exam ii: ch11 Flashcards
theory that is used to explain the likelihood that something will happen, no 100% guarantee
probability theory
theory that requires absolute cut off point; assume that all of the groups in a study used to test a hypothesis are components of the same population
decision theory
cut off point chosen before data to test hypothesis; probability level at which statistical results are judged to be significant
level of statistical significance
null hypothesis rejected when it is true
type 1 error
null hypothesis regarded as true when it is false
type 2 error
conclusion/judgement based on evidence
inferences
application of information acquired from a specific instance to a general situation
generalization
theoretical frequency distribution of all possible values in a population; want to aim for closest to perfect bell curve
normal bell curve
extreme score can occur in either tail of the normal curve
- extreme score = higher or lower than 95% of population
tailedness
assumes that extreme score can occur in either tail of the normal curve
- for nondirectional hypothesis
two tailed test
extreme values occur on a single tail of the curve
- for directional hypothesis
- more powerful than two tailed
one tailed test
(n-1), given other score values established from the sum of these scores
degrees of freedom
what type of statistic is used a lot in nursing studies?
descriptive (summary stats)
type of statistics: describe/summarize
- Measures of central tendency
- Mean, median, modie
- Measures of variability
Range, standard deviation, scatter plots
descriptive statistics
type of statistics: Predictions and generalize findings based on data
Analyze data, test hypothesis, determine causality answer research questions
inferential statistics
freq distributions, measures of central tendency, measures of dispersion
types of descriptive statistics
describes middle of sample, summarizes sample
measures of central tendency
greatest freq (not always center)
mode
sum of scores divided by number being summed (average)
- Most stable + least changed, best to summarize data
mean
midpoint (50th percentile)
median
range, variance, standard deviation, standardized scores
measures of dispersion
high score minus low score, uses only two extremes, sensitive to outliers
range
spread or dispersion of scores
- Calc ONLY at interval or ratio level of measurement
variance
square root of variance, the average difference score
standard deviation
raw scores thata cannot be compared + are transformed into standardized scores
- Common = Z score: provides way to compare scores in similar process
standardized scores
most common standardized score; expresses deviations from the mean in terms of SD units
z-score
two scales = horizontal X axis and vertical Y axis; used to illustrate relationship
scatterplots
nonsymettrical distribution, peak of curve off center → becomes issue when data not normally distributed
skewness
known probability of including the value of the population within an interval estimate
confidence interval
compare procedures used with other stats that could have been used to greater adv
statistical suitability
based on accumulated evidence from many studies; important for verification of theoretical statements
- Basis of a science
- Contribute to scientific conceptualization
empirical generalizations
based on assumption that the data fall into a specific distribution, usually the normal (bell-shaped) distribution
parametric statistical tests
specific data not normally distributed
nonparametric statistical tests
t-test, ANOVA, ANCOVA, MANOVA, pearson’s r
types of parametric test
chi-square, spearman’s rho
types of nonparametric tesrs
type of parametric test: requires interval level of measurement, significant differences between two samples
t-test
type of parametric test: tests for differences between mean, more flexible (examine data from 2+ groups)
ANOVA
type of parametric test: examine effect of treatment apart from effect of 1+ potentially confounding variables (pretest scores, age, education, anxiety levels)
ANCOVA
type of parametric test: measure diffs in group means when 1+ dependent variable
MANOVA
type of parametric test: tests for presence of relationship between 2 variables
- bivariate correlation
Pearson’s r
type of non-parametric test: whether two variables are independent or related
chi-square
type of non parametric test: determine the degree of association between two sets of ranks or at the ordinal level (similar to pearson’s r)
spearman’s rho
results of the study that have been translated and interpreted
findings
associated with importance to nursing body of knowledge
significance of findings
established BEFORE study begins
P = 0.05 (if study done 100 times the chance of making error is 5 out of 100)
P = 0.01 (if study done 100 times the chance of making error is 1 out off 100)
- More difficult
level of significance
finding is UNLIKELY to have occurred by chance or fluke
statistical significance
related to practical importance of findings; value of judgement
- findings can have statistical significance but not clinical
P < 0.05 = null
P < 0.4 = reject the null
clinical significance
Restrictions or problems in a study that may decrease the generalizability of the findings
- Ex: new tool
limitations
meanings of conclusions for the body of nursing knowledge, theory, and practice (more specific than conclusion)
implications
synthesis of the findings, main points should be here
conclusions
what 2 things are studied in quantitative studies?
relationships and causality
what makes a good study
applicable to different studies
what does it mean when the mean score is an extreme value?
Population not likely to be the same as that represented by normal curve
SIGNIFICANTLY DIFFERENT
Means risk of making an error
which tailed test is more powerful
one tailed tests
why is analysis throughout research important
to make sure data is accurate and valid
what type of statistic is used a lot in nursing studies
descriptive statistics
where does data analysis begin
descriptive statistics
why is critically appraising results from quantitative/outcomes studies important
determine if researcher’s interpretations of the results are an appropriate eval of the clinical importance of the study’s findings
which measure of central tendency is the most stable and best for summarizing data
mean
what are the only 2 levels of measurement that can calculate variance
interval or ratio
what is the basis of a science?
empirical generalizations
what is bivarate correlation and what type of test is it part of?
relationship between 2 variables, part of pearson’s r
