Stats Exam 2 Flashcards
What is a descriptive project
describes situation, not carry out hypothesis test
what is a differences project
compares two sets of data, hypothesis relates to differences between two sets
what is a correlation/regression project
emphasis on linking variables, looking at strength and direction of links between two or more variables
what is an association project
looking for links between variables that are categorical
what graph to use of a descriptive project
histograms, density plots, boxplots
what graph to use for differences projects
bar charts and boxplots
what graph to use for a correlation/regression project
scatterplots and line plots
what graph to use for association projects
bar charts and pie charts
What is the outcome of hypothesis testing
likelihood that a hypothesis is correct
what do we test when we test a hypothesis
the null
how does the p-value help make conclusions
tells the probability that the null hypothesis is correct
what conclusions can we draw with a p-value
reject or fail to reject null hypothesis (no change/difference)
code to make a histogram
hist(object)
code to make a boxplot
boxplot(object)
code to scale x axis
xlim=c(number,number)
code to scale y axis
ylim=c(number, number)
code to label x axis
xlab=”title”
code to label y axis
ylab=”title”
code to title graph
main=”title”
code to change graph color
col=”color”, or col=c(“color 1”, “color 2”)
what kind of test would be used for the alt hypothesis that “a cause has an effect” (not specified positive or negative)
two tailed t-test
what kind of test would be used for the alt hypothesis that “a cause has a positive effect”
one tailed t-test
what is popperian philosophy
one observation can disprove something but it takes an uncollectable amount of data to confirm something 100%
what do frequent statistics measure
probability that a null hypothesis is incorrect
what is the level of probability that scientists use for deciding what to do with hypotheses
0.05
study setup that uses a t-test
finding the probability of drawing two samples with different means from a common population
the increasing of what increases t
mean, standard deviation, number of replicates
necessary to determine if a t-test can be used
Shapiro-Wilk test to determine if data is normal (must be greater than 0.05)
r command from a t-test
t.test(object)
what is the difference between 1 and 2 tailed t-tests
1 tailed specifies direction of change, 2 tailed does not specify direction, only that the means are different from the null
when should paired t-tests be used
when sampling units are not independent and not randomly distributed. repeated observations collected for a single variable with two levels (fishing in the same spot on different days)
how do non parametric tests work
ranking data
r code for nonparametric test equivalent to 2 tailed t-test
wilcox.test(object, object)
what is the nonparametric equivalent of a 2 tailed t-test
Mann-Whitney test
what is the nonparametric equivalent of a paired t-test
Wilcoxon matched pairs test
what is preliminary sampling
small study conducted to refine and evaluate sampling size, acceptability, and feasibility
what is dummy data collection
use research and hypothetical data to decide on number of observations needed to detect and effect
what is primary literature investigation
using previous research and data to learn about a system as much as possible and see how they can help your study design
assumptions when using most stat tests
normality, equal variances, linearity, replicate independence
pseudoreplication occurs when which statistical assumption is vilated
replicate independence
what is pseudoreplication
if replicates are somehow tied together, not completely random
what is simple pseudoreplication
only one replicate per treatment and subsamples are collected from each area and treated as replicates
what is sacrificial pseudoreplication
experimental units are replicated and subsamples are treated as independent samples
what is temporal pseudoreplication
single replicate per treatment and subsamples are collected from it repeatedly over time and treated as replicates
what is phylogenetic pseudoreplication
closely related individuals are the units being sampled
what is technical pseudoreplication
different observers or instruments are used for different parts of the experiment
why must converting to categories must be done carefully
could create a misleading p-value
what is a type 1 error
null is true but is rejected
what is a type 2 error
null is false but not rejected
