Internal validity
the causal connectedness between an independent variable and dependent variable in a particular investigation
Statistical conclusion validity
the validity of the inferences we make from statistical tests. what the numbers are telling us, Whether or not something is statistically significant- p<.05
type 1 and type II errors
Type I error
we reject the null but the null is true (we say theres an effect but there isnt)
type II error
accept the null but the null is false (say there is no effect but there is an effect
Construct validity
the variables represent the intended constructs.
the extent to which variables measure what they are supposed to measure.
External validity
ecological validity, generalizability, representativeness
Longitudinal designs
two-wave design, a pretest and a post test
True experiments
Include manipulation of the IV by the investigator and control of extraneous variables by random assignments of participants to conditions
Quasi-experiments
Investigations in which control procedures are instigated but assignment to conditions is not random
case studies, interviews
examples of non-experiments
Developmental design
concerned with the normative changes in individuals and individual differences in the changes
Cross- sectional designs
cohort x time of assessment design
(2 different age groups are studied at the same time
Predict
Control of relevant outcomes
Search for relationships between variables
3 objectives of traditional science
low power
too many statistical test
Violating assumptions of statistical tests (type I and type II errors)
Threats to statistical conclusion validity
Selection bias
Loss of participants (mortality)- decrease in sample size
Location- where the participants are located can create a change in their responses
Instrumentation
Testing
Attitude of participants- manipulation check to see if they're paying atte
Threats to internal validity
Hypothesis guessing (people try to guess the hypothesis)
Evaluation apprehension (people get nervous being evaluated)
Experimenter expectancies (want to prove their hypothesis as correct)
threats to construct validity
Sequential design
combines cross-sectional and longitudinal designs (studying difference between K and 1st graders, then the following year you study the same kids but they are now 1st and 2nd graders)
Time-lag design
measuring 2 groups of people of similar age but measuring one group at one point in time and the other group later in time. attempts to control for time of measurement effects such as historical effects (cultural changes) associated with a particular birt
Time series/ within-subjects
designs that use the results to establish the existence of cause-and-effect relationships. Manipulation of a treatment and see how that treatment impacts people before and after treatment
standardization
reliability
validity
data analysis
ethics
criteria for evaluating scores
Reliability
dependability, generalizability, consistency of scores
concurrent validity: refers to a comparison between the measure in question and an outcome assessed at the same time
predictive validity: compares the measure in question with an outcome assessed at a later time
criterion validity and its 2 types
face validity
Whether the instrument appears to be a valid measure of some construct
content validity
Assesses the degree to which the content of the instrument constitutes a representative sample of some substantive domain
categorical/ nominal, ordinal, and interval/ ratio
Quantitative data
Narrative (individual experience/ sequence)- stories from individuals
Ethnography- context or culture- observation & interviews
Case study- organization- interviews, observations
Phenomenological- people that have experienced a phenomenon- interviews
Grou
Qualitative data
Test-retest (temporal stability)
Equivalent forms method (parallel)- 2 different versions of the same test
Internal consistency (split half)- half of the test measures one construct and the other half measures another
Interrater (interscorer; interobserve
reliability measures