Upgrade to remove ads
Only $35.99/year

Stats Test 2

Terms in this set (72)

What test would you use if:
you had 1 group with a normal distribution?
Paired T-test
What test would you use if:
you had 1 group with a non-normal distribution?
Wilcoxon Signed Rank Test
What test would you use if:
you had 2 groups with normal distribution?
Student's t-test
What test would you use if:
you had 2 groups with non-normal distribution?
Wilcoxon Rank Sum Test
What test would you use if:
you had 3+ groups with normal distribution?
ANOVA
What test would you use if:
you had 3+ groups with non-normal distribution?
Kruskal-Wallis Test
What test would you use if:
you had categorical data with less then 10 people in a group?
Fisher's exact test
What test would you use if:
you had categorical data with more than 10 people in each group?
chi-squared test
How do you know if your data is normally distributed?
means
given standard deviation
How do you know if your data is not normally distributed?
medians
When would you use the Pearson's correlation coefficient?
to determine the strength and association between two variables

data is all normally distributed
When would you use the Spearman's rank-order coefficient?
to determine the relationship between two continuous variables with at least one being not normally distributed

ordinal / ranked data
What % of data is within 1 standard deviation of the mean
68.3%
What % of data is within 2 standard deviations of the mean
95.5%
What % of data is within 3 standard deviations of the mean
99.7%
Standard error equation
standard deviation / sqrt(sample size)
The sample is more representative of the entire population when standard error...
approaches 0
95% confidence interval
SD +/- 2*SE
relationship between correlation coefficient and the coefficient of determination
Coefficient of determination
=
(correlation coefficient)^2
what does the correlation coefficient tell us
about the strength and direction of the relationship
type 1 error
rejects the null when the null is true
type 2 error
fails to reject the null when the null is false
quantitative research
determines the relationship between in IV and DV
descriptive research
-subjects measured once
-establishes associations
experimental research
subjects measured before and after treatment
qualitative research
-used in social sciences
-relies on reasons behind various aspects of behavior
Categorical data types
nominal
ordinal
dichotomous
nominal data
counted
no measurement scale and no rank order

ex: blood type
ordinal data
order / ranked data
more than 2 categories of classification
ex: rate pain 1-10, improved, worse, the same...
dichotomous data
whole numbers
implied favorable direction
ex: well vs sick
standard error definition
measure of the variation of the sample means

estimates the probable error of the sample mean's estimate of the population mean
margin of error
range of values above and below the sample statistic in a confidence interval
summarizing categorical data
frequency
ratio
percentage
proportion
summarizing continuous random variables
mean
standard deviation
min / max
median
mode
unimodal
mean = median = mode
right skewed
mode < median < mean
left skewed
mean < median < mode
upper tailed test
an increase is hypothesized
H1: m > m0
lower tailed test
a decrease is hypothesized
H1: m < m0
amount of "risk" you are prepared to take in erroneously rejecting the null hypothesis
type I error
type I error is denoted by
alpha
P value is significant if it is _____ than alpha
greater than
risk that you miss the truth
type II error
type II error is denoted by
beta
power =
1 - beta
a research study needs a power of
0.8 or greater
what affects power?
effect size
significance level
sample size
population standard deviation
effect size
minimum signal that it is important to detect
if sample size is too small
test is likely to be inconclusive
if sample size is too big
use too many resources and time
larger sample size --> _______ power
greater
the more the data varies, the _________ the sample size needs to be to detect an effect
larger
an interval would be parametric or nonparametric?
parametric
a ratio would be parametric or nonparametric?
parametric
an ordinal scale would be parametric or nonparametric?
non-parametric
a nominal scale would be parametric or nonparametric?
nonparametric
counted data would be parametric or nonparametric?
nonparametric
frequencies would be parametric or nonparametric?
nonparametric
regression (r^2)
-can be simple or multiple
-plots best fit line
ordinary least squares regression
deals with continuous data
logistic regression
-non-parametric
-evaluates occurrence of the event of interest in terms of probability (p)
-binary
institution for animals
IACUC
institutional animal care and use committee
institution for peoples
IRB
institutional review board
interventions involving physical or emotional discomfort or sensitive data require what kind of IRB review
full board review
collection of biospecimens by noninvasive means require what kind of IRB review
expedited review
research with existing documents / record collected for non research purposes in which subjects are identifiable require what kind of IRB review
expedited review
research with de-identified records require what kind of IRB review
they are exempt!
anonymous surveys require what kind of IRB review
they are exempt!
phase I of clinical trial
assesses safety, absorption, distribution, metabolism, excretion, bioactivity, drug-drug interactions in healthy volunteers
Phase II of clinical trial
assesses short term side effects and efficacy in subjects with indications
Phase III of clinical trial
assesses safety and efficacy, basis for labeling, and new formulations in subjects with indications
Phase IV of clinical trial
assess new indications, quality of life, and surveillance in subjects with indications
Students also viewed