Exam 2: Quanitative Flashcards
What are some characteristics of quantitative research ?
- describes with numbers a variety of different variables
- tests relationships and differences between variables
- seeks to explain cause-and-effect relationship between variables
- large samples
- numeric data is summarized and analyzed using statistics
- extraneous variables are controlled for in the research design
What is a independent variable ?
variable that is believed to be the cause or influence on the outcome of interest
What is a dependent variable ?
variable that is hypothesized to depend on or be caused by the independent variable
- outcome variable
- relationship does not imply causation
What is a extraneous variable ?
variables which confuse or confound the relationship between the independent and dependent variables
What is an experimental research design ?
- tests difference in effectiveness of an intervention
- participants are randomly assigned to the treatment or control
- infer causation
- manipulation of independent variable
- control and randomization
What is a quasi-experimental research design ?
- manipulation of the independent variable - lacks randomization or a control group
- infer causation
- researcher wants to test an intervention but for some reason can’t randomly assigned subjects to different groups
What are some advantages and disadvantages of an experimental design ?
Advantages:
- most powerful design
- most confidence in causation
Disadvantages:
- variables must be able to be manipulated
- may be impractical/not feasible/costly
- may be difficult to apply to the “real world”
What are some examples of quasi-experimental designs ?
- non-equivalent control group design
- after-only non-equivalent control group design
- one-group design (pretest-posttest)
- time series
What is a non equivalent control group design ?
- similar to the true/classic experimental design
- only difference is the lack of random assignment
- has a experimental and control group
What is a after-only non-equivalent control group design ?
- are missing the pretest (still has posttest)
- lack of random assignment
What is a one-group design (pretest-posttest) ?
- only 1 group is available for the study
- no control and no randomization
What is a time series ?
- only 1 group is available for the study
- done over a longer period of time
- measures/testing done more than just before/after
What are some advantages and disadvantages of a quasi-experimental design ?
Advantages:
- more practical, feasible, and generalizable
- more adaptable to “real world” application
Disadvantages:
- less confidence in cause-and-effect conclusion
- can’t control for differences within groups
What is a non-experimental quantitative research design ?
- the independent variable is inherently nonmanipulable or there might not be one
- preliminary studies in building a body of knowledge
- hypothesis generating
- collecting information
- not used to establish causation
What are some examples of non-experimental quantitative study designs ?
- descriptive
- correlational
- developmental
What is a descriptive study ?
- accurate information about subjects (characteristics)
- not looking at relationships between variables
- no independent or dependent variables, but lets you think about possible relationships for a future study
- focus on breadth of information more than depth
- aka) exploratory or survey research
- provide background for other studies
- economical and accurate
What is a correlational study ?
- seek to establish relationship(s) between/among variables
- concerned with magnitude and the direction of the correlation between variables
- correlation does not establish causation
How are correlational studies measured ?
with magnitude (how big the relationship is)
- correlations can range from 0 to 1
- the closer the correlation is to 1, the stronger the relationship
- direction= is the relationship positive or negative
- can range from -1 to 1
What is a developmental study ?
- uses a time perspective
- can assess relationships and/or differences between variables
What are some examples of developmental studies ?
- cross-sectional: data collected at one timepoint
- retrospective: using data that has already been collected to link to present events & can’t manipulate IV
- longitudinal: collect data from the same group at different points in time (more than one timepoint=better description of the variable)
What are some advantages and disadvantages to non-experimental designs ?
Advantages:
- efficient/effective
- large amounts of data
Disadvantages:
- least ablet o demonstrate a potential causal relationship
- least amount of control over relationships
What is internal validity ?
the degree of confidence that the relationship or difference the researcher is testing is not influences by other factors or variables
What is instrumentation ?
changes in measurement or observational technique that may account for changes in the obtained measurement (outcome)
What is attrition ?
aka mortality
- loss of subjects over time due to death or dropping out of study
How does testing affect validity ?
testing someone more than once may influence their responses on a follow up post-test
How does diffusion affect validity ?
participants in one group become familiar with the treatment of another group
- they can then either copy that treatment or are just otherwise affected by what they have learned
- no longer “naive” and this changes their behavior
What is external validity ?
