Labs- BIO OCE

contains ANOVA info

what 4 dimensions can most units in biology be transformed into

mass, length, time, and numbers

what are residuals

error amount not explained by the model

how to calculate residuals

data point - model, usually in y direction y-m

why is the distinction between model and error important

helps build better models, test hypothesis, make better predictions, machine learning work, Ai work

how to calculate standard deviation

calculate mean, calculate distance between mean and data, square those points, get total error by adding those, divide error by number of samples, that is variance, square root of variance

what does in situ fluorometer measure

measures chla, FDOM, pigments

how to make a standard curve

measure the signal of a set of known standard solutions, plot those values to find a line of best fit, and use that equation to calculate the concentration of unknown samples

what does it mean to use linear regression to make a standard curve

plotting known analyte concentrations (x) against their measured instrument responses (y, such as absorbance or fluorescence), and using a least-squares statistical method to draw a straight line of best fit (y = mx + b)

Describe the spectrophotometer. What exactly is it measuring and how does it do so?

an analytical instrument that measures the amount of light a substance absorbs or transmits across specific wavelengths of the electromagnetic spectrum

why does lugols need to be in the dark

light exposure causes photochemical decomposition

where is chla found

present in all photosynthetic organisms

where is chlb found

characteristic of ‘green algae’ such as chlorophytes and Prochlorococcus

where is chlc found

typically found in high concentration in diatoms and brown algae

steps to an anova

1. have a null and alternative hypothesis

2. find number of samples in each group (ni) and total sample size (N), number of groups (k), find group mean (xi) and. grand mean (xgrand)

3. calculate sum of squares,

4. calculate degrees of freedom

5. calculate mean squares

6. compute f statistic

7. make a decision to accept to reject null

three sum of squres

sum of squares between, within, and total sum of squares

what does sum of squares between measure

variability between groups and grand mean

what does sum of squares within measure

variability within groups

what does total sum of squares measure

sum between and within

what is grand mean

the overall average of all data observations across every group combined

grand vs group mean

grand is total mean and group is each group

three degrees of freedom

between groups, within groups, total

degrees of freedom between groups

k-1, groups -1, measures the number of group averages that can freely change before the final group's average is mathematically locked in to match the overall grand mean

degrees of freedom within groups

N-k, sample size - groups, total number of individual data points that can freely vary inside each group after that specific group's average is already locked in.

degrees of freedom total

N-1, sample size -1, representing how many scores can freely vary before the overall grand mean is locked in

three types of mean squares

mean square between and within, and total

mean square between

SSb/(k-1), sum squares between/ groups-1, measures variability between different groups relative to grand mean

mean square within

SSw/(N-k), sum squares within/(sample size- groups), measures variability within groups

mean square total

SSt/N-1, sum squares total/(sample size-1), measures variance

f statistic

MSB/MSW, mean square between/mean square within, measures whether means of three of more groups are significantly different from each other

how to look at f statistic

compare to f table to see if it significantw

when is f significant

when your calculated test statistic is greater than or equal to the critical F-value found in the F-table

how to calculate p value from anova

you cannot, must be from a software

how to calculate p value by hand

use a z or t test, use a statistical table to find p value

when is p value significant

when it is less than 0.05, p<0.05

define the p-value of a statistical test

It measures how surprising your data is if nothing unusual or real is happening

smaller p value meaning

Stronger evidence against the null hypothesis, meaning the result is unlikely to happen purely by random chance

If p-value < 0.05

reject the null hypothesis and call the result statistically significant, real effect likely caused the data

if p value > 0.05

fail to reject null, it means that any variation in your results is likely due to random chance alone

how to calculate r2

1- (SSw/SSt)

sum of square within

sum(xij- xj)², xij = jth observation in ith group, xj = mean of ith group

sum of square between

sum*ni*(xi-xgrand)², ni = sample size of ith group, xi = mean of ith group, xgrand = grand mean of all observations

R²

how good your model is at explaining the data

R² = 1, R² =0, R² =0.8

1 = model explains everything, 0 = model explains nothing, 0.8 = model explains 80% of variation