4 - Confidence And hypothesis Flashcards
(15 cards)
What is Poisson distribution in terms of…
n
The variance in relation to the mean
Distribution
- N - numbers/counts in specified intervals
- Variance:mean - variance equal to mean. Mean typically close to 0. As mean increases it looks like flattened normal distribution
- distribution - approx randomly distributed
What factors are needed to construct the confidence interval for either a sample or a population
Sample or population mean
Level of confidence (e.g. 95%)
Standard error of the sample mean OR standard deviation for a population
What is a z score?
Number of standard deviations that a given observation is away from the population mean
Z = (x-mean)/standard deviation
How do you calculate a t score?
T = (observed value - sample mean) / sample standard error
P=0.95 (95.44%) is the probability that an observation within + or - of _____ x standard deviation.
1.96x standard deviation
P=0.99 (___%) Is the probability that an observation is + or - ____x standard deviation
(99. 74%)
2. 58 x standard deviation
What are the confidence limits?
Limits between which estimated parameters are likely to occur. Expressed as a confidence coefficient. Aways 2 numbers, not the range.
What is the confidence interval?
Range of values between upper and lower limits, between which a sample is likely to fall.
For a 95% confidence limit : upper and lower limits of confidence interval are given by:
Mean + or - (standard error x ____)
Where ____ is a critical number (t value)
- 96
1. 96
What does 95% confidence interval mean?
95 times out of 100, the sample mean will be within 95% confidence limits
How do you calculate sample standard error?
Sample standard deviation / square root of sample number
How would you calculate degrees of freedom?
N-1
On a box plot, where are the extreme values in a positive skew? Where is the median?
Extreme to The right so the median is to the left.
What is an alpha level?
Level of significance to reject the null hypothesis
If p < p value for significance
Reject the null hypothesis - there is a difference!