Factorial
How is factorial anova different than one-way BG or WG anova?
Factorial anova has 2 or more IVs
What is the difference between objective and subjective items?
Objective = no judgement required; ex: MC
Subjective = judgment required; ex: SA
What is the difference between descriptive and inferential statistics?
Descriptive statistics describe our sample; inferential statistics infers about the population from our sample data. Both come from the sample but the intent is different.
When plotting results of a factorial anova on a line graph how can we tell, visually, if there is an interaction?
There is an interaction if the lines cross at some point
Describe the research loop.
Circular process that starts with library research followed by RQ formation, research design, data collection, data analysis, RQ testing, and drawing conclusions to add to the literature. Can be applied in three ways: novel RQ, replication, and convergence.
There are three things to interpret in a 2x2 factorial anova. What are they?
main effect A, main effect B, and A*B interaction
Provide an example of a likert-scale item, forced choice item, and multiple choice item.
Likert-style: 1-7 how much do you like puppies?
Forced choice: Do you like cats OR dogs better?
Multiple choice: What is Mickey's dogs name? Pluto, Donald, Daisy, Minnie
Your research question is: There is a significant mean difference in motives for caffeine consumption based on year in college. Identify the target population and provide an example of the sampling frame, selected sample, and data sample.
Target population: college students
Sampling frame: students from UC Wherever
Selected sample: all students enrolled in summer courses who are sent the survey
Data sample: all students who complete the survey
Why is psychometrics important (think: decision theory)?
The practical use of tests involves knowing how to make a binary decision (often times a diagnosis or admission) based on continuous data.
What 3 things are needed for a true experiment?
manipulation of the IV, control of confounds, random assignment
How many cells does a 3x3 factorial anova have?
9
What are the 5 properties of a 'good' measure?
interpretability, norms, validity, reliability, and standardization
What is the difference between critical experiments and converging operations?
Critical experiments: only one 'correct' way to run a study
Converging operations: running different versions of a study and evaluating consistency of results
When creating scales we are trying to measure variables to represent constructs. What is the difference between variables and constructs?
Variables are things we are actually measuring to approximate a construct we are interested in (ex: self-esteem scale is a variable thought to represent the construct of self-esteem).
What is included in the method section of a research paper?
Participants, Materials, and Procedures
When we interpret main effects, what means are we evaluating? When we interpret simple effects what means are we evaluating? When we interpret interactions, what effects are we evaluating?
marginal means; cell means; simple effects
What are the types of reliability we should concern ourselves with when creating scales?
test-retest (getting similar scores over repeated testing), alternate forms (getting similar scores using two versions), internal reliability (consistency of items in the scale), and external reliability (consistency of the scale score)
We found the following results of 3 replication studies and are comparing it to the original study. Decide if there is evidence for replication and why/why not.
1: r(98) = .54, p < .05
2: r(6) = .50, p > .05
3: r(11) = .49, p > .05
4: r(64) = .53, p < .05
There is evidence of replication as the effect sizes are all similar (studies 2/3 are most likely underpowered).
What is the goal of a qualitative review?
To consolidate what is already known about a particular construct.
What are the key terms and variable types used in the following analyses: correlation, chi-square, independent samples t-test, BG ANOVA.
correlation: linear relationship, two quant
chi-square: pattern of relationship, two qual
ind t-test: mean difference, 1 of each but the qual has only 2 levels
BG ANOVA: mean difference, 1 of each but the qual has 3+ levels
Suppose our results look this the following. Interpret the main effects and if there is an interaction present.
Facebook Instagram
Under 30 10 30
Over 30 20 20
Main effect of social media, no main effect of age, interaction (under 30 use IG more than FB but there is no difference for the over 30)
What types of validity should we concern ourselves with during scale construction?
face validity (do the questions 'look' like the measure the construct of interest), content validity (do the questions come from the domain of interest), criterion validity (predictive, concurrent, postdictive), and construct validity (discriminant validity: convergent and divergent)
What are the advantages of a meta-analysis?
allows for comparison across many studies with a single statistic; greater statistical power
Why is a representative sample important?
If our sample is not representative of the population than we cannot be confident that our inferential statistic applies to our population of interest.
RQ: There is a linear relationship between exercise and happiness. We measured 100 people and found: r(98) = .35, p = .025. Interpret this result and what this means (take-home point).
There is a significant, moderate, positive linear relationship between exercise and happiness. As exercise increases, happiness increases. Take-home point: to increase happiness, increase exercise (assuming....?).