Psy 302 notes
Popular in Social Psychology
Popular in Psychlogy
This 5 page Class Notes was uploaded by Meghan Notetaker on Friday January 8, 2016. The Class Notes belongs to PSY 456 at University of Oregon taught by Jordan Pennefather in Winter 2016. Since its upload, it has received 18 views. For similar materials see Social Psychology in Psychlogy at University of Oregon.
Reviews for Psy 302 notes
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 01/08/16
PSY 303 midterm 2 study guide Study guide 1 Categorical variable: two or more categories, no intrinsic ordering to the categories (sex, ethnicity) Quantitative variable: two or more numbers, meaningful numbers (age, weight, subjective stress) Self-report: How likely are you to get distracted during an important task? Observational: response time or accuracy during an attention task Physiological: pupil size, scalp recordings of brainwaves, functional magnetic resonance imaging Reliability of a measure: how consistent the results of a measure are Test-retest reliability: consistent scores every time you use the measure (correlation of two scores, at two different time points, using same participants) Inter-rater reliability: two or more independent observers produce similar scores, most relevant for observational measures (two observers watching the same children should agree about which children look more distressed) Internal reliability: consistent pattern of responses within a measure (e.g. self report anger questionnaire) Split half reliability: degree to which items of a measure are correlated with the other half (response time on odd vs. even numbered items) Cronbach’s alpha: the degree to which every item of the measure is correlated with every other item; alpha value is avg. of all the correlation coefficients TESTING INTERNAL RELIABILITY Construct validity: Subjective Face validity: ask an expert, plausible? If it looks as if it should be a good measure, it has face validity. Content validity: capture all parts of a construct Objective Criterion validity: related to a concrete outcome it should be related to; two methods: correlational evidence: how strongly does it correlate with the concrete outcome? Known group’s evidence: how well does it discriminate among a set of groups whose behavior is well understood? Convergent validity: correlate strongly with measures of the same construct Divergent validity: correlate less strongly with measures of different constructs Measure cannot be more valid than it is reliable Ordinal scale: applies when the numerals of a quantitative variable represent a rank order Interval scale: applies to numerals of a quantitative variable that meet two conditions. 1: numerals represent equal intervals (distances) between levels 2: no “true zero” Ratio scale: applies when the numerals of a quantitative variable have equal intervals and when the value of zero truly means “nothing” Study guide 2 Surveys: pose questions to people, written questionnaires, on the phone, online, personal interviews Observations: watch people and evaluate their behaviors, (naturalistic, structured) Open ended questions: in your opinion, what are the most important problems facing UO today? Advantages: wide range responses, rich; responses that even the researchers did not think about, greater sense of control for respondents. Disadvantages: not efficient, code or categorize; hard to standardize Forced choice format: which of the following is the most important problem facing UO today? A; overall quality of teaching B: inadequate career advising. Advantages: relatively easier to collect, easier to categorize and evaluate. Disadvantages: constrained by the researchers opinions, not as rich Likert scale: rating scale to indicate degree of agreement. E.g.: overall quality of teaching is an important problem facing UO today. Strongly agree, agree, neither agree nor disagree, disagree, strongly disagree Likert type scale: rating scale that does not follow the Likert scale format exactly. Easy to hard (1 2 3 4 5) same advantages/disadvantages as forced choice format Wording problems Leading questions: wording not neutral, leads the respondent. Potential solutions: word every question as neutrally as possible, measure the extent to which wording matters Double barreled questions: asks two questions at once, potential solutions: ask each question separately Negative wording: using words such as never, impossible, etc. in your question/statement. Potential solutions: do not use negative wording, measure the extent to which wording matters Question order: earlier questions can affect the way respondents understand and answer late question. Ex: order 1: “the nurses are friendly” “it takes a long time to see a physician”. Order 2: “it takes a long time to see a physician” “the nurses are friendly”. What to do: prepare different versions of the survey. Response sets: nondifferentiation; answer all questions positively, negatively or neutrally Yea saying: say yes or strongly agree to every item, people apparently have bias to agree with any item. Solution: use reverse worded items Fence sitting: playing it safe by answering in the middle of the scale (e.g. neither agree nor disagree; I don’t know). Solution: take away the neutral opinion. Behavioral observations: three common problems Observer bias: observers’ expectations influence their interpretations of participant behavior. Observer effect: expectancy effect; observer expectations change participant behavior Reactivity: participants react to being watched Blind design: (masked design) observers unaware of the conditions to which participants have been assigned, observers unaware of what the study is about Ways to prevent participant reactivity: blend in; make yourself less noticeable, observe like a casual onlooker, observe without being seen (one way mirror). Wait it out, let participants get used to your presence (children in a classroom). Measure the behavior’s results. Sampling: Population: group of people that the researchers are interested in learning about Sample: specific group of people who participated in the study Biases in sampling: biased=unrepresentative Convenience sampling: use those who are readily available to participate (those who are easy/convenient to contact) Self-selection: sample includes only those who volunteer (mostly online surveys) Random sampling: every member of population of interest has an equal chance of being selected Simple random sampling: randomly select from the population; no constraints (e.g. randomly select names from a bag, use random number generator). Advantage; high external validity. Disadvantage: may be very hard and/or time consuming Cluster sampling: randomly select clusters within a population, and then use all individuals in selected cluster (e.g. selecting high school students in Lane County, get a list of all high schools in the county, randomly select 3 schools, include all students from those schools) Multistage sampling: randomly select clusters within a population, then a random sample of people within those clusters (e.g. selecting high school students in Lane County, randomly select three schools, random selection of students from each school) Stratified random sampling: select particular demographic categories intentionally, then select from each of the categories Oversampling: intentionally over represent one or more groups Systematic sampling: every third person in a list of potential participants, more common for generalizing to settings Study guide 3 Bivariate association Studies: relationship between two measured variables. Ex: relationship between exercise and depression, link between parental language use and vocab development in 2 year olds. Association does not equal causation Research: degree of linear relationship between two variables, strength of the association: r value = correlation coefficient Positive correlation: increase in value of one variable is associated with the increase in the value of another variable. Negative correlation: increase in value of one variable is associated with the decrease in the value of another variable; decrease in value of one variable is associated with increase in value of another variable Curvilinear association: increase in value of one variable is associated with both increases and decreases in the value of another variable Effect size: strength of the association (r) Statistical significance (p value): probability that the sample’s association came from a population in which the association is zero, likelihood of observing a value (e.g. correlation coefficient, difference in means Outliers: extreme scores; can affect results, especially with small sample sizes. What to do: good practice, always report the presence of outliers. Many options: remove from analyses, report analyses both with and without outliers Restricted range: when there is not a full range of scores on one of the variables, underestimates the association Temporal precedence issue: directionality problem, which one came first? Third variable problem: is there a third variable that is associated with A and B independently? Ex: meeting one’s spouse online associated with happier marriage Moderation; the relationship between two variables changes depending on the level of another variable. (Interaction) Mediation: associations between two variables are explained by a third variable Mediation vs. third variable problem: mediation explains why two variables are correlated; third variable problem is associated with both variables but does not explain the reason for the association
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'