Week 11 Flashcards
Define the null hypothesis and alternative hypothesis
Alternative is when you predict that change will occur, p is less than 0.05 therefore we conclude there is a real relationship in the population
Null is where you predict nothing will happen, p is greater than 0.05, not sufficient evidence of a relationship between our measures
Define the t value
An indication of the signal to noise ratio between groups being tested
the difference between the mean or average scores of two groups, while taking into account any variation in scores.
Define a degree of freedom
Total number of measurements in your sample minus the number of groups you have
df = (n1 +n2) -2
A t-test is the most basic statistical test - define the two types
Independent groups - sample involved two different unrelated categories
Paired sample - measures from the same group taken (repeated) twice like before and after
What is Ho and Ha
Ho - null hypothesis
Ha - alternative hypothesis
Define the power of a study
The probability that the study will detect a difference in measurement between the two groups if such exists
Define a type 2 error
There are real systematic differences between groups that are not detected
Failing to reject the null hypothesis when it is false
Often due to a small sample size
Define a type 1 error
True null hypothesis is rejected (author concludes there is a difference when in fact no difference exists
Define the p value simply
the probability that the observed result occurred by chance alone
Define the null hypothesis and alternative hypothesis
Alternative is when you predict that change will occur, p is less than 0.05 therefore we conclude there is a real relationship in the population
Null is where you predict nothing will happen, p is greater than 0.05, not sufficient evidence of a relationship between our measures
What do you need to calculate a confidence interval
sample size
standard deviation
level of confidence
What happens if the lower limit of a confidence interval crosses the vertical line
It means the point has no effect and that the findings from such a trial are not statistically significant
What happens to the confidence interval if your sample size decreases
It becomes wider as we are less certain about the true size
Define confidence intervals
Either side of the point estimate that tells you how much a point estimate may vary in the population
In Ryan terms if your value is within the confidence interval than we are 95% sure that it will occur
PLZ GO BACK TO NOTES TO UNDERSTAND ODDS AND RISK RATIO TAKE TIME AS CONFUSING
thanks xx
In odds ratio and relative risk ratio define the meaning of each range
0 - less than 1 - reduced risk of even
Greater then 1 - increased risk of event
=1 - No effect - risk of event is no better than chance
0 - Event will never happen
Define the Standard Mean Deviation/Cohen’s D
Expresses the absolute change relative to the standard deviation
(Mean 1 - Mean 2) divided by pooled standard deviation
0.2 - represents a small change
0.5 - represents a moderate change
0.8 - represents a large change
