Flashcards in BIO 330 Deck (379)
Loading flashcards...
61
CV used for
relative measures- comparing data sets
62
sampling distribution
probability distribution of all values for an estimate that we might obtain when we sample a population, centred at true µ
63
values outside of CI
implausible
64
how many quadrats to use
till cumulative number of observations asymptotes
65
law of total probability
P[A] = Σ P[B].P[A I B]
for all B_i 's
66
null distribution
sampling distribution for test statistic, if repeated trials many time and graphed test statistics for H_o
67
Type I error
P[Reject Ho I Ho true] = alpha
68
reject null
P-vale < alpha
69
Type II error
P[do not reject Ho I Ho false]
70
Power
P[Reject Ho I Ho false]
increases with large n
decreases P[Type II E]
71
test statistic
used to evaluate whether data are reasonably expected under Ho
72
p-value
probability of getting data as extreme or more, given Ho is true
73
statistically significant
data differ from H_o
not necessarily important- depends on magnitude of difference and n
74
why not reduce alpha
would decrease P[Type I] but increase P[Type II]
75
continuous probability
P[Y = y] =
0
76
sampling without replacement
ex. drawing cards
(1/52).(1/51).(1/50)
77
Bayes Theorem
P[A I B] = ΣP[B I A].P[A] / P[B]
78
P-value > alpha
do not reject Ho
data are consistent with Ho
79
meaning of 'z' in standardization
how many sd's Y is from µ
80
standardization for sample mean, t =
Ybar - µ / (s / sq.rt. n)
81
CI on µ
Ybar ± SE.tcrit
SE of Ybar
t of alpha(1 or 2), degrees of freedom
82
1 sample t-test
compares sample mean from normal pop. to population µ proposed by Ho
83
why n-1 account for sampling error
last value is not free to vary if mean is a specified value
84
1 sample t-test assumptions
data are a random sample
variable is normally distributed in pop.
85
paired t-test assumptions
pairs are a random sample from pop.
paired differences are normally distributed in the pop.
86
how to tell whether to reject with t-test
if test statistic is further into tails than critical t then reject
87
2 sample design compares
treatment vs. control
88
2 sample t-test assumptions
both samples are random samples
variable is normally distributed in each group
standard deviation in two groups ± equal
89
degrees of freedom
1 sample t-test: n - 1
paired t-test: n - 1
2 sample t-test: n1 + n2 - 2
90