chi - squared test
x^2 = sum of (O-E)2 / E
o is observed frequency - so eg 2 robins at site 1
How to work out frauency using chi squared
1- Expected frequency - row total X column total / overall total
2 - X^2
3- degree of freedom (rows minus 1 times column minus 1)
4 - use probability table to get critical value
5- If chi squared value is more than critical value then there is a significant difference -> can reject null hypothesis.
If less than critical value u can reject.
unpaired student t-test
looking at comparing means for 2 diff groups of individuals. W independent data, interval
degree of freedom for unpaired t test
df = n1 + n2 - 2 (number of values in data set)
t- test formula values
X1 = mean of set 1 or 2
s = standard deviation
n = total number of data
ignore any minus signs
paired t- test
when data is taken from the same individuals, both groups are the same size, each measurement is one group paired with measurement in other group.
values in paired/ related student r-test
D = mean of the differences, n is number of pairs, s is standard deviations of differences
degree of freedom for related/ paired t-test
df = n-1
when can a null hypothesis be rejected
if the value is greater or equal to critical value
correlation coefficient
1 is the strongest positive.
0 is none
-1 is strongest negative.
when investigating weather 2 continuous variables are related
% error
PE = uncertainty / measured value x100
used when discussing reliability or accuracy.
if measurement taken twice the uncertainty is doubled.
