Week 5 (z scores, confidence intervals, null hypothesis testing) Flashcards
When are standard scores used?
When we are really interested in describing the relative position of an individual score within a population
What do standardised scores facilitate?
They make it easy to determine how extreme or unusual the school is and makes it easy to compare data from different scales
Do Z scores help to normalise data shape?
No they aren’t a way to make the data shape more normal their calculated from data that is already normally distributed in the raw form numbers are expressed in a different measurement scale
How do you convert a raw score to a Z score?
You subtract the mean from the individual score and then divided by standard deviation
Z=(score-mean)/SD
Converting Z scores to raw scores?
Raw score = Z x SD + Mean
What describes the distribution of a sample?
Mean ± SD describes distribution of a sample
Does sample size affect SD?
No. Sample size does not systematically affect SD
What is standard error?
M ± SE describes sampling distribution
It is the expected distribution of statistics if the sampling repeated many times
Is SE affected by sample size?
Yes. There is an inverse relationship. As sample sizes get bigger, SE decreases
Smaller samples = more variation
What is a confidence interval?
Indicates precision of estimate
Proposes a range of plausible values for an unknown parameter, a likely range
Are CI’s or SE wider?
CI are roughly 2x as wide
Are CI’s affected by sample size?
Yes they are calculated from SE
(SE + z score)
As sample gets bigger, CI get smaller
What do narrow confidence intervals suggest?
High precision
What do wide confidence intervals suggest?
Low precision
Can CI’s be used to describe sample distribution?
No they can’t and shouldn’t be. SD should be used to do so
