hypothesis testing Flashcards
how many steps are in hypothesis testing
5
what are the steps to hypothesis testing
- research alt hypoth ; H1
- create null H0
- construct sampling distribution of stat on assumption null is true
- collect data from random sample of pop and compute sample stat
- compare sample stat to critical value of distrib
what does hypothesis testinf involve
determining how likely obtained result would occur if null hypoth was true
what is probability represented by
p value
what does the probability estimate
likelihood of obtaining result if null is true
what makes the p value significant so it can reject the null
if specific critical value obtained
what is a 1T test
specifies direction of deviation from null hypoth
to obtain p of .05 or less; specified side of null hypoth distrib must be .05 or less
what is a 2T test
does not specify direction of deviation from null; both sides of null considered
to obtain p of .05 or less, eah side must be .025 or less
what is H0 equal to
The null hypotheses will be X is equal to mu.
the individual observation is from the (specified) population.
what is H1
The experimental hypothesis will be X is not equal to mu
the individual observation is not from the (specified) population.
what is the APA format for a simple hypoth
We can(not) conclude that the individual observation did not come from a population of \_\_\_\_\_, z =\_\_\_\_, p (ns) or < \_\_\_.
what is the process of hypoth testing under normal distrib
- State hypotheses
- Convert score to Z score
- Compare the obtained value with the critical value of ±1.96
- Make statistical decision
- Write up results in APA format
what is a simple hypothesis
Normal distribution for sample statistics (mean) when o is known
use central limit theorem to obtain sampling distrib of mean when H0 true
what are the steps of simple hypotheses testing
- State hypotheses
- Compute Standard Error
- Convert sample mean to a z-score:
- Compare the obtained value with the critical value of ±1.96
- Make statistical decision
- Write-up the finding in APA format
what happens when we do not know the population SD
adjust formula to compute one sample t test
how do you compute t
The standard error for a one sample t-test is s divided by the square root of the sample size.
t is equal to (x bar minus mu) divided by the standard error.
give steps for sample mean when SDfor pop is unknown
- State hypotheses
- Compute Estimated Standard Error:
- Convert score to t score:
- Compute Degrees of Freedom: Number of independent
information remaining after estimating one or more parameters; (df): N-1 - Look tcritical and make statistical decision
- Write up finding in APA format
when is t significant and what can you do if it is
t (tobtained) is larger than tcritical, then you can reject the null hypothesis. compare the tobtained with the tcritical to make your statistical decision
what is the difference between a z and t test
z has SD known whilst t SD isunkown
what is the confidence interval
range in which think true pop mean will be within specified level of confidence
what are the steps for computing confidence interval on the mean
- Compute the sample mean and standard deviation
- Compute estimated standard error:
- Compute degrees of freedom: N-1
- Look up critical t
- Compute CI: ± tcritical( )
- Write up finding
what is a type 1 error
Probability of rejecting H0, given that it is true
Designated as α (.05, .01)
–
we can only make a type I error WHEN the null hypothesis is true.
what is a type 2 error
- Probability of failing to reject H0, given that it
is false
Designated as β
what happens if we minimise type1 error by moving critical value left
increase type 2 error
define power
Ability to find statistically significant results
involving sample size, effect size and alpha level
what happens if you increase effect size
power increases
wht happens if you increase sample size
SE reduced; poower increases
what are assumptions of parametric data
- Normally distributed data
- Data within the population is normally
distributed- Homogeneity of variance
- Variances should be the same
- Interval data
- Distance b/t scale points should be equal
- Independence
- Behaviour of one P doesn’t influence another’s
- Homogeneity of variance