Lecture 2: sampling and confidence

Question 1

Define statistic, parameter and estimator

Answer 1

Statistic is the function we use to estimate the parameter in the population

Statistic is the estimator of the model parameter

Estimator of the population mean is the statistic the sample mean

Greek characters for the population
Latin characteristics for the statistics

Question 2

What is random variation/sampling error?

Answer 2

Due to randomness and changes in who is included in the sample

The statistic and estimated value for a population parameter will vary

Question 3

What is the difference between random error and systematic error?

Answer 3

Random error is due to unknown factors and can go in either variation

Systematic error or bias is consistent and repeatedly under/overestimates the true value - as its due to known factors it can be traced and reduced

Question 4

What is the sampling distribution?

Answer 4

A distribution of point estimates for a population mean computed using different samples

• a normal distribution is created
• called central limit theorem

Question 4

What is the sampling distribution?

Answer 4

A distribution of point estimates for a population mean computed using different samples

• a normal distribution is created
• called central limit theorem

Question 5

In normal distribution how observations are +/- one SD from the mean?

Answer 5

68%

90% are +/- 1.65 SD from mean

95% are +/- 1.96 SD from the mean

99% are +/- 2.58 SD from the mean

Normal distribution can also be called Gaussian distribution

Question 6

What is the equation for variance and SD

Answer 6

Variance = variability squared / N

SD = square root of (variance)

Standard error = variance / square root of sample size

Question 7

What is the equation for the confidence intervals?

Answer 7

mean -/+ 1.96 x (SD / square root of sample size)

Therefore if mean and standard error remain the same to get a wider CI - smaller sample size