Greg Chism
October 25, 2022
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
Objective: Evaluate the effectiveness of cognitive-behavior therapy for chronic fatigue syndrome.
Participant pool: 142 patients recruited from referrals by primary care physicians and consultants to a clinic that specializes in chronic fatigue syndrome.
Actual participants (N): 60 of 142 patients entered the study, some were excluded for various reasons and others refused to participate
Distribution of patients with good outcomes at 6-month follow-up. 7 patients dropped out of the study: 3 from treatment, 4 from control.
Are the results generalizable to all with chronic fatigue syndrome?
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
The number of standard deviations a value falls above or below the mean - e.g., 2 standard deviations: \(Z = 2\)
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
What is the likelihood that your data is different that expected from random variation?
What is the likelihood that your data is different that expected from random variation?
F-Distribution
Probability of observing data at least as favorable to the alternative hypothesis, if the null hypothesis is not true.
Translation: The relationship that we find is not from random chance.
A Type I Error is rejecting the null hypothesis when \(H_0\) is true
A Type II Error is rejecting the null hypothesis when \(H_A\) is true
Power = 1 - Type II Error
https://rpsychologist.com/d3/nhst/
Statistical Power: The probability of accepting the alternative hypothesis if it is true.
https://rpsychologist.com/d3/nhst/
https://rpsychologist.com/d3/nhst/
https://rpsychologist.com/d3/nhst/
Data and Sampling
Distributions
Central Limit Theorem
Statistical Significance & Power
Hypothesis Testing
Can we make any inference about the new data point?
No, the point is outside of the bounds of our regression model.
This would be extrapolation, which is possible, but not advised.
Instead you’d have to run a new model.
Is height (cm) related to Blood Glucose (mmol/L)?
Maybe…
There’s no information on how the data was collected or if this is experimental at all.