Unit 6.1: Statistical Analysis
No description yet. This curated deck is ready for daily study.
Card preview (9)
What are error bars?
Graphical representation of data variability
What do error bars represent?
• The mean, and how variable data is with regards to the mean • Besides mean, can depict range/standard/deviation/error deviation/other estimates of variability
What is the mos useful thing error bars represent? How can they be interpreted?
• Standard Deviation • When estimate of variability widespread -> mean unreliable/highly variable • When estimate of variability narrowly spread -> mean more reliable
What happens if we assume a test has normal distribution? How is it depicted?
• 68% of data lies between +- 1 Standard Deviations of the mean • 95% of data lies between +- 2 Standard Deviations of the mean • Depicted by Gaussian function (Campana de Gauss)
How is Standard Deviation useful for comparing means and spread of data between two or more samples?
• More standard deviation -> wider variation / less standard deviation -> narrower variation • Overlapping of data = suggests data do not have significant differences
What can Standard Deviation not be used as?
• As an inferential Statistic • Cannot make predictions solely based on Standard Deviation (use suggests…)
What it as t-test? How can we interpret its results?
• Method to determine if there is a difference between the means of two populations • Test will give a p-value = probability of no difference existing between two values • H0 = no significant difference between two samples • H1= There is a significance difference • If P-Value > SL -> Accept H0, If P-Value </= SL -> Reject H0 • If T-Value < CV -> Accept H0, If T-Value > CV -> Rehect H0
What can correlation not be used as proof of?
Proof of cause