size of majority, anonymity - In general, basic processes stand tests of time compared to social processes