The normal distribution and estimation Flashcards Preview

Stats intercal > The normal distribution and estimation > Flashcards

Flashcards in The normal distribution and estimation Deck (12)
Loading flashcards...
1

What is the relative frequency density graph?

Like a histogram but y axis is now density. the area under the histogram is 1. density is defined as the proportion of observations per unit interval of the x-axis

2

What is probability density function?

With an increasing sample size the width intervals on the histogram - with an infinite sample size the line is smooth.
The area under the curve is one
The AUC can be used to calculate the probability within a range of values of interest

3

What is the normal distribution?

Describes a family of bell-shaped probability functions which are totally specified by two parameters

4

What are the parameters of normal distribution?

mean value and standard deviation

5

What is variance?

SD squared

6

What happens with an N(0,1) distribution?

the x-axis is usually labelled z and the cumulative distribution function (c.d.f) denoted ... is the probability that a value chosen at random from N(0,1) distribution is less than or equal to z

7

How do you assess the normality of data?

the mean and median should be approximately equal
the box plot should be symmetric
95% of the data should lie within 2SDs of the mean and almost all within 3SDs
Normal probability plot should be linear

8

What is the standard error?

SE = (population SD) / square root (n)

9

How do we estimate the population mean?

If we assume that the distribution of the sample means in Normal, then we expect 95% of the sample means to lie within 1.96 standard errors of the true population mean

10

What is the 95% confidence interval?

the sample mean +/- (1.96 X SE) and will contain the population mean in 95% of the samples

11

What can we say about a 95% confidence interval?

The limits of a 95% confidence interval contain the true population parameter with 95% confidence

12

What does the width of the confidence interval indicate?

how precisely we have measure the mean - the narrower the interval, the better the precision