Week 5 quiz Flashcards
quantitative research
explains conclusions by collecting numerical data that are analyzed using statistics
control
ability to manipulate, regulate, or statistically adjust for factors that can affect the DV
manipulation
treatment, implementation or IV in a study to determine it’s effect on the DV
bias
influence that distorts the results of the study
-identify possible sources of bias is critical
-randomization and binding
randomization
true experimental studies include some type of random sampling
-ensures equivalence of groups, eliminates key threat to internal validity
random sampling
each person in population has an equal chance of being selected
random assignment
equal change of being assigned to treatment or control group
single bind
participants do not know which study group they are in
double bind
binding both experimenters and participants
experimental research
(true) experimental research
quasi - experimental research
non-experimental research
observational , no manipulation
correlational / descriptive research
experimental design (randomized control trials)
examines differences b/t treated and untreated subjects
experimental design MUST include
manipulation
control
randomization
experimental design characteristics
-large number of participants from diverse areas
-strict guidelines for study inclusion
-random group assignment
-homogeneity b/t intervention and control
-consistent implementation
-same DV is measured in intervention and control group
experimental design strength
powerful in examining cause and effect relationship
-level II evidence
experimental design weakness
may be complicated to develop and expensive
difficult or impractical for certain clinical settings b/c of ethics
quasi experimental characteristics
-manipulation of IV
-lack of randomization or control
-practical, less expensive, generalizable
-more adaptable to real world setting
-level III evidence
nonexperimental design
-IV is NOT manipulated
-subjects are not random
-no control
-cannot make claims on cause and effect
nonexperimental use
describe phenomenon in detail
explaining or predicting relationships
longitudinal designs
gather data about subjects at more than one point in time
-experimental OR non-experimental
-prospective OR retrospective
longitudinal strengths
assess change in variables overtime
longitudinal weakness
-data collection may take a long time
-testing effects may be a threat
-mortality is a significant threat owing to the increased potential for attrition
cross-sectional design
collects data about IV and DV at the same time
-difficult to determine cause and effect
-non-experimental
cross sectional strengths
-less time consuming and expensive
-large amounts of data can be collected at one point
-confounding variable of maturation resulting from elapsed time
cross sectional weakness
interrelationships not established
data collection plan
set timeline
-one point or repeated measures ?
-most likely to be available
determine data collection methods
-questionnaires
-observation
-scales
develop data management plan
data analysis process
- prepare the data and enter into computer
- clean data file
- run descriptive statistics
- run inferential stats to test hypothesis
categorical measurement
nominal and ordinal
nominal
different in name only
-cannot rank or order
ordinal
can be ranked or ordered but still in categories
Mild
Moderate
Strong
continuous measurement
interval and ratio
interval
fixed unit of measurement without a meaningful zero
degrees
fahrenheit
ratio
fixed unit of measurement WITH meaningful zero
dollars
age
years of education
levels of measurement
ratio
interval
ordinal
nominal : weakest
statistical analysis
descriptive statistics and inferential statistics
descriptive statistics
explain characteristics of variables
-numbers that summarize data
mode
most frequently occurring value in data set
median
center of data set
mean
mathematic average
standard deviation
square root of variance
range
difference b/t two extreme scores
percentage
number or ratio expressed as fraction of 100
inferential statistics
makes predictions about population based on a sample
purpose of inferential stats
test hypotheses
make decisions about whether findings can be applied to population
null hypothesis
opposite of what you’re testing
alternative / research hypothesis
claim / expected results you’re testing
alpha
significance level
need to identify before running statistics
P value
probability value
probability of obtaining results at least as extreme as the result actually observed under assumption null hypothesis is correct
stats clinical use
analyze research collected by nursing staff
reading and critiquing published research
examining outcomes of nursing process
evaluation
examination of admin data
demonstrating a problem or need
