Research methods
Terms
undefined, object
copy deck
- .15=
- large effect
- scientists 2 jobs
- discover and document phenomenon; explain phenomenon
- experimental group
- participants to whom some sort of active manipulation is given
- public verification
- research must be conducted so that others can observe,replicate and verify it
- MANOVA
- allows us to examine more than 1 IV at a time
- things that contribute to measurement error
- situation factors,transient states, stable traits, measure characteristics, actual mistakes
- DF within groups
- N-K (# of part. -# of groups)
- inferential statistics
- let researchers determine how the results of their participants can generalize to other populations
- demand characteristics
- subtle clues to participants that tell them how they should behave in an experiment-solution, use a double blind procedure
- positive skew
- way more scores toward the lower end of the distribution
- pre-test sensitization
- having practice of taking the initial test can affect scores on the post-test
- applied research
- find solutions to real world problems
- outliers
- any score that is at least 3 SD's from the mean
- differential attrition
- people in 1 group drop out at a higher rate than people of the other group
- evaluation research
- program evaluation
- effect size
- assesses the strength of a relationship between 2 variables
- scientific method
- theory,hypothesis,test,re-theorize,re-hypothesize,retest
- 3 goals of behavioral research
- predict,describe and explain behaviors and mental processes
- representative sampling
- ensure participants are representatives of the entire population that you're describing
- 3 criteria of scientific research
- systematic research, public verification, solvable problems
- t-comp > t-crit
- reject the null
- experimenter expectancies
- experimenters biases affect their interpretation of the data
- DF
- # participants -2
- one tailed test
- the experimental hypothesis specifies a direction
- placebo effect
- physiological or psychological change that occurs just b/c it has been suggested. solution- placebo control group
- true score..
- score the person would have obtained if our measures were perfect
- Factorial ANOVA
- more than one IV each w/two or more levels
- control group
- participants who receive no manipulation
- reasons research is important
- help us evaluate common ideas/beliefs, helps keep you up to date on information that is relevant to your life, makes you an informed consumer,helps you become an authority/expert in certain topics
- what makes an experiment
- researcher must manipulate or vary at least one IV, must assign people to groups in a way that ensures their initial equivalence,must control for extraneous variables
- between groups design
- different people are in different groups/levels of the IV
- longitudinal
- the same group of participants is questioned more than once
- systematic empiricism
- observing the world in an organized,controlled and consistent matter
- increasing reliability of a measure...
- standardize administration, thoroughly train observers, clarify funky instructions and questions,minimize errors in coding data
- variability
- variance, sum of squares, range, standard deviation
- one way ANOVA
- IV with more than 2 levels
- cross-sectional
- a group of participants is surveyed at 1 time only
- reducing bias
- make sure items apply to all people who may take your test
- 3 criteria for causal variable
- cause must come before effect- there must be correlation- the causal relationship must be nonspurious
- meta-analysis
- research procedure that combines results from many studies-specifically looks are systematic variance across studies-research must compute effect size from each study-the effect sizes are averaged to determine an ultimate strength of relationship
- error variance
- variance in people's behavior that can't be accounted for; variance that is unrelated to the other variable in the study
- A hypothesis is deduced
- from a theory
- discussion
- put your results in a larger context
- .01=
- small effect
- interrater reliability
- consistency among two or more researchers who observe and record participants' behavior
- matched subject
- find a pair or set of participants who are similar on many important characteristics, then randomly assign one member of each pair or set to a different IV level
- random groups
- you randomly assign participants to the levels of the IV
- observed score=
- true score+ measurement error
- measurement error..
- the deviation of a person's observed score from their true score
- t-test-
- an inferential statistic that tells if there is a difference between the means of two groups
- withing groups design
- the same people are put into different groups/levels of the IV
- t-comp< t-crit
- fail to reject the null
- operational definitions
- defines a concept by saying precisely how it is measured in a given research study
- abstract
- comes first in the paper but is written last. gives an overview of the entire study
- maturation
- participants development over time influences results
- 3 types of one way designs
- random groups, matched subject, repeated measures/withing subjects
- repeated measures/withing subjects
- each participant experiences each level of the IV at a different time period
- negative skew
- way more scores on the higher end of the distribution
- introduction
- do a literature review and tell readers what your research question and hypothesis are
- basic research
- understand psychological processes; gain knowledge
- systematic variance
- variance in people's behavior that can't be accounted for; variance that is unrelated to the other variable in the study
- test-retest reliability
- consistency of participants responses on a measure over time
- type I error
- we claim that the IV had an affect on the DV but it did not
- confounds
- fatal flaw of an experiment
- .06=
- medium effect
- biased assignment
- participants in 1 group are systematically different than participants in other groups at the study's onset-solution-random assignment
- inter-term reliability
- consistency among the items of a scale
- two tailed test
- the experimental hypothesis does not specify a direction
- 4 types of research
- descriptive, correlational, experimental,quasi-experimental
- solvable problems
- only investigate questions that can be answered given current knowledge and research techniques
- history effects
- an outside event occurs during the experiment that influences the results
- descriptive statistics
- summarize the behavior of the participants in your study
- DF between groups
- K-1 (k=# of groups)
- threats to internal validity
- biased assignment, differential attrition, pre-test sensitization,history effects, maturation
- type II error
- we say that there is no affect of the IV on the DV but there is
- successive independent samples
- 2 or more different groups of participants answer the same questions at different time points
- method and results
- step-by-step of the experiment, report statistical information that results from your data analysis