the degree of confidence that the findings of the study are applicable to different groups of people or environments
- aka generalizability
- no study can be fully generalizable
What is selection effects ?
results of findings are not generalizable to other populations
What are reactive effects of external validity threats ?
subjects responses to being studied
- subjects may respond to the investigator just because they are being studied (hawthorne effect)
Wha is measurement effects ?
if you are doing a pretest, your study findings may not be applicable to other populations
What is the Hawthorne effect ?
people change their behaviors when they know they’re being studied
What are some characteristics of observational measures ?
directly observing behaviors of people under certain conditions
- requires a standardized and systematic plan for observing and recording data
What are some pros and cons of observational measures ?
Pros:
- direct assessment of your outcome of interest vs. asking about the behavior
- provides more depth and variety then self-report measures
Cons:
- Hawthorne effect
- reliability
- privacy/ethics
What are some pros and cons of self-report measures ?
Pros:
- easy to administer, data less complex but easier to analyze
- used for when you are collecting outcomes that cannot be observed
Cons:
- individuals may be telling you what they want you to hear
- respondent burden: surveys can be long and tire out your participant
What are some types of self-report measures ?
- interview: face-to-face, phone, virtual
- questionnaire: open vs. closed questions
- scales: visual analog, Likert scale
What are some pros and cons of using existing data for data collection ?
Pros:
- accessibility
- no subject burden/cost effective
Cons:
- data often incomplete or inaccurate
- can’t measure things how you would prefer
- privacy
What is sampling ?
process of selecting a portion of the designated population for study in a research investigation
- purpose is to use information for a portion of the population to represent an entire population
- if your sample doesn’t represent your population then this affects generalizability
What is random (probabilty) sampling ?
each person in the accessible population has a statistically equal chance of being selected in your sample
- few studies have access to an entire population
- costly, time consuming, and resource intensive
What is non random (non-probability) sampling ?
unequal chance of being selected, so can be less generalizable
- far more common since its more feasible/realistic
What are some types of random sampling ?
- simple random
- stratified random
- cluster
What is simple random sampling ?
each member has an equal chance of being selected and each member selected independently
- Ex.) draw names from hat, table of random numbers, computer generated
What is stratified random sampling ?
total population is divided into strata/subgroups then are randomly selected from those groups to create smaller groups which comprise your sample
- goal: achieve a greater degree of representativeness
- ex.) people divided by age and then randomly assigned
What is cluster sampling ?
you divide the population into small clusters/groups and then the clusters are randomly selected to create a sample
- difference from stratified is you’re randomly selecting sites/groups before randomly selecting people
What are some types of non-random sampling ?
- convenience
- quota
- purposive
- network
What is convenience sampling ?
you create a sample from people easily accessible to you
- higher risk of selection bias
What is quota sampling ?
you set a quota for a certain characteristic and then you keep selecting until you have our desired number
What is purposive sampling ?
conscious selection of subjects you want based on more complex criteria
- very common in qualitative studies where the researcher is intending to recruit pt’s that are typical of the population
- may be bias on the part of the researcher regarding who is targeted and the reason
What is network sampling ?
- subjects are asked to refer others
- aka snowball sampling
- often used when samples are difficult or impossible to locate
- social media or in-the-field sampling is often used to connect to communities
What is content validity ?
experts say you’re measuring what you think you’re measuring
What is criterion-related validity ?
degree to which a participant’s performance on an instrument and their actual behavior are related
- whether a test can accurately predict something it’s supposed to measure, based on a comparison with another related test or real-world outcome.
What is construct validity ?
fits into the theoretical model you have developed based on your review and appraisal of current literature
- ensuring that the test is truly assessing what it claims to assess, rather than something else.
What is reliability ?
extent to which the instrument yields the same results on repeated measures (consistency)
- concerned with stability, homogeneity, and equivalence
What is test-retest method (stability) ?
the same test (measurement) is administered 2 times
- same subjects each time
- correlate results
What is homogeneity (internal consistency) ?
do all the items or question represent or measure the same concept
What is equivalence ?
agreement between alternate forms of an instrument
- both versions of a test should score the same
What is a power analysis ?
statistical calculation for sample size based on the minimum sample needed to detect a significant effect in the study