week 5 SCM (confidence intervals)

Question 1

what does a confidence interval describe?

Answer 1

• the uncertainty about a statistical estimate
• it describes the range of values or intervals within which the true population parameter will be expected
• it gives an average estimation with a given probability such as 95%. This means that 95% of the time the CI rate would capture the true population parameter

Question 2

what does it tell you if the confidence intervals are overlapping/ not overlapping?

Answer 2

• if they are overlapping the samples are likely very similar to each other
• if they are not overlapping the samples are probably taken from different populations

Question 3

how to calculate the confidence interval with a 95% population parameter

Answer 3

CI= mean +/- 1.96 x SEM

1.96 because 95% of z scores fall between -1.96 and 1.96

SEM because this shows us the standard deviation of the sampling distribution of means (how much the sample mean is likely to vary from the population mean)

Question 4

what can you assume when the number of samples is greater than 30?

Answer 4

a normal distribution

Question 5

what measure of the sampling distribution would you use when the sample is smaller than 30?

Answer 5

a t-distribution

a t-distribution is slightly broader, with a width that is widest for smaller sample numbers, adjusting confidence intervals to make them slightly wider

Question 6

how would you calculate confidence intervals from t-distribution

Answer 6

we use adjusted critical values called a critical t or tcrit

we find this value in a table with degrees of freedom (n-1)

CI= mean+/- t-crit x SEM

Question 7

when would you use a one tailed statistical test vs a two tailed statistical test?

Answer 7

when you have a hypothesis with a specific direction that you want to test

e.g if you wanted to test is more effective than a different drug you could use a one tailed test, but if you also wanted to include the possibility that it was less effective, you would use a two tailed test

Question 8

does SEM increase or decrease as sample size increases?

Answer 8

decreases

Question 9

do confidence intervals get bigger or smaller as sample sizes increase?

Answer 9

they get smaller