A. Why do we sample?

B. What is inference?

C. Because the sample is just a portion of the population, the sample ________ (will or won't- choose one) be exactly the same as the population.

D. Variation due to sampling is called _____________, and even perfectly random samples are subject to this.

E. A _______ is the best way to give an estimate of the population, not one single value.

F. For a population average (ex. Average of all apples in the orchard), what is used as the best estimate of the population average?

G. We use a confidence interval to express _______________.

H. The width of the confidence intervals depends on 2 things. What are they?

I. If you have high variation in your population, we would be _______ sure (more or less - choose one) that the sample mean was close to the population mean.

J. Greater variation within a population leads to a _______ (wider, narrower - choose one) confidence interval.

K. Which size sample is a better estimate of the population?

L. Small samples lead to ____ confidence intervals (widerarrower - choose one).

M. What happens to the effect of sampling error with larger samples?

N. Are perfectly random samples subject to sampling error?

O. Which one matters more: sample size or population size?

*even when we have random samples, it is not completely reliable can be fixed by changing frame or method Potential sources of error statistics about a population in estimating a population distribution using a sample This is the We use samples to estimate proportion and mean (parameters). ONLY error covered by the Sampling margin error Non-sampling error of error errors not related to the act of selecting a sample which is found by the Because the Questionnaire formula: sample is not Poor sampling or Behavioural the whole method measurement effects sample is important population error voluntary response response error, such as lying or which is when Errors in mechanical not remembering people are correctly. asked to call in tasks, such as: or click an arithmetic nonresponse happens online poll. or_entering when some people who are responses in computecontacted refuse to cooperate. wording of questions can undercoverageoccurs when some influence/bias the groups in the population are left out in the results of a sample process of choosing the sample, such as using a survey. telephone directory _for a telephone survey, so unlisted households get missed. really reduce error: need randomization ex:homeless pplor ppl who move around a lot and need to stratify or cluster