statistical inference

we can use to describe the error associated with our sample mean

Standard Error of the Mean (SEM)

inferential statistics

stats that estimate population characteristics from samples and use these estimates to determine the probability of observing some relationship between our variables of interest

the accuracy with which samples estimate population characteristics depends on...

measurement error and sample size

the frequency distribution of a continuous trait in a population is assumed to approximate what

a normal curve

95% of all values ina normal curve will fall within what value

1.96

the frequency distribution of a continuous trait in a sample is assumed to approximate what curve

a t-distribution

t-distribution

a flatter normal curve

the smaller the sample size, the _____ the curve

wider/flatter

Degree of freedom (df)

the number of values that are free to vary (n-1)

Confidence intervals (CIs) for 95% indicates

the range of values that we are 95% confident contains the estimated population parameters

the guarantee that the mean and SD of our sample almost never term match the population

sampling error

null hypothesis

studies that prove that the independent variable does not affect the dependent variable

When is teh alternate hypothesis accepted

if we disprove the null

alternative hypothesis

the independent variable affects the dependent variable

p-value

the probability of observing some effect under the null purely by chance

What values is the p-value based on

magnitude of an effect, variability of effect, sample size

When interpreting a p-value, how should it be read

it is the probability that something occurred by chance and null is true

directional hypothesis

the data only follows one direction (1 tailed)

nondirectional hypothesis

data is unpredictable (2 tailed)

type I errors

saying a relationship exists when it actually doesnt (false positives)

Type II errors

mistakenly accept the null hypothesis. (misses)

significance level (alpha)

the p-value that we are willing to accept in order to reject the null

What value is alpha normally

0.05

what is used to define the critical value

alpha and the sample size

critical value

represents how many standard error that 2 scores must be from on another to be considered different

the rate of Type I errors is set by

significance level

the rate of type II errors is set by

power

what are type II errors set by

beta=0.2

power =

1-beta (almost always .2)

power analysis function values

alpha, variance, sample size, effect size

minimal detectable effect (MDE)

the smallest effect that a statistical test can detect

central limit theorem

as we sample a population, the frequency distribution for sample means will approximate a normal distribution

parametric statistics

stats that assume we can estimate population parameters from our sample

what are the assumptions made for parametric stats

-data is derived from populations that approximate a normal curve
- each data point is an independent observation
-the variance in our samples are fairly close to one another
-the data are measured on interval or ratio scales

when parametric assumptions are violated, what do we have

nonparametric statistics

nonparametric statistics

do not try to establish population parameters