stats + scientific method Flashcards
what is discovery science
describing nature
qualitative data
observations
what is hypothesis-based science
explaining nature
quantitative data
experiments
steps of scientific method
- make observations
- ask questions
- form a hypothesis
- make a prediction
- test prediction
- interpret data
- draw conclusions
what does testing the prediction involve
plan experiments
perform experiments
collect data and analyze
what is a theory
well-substantiated explanation for a set of verified hypotheses
what is an independent variable
variable that is being manipulated
location
what is a dependent variable
results from the change to the independent variable
presence of mutations
what are the types of statistics
descriptive and inferential
what are descriptive stats
central tendency
measure of variance
what is central tendency
mode mean and median
what is the mode
most common number in a set of numbers
what is the median
the middle value in a set of numbers
what are the measures of variance
variance and standard deviation
what are inferential stats
error bar analysis
unpaired/independent t test
what are the characteristics of a normal distribution
all central tendencies are the equal
bell shaped curve
most of the values and in the middle
curve is symmetrical
what happens when the mean is standardized to zero
majority of points will be grouped in the middle
what is variance
relative distance between data points and mean
S^2
what is standard deviation
how far apart are the numbers in the sample
SD
what is the main advantage of using standard deviation compared to variance
standard deviation is in the same units as the data points
what is standard error
how accurate your results are compared to the population
SE
what does having a higher S^2, SD or SE mean
there is more variability in the sample and we can be less confident in the results
n < 10, gap is 2
likely significant, p is around 0.05
n > 10, gap is 2
significant, p is around 0.01
n > 10, gap is greater than 2
significant, p < 0.05
n > 10, gap is 1
likely significant, p around 0.05
n < 10, gap is greater than 2
significant, p < 0.05
n < 10, gap is 1
not significant, p > 0.05
gap is 0
not significant
when can you not use an unpaired t test
before and after of the same group
no means
comparing more than two groups or only one
t test steps
- state hypotheses
- write down data
- write number of replicates
- calculate means
- calculate variance
- calculate standard deviation
- calculate t value
- pick a significant value (0.05)
- figure out if hypothesis is one or two tailed
- find t critical
- compare values
p value > alpha
no significance
alpha > p value
significant
how does the hypothesis being two tailed affect the alpha value
alpha becomes 0.025 for each side
t crit < t cal, negative one tailed
not significant
t crit > t cal, positive one tailed
not significant
t crit < t cal, positive one tailed
significant
t crit > t cal, negative one tailed
significant
t crit > t cal, two tailed
significant
t crit < t cal, two tailed
significant