exam 1 Flashcards
sample error definition
deviation between population and sample
three major causes of sample error
- small sample size
- sample bias
- non-independence
small sample size description
when you take a small sample, you may miss variation & it is always smaller than population size.
sample bias description
doesn’t represent entire population
non-independence description
- when you measure the same person or thing more than once
- or pseudo-replication
confidence interval
more obvious than standard error
standard error of mean
standard deviation/sample size
Confidence intervals
gives a range around the sample mean in which the population mean should fall.
if the mean is 65 days and the confidence interval is 25 days. where does the true population mean fall?
- 65-25=40
*65+25=90
observational studies
- look for patterns in naturally occurring variation to test a hypothesis
- always correlational
- should be 1st step
experimental studies
- use carefully controlled experiments to test a hypothesis
observational studies benefits and costs
- benefits: generally easier, faster, and cheaper. you don’t have to manipulate anything
- costs: knowing two variables are related does not tell you how or why. cannot understand cause and effect
blank
control for the absorptions
wild-type
this is whats common in that population
sham
acts as placebo
- still does the same procedure without adding anything new
vehicle
- administered using other mediums
- could be the saline making an effect and not the drug in an IV
placebo
- prevents patient from knowing if they are getting a drug or fake drug
- researchers also do not know
placebo effect
cortisol is a immunosuppressant
- it can effect the immune system just by thinking
variable
- anything you are measuring or manipulating
independent variable
cause
- factor
- predictor
- explanatory variable
dependent variable
effect
covariate
- like independent variable but you aren’t worried about how it changes itself
- ex. light changes so we worry about levels of light that effects the plant
response variable
- will create variation
continuous
- every value within the designated range is possible
- can be plot on a scale
- independent variable
categorical
- not numerical; discrete groupings
- cannot use mean
- dependent variable
ordinal
somewhere in between; categorical with an ordering structure
- low, med, high
- things that cannot be used in categorical, can be used here
replication
- you need to have a sample size bigger than 1
- minimizes sample error
power
the likelihood of rejecting a truly false null hypothesis given a specific sample size and effect size
- should be able to reject null hypothesis at least 80%
pseudo replication
- researchers think their data is independent and treat it like such
- has really bad controls
- fake
what is central tendency?
the average
what are the 3 major measures of central tendency we have discussed so far?
mean, median, mode
mean
common average
median
number that divides high and low on the scale
mode
the most common value that appears
difference between inferential and descriptive statistics?
inferential: used to test hypothesis. last step will be calculating p-value
descriptive: used to summarize data and point out important trends, what the avg is like
what is the probability of flipping a coin 3 times and getting heads all three times?
12.5%
what is the probability of flipping a coin 3 times and getting heads atleast once?
7/8
compute the descriptive statistics for the set of numbers below
2,8,10,15,15,15,30
mode: 15
median: 15
mean: 13.57
summation: 453.73
variance: 75.62
Standard deviation: 8.69
standard error:3.28
a drug trial is underway to determine whether or not a drug slows the growth of cervical tumors. what is the null hypothesis for the drug?
the drug has no effect on the growth of the tumor
what does the p-value directly quantify?
- the probability of a false positive if the null is rejected
of all the emasurements of central tendency, the ___. is the most sensitive to outliers/ zeroes
mean
