1/55
Looks like no tags are added yet.
experiment
manipulate IV and DV and measure DV, random assignment and controlled environment
independent variable
manipulated, cause of dependent
dependent variable
measured and dependent on IV
condition
a version of the experiment conditions that participant receives, treatment differs between groups while everything else stays constant
random assignment
participants have an equal chance of being assigned to each group and assignment is random
experimental control
variables that stay constant while only one changes
confounding variable
alternative factor that varies with IV and could explain outcome
internal validity
how confidently can researchers attribute an observed effect to the experimental manipulation rather than alternative explanations
external validity
how well might the findings generalize to other people, settings, products, messages, or situations
history
outside event influences participants outcomes during a study
maturation
participants change naturally over time, independent of experimental treatment, ex. fatigue
testing effects
taking a measurement influences participants responses to a later measurement
instrumentation
when the measurement instrument or process changes a study
selection bias
groups differ systematically before the experimental treatment
attrition
participants drop out of a study before completing it
regression to the mean
when unusually high or low measurements tend to be closer to the average when measured again, partly because of random variation
t-test
2 groups, continuous (interval or ratio) data, statistical test used to evaluate whether a difference between 2 means is sufficiently large relative to the variability
ANOVA
3+ groups, continuous (interval or ratio) data, statistical test used to evaluate whether a difference between 2 means is sufficiently large relative to the variability
null hypothesis
groups have equal means
alternative hypothesis
population means differ
one sample t-test
compares one sample mean with specific reference value
independent samples t-test
compares the means of two separate, independent groups
paired samples t-test
one group measured at 2 times, two means and standard deviations compared
t
difference between mean values/standard deviation from mean (standard error)
cohen’s d
measures the difference between two means in terms of standard deviation units
between subjects design
different participants for each condition
within subjects design
same participants experience every condition
multi message design
using multiple messages to test the same variable
grand mean
mean score from all groups
sampling
process of selecting a subset of individuals, groups, or other units from population to participate in a study
probability sampling
participants are selected using a random process, everyone has the same probability of being chosen
simple random sample
assign each member an identifier, use random number generator, have list of population members, can be hard/expensive to do
systematic sampling
select participants at regular intervals from an ordered list, can have bias if the list has a pattern
stratified random sampling
divide population into meaningful subgroups (strata) and randomly select participants from each subgroup
cluster sampling
population divided into naturally occurring groups (clusters), randomly select clusters, collect data from everyone in chosen clusters, people self-identify into groups
nonprobability sampling
participants selected without random selection, can’t calculate sampling error, most common
convenience sampling
select participants readily available to the researcher, quick and easy but biased
purposive sampling
deliberately selecting participants because they have specific characteristics, experiences, or knowledge relevant to the research question
quota sampling
recruiting participants until predetermined numbers for particular subgroups are reached, attempt to resemble population on selected characteristics but this doesn’t always happen
snowball sampling
asking existing participants to refer or recruit additional participants who meet the study criteria
sampling bias
when the way a sample is selected systematically favors certain members of the population over others
representativeness
how well the characteristics of a sample reflect those of the target population
expected effect size
how large of a relationship or difference do you expect to find
statistical power
how likely do you want your study to be to detect an effect if one exists
significance level
what threshold will you use to determine statistical significance
regression analysis
examine how one or more IVs are associated with a dependent variable
regression
estimates how an outcome changes in relation to predictors
residual
error, actual value - estimated
linearity
the relationship between the dependent variable and independent variables should be linear
independence
the observations should be independent of each other
normality
data should be normally distributed
equal variance
standard deviations are similar values
no or litter multicollinearity
this assumes that the independent variables are not highly correlated with each other
R²
how much of the variation in the outcome variable is explained by the predictors included in the regression model
adjusted R²
accounts for the number of predictors in the model relative to the sample size
F-test
determines whether the differences between groups are bigger than we’d expect from random chance alone, by comparing between-group variance to within-group variance.